 

LIBRARY
Michigan State
University

 

 

 

PLACE iN RETURN BOX
to remove this checkout from your record.
TO AVOID FINE-S return on or before date due.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

0:7 1 62mm

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1/93 WWW“

THE USE OF OZONATION AND BIOLOGICAL FLUIDIZED BED TREATMENT
FOR THE CONTROL OF NOM IN DRINKING WATER

By

Alexander Anatoly Yavich

A DISSERTATION
Submitted to
Michigan State University

in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Civil and Environmental Engineering

1998

ABSTRACT

THE USE OF OZONATION AND BIOLOGICAL FLUIDIZED BED TREATMENT
FOR THE CONTROL OF NOM IN DRINKING WATER

By

Alexander A. Yavich

Bench-scale and pilot-scale studies were conducted using Huron River water
taken at the Ann Arbor Water Treatment Plant. The major water quality parameters that
were monitored in this study included TOC, BDOC, UV-254 absorption, THMFP, AMW
distribution, humic and non humic fractions of NOM, turbidity, pH and alkalinity.

A combination of factors, including ozone dose, hydraulic retention time, and the
concentration of dissolved ozone was identiﬁed, which controlled both the destruction of
organic carbon and the production of low-molecular weight organic compounds and
biodegradable organic matter during ozonation. The relationships between properties of
THM precursors, including TOC, UV-254, humic substances, and THMF P were
established and used to investigate the efficacy of the ozonation and biological ﬂuidized
bed treatment (FBT). A mathematical model that described the transformation of NOM
during ozonation was developed and veriﬁed over a range of ozone doses of up to 3
mg/mg C, temperatures of up to 25°C, and hydraulic retention times of up to 20 minutes.

The biodegradable organic matter in Huron River water consisted of rapidly and
slowly biodegrading fractions (“fast” and “slow” BDOC). The biodegradability of raw
and treated waters was characterized by the maximum biodegradation rate of “fast”

BDOC (Rm), the minimum biodegradation time that required to eliminate “fast” BDOC

(EBCTmin), and by the minimum concentration of “slow” BDOC that remained in the
water after biodegradation at EBCTmin (BDOCslow).

The following treatment processes that included ozonation and PET were
investigated: (1) single-pass ozonation/FBT; (2) ozonation/FBT with recycle; (3) single-
pass FBT/ozonation with bioﬁltration; (4) recirculating FBT/ozonation with bioﬁltration;
(5) single-pass FBT/ozonation/bioﬁltration with the addition of acetate to the PET
column (“stimulat ” PET); and (6) recirculating “stimulated” FBT/ozonation with
bioﬁltration. Among these processes, the recirculating “stimulated” FBT/ozonation
process followed by bioﬁltration was most efﬁcient in terms of the removal of NOM
relative to ozone consumption and biodegradation time. The removal of NOM was
comparable to that achieved at the Ann Arbor Water Treatment Plant, that uses lime
softening, ﬂocculation/sedimentation, ozonation and GAC ﬁltration. For a design
capacity of 1 MGD, the cost of treatment by the FBT/ozonation process followed by
bioﬁltration was estimated to be at least 40 percent lower than that by a conventional

ﬂocculation/sedimentation process with ozonation and GAC ﬁltration.

€0pyright by
Alexander Anatoly Yavich

1998

iv

ACKNOWLEDGEMENTS

I wish to express my sincere thanks to Dr. Masten for the opportunity to start and
complete this research and for her guidance and support during all these years I have
been in graduate program. I also wish to thank Dr. Rajan and Dr. Hickey for the time they
spent discussing with me this work. Those discussions were extremely helpful. I am
grateful to my committee members Dr. Davies, Dr. Davis, once again Dr. Rajan, and Dr.
Zabik for technical and moral support and for the time they invested in reviewing this
work. My thanks are also extended to Dr. Amy and Dr. Collins for support of the ideas
that were realized in this research.

I am thankful to Jeff Cook and Dan Wagner who helped me in building and
operating the pilot-scale system. My thanks are also due to my fellow graduate students
Ajay Kasarabada, Jullie Mellema, and Yvonne Chang who took part in this research. I
also would like to thank Janice Scadsen of the Ann Arbor Water Treatment Plant for
technical support. I am grateful to the EPA Fellowship ofﬁce and NSF for ﬁnancial
support.

The completion of this work would have not been possible without help of my
wife, Yelena, my daughter Dina, and my parents. I wish to thank them for their love and

patience.

TABLE OF CONTENTS

TABLES ............................................................................................................................ ix
FIGURES ........................................................................................................................... xi
ABBREVIATIONS .......................................................................................................... xv
1. INTRODUCTION .......................................................................................................... 1
1.1. Signiﬁcance .............................................................................................................. 1
1.2. Objectives ................................................................................................................ 4
1.3. Project Otline ........................................................................................................... 6

2. BACKGROUND ............................................................................................................ 7
2.1. Characterization of NOM ........................................................................................ 7
2.1.1. Humic Substances ............................................................................................. 8
2.1.2. Organic Carbon Content ................................................................................. 11
2.1.3. UV Absorption ................................................................................................ 11
2.1.4. THMFP ........................................................................................................... 12
2.1.5. Molecular Weight Distribution ....................................................................... 13
2.1.6. BDOC ............................................................................................................. 14

2.2. Review of Treatment Processes ............................................................................. 16
2.2.1. Slow Sand Filtration ....................................................................................... 16
2.2.2. Chemical Coagulation ..................................................................................... 17
2.2.3. Carbon adsorption ........................................................................................... 19
2.2.4. Membrane separation ...................................................................................... 20
2.2.5. Ozonation ........................................................................................................ 22
2.2.6. Combined ozonation/biotreatment .................................................................. 25
2.2.7. Proposed Ozonation/FBT Process .................................................................. 29

3. MATERIALS AND METHODS .................................................................................. 33
3.1. Water Source .......................................................................................................... 33
3.2. Experimental Systems ............................................................................................ 34
3.2.1. Bench-Scale Ozonation System ...................................................................... 34
3.2.2. Ozonation/PET System ................................................................................... 36
3.2.3. PET/Ozonation System ................................................................................... 40
3.2.4. FBT System .................................................................................................... 41
3.2.5. Bioﬁltration system ......................................................................................... 41

3.3. Analytical Methods ................................................................................................ 44
3.3.1. Organic carbon ............................................................................................... 44
3.3.2. UV-254 ........................................................................................................... 45
3.3.3. Humic Substances. .......................................................................................... 45
3.3.4. AMW Distributions ........................................................................................ 46
3.3.5. THMFP. .......................................................................................................... 47
3.3.6. BDOC ..‘. .......................................................................................................... 48
3.3.7. Other parameters ............................................................................................. 48

3.4. Quality Assurance and Control .............................................................................. 49
3.4.1. Data Management ........................................................................................... 49
3.4.2. Sample Collection and Analysis ..................................................................... 50

vi

3.4.3. Data Analysis .................................................................................................. 51

4. PRELIMINARY EVALUATION OF THE OZONATION/FBT PROCESS .............. 53
4.1 . Introduction ............................................................................................................ 53
4.2. Potential of Ozonation/FBT System ...................................................................... 54
4.3. Effect of Ozonation and FBT on Transformation of NOM ................................... 63
4.4. Fractionation of NOM by Ultraﬁltration ............................................................... 68
4.5. Conﬁrmation Study ................................................................................................ 75
4.6. Ozonation/FBT with Recycle ................................................................................ 80
4.7. Summary ................................................................................................................ 82

5. SURROGATE CHARACTERIZATION OF NOM ..................................................... 84
5.1 . Introduction ............................................................................................................ 84
5.2. UV-254 and TOC ................................................................................................... 86
5.3. HS and UV-254 ...................................................................................................... 94
5.4. THMF P and TOC .................................................................................................. 97
5.5. THMFP and UV-254 ........................................................................................... 100
5.6. Summary .............................................................................................................. 101

6. OZONATION OF NOM ............................................................................................. 102
6.1. Introduction .......................................................................................................... 102
6.2. Effect of Operational Parameters ......................................................................... 104

6.2.1. Effect of Gas F low Rate ................................................................................ 104
6.2.2. Effect of Hydraulic Retention Time ............................................................. 110
6.2.3. Effect of Ozone Dose on BDOC ................................................................... 123
6.3. Ozonation Kinetics ............................................................................................... 126
6.3.1. Introduction ................................................................................................... 126
6.3.2. Model Development ...................................................................................... 130
6.3.2.1. Model Based on the Speciﬁc Ozone Consumption Rate ..................... 130
6.3.2.2. Model Based on HS and nonHS. ......................................................... 133
6.3.3. Rate of Ozone Consumption ......................................................................... 133
6.3.3.1. Spectroscopic Method .......................................................................... 134
6.3.3.2. Method Based on Ozone Consumption Model .................................... 141
6.3.3.3. The Use of Ozone Consumption Rate to Evaluate the Efficacy of
Ozonation .......................................................................................................... 144
6.3.4. Reaction of Ozone with Humic Substances .................................................. 152
6.3.4.1. Stoichiometric Coefficient ................................................................... 152
6.3.4.2. Reaction Rate Coefficient ..................................................................... 155
6.3.5. Reaction of Ozone with Nonhumic Substances ............................................ 159
6.3.5.1. Stoichiometric Coefﬁcient ................................................................... 159
6.3.5.2. Reaction rate coefﬁcient ....................................................................... 161
6.3.6. Mass Transfer Coefﬁcient ............................................................................. 164
6.3.7. Ozonation Model .......................................................................................... 165
6.3.8. Model Validation .......................................................................................... 167
6.4. Summary .............................................................................................................. 180

7. BIODEGRADATION OF NOM ................................................................................ 183
7.1 . Introduction .......................................................................................................... 1 83
7.2. Approach .............................................................................................................. 184
7.3. FBT vs. Bioﬁltration ............................................................................................ 186

vii

7.4. Effect of Temperature .......................................................................................... 191

7.5. Biodegradation of Ozonated Water ..................................................................... 195
7.6. Summary .............................................................................................................. 198

8. EVALUATION OF COMBINED OZONATION AND FBT PROCESS ................. 199
8.1. Introduction .......................................................................................................... 199
8.2. Single Pass FBT/Ozonation Process .................................................................... 202
8.3. FBT/Ozonation Process with Recycle ................................................................. 212
8.4. Stimulated FBT/Ozonation Process ..................................................................... 219
8.5. Stimulated FBT/Ozonation Process with Recycle ............................................... 224
8.6. BIODEGRADATION KINETICS ...................................................................... 228
8.7. Summary .............................................................................................................. 232
9. ECONOMIC ANALYSIS .......................................................................................... 234
10. CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE RESEARCH 241
10.1. Conclusions ........................................................................................................ 241
10.1.1. General Conclusions ................................................................................... 241
10.1.2. Surrogate Characterization of NOM ........................................................... 242
10.1.3. Ozonation Studies ....................................................................................... 243
10.1.4. Biodegradation Studies ............................................................................... 245
10.1.5. Pilot-Scale Study of the Combined Ozonation and FBT Process ............... 246
10.2. Recommendations for Future Research ............................................................. 248
APPENDICES ................................................................................................................ 250
Appendix A. List of formulas ..................................................................................... 251
Appendix B. Annual cost analysis .............................................................................. 254
Appendix C. Control charts ........................................................................................ 262
Appendix D. Spreadsheet Formats ............................................................................. 267
REFERENCES ............................................................................................................... 275

viii

Tables

3.1
3.2
4.1
4.2

5.1
6.1
6.2
6.3
6.4
6.5
6.6
6.7

6.8

7.1

7.2
7.3
8.1
8.2
8.3
8.4

TABLES

Typical quality characteristics of Huron River water ......................... 33
QA/QC summary .................................................................. 52
Performance of the ozonation/FBT system ..................................... 62

Comparison of recycle and single pass modes of ozonation/FBT

operation .............................................................................. 81

Characteristics of surrogate parameters ......................................... 84

Experimental conditions for bench- and pilot-scale systems ................. 112
Effect of quenching on UV-254 and TOC after ozonation ................... 135
Experimental data for VnonHS ...................................................... 160
Experimental data for knonHS ....................................................... 162
Experimental data for k1,, ....................................................... 164
Kinetic parameters for ozonation model ....................................... 166

Experimental and calculated data for ozonation of raw water (ozone

dose 1 mg/mg C, HRT = 7 minutes) ........................................... 178
Experimental and calculated data for ozonation of raw water (ozone

dose 1 mg/mg C, HRT = 20 minutes) ........................................ 179
Operational biodegradation characteristics for FBT and bioﬁltration

effluents ............................................................................. 190
Effect of temperature on biodegradation characteristics of raw water ...... 195
Effect of ozonation on biodegradation characteristics of water 197
Efficiency of FBT/ozonation/bioﬁltration and ozonation/FBT .............. 211
Biodegradation characteristics of FBT/ozonation efﬂuent ................... 212
Biodegradation characteristics of stimulated F BT/ozonation effluent. . .. 224

Biodegradation characteristics of recirculating stimulated
FBT/ozonation effluent ............................................................ 228

ix

9.1 Values for major cost, design and operating parameters ...................... 236
9.2 Cost breakdown for the combined ozonation and FBT process ............. 239
9.3 Costs of selected treatment systems ............................................. 239

Figures

2.1
3.1
3.2
3.3
3.4
3.5
3.6
4.1
4.2
4.3
4.4
4.5
4.6
4.7
4.8
4.9
4.10

4.11
4.12
4.13
4.14

4.15

5.1

FIGURES

Comparison of F BT and bioﬁlter operations .................................. 30
Schematic of the bench-scale ozonation system .............................. 35
Schematic of the ozonation/FBT system ....................................... 38
Schematic of a downﬂow ozone contactor ..................................... 39
Schematic of the FBT/ozonation system ....................................... 40
Schematic of the recirculating FBT system .................................... 42
Schematic of the bioﬁltration system .......................................... 43
Simpliﬁed schematic of the ozonation/FBT system .......................... 56
Time plot of TOC removal during acclimation period ....................... 57
Removal of TOC from untreated Huron River water by FBT ............... 58
Variations in TOC concentrations with time and ozone doses . . . . . . . . 60
Variations in TOC concentrations with time and ozone doses .............. 61
Effect of ozone dose on TOC removal after ozonation and F BT ........... 65
Effect of ozone dose on UV-254 removal aﬁer ozonation and FBT ....... 66
Effect of ozone dose on HS removal after ozonation and F BT ............. 67
AMW distribution after FBT .................................................... 71
AMW distribution after ozonation and FBT

(ozone dose—0.5 mg/mg C) ....................................................... 72
AMW distribution after ozonation and FBT

(ozone dose—1.0 mg/mg C) ....................................................... 73
AMW distribution after ozonation and FBT

(ozone dose—2.0 mg/mg C) ....................................................... 74
Effect of ozone dose on TOC removal after ozonation and FBT ........... 77
Effect of ozone dose on UV-254 removal after ozonation and FBT ....... 78
AMW distribution after ozonation and F BT

(ozone dose—7.0 mg/mg C) ....................................................... 79
Seasonal variations of TOC and UV-254 in Huron River water ............. 87

xi

5.2
5.3
5.4

5.5
5.6
5.7
5.8
5.9
5.10
5.11
6.1
6.2
6.3
6.4
6.5
6.6
6.7
6.8
6.9
6.10
6.11
6.12
6.13
6.14
6.15
6.16
6.17

6.18

Correlation between TOC and UV-254 for ozonated water .................
Correlation between TOC and UV-254 for biotreated water ...............
Correlation between TOC and UV-254 for ozonated and

biotreated water ...................................................................
Ozone-dose averaging linear correlation between TOC and UV-254
Exponential relationship between TOC and UV-254 .........................
Correlation between HS and UV-254 ..........................................
Correlation between average values of HS and UV-254 .....................
Correlation between TI-IMF P and TOC .........................................
Correlation between average values of THMFP and TOC ...................
Correlation between average values of THMFP and TOC ...................
cgin requirements for various gas ﬂow rates ....................................
Dissolved ozone concentrations for various ﬂow rates .......................
Effect of ozone dose on TOC concentrations for various gas ﬂow rates...
Effect of ozone dose on UV-254 for various gas ﬂow rates .................
Effect of ozone dose on TOC concentrations for various HRTs ............
Effect of ozone dose on UV-254 for various HRTs ...........................
cgin requirements for various HRTs .............................................
Dissolved ozone concentrations for HRTs ......................................
Efﬁciency of ozone utilization in the bench-scale ozone contactor. . . . .
cgin requirements for bench- and pilot-scale systems ..........................
Dissolved ozone concentrations for bench- and pilot-scale systems ........
Removal of TOC during ozonation in bench- and pilot-scale systems... ..
Removal of UV-254 during ozonation in bench- and pilot-scale systems.
Removal of UV-254 during ozonation at low HRTs ..........................
Effect of ozone dose on BDOC concentration .................................
Kinetic curve of ozone depletion based on UV-254 ...........................
Kinetic curve of ozone depletion based on UV-254 using

base-line correction ................................................................
Correlation between dissolved ozone concentrations measured

spectrophotometrically and by Indigo method .................................

xii

88
89

9O

92

93

95

96

98

99

100
106
107
108
109
114
115
116
117
118
119
120
121
122
124
125
138

139

140

6.19
6.20
6.21
6.22
6.23
6.24
6.25
6.26
6.27
6.28
6.29
6.30
6.31
6.32
6.33
6.34
6.35
6.36
6.37
6.38
7.1
7.2
7.3
7.4
7.5
7.6
7.7
7.8
8.1
8.2
8.3

Correlation between ko3s determined by different methods ................
Effect of ozone dose on K» for various gas ﬂow rates ........................
Correlation between UV-254 and km for various gas ﬂow rates ............
Correlation between TOC and kg; for various gas flow rates ................
Effect of ozone dose on km for various HRTs .................................
Correlation between UV-254 and kg; for various I-IRTs ......................
Correlation between TOC and km for various HRTs .........................
Transformation of organic matter during ozonation ...........................
Relationship between HS and km, for various temperatures .................
The plot of Arrhenius’ equation for the reaction of ozone with HS. .. . .
The plot of Arrhenius’ equation for the reaction of ozone with HS ........
Rate of ozone utlization ...........................................................
Experimental and calculated data for cgout (t=20°C, HRT=7 min) ..........
Experimental and calculated data for c], (t=20°C, HRT=7 min) .............
Experimental and calculated data for HS (t=20°C, HRT=7 min) ............
Experimental and calculated data for nonHS (t=20°C, HRT=7 min) .......
Experimental and calculated data for cg°ut (F20°C, HRT=20 min) ........
Experimental and calculated data for c1, (t=20°C, HRT=20 min). . . .
Experimental and calculated data for HS (t=20°C, HRT=20 min) ..........
Experimental and calculated data for nonHS (t=20°C, HRT=20 min). . .
Biodegradabale and nonbiodegradable fraction of NOM .....................
Removal of UV-254 by FBT and bioﬁltration .................................
Removal of TOC by FBT and bioﬁltration .....................................
Removal of BDOC by FBT and bioﬁltration ...................................
Removal of TOC by FBT for various temperatures ...........................
Removal of UV-254 by F ET for various temperatures ........................
Removal of BDOC by FBT for various temperatures .........................
Biodegradation of ozonated samples ............................................
Simpliﬁed schematic of the FBT/ozonation system ...........................
Removal of UV-254 during ozonation of raw water and FBT effluent. . ..

Removal of TOC during ozonation of raw water and F BT effluent ........

xiii

143
146
147
148
149
150
151
154
157
158
163
169
170
171
172
173
174
175
176
177
185
187
188
190
192
193
194
197
204
208
209

8.4
8.5
8.6
8.7
8.8
8.9
8.10
8.11
8.12

8.13
8.14
8.15

8.16
9.1
9.2

AMW distribution of NOM aﬁer FBT and ozonation ........................ 210

Biodegradation kinetics of the FBT/ozonation effluent ...................... 21 1
Simpliﬁed schematic of the FBT/ozonation system with recycle... . . . . . 213
Effect of recycle on the removal of TOC by FBT/ozonation ................ 214
Effect of ozone dose on the removal of TOC by FBT/ozonation ............ 216
Effect of recycle on the removal of UV-254 by FBT/ozonation ............ 217
Bioﬁltration of the FBT/ozonation effluent .................................... 218
Effect of acetate dose on the removal of TOC by F BT/ozonation .......... 221
Bioﬁltration of effluents from FBT/ozonation and enhanced

FBT/ozonation ..................................................................... 222
Biodegradation kinetics of enhanced FBT/ozonation effluent ............... 223
Bioﬁltration of effluents from enhanced F BT/ozonation with recycle. . . .. 226
Biodegradation kinetics of efﬂuents from enhanced F BT/ozonation with

recycle ............................................................................... 227
Relationship between BDOC concentration and biodegradation rate ...... 231
Simpliﬁed schematic of the FBT/ozonation process with recycle ........... 235
Effect of recycle on treatment cost ............................................... 240

xiv

SDWA
DBP
THM

THMFP

TOC
UV-254
HS
NonHS
AMW
FBT

EBCT

MWCO
mg C/L

RSD

ABBREVIATIONS

Safe Drinking Water Act
Disinfection by-products
Trihalomethanes

Trihalomethane formation potential
Haloacetic acids

Total organic carbon

Ultraviolet absorbance measured at 254 nm
Humic substance concentration
Nonhumic substance concentration
Apparent molecular weight

F luidized bed treatment

Empty-bed contact time

Hydraulic retention time

Molecular weight cutoff

Milligram carbon / L

Relative standard deviation

XV

1. INTRODUCTION

1.1 . SIGNIFICANCE

A signiﬁcant portion of the water treatment industry will face major technical and
ﬁnancial hurdles in meeting the disinfection by-product (DBP) requirements of the Safe
Drinking Water Act (SDWA) reauthorization. Under the Disinfectants/Disinfection By-
products (D/DBP) rule, the United State Environmental Protection Agency (U .8. EPA)
has recently approved legislation to reduce the maximum allowable trihalomethane
(THM) concentrations in ﬁnished waters from 100 to 80 ug/L (Arrora, LeChevallier, and
Dixon, 1997; Pontius, 1995) I. This criterion will be difﬁcult to meet for many water
utilities, especially those that use surface water supplies. In compliance with the SDWA
amendments the US. EPA would also be required to identify technologies that are
technically and economically feasible for small systems serving up to 10,000 people
(National Research Council, 1997). Therefore, a major concern of water utilities is
identifying suitable economical treatment processes that will reduce the levels of DBP
precursors in ﬁnished drinking waters.

The addition of ozone prior to chlorination has been identiﬁed as an important
pretreatment step in the minimization of THM formation to meet the current and future
US. EPA standards (U .8. EPA Drinking Water Regulations, 1997). Past regulatory

actions by the US. EPA resulted in an increase in the acceptance of ozone by US water

 

' The US. EPA is currently considering the revision of the Disinfection/Disinfection Byproducts Rule
(D/DBP), which may result in chages in the maximum allowable THM concentrations in drinking water
(Federal Register, V. 63, No. 61, pp. 15673-15692, March 31, 1998).

1

utilities. Since the early 19808, the number of US. drinking water treatment plants

employing ozone has risen rapidly from about 10 to more than 200. The number is

expected to reach at least 300 plants by the year 2000 (U .8. EPA Drinking Water

Regulations, 1997). It is anticipated that current and future EPA regulatory actions will

increase the acceptance of ozone even further (U .8. EPA Drinking Water Regulations,

1 997).

This research project investigated the use of ozonation in combination with

biological-ﬂuidized—bed treatment (FBT) to control natural organic matter (N OM) and

THM precursors in drinking water. The proposed ozonation/FBT process has several

potential advantages over conventional ozonation/bioﬁltration processes, including:

(1)

(2)

(3)

The ozonation/FBT process uses a ﬂuidized bed reactor (in place of
bioﬁlters) in which clogging is virtually impossible, and, therefore, no
additional pretreatment is required.

The efﬁciency of the removal of biodegradable organic matter by FBT is
expected to be greater than that obtained by bioﬁltration. This is because the
attached biomass concentration is expected to be greater in the FBT system
as compared to those in the conventional bioﬁlter due to greater speciﬁc
surface area and no-clogging operation. Also, because of greater speciﬁc
surface area, if activated carbon is used as a support medium, the use of FBT
will result in signiﬁcant saving in activated carbon.

The ozonation/FBT process can be operated in a cyclic mode, in which the
water alternately passes through ozonation and biodegradation steps. In this

way, ozonation is expected to degrade only those compounds that are not

biodegradable, thus reducing ozone consumption and, therefore, capital and
operating and maintenance costs.

(4) The ozonation/FBT units will be standard-size units. The process
performance of the unit can be optimized for different sources of water
supplies. The system is also expected to meet future effluent standards
through process parameter optimization without process expansion or
modiﬁcation.

(5) Finally, FBT systems have a considerably smaller footprint than bioﬁlters,
and therefore may be very useful for applications where space for retroﬁtting
or expansion is limited.

Over the last ten years, the biological component of this proposed system has
been developed and used in several applications for the biological treatment of process
waters and for the bioremediation of volatile organic compounds and chlorinated solvents
in groundwaters (Hickey, Wagner, and Mazewski, 1991). Because of low operational
costs (power, GAC, oxygen), these systems have proved a competitive advantage over
aqueous carbon adsorption and air stripping/vapor carbon adsorption in bioremediation
applications. These systems can also be monitored and controlled from remote locations
using advanced telemetry, thus reducing staff'mg requirements and O&M costs (Hickey,
Wagner, and Mazewski, 1991).

Skid-mounted biological FBT systems (with single pass ﬂow capacities ranging
from 30 gpm to 1850 gpm) are available off-the-shelf, from a leading US. manufacturer of

water and wastewater treatment equipment/processes (Envirex, Ltd. ). These systems could

serve communities ranging form 500 to 50,000 people. Using multiple FBT units in one
system, their applications can be further expanded.

If technical and economical merits are established, the ozonation/FBT process
will be a very attractive alternative to both conventional coagulation and
ozonation/bioﬁltration processes, especially for small water treatment utilities and for

larger utilities where space for retroﬁtting or expansion is limited.

1.2. OBJECTIVES

The ultimate goal of the project was to develop an efﬁcient and simple processing
system that could be used in both small and larger drinking water treatment facilities to
control THM precursors. An additional goal was to better understand the transformation
of NOM during ozonation and biotreatrnent to allow for the optimization of the
developed process. The speciﬁc aims of this project were:

(1) To assess the impact of ozonation and F BT on the properties of NOM, including
total organic carbon (TOC), ultraviolet absorption measured at 254 nm (UV -254),
humic and nonhumic content, molecular weight transformations, and
trihalomethane formation potential (THMF P).

(2) To establish the relationships between various properties of NOM and to assess
the role of these parameters in the evaluation and optimization of the proposed
ozonation/FBT process.

(3) To test the hypothesis that the efﬁciency of FBT increases with ozone dose.

(4)

(5)

(6)
(7)

(8)

(9)

(10)

(11)

To investigate the effect of control variables, including ozone dose and retention
time on the efﬁcacy of ozonation and F BT for the removal of NOM.

To determine if the humic and nonhumic fractions of NOM can be used to model
the kinetics of ozone reactions with NOM. If so, to develop a kinetic model of
ozone reactions with NOM using these parameters.

To compare the efﬁciency of FBT to that of bioﬁltration for the removal of NOM.
To test the hypothesis that the recycle of a portion of system effluent back to the
ozonation stage increases the efﬁciency of the proposed ozonation/FBT for the
removal of NOM relative to ozone consumption.

To test the hypothesis that the use of FBT prior to ozonation increases the
efﬁciency of the combined ozonation/FBT process for the removal of NOM
relative to ozone consumption.

To assess the role of bioﬁltration as the ﬁnal stage of the ozonation/FBT process
for the control of THM precursors and biodegradable organic matter in ﬁnished
water.

To assess the impact of adding an easily biodegradable carbon source to the F BT
system in order to increase process efﬁciency in terms of the removal of NOM
and biodegradation time.

To estimate the cost of treatment by combined ozonation and FBT and compare
this cost with the cost of conventional treatments.

Several other hypotheses were developed and tested in the course of this study.

They will be discussed in the appropriate chapters or sections.

1.3. PROJECT OUTLINE

Chapter 2 provides the background information relevant to the project. This
includes characterization of NOM and its impact on water quality and treatment and the
processes available for the treatment of drinking water. Additional background
information relevant to the topics of the discussion is provided in each chapter that
follows. Chapter 3 describes experimental systems and analytical methods used in this
research project. Chapter 4 evaluates the potential of the ozonation/FBT process to
remove NOM and to control THM precursors. Chapter 5 identiﬁes surrogate parameters
that are subsequently used for the investigation of the ozonation and biodegradation of
NOM and for the optimization of the proposed ozonation/FBT process. Studies of the
ozonation and biodegradation of NOM are presented in Chapters 6 and 7, respectively.
Chapter 8 investigates several approaches for the optimization of the proposed process.
Chapter 9 presents a preliminary economic analysis of the proposed process and
compares it with that of conventional treatment processes. Conclusions and

recommendations for future research are given in Chapter 10.

2. BACKGROUND

This chapter presents background information relevant to the matter of this
dissertation. Section 1.2 provides general information about natural organic matter (N OM)
and methods for its characterization. Section 2.3 describes the impact of NOM on water
quality and treatment. Section 2.4 discusses treatment methods for the removal of NOM

ﬁom water.

2.1. CHARACTERIZATION OF NOM

Natural organic matter (N OM) is a complex matrix of organic materials present in
natural waters. A major concern of water utilities with respect to NOM is in that that it
can impart color, tastes, and odors to the water and react with disinfectant to form
disinfection by-products.

NOM consists of humic and nonhumic fractions. The humic fraction is more
hydrophobic than the nonhumic fraction (Manahan, 1991; Liao et al., 1982, Schnitzer and
Kahn, 1972). The nonhumic fraction includes hydrophilic acids, proteins, aminoacids,
and carbohydrates (Manahan, 1991). Nonhumic substances have been considered to
present less of drinking water quality problems than humic substances (Bablon et al.
1991a).

NOM can be characterized by nonspeciﬁc parameters based on their ability to

absorb UV light, by their organic carbon composition, or by their potential to form

trihalomethanes (THMFP). A molecular weight characterization of NOM provides
valuable information on the transformation of NOM and is a powerful tool in evaluating
potential treatment strategies. NOM can also be characterized by its biodegradability
potential. The latter characterization can provide information on the efficacy of biological

treatment or to predict the possibility of bacterial growth in the distribution system.

2.1.1. Humic Substances

Humic substances, which comprise 40-60% of the dissolved organic carbon
(DOC) in natural waters, are the largest fraction of NOM in water. They are normally
present in concentrations > 1 mg of C/L (V ik and Eikebrokk, 1989).

Humic substances represent a unique category of natural products in which the
essence of the materials appears to be its heterogeneity. Although these materials are known
to result ﬁ'om the decomposition of biological material, the precise biochemical and
chemical pathways by which they are formed have not been elucidated. Because of their
complex, multicomponent nature, humic substances cannot be described in speciﬁc
molecular terms. Thus, when dealing with humic substances, researchers are forced, by
circumstances, to express the amount of material in terms of organic matter or as dissolved
organic carbon (MacCarthy & Suffet, 1989).

Humic substances are deﬁned operationally by the methods used to extract them
from water or soil. Typically they are divided into more soluble fulvic acids and the less

soluble humic acids, with fulvic acids predominating in most waters (Manahan, 1991). The

humic acid and fulvic acid do not refer to single compounds, but to a wide range of
compounds of generally similar origin.

Aquatic humic acids differ from aquatic fulvic acids in average molecular weight,
elemental and functional group composition, and in other characteristics. Aquatic fulvic
acid usually has a molecular weight of less than 2000 daltons (Thurman, 1985). Humic acids
have a molecular weight of 2000-5000 daltons or greater (Thurman, 1985), and is, therefore,
considered colloidal. As, in natural waters, humic acid is associated with various metal ions
and clays, it may have a larger molecular weight than that determined with puriﬁed free
acid.

Humic substances are polyelectrolytes with various functional groups attached to the
hydrocarbon skeleton (Manahan, 1991). Trussel and Umphres (1978) described the
structure of humic substance as a huge amorphous mass containing a polyheterocondensate
center to which certain functional groups, such as carboxyl acid, phenolic groups, alcohols,
methoxyl groups, carbonyl groups, ethers, and esters are attached. The carboxyl and
phenolic functional groups are primarily responsible for the acid-base and complexation
characteristics of humic substances (Manahan, 1989). Humic acid contains fewer carboxyl
and hydroxyl functional groups than does the fulvic acid fraction, making it less soluble in
water (V ik & Eikebrokk, 1989). Liao et al. (1982) concluded that humic acid contains
longer fatty acid chains than does ﬁrlvic acid. This ﬁnding may help to explain the
hydrophobic properties of humic acid. Thurman (1985) found that phenolic content of
humic acid is greater than that of fulvic acid, and there are more color centers on the humic

acid molecule.

Humic substances can be measured by separating them from nonhumic materials
by the adsorption on XAD-8 resin. In fact, the ability of humic substances to adsorb on
XAD-8 resin represents their operational deﬁnition.

The main physicochemical properties of humic substances, which arise from the
presence of several functional groups, are a strong reactivity with halogens, complexation
with metals, and the association of humic substances with organic micropollutants (Bablon
et al. 1991a). Humic substances also produce aesthetically undesirable problems such as
color. Humic substances can reduce the capacity and effectiveness of activated carbon by
competing with other organic contaminants for active sites on activated carbon. Humic
substances may reduce the bioavailability of hydrophobic substances, thereby, reducing the
rate at which the biodegradation of organic compounds occurs (MacCarthy and Suffet,

1 989).

Among the adverse effects that humic substances have on water quality and
treatment, the most important is the reaction of humic substances with chlorine. This
reaction results in the formation of chlorinated organic compounds, such as chloroform,
other low-molecular-weight disinfection by-products (e. g., haloacetonitriles, haloacids,
haloaldehydes, haloketones, chlorophenols, chloropicrin, and others) and complex high-
molecular-weight chlorinated compounds (Stevens et al., 1989). In fact, it is owing to the
discovery of trihalomethanes (I'HMs) in water supplies, that humic substances have been
given special attention. It is now generally accepted that these suspected carcinogens can
be formed in the presence of hurrric substances during the disinfection of water by
chlorination (Manahan, 1991). Trussell and Umphres (1978) suggested that the

halogenation of aquatic humic substances occurs via alternate hydrolysis and

10

halogenation steps. Several factors, such as the nature of humic substances, chlorine

dose, and pH, have been found to affect the formation of THMs.

2.1.2. Organic Carbon Content

As a result of their heterogeneous and ill-deﬁned character, there are no direct
analytical techniques to measure NOM. One way is to deﬁne NOM by its organic carbon
content.

About 50 percent of the organic carbon in surface water of the United States
consists of humic substances. The humic substances are from 40 to 65% carbon by
weight (Manahan, 1990; McCarthy and Suffet, 1989). Many workers use the terms
organic carbon contents and humic substances interchangeably (McCarthy and Suffet,
1989). Although such usage is not entirely correct, in general there is a good correlation
between organic carbon contents and humic substance concentration (Krasner et al.,

1 996).

2.1.3. UV Absorption

Organic compounds that are aromatic or have conjugated double bonds absorb
light in the ultraviolet (UV) wavelength region. UV absorption, thus, is a good technique
for measuring the presence of NOM, such as humic substances, because they contain
aromatic moieties and are the dominant form of organic matter in natural waters

(Edzwald, Becker, and Wattier, 1985). Strong correlation may exist between UV

11

absorption and organic carbon content and THM precursors (Eaton, 1995; Edzwald,
Becker, and Wattier, 1985). In some cases, UV-254 absorbance may be a better indicator
than TOC concentration for DBP formation (N ajm et al., 1994; Edzwald, Becker, and
Wattier, 1985). UV absorption has also been used to evaluate organic removal by various
water treatment processes (Eaton, 1995; Edzwald, Becker, and Wattier, 1985). The use of
spectroscopic measurements coupled with organic carbon monitoring can provide
information about the nature and variability of NOM. Speciﬁc absorption, the ratio of UV
absorption to organic carbon concentration, has also been used to characterize natural
organic matter (Krasner et al., 1996; Eaton, 1995).

Under the Information Collection Rule (ICR), water utilities across the United
States will be required to monitor UV absorption in an attempt to correlate this parameter
with DBP precursors. A wavelength of 254 nm has been used to characterize the aromatic
portion of NOM (Masten, 1991). This wavelength has also been selected for the purpose

of ICR (Eaton, 1995).

2.1.4. THMFP

THMs (chloroform, dichlorobromomethane, chlrodibromomethane, and bromoform)
are the dominant DBP species formed during chlorination. THMs are increasingly regulated
by US. EPA and is one of the major parameters that needs to be controlled by a treatment
process. Therefore methods to determine the potential for forming THMs are useful in
evaluating water treatment processes or water sources or for predicting THM concentrations

in a distribution systems.

12

There has been extensive research relating to factors that inﬂuence the creation of
THMs. Trussel and Umphres (1978) reported that factors that might be responsible for the
formation of THle include (1) TOC concentration, (2) inorganic chemistry of water (e.g.,
pH, bromide concentration), (3) contact time of chlorine with NOM, and (4) chlorine/TOC .
ratio. THM formation is stimulated by elevated temperature and alkaline pH and by
increasing concentration of ﬁee chlorine residual. A longer contact time generally increases
THM formation. Therefore, in order to be able to analyze and compare the results from
different sources, the control of such variables as temperature, reaction time, chlorine dose
and residual, and pH is essential. The standard reaction conditions have been identiﬁed as
follows: ﬁ'ee chlorine residual at least 3 mg/L and not more than 5 mg/L at the end of 7 day
reaction period, with sample incubation temperature of 25°C, and pH controlled at 7 with
phosphate buffer (Standard Methods, 1992).

It should be noted that THMFP determined under standard reactions conditions is
not intended to simulate water treatment processes but are most useful for estimating the
concentration of THM precursors, as well as for evaluating water-treatment options for

reducing levels of THM precursors in the raw water.

2.1.5. Molecular Weight Distribution

The performance of many water treatment processes is related to size distribution
of NOM (Jackson, Hong, and Summers, 1993; Collins, Amy, and Steelink, 1986).
Different processes have been cited as being amenable to removing certain size

molecules. For example, higher molecular weight matter was found to be generally more

13

amenable to the removal by coagulation (Amy at al., 1992, Chadik and Amy, 1983). A
water source with medium molecular weight materials can be effectively treated by
activated carbon adsorption (Pirbazary et al., 1989). High molecular weight materials
may be difﬁcult to treat by activated carbon adsorption, because large molecules are
unable to enter smaller pores of activated carbon (Pirbazary et al., 1989). Ozonation
partially oxidizes higher molecular weight compounds leading to the formation of lower
molecular weight compounds, which are more biodegradable (Amy, Kuo, and Sierka,
1988). Other researchers (Anderson, Johnson, and Christrnan, 1986) observed loss of
carbon from high and low molecular weight ﬁactions during ozonation.

In this study, molecular weight fractionation was used to evaluate the
ozonation/FBT process and to better understand the transformation of NOM during

ozonation and biotreatment.

2.1.6. BDOC

Biodegradable organic carbon (BDOC) is the portion of organic carbon in water
that can be mineralized by heterotrophic microorganisms. One characteristic of ozonation
is its tendency to increase the biodegradability of aquatic organic matter. Increased
biodegradability can lead to improved performance of subsequent biological processes.
However, if left uncontrolled, it can also encourage the regrowth of bacteria in
distribution systems (Reckhow et al., 1993).

Several methods for measuring potentially biodegradable organic matter have

been developed in recent years. A method for measuring assimilable organic carbon

14

(AOC) was developed by van der Kooij, Visser, and Hijnen (1982). However, this
method attempted to measure bacterial regrth potential rather than a change in organic
carbon.

Another method was developed by Servais and colleagues (Servais, Billen, and
Hascoet, 1987). The method includes the inoculation of a sample with a mixed bacterial
culture and measuring DOC at the beginning and end of incubation period. The sample is
incubated at room temperature for approximately four weeks. This methods gives no
information on the kinetics of biodegradation.

Fixed-bioﬁlm biodegradation to measure BDOC was introduced by J oret and
Levy (1988) and Morgen, Scarpino, and Summers (1990). The ﬁrst method (Joret and
Levy, 1988) involves the addition of sand with acclimated bioﬁlm to water samples and
incubation at room temperature under aerated conditions. Daily measurements of DOC
are made until there is no further change. BDOC is taken as the decrease in DOC. The
method raises the possibility of release of organic matter from biomass.

The second method (Morgen, Scarpino, and Summers, 1990) involves
recirculation of the sample through a bioﬁltration column. The methods is represantative
of biotreatrnent processes and is currently widely used for measuring BDOC when the
concern is the reduction of DBP formation potential through a biological processes
(Cipparone, Diehl, and Speitel, 1997; Huck, 1990). The procedure will be discussed in

greater detail in Section 3.3.6.

15

2.2. REVIEW OF TREATMENT PROCESSES

Since the late 1970s much work has focussed on the development of processes to
reduce the formation of THM precursors in drinking waters. Among the processes which
have been most extensively investigated are slow sand ﬁltration, coagulation, activated
carbon adsorption, membrane separation, ozonation, and ozonation combined with
biological granular activated carbon adsorption (GAC) or bioﬁltration. The review of these

major processes is given below.

2.2.1. Slow Sand Filtration

Slow sand ﬁltration was the original form of water treatment used in nineteenth
century and is now considered a low-technology approach to water treatment (National
research Council, 1997). Although not originally conceived as a biological process, its value
in that role has been recognized over the years. Removal mechanisms in slow sand ﬁltration
are both physical and biological: biological action breaks down some organic matter, and
some inert suspended particles are physically removed from the water (Cleasby, 1990).

The process usually involves passage of raw water through a bed of sand at low
velocity (usually less than 0.4 m/h). With extended use of the ﬁlter, a biological ecosystem
grows in the sand bed. On the top of the ﬁlter media, a biologically active organic layer (so-
called Scmutzdecke layer) builds up and assists ﬁltration. The water then enters the top layer

of sand, where more biological action occurs (Cleasby, 1990).

16

The use of this process is limited to waters having low turbidity and organic carbon
content (Cleasby, 1991; Cleasby et al., 1984; Fox et al., 1984). Slow sand ﬁltration is not
very effective at removing THM precursors or color. Therefore, only high quality waters
(low in turbidity, algae, and color) are suitable for application to slow sand ﬁlters without

pretreatment.

2.2.2. Chemical Coagulation

Evidence available in the literature indicates that chemical coagulation removes
DBP precursors and lowers level of DBPs formed upon chlorination of settled water (N ajm
et al., 1994; Chadik and Amy, 1983; Semmens and Field, 1980). Alum has commonly used
for the removal of organic matter, including DBP precursors (Najm, 1994). Alum
coagulation at a dosage of 50 mg/L has been reported to reduce THMFP at values of up to
70 percent (Reckhow and Singer, 1990). The use of coagulation with metal salts and
polymers has also been proven successful for the removal of humic substances (Baboock
and Singer (1979), Vik and Eikebrokk (1989), and Bottero and Bersillon (1989)). Baboock
and Singer (1979) showed that up to 90% of TOC can be removed by the coagulation of
humic acid, whereas the coagulation of fulvic acid removed only 20% of TOC. Similar
results have been reported by Perdue and Lyttle (1983). Vik and Eikebrokk (1989) pointed
out that the main parameters affecting the coagulation of humic substances were the initial
concentration of humic substances, the coagulant dosage, and the pH at which the
coagulation were carried out. They also showed that the removal of humic substances by

alum coagulation was efﬁcient only over a narrow range of pH (4.5-5). Glazer and Edzwald

l7

(1979) achieved approximately 40% removal of TOC by the coagulation of humic acid
using cationic polyelectrolytes.

Although coagulation is one of the most common processes used in water treatment,
one must recognize the problems associated with this process. The effectiveness of
chemical coagulation for DBP precursors removal strongly depends on the type and
concentration of organic material present in the raw water, the coagulant type and dosage,
and pH (Najm, 1994).

Coagulation of humic substances has been shown to be effective only over a narrow
range of pH, with optimum-removal occurring at a pH in the range of 5-6 (Reckhow and
Singer, 1990; Dempsey, Ganho, and O’Melia, 1984; Chadik and Amy, 1983). Higher pH
would require considerably higher coagulant dosages for effective coagulation (V ik and
Eikerbrokk, 1989). The coagulation of humic substances with alum, which has been shown
to be most effective coagulant, may cause a particle separation problem (V ik and
Eikerbrokk, 1989). Turbidity created by colloidal destabilization may be difﬁcult to
remove.

Morris and Knocke (1984) showed that temperature had a great impact on
coagulation with alum. In some typical tests, residual turbidity increased from 0.5 NTU at
25°C to 2.4 NTU at 5°C. Some of the reasons suggested for the decreased efﬁciency of
coagulation at lower temperatures is increased viscosity and its effect on sedimentation
(Arrnirtharajah and O’Melia, 1990). Among the most challenging conditions for treatment
by coagulation are very cold water, approximately 5°C and colder, and turbidities of
approximately 10 NTU and lower. When the amount of particulate matter in the water is

low, sedimentation is not very effective (National Research Council, 1997). Another

18

difﬁcult condition for water treatment is the combination of high color and moderate to high
turbidity. The pH that is best for color removal may be different from the pH that is best for

turbidity removal (National Research Council, 1997).

2.2.3. Carbon adsorption

Historically, activated carbon adsorption was used primarily to remove tastes and
odors from water, but its use as an adsorbent for toxic and carcinogenic compounds has
increased steadily and is a primary application (National Research Council, 1997).

Activated carbon adsorption has also been used for the removal of humic substances.
Kaastrup and Halmo (1989) found that the nature of humic substances and the pore-size
distribution of the activated carbon were the most important factors in determining the
absorbability of humic substances. The authors noted the poor absorbability of humic
substances, which was reﬂected in unfavorable adsorption isotherms, slow adsorption
kinetics, and immediate breakthrough of color-imparting substances from the adsorption
column. The absorption capacities for smaller fulvic acid molecules were seen to be higher
than that observed for humic acid. McCreary and Snoeyink (1980) obtained similar results.
The adsorption capacities for humic acid were observed to be directly related to the ﬁaction
of large pores in the carbon. They also found that the adsorption rate of humic acid was
inversely related to carbon-pore size. By studying the adsorption of humic and fulvic acids
onto activated carbon, McCreary and Snoeyink (1980) found that the lower-molecular

weight species were more readily adsorbed. Lee et a1. (1981) did not achieve satisfactory

results for the removal of humic acid when activated carbon was used alone.

19

Competitive adsorption is an important consideration in the design of an activated
carbon system for drinking water treatment NOM in water may compete with contaminants
for adsorption sites on the carbon, thus increasing the amount of carbon needed to remove
the target contaminants. Pirbazary et al. (1989) showed that humic substances reduced the
adsorption capacity of the activated carbon for other organic pollutants normally present in
water. Competing organic contaminants can also displace the contaminants already
adsorbed to the carbon.

Lykins and Clark (1989) showed that activated carbon adsorption alone was unlikely
to be the process of choice for the removal of humic substances and pretreatment would be
required to decrease the molecular weight of humic substances before activated carbon
could be applied effectively. GAC use also requires removing particulate matter ﬁ'om

untreated water to avoid clogging of the treatment column.

2.2.4. Membrane separation

Once considered a viable technology only for desalination, membrane ﬁltration is
increasingly employed for the removal of NOM. Several membrane ﬁltration
technologies appropriate for water treatment are distinguished by their nominal molecular
weight cutoff (MWCO), which is an estimate of the smallest size molecule that will be
retained by the membrane in a ﬁltration process. By these guidelines, membrane ﬁltration
technologies are classiﬁed as employing microﬁltration, ultraﬁltration, or nanoﬁltration

(National Research Council, 1997).

20

Microﬁltration uses membranes with an MWCO of greater than 100,000 daltons

and is effective for the removal of some bacterial species, but is not capable of removing

most humic materials. Ultraﬁltration uses membranes with an MWCO of approximately

10,000 to 100,000 daltons and can remove some humic materials. Nanoﬁltration

membranes have an MWCO of 1,000 to 10,000 daltons and are very effective towards the

removal of humic materials (National Research Council, 1997).

Although membrane ﬁltration is effective for the removal of THMFP and greatly

reduces the need for disinfectant, several potential disadvantages with respect to the

application in water treatment must be recognized (National Research Council, 1997;

Weiesner et al. 1994):

Membrane ﬁltration is sensitive to particulate matter, turbidity, and some NOM
that can foul the membrane. To avoid this, water must be pretreated by a more
conventional treatment process prior to polishing by membrane ﬁltration.

As nanoﬁltration also removes alkalinity, the product water can be corrosive and
adding alkalinity may be needed to reduce corrosivity.

Membrane ﬁltration may exert high operating costs associated with module
replacement costs, pressure requirements, and cleaning.

Choosing the most appropriate membrane material for a given application still
remains an art, since many different factors, including both water quality and
operational conditions, can greatly affect the membrane performance.

Unlike conventional treatment processes, in which approximately 5 to 10 percent
of the water is discharged as waste, membrane processes produce waste streams

amounting to as much as 15% of the total water volume.

21

2.2.5. Ozonation

Since early 19808, the number of US. drinking water treatment plants employing
ozone has risen rapidly from about 10 to more that 200. The number is expected to reach
at least 300 plants by the year 2000 (U .S. EPA Drinking Water Regulations, 1997). The
increasing acceptance of ozone in US. water treatment plants has been affected
positively by past EPA regulatory actions. It is anticipated that current and ﬁrture EPA
regulatory actions will increase the acceptance of ozone even further (US. EPA Drinking
water Regulations, 1997).

Ozonation, at doses used in conventional water treatment practice (1 mg
ozone/mg C or less), is anticipated to bring about only slight reduction (< 10 percent) in
organic carbon (Amy, Kuo, and Sierka, 1988). Masten (1991) and Bablon et al. (1991b)
also observed only a minor reduction in the concentration of organic matter during
ozonation of humic substances. However, it has been established that ozone oxidizes
humic substances leading to the formation of lower molecular weight species. . Flogstad
and Odergaard (1985) showed that, in drinking water, the concentration of high-molecular-
weight humic substances was reduced after ozonation and a nearly corresponding amount of
low-molecular-weight compounds was produced. They also showed that increased ozone
dosage led to an increase in the biological degradability. There have, however, been some
conﬂicting reports relating to the transformation of molecular weight fractions. Edwards and

Benjamin (1992), for example, observed the reduction in low-molecular weight compounds

22

aﬁer ozonation at doses less than 1 mg/mg C. Similar results were obtained by Anderson,
Johnson, and Christrnan (1986).

Meijers (1976) showed that waters containing humic substances could be efﬁciently
using ozone. The degradation of color has been found to be almost total when ozone
dosages between 1 and 3 mg ijg C were applied (Masten, 1991; Killops, 1986; Watts,
1985). This suggests that ozone reacted efficiently with the aromatic portion of the humic
substances. The effect of ozone on the aromatic portion of the humic substance has also
been shown to occur by following the decrease in the UV absorbance during and after
ozonation (Masten, 1991; Kruithof et al., 1989; Reckhow et al., 1986; Anderson, Johnson,
and Christman, 1986). The decrease in the UV absorbance is also an indicator of the
changing character of humic substances during ozonation. For example, Masten (1991)
showed that during ozonation of humic acid the greatest change in the UV absorbance
spectra occurred at a wavelength of 254 nm, indicating a signiﬁcant change in the
aromaticity of the humic acid. Legube et al. (1989) showed that ozone consumption
increased with increasing the initial UV absorbance of the humic substances, which may be
attributed to both the higher initial concentration of humic substance and the higher content
of aromatic ﬁmctional groups. Masten (1991) also observed similar results.

Complete ozonation of hmnic substances to carbon dioxide and water is too
expensive to be feasible for water treatment. At typical ozone dose-to-carbon ratios, the
oxidation of humic materials is incomplete and some ozonation by-products are formed.
The by-products, which are formed, may result either from the direct reaction of the humic
substance with molecular ozone or from indirect reactions involving free radicals formed

during the decomposition of ozone. The nature of parental organic matter can vary widely,

23

as can the pH and the aqueous concentration of free radical scavengers, such as bicarbonate
and carbonate ions. For this reason, the character and concentration of ozone by—products
will vary significantly from water to water (Bablon et al., 1991b). It appears, that aldehydes
are ubiquitous products formed ﬁ'om ozonation of NOM (Glaze et al., 1989). Other
compounds identiﬁed in ozonated waters and humic substance solutions include mono- and
dicarboxylic acids, mono- and diketones and alkanes (Glaze et al., 1989; Killops, 1986;
Lawrence et al. 1980). Watts (1985) also identiﬁed aromatic acids in ozonated humic
substance solutions.

Several investigators have studied the effect of ozonation on the reactivity of humic
substances with chlorine. Using an aquatic fulvic acid, Reckhow et al. (1986) showed that
ozonation at a neutral pH (phosphate buffer solution) resulted in a decrease in THMFP, total
organic halides (T OX), trichloracetic acid, and dichloroacetonitrile. Using fulvic acids
extracted from numerous surface waters, Legube et al. (1989), obtained the results
comparable to those obtained by Reckhow et al. (1986) under similar reaction conditions.
They suggested that the reaction of molecular ozone with humic substances could lead to the
degradation of aromatic sites, which are thought to be the precursors of organochlorinated
by-products. On the contrary, hydroxyl radicals appear to react with humic substances,
resulting in the hydroxylation of sites. As a result of that, new THM precursors may form
(Legube at al., 1989). As the presence of bicarbonate ions in the system will tend to
scavenge the hydroxyl radicals, the formation of new THM precursors will be reduced and
the degradation of existing ones will be increased by the action of molecular ozone
(Reckhow et al., 1986). Although the effect of ozonation on the reactivity of NOM with

chlorine is not well understood, there has been a general agreement regarding the effect of

24

ozonation of the reduction of THMFP. Amy, Kuo, and Sierka (1988) observed 15-40%
reduction in THMFP during ozonation at doses up to 2.5 mg/mg C. Reckhow and Singer
(1990) observed a 15% THMFP reduction at an ozone dose of 1 mg/mg C. Similar results of
THMFP were reported by Singer and Chang (1990), who conducted tests at ten water

utilities.

2.2.6. Combined ozonation/biotreatment

It is generally recognized that ozonation converts nonbiodegradable organic matter
to biodegradable organic carbon (BDOC) by breaking the structure of NOM and enhancing
the transformation of higher molecular weight compounds to lower molecular weight
compounds. The removal of biodegradable organic matter during subsequent treatment is
particularly important because this matter has been linked to the growth of microorganisms
within distribution systems (Ahmad, 1998).

Biological treatment is one method of removing biodegradable organic matter. As
was stated earlier, the primary interest in biological processes stems from the recognition
that after ozonation, biological activity is stimulated whether it is a speciﬁc design
objective or not. All biological processes used in drinking water treatment are the bioﬁlm
type. Most commonly used systems in water treatment are biological ﬁlters or GAC ﬁlters.
Slow sand ﬁltration is still used at some plants in European countries and is able to
adequately remove organic matter (Ahmad, 1998).

An increase in the biological degradability of NOM after ozonation is described

elsewhere (Goel, Hozalski, & Bouwer, 1995; Flogstad and Odergaard, 1985). Cipparone,

25

Diehl, and Speitel (1997) established that the rate and extent of biodegradation of organic
matter increased with ozone dose. Shukairy and Summers (1992) conducted batch
ozonation and biodegradation studies to remove organic halides from surface waters and
groundwaters containing humic substances. The authors reported that the combined
treatment removed 47% of purgable organic halogen formation potential (POXFP) and
dissolved organic halogen formation potential (DOXF P) from Ohio River water versus
25 and 10%, respectively, when ozonation was used alone. For groundwater humic
substances, the removal of POXFP increased from 40% (ozonation alone) to 80%
(combined ozonation and biodegradation) and the removal of DOXFP increased from 25
to 70%. In all experiments, a dosage of 2 mg/mg TOC was applied. In further studies,
increased ozone dosage was shown to result in an increased removal of THMFP
(Shukairy and Summers, 1996). A similar effect was established by Speitel et al. (1993).
The study showed that the percentage removal of THMFP in natural water by ozonation
and biodegradation steadily increased with ozone dosage, up to a maximum of about 60%
at a dose of 5 mg/mg TOC. The contribution of ozonation to the removal of THMFP was
about half that of biodegradation at ozone doses of 2 mg/mg TOC. At an ozone dose of 3
mg/mg TOC, the contributions of ozonation and biodegradation were approximately
equal, whereas at an ozone dose of 5 mg/mg TOC, the contribution of ozonation was
about twice that due to biodegradation (Speitel et al., 1993).

The general approach to the removal of THM precursors from drinking waters by
combined ozonation and biotreatrnent has been to ﬁrst oxidize humic substances and
other NOM to a biodegradable form, and then to use biological treatment to remove any

biodegradable organic matter. It was shown that the production of ADC and BDOC starts

26

at low levels of ozone dosage and increases with ozone dosage (Speitel et al., 1993). The
maximum removal of THMFP and HAAF P by ozonation followed by biodegradation
was observed at an ozone dosage of approximately 5 mg/mg C, which was the maximum
ozone dose used in the experiments (Speitel et al. 1993). The high dosage of ozone
required can be attributable to the fact that a portion of ozone is consumed by the
ozonation products that could have been degraded biologically. This conclusion is
supported by the data obtained by Speitel et al. (1993), which showed that at large ozone
doses, THMFP removal occurs largely through ozonation, rather than biodegradation.

Combining ozonation and biological activated carbon (BAC) has been
successfully used at many water works in Great Britain, France, Germany, and other
European countries (Ahmad, 1998; Rachwal, Foster, and Holmes, 1992; Venteresque,
Bablon, and Jadar-Hecart, 1991; Sontheimer and Hublele, 1986) for more than two
decades. In recent years, there has also been increasing interest in the combined
ozonation/bioﬁltration in the United States (Price et al., 1993; Malley et al.1993; Krasner
et al., 1993).

The Swinford Water Treatment Works (Oxford, England) built a demonstration
plant that incorporated preozonation, chemical coagulation/ﬂocculation, rapid gravity
ﬁltration, main ozonation and biologically active GAC to treat Thames River water
having TOC from 2 to 7 mg/L (Rachwal, Foster, and Holmes, 1992). The total ozone
dose for preozonation and main ozonation was approximately 1 mg/mg C. Nominal
contact times of 5 and 15 minutes was used for preozonation and main ozonation,

respectively. The GAC ﬁlters had an EBCT of 15-30 minutes. Overall TOC removal

27

through the plant was approximately 40%. The process was capable of eliminating of
approximately 65% of THMFP, which correlated well with UV-254 absorbance.

The Metropolitan Water District of Southern California (MWD) conducted a
pilot-scale study of ozonation and high-rate ﬁltration using California state water and
Colorado River water with TOC levels of 3.5 and 2.5 mg/L, respectively (Krasner et al.,
1993). The pilot plant included ozonation at a dose of approximately 0.5 mg/mg C,
coagulation with alum, followed by ﬂocculation and sedimentation, and high-rate
bioﬁltration using GAC-sand. The bioﬁlter was operated at a low EBCT of 4.2 minutes.
The pilot-plant testing showed that the process did not adequately remove TOC. Nearly
90% of TOC passed through the ﬁlter. However, the process was effective towards the
removal of aldehides, with nearly 80% of formaldehyde, glyoxal, and methylglyoxal that
wasfonned during ozonation being removed through bioﬁltration. It should be noted that
the temperature during pilot plant testing varied from 15 to 25°C.

East Bay Municipal Water District (EBMUD) conducted similar pilot testing at
the Sobrante Filter plant using water having a TOC concentration of 2.2-3.2 mg/L (Price
et al., 1993). The pilot plant testing consisted of taking water aﬁer
ﬂocculation/sedimentation from the full-scale facility, ozonating it at a dose of
approximately 0.7 mg/mg C, and passing it through a GAC/sand ﬁlter. The ﬁltration rate
corresponded to an EBCT of 4.5 minutes. The temperature was 10-22°C during the
experiments. The pilot testing showed that ozonation removed approximately 18% of
TOC and 30% of THMFP, however, very little removal of TOC and THMFP was

observed during bioﬁltration.

28

Pilot-plant studies of combined ozonation and slow sand ﬁltration were conducted
at the Andover Water Treatment plant (Andover, Mass.) (Malley et al., 1993). The raw
water had a low turbidity (approximately 1 NTU) and an average TOC concentration of
4.3 mg/L, a UV-254 absorbance of 0.1 cm'l, and a THMFP of 200 rig/L. An ozone dose
of approximately 2 mg/mg C was used in all experiments and the EBCT in the bioﬁlter
was 2.8 hours. The pilot-plant testing was conducted at a temperature of 23-27°C. The
removal of TOC and THMFP in the entire treatment process was 30-70% and 50-68%,
respectively. It was shown that ozone increased the BDOC content in water. However,
ozone also reduced ﬁlter run times as a result of increased clogging of the ﬁlter media via

bioﬁlm development.

2.2.7. Proposed Ozonation/FBT Process

The use of biological ﬁltration systems utilizing GAC and other media has been
well documented. Multi-step physical treatment and biological ﬁltration are used in many
European countries. It appears that the approach to including biological processes in
drinking water treatment practice in this country is to combine bioﬁltration with an
existing physiochernical unit operation.

In this project, a different approach to the problem has been proposed and
investigated. It includes using ozonation in combination with biological-ﬂuidized-bed
treatment (FBT).

Figure 2.1 demonstrates the differences in operation of a bioﬁlter and FBT

column. In bioﬁlters, water ﬂows downwards through a packed medium, which is

29

colonized by rnicrorganisms. Biological action breaks down some organic materials, and
some inert suspended particles are physically removed from the water. In FBT columns, a
bed of small particles is ﬂuidized by the upward ﬂow of water. Very high speciﬁc surface

area can be achieved without introducing the problem of clogging.

lnfiuent Efﬂuent

 

 

Efﬂuent lnﬂuent

 

Bioﬁlter FBT

Figure 2.1 Comparison of FBT and bioﬁlter operations

30

The proposed ozonation/FBT process has several potential advantages over
conventional ozonation/bioﬁltration processes, which are:

0 The ozonation/FBT process uses a ﬂuidized-bed reactor (in place of bioﬁlters) in
which clogging is virtually impossible, and, therefore, no additional pretreatment
may be required.

0 The efﬁciency of the removal of biodegradable organic matter by FBT is expected
to be greater than that obtained by bioﬁltration. This is because the attached
biomass concentration is expected to be greater in the FBT system as compared to
those in the conventional bioﬁlter due to greater speciﬁc surface area and no-
clogging operation. Also, because of greater speciﬁc surface area, if activated
carbon is used as a support medium, the use of FBT will result in signiﬁcant
saving in activated carbon.

0 The process performance of the unit can potentially be optimized for different
sources of water supplies. The system is also expected to meet future efﬂuent
standards through process parameter optimization without process expansion or
modiﬁcation.

0 Finally, F BT systems have a considerably smaller footprint than bioﬁlters, which
may be very useful for the application where space for retroﬁtting or expansion is
limited.

Over the last ten years, the biological component of this system has been
developed and used in several applications for the biological treatment of process waters
and for the bioremediation of volatile organic compounds and chlorinated solvents in

groundwaters (Hickey, Wagner, and Mazewski, 1991). The low operational costs of these

31

systems (power, GAC, oxygen) have proved their competitive advantage in
bioremediation applications. These systems can also be monitored and controlled from
remote locations using advanced telemetry, thus reducing manpower requirements and
O&M costs. Integrated ozonation-biotreatment that used a FBT system has also been
investigated for the treatment of bleaching effluents from pulp and paper industry
(Heinzle et al., 1992).

Commercial ﬂuidized bed systems are currently marketed by Ecolotrol, Biothane,
DorrOliver, and Envirex. Skid-mounted biological F BT systems (with single pass ﬂow
capacities ranging from 30 gpm to 1850 gpm) are available off-the-shelf. These systems
could serve communities ranging form 500 to 50,000 people, which can be further expanded
using multiple F BT units in one system.

If technical and economical merits are established, the ozonation/FBT process
will be a very attractive alternative to both conventional coagulation and
ozonation/bioﬁltration processes, especially for small water treatment utilities and for

larger utilities where space for retroﬁtting or expansion is limited.

32

3. MATERIALS AND METHODS

3.1. WATER SOURCE

The study was conducted with Huron River water collected at the pump station to
the Ann Arbor Water Treatment Plant. Typical characteristics of Huron River water are
summarized in Table 3.1. Huron River water contains relatively high concentrations of
TOC and THMFP and, hence, was well suited to evaluate the ability of the
ozonation/FBT system to remove TOC and THM precursors from water and to sustain

turbidity without pretreatment. The raw water has a low bromide concentration.

Table 3.1 Typical Quality Characteristics of Huron River Water

 

 

 

 

 

 

 

 

Parameter Raw water

TOC, mg/L 5.3 — 8.3
pH 7.8 - 8.2
Alkalinity, mg/L as CaCO3 350420
Turbidity, NTU 0.9 — 3.4
UV absorbance @ 254 nm, cm" 0.152 - 0.228
Bromide, rig/L 70
THMFP, ug/L 340 - 460

 

 

 

 

It should be noted that it was essential to focus our efforts on a single water
source, rather than using multiple water sources. The latter approach would have altered

our objective of better understanding the transformation of NOM during ozonation and

33

FBT to simply a feasibility study. The study of the applicability of the process to various

water sources is the next logical step upon completion of this project.

3.2. EXPERIMENTAL SYSTEMS

3.2.1. Bench-Scale Ozonation System

A schematic of the system used to study the ozonation of NOM is shown in
Figure 3.1. The water was pumped by a peristaltic pump (feed pump) from a holding tank
into the ozone reactor. The ozone contactor was a 360 mL water jacketed gas-washing
bottle, which was Operated as a completely-stirred tank reactor (CSTR). The reactor was
equipped with a fritted glass gas diffuser near the base, a vent-gas exit port, and water
entrance and exit ports. The diffuser was used to sparge the ozone/oxygen mixture into
the contactor. Additional mixing was provided with a magnetic stirring bar. All ﬁttings
and tubings were made of either PTFE or stainless steel.

A second pump was installed downstream of the ozone contactor, which was
synchronized with the feed pump in order to maintain a constant liquid level in the
contactor. This effluent pump pumped water either into an efﬂuent collection tank or into
a purging ﬂask from where water ﬂowed by gravity into the effluent collection tank. The
purging ﬂask was used to quench the ozone reaction by purging the reactor efﬂuent with
helium to remove unreacted ozone. The residual ozone in the gas phase was destroyed by

passing the gas through a potassium iodide solution.

34

Ozone was generated from pure, dry oxygen using an air-cooled Ozotech Model
OZlPCSN ozone generator (Ozotech, Inc., Yreka, Ca). The gas ﬂow rate to the reactor
was controlled by a Sierra Instrument Model 900 mass ﬂow controller (Sierra
Instruments, Inc., Monterey, CA), installed upstream of the ozone contactor. A check
valve was installed downstream of the ﬂow controller to prevent water in the ozone
contactor from cycling back into the ﬂow controller and the ozone generator. Inﬂuent and
effluent ozone gas concentrations was measured spectrophotometrically at 254 nm using
0.2-cm ﬂow-through quartz cells with the Milton Roy Spectronic Genesys-S
spectrophotometer (Milton Roy, Inc., Rochester, NY). An extinction coefﬁcient of 3000
LM'lcm'l was used to convert absorbance reading mg O3/L. Aqueous ozone

concentrations were measured using the Indigo method (Standard Methods, 1992).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Oil-gee tag '\1
Ozone
Purim W“
Tank
EﬂhIent \—
> 4 MUN“
t i
Eiiiuent C D
m 4
Off-gas
Gum
Ozone Germ E >' 0mm
Flow W
Controller

Oxygen

Figure 3.1. Schematic of the bench-scale ozonation system

35

3.2.2. Ozonation/FBT System

A schematic of the pilot-scale ozonation/FBT system is shown in Figure 3.2. The
system included an ozone contactor, retention tank, and a FBT column. The water was
pumped by a peristaltic pump from a holding tank into the ozone contactor. The ozone
contactor was a downﬂow injector-type bubble contactor. The contactor consisted of a
downﬂow vertical tube and an expanding cross section hood, as shown in Figure 3.3.
Water and gas entered at the top of the contactor and ﬂowed concurrently downward
through the tube into the expanding hood, and exited at the bottom of the contactor. The
contactor was designed so that the inlet velocity was greater and the exit velocity was
lower than the buoyant velocity of the bubbles. Thus, the bubbles were trapped inside the
hood. At the bottom of the downward tube, the velocity changed rapidly which created
signiﬁcant swirling and turbulence at the top of hood. This resulted in a high gas-liquid
interfacial area and a high mass transfer rate.

A retention tank was installed downstream of the ozone contactor. The pressure
head that developed in the hood of the contactor moved water out of the contactor into
the middle of a retention tank. The retention tank was a 1L water jacketed glass column.
Periodically, gas, which had accumulated in the contactor, was released through the

retention tank. This gas was analyzed for ozone and then passed through a potassium
iOdide trap to destroy residual ozone prior to discharge of the gas. The tank was also used
t0 collect samples of ozonated water. Since the ﬂow rate of inﬂuent ﬂow was
Signiﬁcantly lower that that required to maintain sufficient turbulence in the ozone

c(ltltactor, an additional pump was installed (not shown in Figure 3.2) to recirculate water

36

 

from the bottom of the retention tank back to the ozone contactor. A recycle pump was
installed to provide the recycle of a portion of system efﬂuent back to the ozone
contactor.

From the retention tank, water ﬂowed by gravity into a small recirculation tank
from where it was recirculated through the FBT column with a peristaltic pump at a rate
sufﬁcient to ﬂuidize the bed. The FBT column was a jacketed glass column 2 inches in
diameter and 60 inches in height. Non-activated carbon, termed “baker product” by the
manufacturer (Calgon Co., Pittsburgh, Pa), was the same material as granular activated
carbon (GAC) except that it had not undergone the activation step in the manufacturing
process and had, therefore, little sorptive capacity (Voice eta1.,1992). It was essential to
use a non-adsorbent material as a support media in the F BT column in order to separate
microbial removal of BDOC from sorptive removal.

During the ﬁrst stage of the study, the height of the ﬂuidized bed was maintained
at 35 inches. This corresponded to an EBCT of approximately 180 minutes at a water
ﬂow of 10 mL/min. Later, two-third of the acclimated material was removed from the
column, a portion of which was transferred to the bioﬁltration column, described in
Section 3.5. The remaining ﬂuidized bed provided an EBCT of 30 minutes at a water
ﬂow of 20 mL/min.

Ozone was generated from pure, dry oxygen using an Ozotech Model OZlPCSN
ozone generator (Ozotech, Inc., Yreka, Ca). The gas ﬂow rate to the reactor was
controlled by a Sierra Instrument Model 900 mass ﬂow controller, which was installed
between the ozone generator and the ozone contactor. A check valve was installed

downstream of the ﬂow controller to prevent water in the ozone contactor from cycling

37

back into the ﬂow controller and the ozone generator. Inﬂuent and efﬂuent ozone gas
concentrations were measured spectrophotometrically at 254 nm in 0.2-cm ﬂow-through
quartz cells with the Milton Roy Spectronic Genesys-S spectrophotometer and
amperimetrically with the Orbisphere MOCA ozone analyzer (Orbisphere Laboratories,
Geneva, Switzerland). Aqueous ozone concentrations were measured amperimetrically
with the Orbisphere ozone analyzer.

It should be noted that since the design capacity of the system was greater than
the operating inﬂuent ﬂow, additional recirculation loops were provided in the FBT
column and ozone contactor in order to maintain the ﬂuidization of the bed in the FBT
column and the inlet velocity in the ozone contactor, respectively. Since these
recirculating ﬂowrates were several orders of magnitude greater than the inﬂuent
ﬂowrate, both FBT column and the ozone contactor could be considered as completely-

mixed reactors.

38

 

 

 

 

 

 

 

Waste Gas

 

 

 

 

 

Ozone

Contaetor

Feed Pump

Jl'\

 

 

 

 

 

 

 

 

 

 

 

 

Liquid-maria

Sampling Pump

 

 

 

U

E]

O
O

 

 

 

 

[2?] Flow Controller

Ozone
Generator

ROIOMIOH

Tank

 

7

F BT Pump

 

 

 

 

 

 

FBT

 

 

 

4
l
‘L

 

 

Recycle

Pump £02

 

 

-Water

—Ozone~

 

 

 

 

39

Figure 3.2. Schematic of the ozonation/FBT system

L——e
entrained gas/liquid mixture

Figure 3.3. Schematic of a downﬂow ozone contactor

3.2.3. FBT/Ozonation System

The F BT/ozonation system, shown in Figure 3.4, was essentially the same as that
described in Section 1.2.2 with the exception that raw water was pumped ﬁrst to the FBT
column. The water from the F BT column ﬂowed by gravity to the ozone contactor. Water
from the ozonation retention tank ﬂowed by gravity to a holding tank from where it was
directed to the bioﬁltration column described in Section 3.2.5. A recycle line was

provided to recycle water from the ozonation system to the FBT column.

Waste Gas

 

_

 

 

111113

 

 

——1

El 4 v

o o T
1 Spectrophotometer
[I]
O O

Liquid-phasio

Ozone \_ Drain
m1; 4a“

Sampling Pump

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

KN

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

AL Ozone Retention Holding
Wind" EEISJFIow Controller Tank Tank
[:21 To bioﬁiter
O
0 Ozone
l] Generator FBT Pump r
W" ° '\

 

 

 

 

 

 

 

 

Raw _
W31“ Q Feed Pump

Figure 3.4. Schematic of the FBT/ozonation system

40

3.2.4. FBT System

The biodegradation of NOM was studied using a recirculating FBT system shown
in Figure 3.5. The recirculating FBT system used the same FBT column employed in the
FBT/ozonation system. In this conﬁguration all lines connected to the ozonation system
were shut off.

Water was continuously pumped from a holding tank to the bottom of the FBT
column by a peristaltic pump. The ﬂuidization of the bed was maintained by recirculating
water from the top of the reactor to the bottom by a second peristaltic pump. The efﬂuent

ﬂowed by gravity into the collection vessel from where samples were taken for analyses.

3.2.5. Bioﬁltration system

The bioﬁltration system (see Figure 3.6) included a glass chromatography column
with a diameter of 2.5 cm and a total volume of 100 cm3, a peristaltic pump, and feed and
collection reservoirs. All ﬁttings and tubing were of stainless steel or PTFE. An
acclimated bioﬁlm media were taken from the FBT system and packed into the
bioﬁltration column. The system was operated at room temperature.

The system operated in either single-pass or recalculating modes. In the single-
pass mode, water was passed from the feed tank through the bioﬁlter and collected in the

collection reservoir. The EBCT in the coltunn was controlled by varying the ﬂow rate of

41

the inﬂuent ﬂow. The samples were taken for analyses from the system inﬂuent and
effluent.

When the recirculating mode was employed, water was circulated from the feed
reservoir through the column and back to the reservoir. The operation is described in
more detail in Section 3.3.6. Between the experiments, the bioﬁlm was maintained by

recirculating ozonated water through the ﬁlter.

 

 

 

FBT

 

 

Recirculation Pump Efﬂuent Tank

Influent f = )

lnﬂuent Pump

 

 

 

 

 

 

 

Samples

 

Figure 3.5. A schematic of the recirculating FBT system

42

 

 

 

 

 

 

Bloiilier Bioﬁltel’

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Tank Eli‘luent Tank Samples : Pl

(a) (b)

Figure 3.6. A schematic of the bioﬁltration system

43

Samples

3.3. ANALYTICAL METHODS

3.3.1. Organic carbon.

Total and dissolved organic carbon were determined using a Dohrmann Model
DC-190 carbon analyzer (Dohrmann, Santa Clara, Calif) that uses the combustion
method and an 01 Analytical Model 1010 analyzer that uses the UV/persulfate method
(Standard Methods, 1992). A 5 mL sample loop was employed in the 01 Analytical
analyzer. A 40 mL sample was used with the OI Analytical analyzer, and a 10 mL sample
was used with the DC-l90 analyzer. When it was necessary to measure dissolved organic
carbon (DOC), samples were preﬁltered through a 0.45 pm glass-ﬁber ﬁlter before
analysis.

All samples were acidiﬁed to pH less than 4.0 and purged with high purity helium
for three minutes. Each run was performed with four replicates. The ﬁrst replicate was
usually greater than the three subsequent replicates and, when it was the case, the ﬁrst
replicate was eliminated. The average relative analytical accuracy was approximately
1.5% for the DC-190 analyzer and 0.6% for the 01 Analytical analyzer. To ensure the
reliability of results, the blank (Milli-Q water) and 2.5, 5, 7.5, and 10 ppm standards were

run before running the samples.

44

3.3.2. UV-254

UV absorption was measured at a wavelength of 254 nm with a Milton Roy
Genesis-5 spectrophotometer (Milton Roy, Inc., Rochester, NY) using a 1 cm quartz cell.

Samples of treated and untreated natural waters were analyzed without pH adjustment.

3.3.3. Humic Substances.

The concentration of humic substances in water samples was measured by
adsorption on XAD-8 resin using a slightly modiﬁed procedure of Method 5510C
(Standard Methods, 1992). The resin was packed in a 1.0 x 15-cm glass column. The
resin preparation and cleaning procedures were followed as described in the Standard
Methods. The fractionation procedure was as follows. A 100 mL sample acidiﬁed to pH 2
with concentrated phosphoric acid, was pumped onto the top of the resin bed at a rate of
2-3 lemin. The efﬂuent was collected at the bottom of the column. The ﬁrst 3 mL of
the column effluent were discarded. After the sample passed through the column, 3 mL of
Milli-Q water were pumped onto the column to displace the portion of the sample
remaining in the column and to remove water soluble organic acids that may be retained
by the resin. The column efﬂuent was analyzed for DOC, which represented the non-
humic fraction of the dissolved organic matter. The column containing adsorbed organic
materials was back eluted with 100mL of 0. IN NaOH at a rate of 0.8-1 .0 mL/min. The
eluent was immediately acidiﬁed with concentrated phosphoric acid to a pH of 2-4. Affer

collection, the eluent was purged with high-purity helium to remove inorganic carbon,

45

and analyzed for DOC. The organic content of the eluent represented the humic substance
concentration. The average mass balance for humic and nonhunmic fractions of NOM in

Huron River water ranged from 92 to 99% with an average of 95%.

3.3.4. AMW Distributions

Apparent molecular weight (AMW) distributions were determined by an
ultraﬁltration membrane technique. Molecular weight separations were performed using a
pressurized cell apparatus (Amicon Corp, Danvers, Mass.) using 200-mL stirred cells and
YM and YC series ultraﬁltration membranes characterized by nominal molecular weight
cutoffs of 500, 1000, 3000, 10000, and 30000 AMW units (Collins and Vaughan, 1993;
Amy et al., 1992). The ultraﬁltrations were performed in parallel, therefore, the
procedure did not provide a discrete AMW fraction, but rather the accumulation of
organic matter below the given AMW cutoff of the membrane. The UF membranes were
prepared according to the manufacturer’s recommendations by rinsing with Milli-Q water
for one hour and alternate soaking in a 0.1N NaOH solution and Milli-Q water at least
twice. The cleaning removed the glycerine and any absorbing materials coating the UP
membranes. Prior to sample ﬁltration, 200 mL of Milli-Q water was ﬁltered and tested
for TOC to assure that there was no contamination. To avoid a dilution effect from the
rinse water, the ﬁrst 20 mL of the permeate was wasted, then the sample was collected.
All permeates obtained were characterized according to organic carbon content. The

detail analytical protocols are presented in Mellema (1998).

46

3.3.5. THMFP.

THMs were determined using EPA Method 3810 (US EPA, 1988) and the
Standard Method 6232 (Standard Methods, 1992). Chlorine demand and THMFP were
determined using Standard Methods 2350 and 5710, respectively (Standard Methods,
1992).

Method 3810 is a static headspace technique for extracting volatile organic
compounds from samples. It is a simple method that allows for a large number of samples
to be screened in a relatively short period of time and is well suited for comparative
analysis (Kolb, Auer, and Pospisil, 1983).

A Model 8700 Autosystem gas chromatograph (Perkin Elmer Corporation,
Norwalk, Connecticut) with an HS 40 automated head space sampler (Perkin Elmer
Corporation, Norwalk, Connecticut) was used. The gas chromatograph was equipped
with a 63Ni electron capture detector (ECD) with a 30 m x 0.53 mm DB-624 column (J &
W Scientiﬁc, Folsom, California). The detector temperature was set at 350 °C. The
injector temperature was 200°C. The column temperature was initially set at 80 °C, then
increased at a rate of 10 °C/min to 200 0C, held for 0 minutes. The ﬂow rate of the carrier
gas (H2) is 10 mL/min.

For quality control, samples were periodically analyzed using the liquid-liquid
extraction method. Detail analytical protocols for the head-space and liquid-liquid
extraction methods and quality control procedures are presented in Kasarabada (1997).

The methods produced similar results and were found to be statistically equivalent

(Kasarabada, 1997).

47

3.3.6. BDOC

The BDOC was determined by recirculating raw or treated water through the
bioﬁltration column described in Section 3.2.3. The procedure described by Cipparone,
Diehl, and Speitel (1997) was followed except for slight modiﬁcations. Brieﬂy, before
starting the sample analysis, 2 L of the water sample were pumped through the column
and discarded. Upon completion of this step, the volume of the water in the feed reservoir
was adjusted to 4 L and recirculation of the sample began at a rate of 10-12 mL/min. The
TOC concentration was monitored until no further removal of TOC was observed, which
indicated the completion of the biodegradation. Biodegradation was usually completed
after 5 or 6 days, but recirculation was continued for seven days in most cases to ensure
complete biodegradation. The BDOC concentration was calculated as the difference
between the TOC concentration in the sample before the analysis and that in ﬁnal

biodegradation sample.

3.3.7. Other parameters

Other physical-chemical parameters, including pH, alkalinity, turbidity were

determined according to Standard Methods (1992).

48

3.4. QUALITY ASSURANCE AND CONTROL

3.4.1. Data Management

Considering the large volume of data collected in this research, it was important
to develop and maintain a cross-referenced database for all data, interfaced with graphical
and statistical software. The spreadsheet formats are presented in Appendix D. The main
spreadsheets for the pilot ozonation/FBT system and for bench-scale ozonation system,
shown in Table DJ and D.2, included the following information:

0 date of the collection of water batch

0 date of sample collection (for the pilot system)

0 date of bench experiments (for the bench system)
0 mode of operation (for the pilot system)

0 the operating conditions

0 analytical data.

The measured parameters (e.g., gas ﬂow, ozone concentrations, temperature) and
parameters that were analyzed immediately upon collection, including pH, alkalinity,
turbidity, UV-254, were hand written on a blank spreadsheet and then transferred to an
appropriate electronic ﬁle, which automatically calculated ozone dose, retention time (for
the bench system) and EBCT (for the pilot system). Other analytical data could be
accessed through the appropriate spreadsheets (see Tables D.3-D.5), which shared
information with the main spreadsheet.

Since many analytical parameters, including TOC, HS, nonHS, and MWD,
involved TOC measurements, a separate spreadsheet was maintained for all TOC data

49

collected on a speciﬁc day. The TOC spreadsheet shown in Table D6, contained
complete sample description and could be accessed by the date of analysis, which in all
cases was performed on the day of sampling.

All graphs presented in this research are interfaced with appropriate spreadsheet
data, the reference of which is displayed by clicking on the graph. All raw data were ﬁled

and could be accessed by the date of analysis.

3.4.2. Sample Collection and Analysis

Samples were collected in the appropriate containers and were labeled to provide
complete sample identiﬁcation. To minimize cross-contamination, containers were
designated to speciﬁc types of samples by the type of analyte and sampling location. To
avoid contamination, samples for TOC analyses were collected to TOC vials that were
also designated to speciﬁc locations (e. g., samples collected ﬁ'om the retention tank were
colected in the same set of vials). All glassware was cleaned according to Standard
Methods (1992). Most samples were analyzed within several hours after collection and all
samples were analyzed with 24 hours after collection.

Proper instrument operation was assessed before each sample analysis and blanks
were run to ensure that system components were clean. To ensure the quality of the
results the standards with known concentration of compounds were run before each
analysis. All standards solutions were stored for no longer than recommended by the
manufacturer. For TOC analyses using the Dorhmann analyzer, the calibration was

conducted before each run. The calibration of the 01 Analytical analyzer was performed

50

when fresh standards were prepared. The standards of 0, 2.5. 5, and 7.5 mg/L were run
before each set of analysis and proper operation was assessed by maintaining a control
chart. For humic substance and MWD analyses, the blanks were run through the XAD-8
columns and ultraﬁltration membranes to ensure that the systems were clean. The

glassware used in these analyses was designated to speciﬁc samples.

3.4.3. Data Analysis

Analytical quality control procedures were applied to properly qualify the results
obtained from this research. Statistical tests for comparison of variances and means were
based on 95% conﬁdence levels. Up to ﬁve replicates were used for assessment of
precision at different levels of processing/analyses. Accuracy controls included
comparison to standards or spikes, where appropriate. The summary of QA/QC results

are presented in Table 3.8.

51

Table 3.8. QA/QC Summary

 

 

 

 

 

 

 

 

 

 

 

 

Parameter Limit of Detection Accuracy Precision
% recovery Range or % RSD
TOC, mg/L 0.07 98-101 0.8 rsd
UV-254, 1/cm 0.002 - 0.001- 0.002 range
THMs‘, rig/Ll 5 85-120 10941.1 rsd
BDOC, mg/L 0.2 - -
HS, mg C/L n/d 952t3 2.4 :1: 1.5rsd
MW500, mg C/L n/d 95i2.7 4.0 i 3 rsd
‘ Kasarabada (1997)

52

 

4. PRELIMINARY EVALUATION OF THE OZONATION/FBT PROCESS

4.1 . INTRODUCTION

The objectives of the study presented in this section were to
o assess the impact of ozonation and F BT on the properties of NOM, including
TOC, UV-254, humic substances, molecular weight distribution, and THMFP.

o evaluate the ability of the system to sustain turbidity without pretreatment.

o assess the role of ozonation and F BT in the transformation of NOM.

o evaluate the effect of the recycle of a portion of system effluent on the

efﬁciency of the ozonation/FBT process.

The hypotheses that were tested in this portion of the study are described in each
section that follows. Section 4.2 presents the results of the experiments that assessed the
impact of ozonation and FBT on the properties of NOM. These experiments also tested
the hypothesis that the efﬁciency of the ozonation/FBT process would increase with
ozone dose. Sections 4.3 - 4.5 present the results of the experiments that assessed the
role of ozonation and FBT in the transformation of NOM. These experiments also
attempted to reconcile conﬂicting reports regarding the effect of ozonation on the
production of biodegradable organic matter and low-molecular weight organic matter

(less than 1,000 daltons).

53

4.2. POTENTIAL OF OZONATION/FBT SYSTEM

The experiments described in this section assessed the impact of the ozonation/
FBT process on the properties of NOM. It was expected that ozone would oxidize NOM,
rendering it more amenable to the removal by subsequent biodegradation. As such, the
efficiency of the ozonation/FBT system towards the removal of NOM was expected to
increase with ozone dose.

This study was conducted using the pilot-scale ozonation/FBT system described
in Section 3.2 (see Figure 3.2.2). A simpliﬁed schematic of the treatment train with
sampling locations is shown in Figure 4.1. The experiments were conducted at room
temperature. Ozone dose was varied from 0 to 2 mg/mg C. The parameters that were
monitored in this study included TOC concentration, UV-254 absorbance, HS, turbidity,
chlorine demand, THMFP, and AMW distribution. pH and alkalinity were also measured
along the treatment train. Sampling was started after four months of acclimation period
(from July to November 1996). During that period water was ozonated with an ozone
dose of 2 mg/mg C. As can be seen in Figure 4.2, no removal of TOC in the FBT column
occurred during the ﬁrst two months of the acclimation period. During the next two
months, the removal steadily increased reaching a plateau at the end of November.

The experiments described in this chapter were conducted in the period from
November 1996 to March 1997. For each ozone dose, the system was operated for
approximately one month. After all three dosages were evaluated, experiments were
repeated using ozone doses of 1 and 2 mg/mg C for another two to three weeks. This

experimental design was used to determine if seasonal variations of water quality and

54

changes in bioﬁlm growth during prolonged experimental time would affect the process
performance.

The FBT alone resulted in the maximum removal of about 16 percent of organic
carbon from untreated Huron River water. As can be seen in Figure 4.3, the TOC removal
steadily increased with EBCT reaching a plateau at an EBCT of approximately 180
minutes. Subsequently, the ozonation/FBT system was operated at a ﬂow rate of 10
mL/min, which corresponded to an empty-bed contact time (EBCT) in the FBT column
of 180 minutes. This ﬂow rate was selected for the following reasons:

(1) several operational conditions could be investigated using the same batch of water;
(2) the biodegradability potential of Huron River water could be evaluated; and

(3) the transformation of NOM during FBT could be better understood.

55

Raw water

 

7

Ozonation

 

 

 

 

 

 

 

Sample

Sample

 

FBT

Efﬂuent

 

 

 

 

 

Sample

Figure 4.1. Simpliﬁed schematic of the ozonation/FBT system

56

300

 

250*

200

 

 

TTXDnmnowm,%i
a
O

.3
,o
O

 

50

 

 

 

 

-50

 

 

Date

Figure 4.2. Time plot of TOC removal from ozonated water by F BT during acclimation

period (ozone dose - 2 mg/mg C, EBCT — 180 min)

57

TOCmmmMﬂt

m

w

.5
A

.3
N

.s
O

on

 

 

 

 

 

 

 

 

a) 1m) 1K) an 25) an an
EBCT.mm

Figure 4.3. Removal of TOC from untreated Huron River water by F BT

58

Figures 4.4—4.5 show time plots of TOC and UV-254. The results for THMF P
were presented by Kasarabada (1997) and are summarized in Table 4.1. TOC followed
expected trends, with both ozonation and F BT contributing to the reduction of TOC (see
Figure 4.4). The TOC removal increased with ozone dose. The lowest TOC concentration
after ozonation/FBT was approximately 3 mg/L, which was achieved at an ozone dose of
2 mg/mg C. The results for winter and spring testing were essentially the same. Figure
4.5 shows a time plot of UV-254. The UV-254 reduction occurred mostly through
ozonation, rather than through biodegradation. The removal of UV-254 increased with
ozone dose. When no ozone dose was applied FBT resulted in the reduction in UV-254

from 0.160 to approximately 0.135 cm".

59

 

7.0 2.5

 

 

 

 

   

 

 

 

 

 

 

 

#20
.1 .
2’ o
'- a:
' ' 1.58
‘40-» '
i '2 E
, = i
:3.0 ‘1
8 I, o...0...o...o p---o---o 11.0 g
N
3 o
l-
2.0 f a
.b...o...o...o...o" ‘1‘ 5 “0.5
1.017WW‘WW,___
0.0 11111111111411111519111 0.0
95 q. 91 ’\ o '5 '5 K to '1. 9 N '5 i» 65 ’\
ﬁwffeedseeetrr assess
Date

 

f—hRawwater +Ozonation +Ozonation/FBT --e--Ozon:d£sej

 

Figure 4.4. Variations in TOC concentration with time and ozone doses in raw water and

along the treatment train (FBT EBCT = 180 minutes)

60

02 250

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.18 ZN
0.16 —W v 12.00
0.14 L 1
1 150 (3”
s - §
91' .1
5 . : 1.00%
f O
L r 0.50
0.02 4‘.
O 4 1 1 1 1 1 1 1 1 1. 1. 1 1 1 1 1 1 1 ee 1 é 1 .e 0.00
9 '\ Q ‘5 '5 '\ Q ’\ h h ’\
"\@\"<‘9\‘<ﬂy <19 0‘09 "°~\" 4‘ '5” 6” 1S” 3%" 43%" «it $4941 41’»?
Date
__1

[:o—Raw water +Ozonation +Ozonation/FBT --o--Ozoned0391

 

 

Figure 4.5. Variations in UV-254 with time and ozone doses in raw water and along the

treatment train (FBT EBCT = 180 minutes)

61

Table 4.1 summarized the effect of ozonation/FBT on major water quality

parameters. The removal of TOC increases with ozone dose reaching approximately 48

percent at a dose of 2 mg/mg C. FBT alone resulted in a reduction in turbidity to some

extent. The removal of turbidity increased with ozone dose reaching nearly 80 percent at

a dose of 2 mg/mg C. It should be noted that breakthrough of turbidity was not observed

during the operation of the ozonation/FBT system. FBT alone resulted in approximately

20 percent decrease in the concentration of UV-254 absorbing compounds. This suggests

that F BT was able to break down some aromatic portion of NOM. As was the case with

TOC and turbidity, the removal of UV-254 increased with ozone dose reaching nearly 80

percent at a dose of 2 mg/mg C. It is worth noting that FBT alone was able to remove

approximately 16 percent of TOC and UV254-absorbing compounds. The removal of

THMFP also increased with ozone dose. The difference in the concentration of THMFP

for ozone doses of 1 and 2 mg/mg C was not statistically signiﬁcant at 5 percent level.

Table 4.1. Performance of the ozonation/FBT system

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Raw water System efﬂuent
Parameter (system Ozone dose, mg/mg C
inﬂuent) 0 0.5 1.0 2.0

TOC, mg/L 6.041032 4.821002 4.441014 3.711010 3.041012
Turbidity, NTU 1.331024 1.051015 0771020 04910.16 02810.11
UV-254, cm'I 0.16610007 0.13910009 009110.005 005110008 003510003
C12 demand, mg/L 6.951136 5.451049 2.471041 1.53105 1.71041
THMFP, ug/L 403146 351129 164126 135134 116128

 

62

 

4.3. EFFECT OF OZONATION AND FBT ON TRANSFORMATION OF NOM

The results of the experiments described in the previous section agreed with the
hypothesis that the efficiency of the ozonation/FBT system to remove NOM increased
with ozone dose. The ozonation/FBT was also able to control turbidity. The next
important question that needed to be answered was whether or not the efﬁciency of FBT
would increase with ozone dose.

There have been conﬂicting reports on the effect of ozone doses on the production
of biodegradable organic. Miltner et al. (1992) who studied the ozonation of Ohio River
water reported that BDOC increased with an increase in ozone dose to up to 3 mg/mg C.
Volk et al. (1993) determined that the ozonation of Seine River and Oise River waters
resulted in a maximum production of BDOC at a dose of 0.5 mg/mg C. BDOC then
decreased with higher ozone dosages. The authors also showed that an addition of
hydrogen peroxide resulted in a decrease in the production of BDOC compared to
ozonation alone. Cipparone, Diehl, and Speitel (1997) observed an increase in BDOC
concentration during ozonation of Lake Austin water. BDOC leveled off at a dose of 1
mg/mg C. This study attempted to reconcile these differences.

Figure 4.6 summarizes TOC removal for all ozone doses, categorized by the type
of removal. Ozonation at an ozone dose of 0.5 mg/mg C resulted in the removal of
approximately 16 percent of TOC. The following F BT resulted in the removal of
additional 12 percent of TOC, compared to 16 percent of TOC that was removed from
nonozonated water (the difference was statistically signiﬁcant at 5 percent level). The

removal of TOC via ozonation increased with ozone dose and appeared to reach a plateau

63

at a dose of approximately 1 mg/mg C. Ozonation at doses of less than 1 mg/mg C did
not result in an increase in the production of BDOC. Ozonation at a dose of 2 mg/mg C
resulted in an increase in the production of BDOC, which was removed by the following
FBT. The removal of TOC after ozonation and FBT reached a maximum of 50 percent at
an ozone dose of 2 mg/mg C, with approximately 20 percent of that being removed via
FBT.

As can be seen in Figure 4.7, FBT alone resulted in an approximately 16 percent
reduction of UV-254 in the raw water. However, when ozone was applied, the removal of
the UV-254 occurred mostly through ozonation rather than the FBT. This is not
surprising considering the fact that ozone is especially effective towards the oxidation of
UV254-absorbing organics, i.e. the aromatic fraction of NOM (Masten, 1991). The
removal of UV-254 during ozonation increased with ozone dose. Ozonation with an
ozone dose of 1 mg/mg C resulted in the removal of 65 percent of UV-254. An increase
in an ozone dose to 2 mg/mg C contributed only an additional 10 percent to UV-254
removal. Only slight removal of UV-254 was observed aﬁer the FBT of ozonated water.

The HS concentrations decreased with an increase in ozone dose. The FBT, as
expected, did not affect HS concentration. Figure 4.8 shows that the most noticeable
changes in HS concentrations occurred when ozone dose increased to 1 mg/mg C, with
only minor changes occurring with an increase in ozone dose from 1 to 2 mg/mg C. This

correlates very well with the UV-254 trends.

64

 

 

 

 

 

 

 

 

 

-1O
Ozone dose. mg/mg C

 

3.50 A -

 

 

 

 

 

 

 

 

0.50

 

 

0.00 . 1 1 1 471—.
o 0.5 1 1.5 2 2.5

Ozone dose. mglmg C

 

f "+_‘__B‘23?a_1733 ”+7 ‘FEi +6 2 9 BTU 03/161? __1

Figure 4.6. Effect of ozone dose on the removal of TOC after ozonation and FBT

(EBCT = 180 min, RT in ozone contactor =100 min, liquid ﬂow =10 mL/min, gas ﬂow
15 mL/min)

65

 

 

 

 

 

 

 

k

 

UV-254 removal, %

\
1

 

 

 

 

I + i
2

Ozone dose. mg/mg C

 

[+Ozonation +FBT+Ozonation7PRTi

 

Figure 4.7. Effect of ozone dose on the removal of UV-254 after ozonation and FBT

(EBCT = 180 min, RT in ozone contactor =100 min, liquid ﬂow =10 mL/min, gas ﬂow
15 mL/min)

66

4.1K)

 

3.50

 

 

 

3.00 1

 

 

2.50

 

 

 

HS. mall.
10
8

 

1.50 ._

 

1.00

0.50

 

 

 

 

 

 

 

0.00

 

o 0.5 1 1.5 2 2.5
Ozone dose. mglmgC

 

{+Ozonation +Ozonation/_:B__:1

 

Figure 4.8. Effect of ozone dose on the removal of HS after ozonation and FBT

(EBCT = 180 min, RT in ozone contactor =100 min, liquid ﬂow =10 mL/min, gas ﬂow
15 mL/min)

67

4.4. FRACTIONATION OF NOM BY ULTRAFILTRATION

The results of the experiments described in Section 4.3 showed that under
operational conditions studied ozone appeared to oxidize some organic matter that could,
otherwise, be degraded biologically. This ﬁnding was further explored using fractionation
of NOM by ultraﬁltration, which provided additional information on the transformation
of organic matter during ozonation and FBT.

There have been conﬂicting reports on the transformation of molecular weight
fractions during ozonation. Several investigators observed the destruction of high-
molecular weight compounds and a corresponding increase in lower molecular weight
compounds upon ozonation (Hongve et al., 1989; Flogstad and Odergaard, 1985;
Veenstra et al., 1985), while others saw a reduction of low molecular weight compounds
(Amy, Kuo, and Sierka, 1988; Anderson, Johnson, and Christrnan, 1986). No correlations
between these changes and ozone dose was found. This study attempted to reconcile
these differences.

The results are described in detail by Mellema (1998). Brieﬂy, Figures 4.9-4.12
show AMW distribution of organic matter in raw water and in water after ozonation and
FBT for various ozone doses. The FBT of raw water appears to reduce, to some extent,
higher molecular weight organic matter (greater than 1,000 daltons), except for the
fraction with a molecular cutoff greater than 30,000 daltons (see Fig. 4.9). The low
molecular weight compounds (AMW < 1,000) were only slightly reduced after FBT. This
is thought to be because the FBT broke down some higher molecular weight compounds

into smaller constituents, thus, contributing to low molecular weight fractions. An

68

increase in the concentration of organic carbon in the fraction with an AMW of greater
than 30,000 daltons has yet to be explained.

As expected, ozonation with an ozone dose of 0.5 mg/mg C resulted in a decrease
in the concentration of higher molecular weight organic matter (see Figs. 4.10). However,
a corresponding increase in the concentration of organic carbon in lower-molecular
weight fractions, which were thought to be potentially biodegradable, was not observed.
This suggests that along with higher molecular weight compounds, ozone simultaneously
oxidized low molecular weight compounds, thus reducing the amount of substrate
available to microorganisms during subsequent F ET. This correlates well with the results
of the experiments presented in Section 4.3, which showed a decrease in BDOC
concentration after ozonation (see Figure 4.5).

Ozonation with an ozone dose of l mg/mg C resulted in further removal of higher
molecular weight compounds (see Fig 4.11). Once again, an increase in the concentration
of low molecular weight fractions was not observed. Based on the results of earlier
experiments (see Fig. 4.10) and the TOC trends (see Fig. 4.6), this already was expected.

Ozonation with an ozone dose of 2 mg/mg C did result in an increase in the
concentration of low-molecular-weight organics (see Fig. 4.12). The result correlates well
with the data presented in Figure 4.6, which showed an increase in BDOC when ozone
dose was increased from 1 to 2 mg/mg C.

As was the case with raw water, the FBT of ozonated water broke down a portion
of high molecular weight organic matter (see Figures 4.9-4.11). It should be noted that
this partial removal of high molecular weight compounds could not be attributed only to

the adsorption of these materials to the bioﬁhn support media. This is because very little

69

TOC removal in the FBT column was observed during acclimation period (see Figure

4.2).

70

Organic carbon, mg/L
0
a
‘ 1

< 500

 

 

 

 

 

500 - 1000 1000 - 3000 3000 - 10000 10000 - 30000 > 30000
AMW fractions

flRawVaTéj gnFBTJ

Figure 4.9. AMW distribution after F BT

(EBCT - 180 min)

71

Organic carbon. mg/L

 

 

 

500-1000 1000-3000 3000-10000 10000-30000
MW fractions

 

I lRawwater I020nation DOZOnation/FET 1

 

Figure 4.10. AMW distribution after ozonation and FBT

(ozone dose -0.5 mg/mg C, EBCT — 180 min)

72

Organic carbon. mg/L

 

 

 

 

 

 

 

 

 

Figure 4.11. AMW distribution after ozonation and FBT

(ozone dose -1 mg/mg C, EBCT — 180 min)

73

Organic carbon, mg/L

  

   

 

1000-3000 3000-10000 10000-30000
MW fractions

lRawwaterIOzorEtionDOzonation/FBT

Figure 4.12. AMW distribution aﬁer ozonation and FBT

(ozone dose —2 mg/mg C, EBCT — 180 min)

74

 

>30000

4.5. CONFIRMATION STUDY

The results of the experiments presented in Sections 4.3 and 4.4 showed that TOC
removal during ozonation increased with ozone dose and appeared to reach a plateau at a
dose of approximately 1 mg/mg C. Ozonation with ozone doses of 0.5 and 1 mg/mg did
not appear to result in an increase in the concentration of low molecular weight
compounds. A decrease in BDOC concentration was observed. An increase in ozone dose
from 1 to 2 mg/mg C did result in an increase in the concentration of both low molecular
weight compounds and BDOC. From the observation of the trends shown in Figures 4.6,
4.10-4.12, it appears that when the removal of TOC during ozonation leveled off, ozone
was utilized to produce BDOC and low molecular weight compounds rather than to
mineralize organic carbon. Based on these ﬁndings, the following hypotheses were
formulated:

0 under the operational conditions studied and for ozone doses greater than 1
mg/mg C, increasing ozone dose would not affect TOC removal;

0 under the operational conditions studied and for ozone doses greater than 1
mg/mg C, the production of biodegradable organic matter would increase with
ozone dose; and

0 under the operational conditions studied and for ozone doses greater than 1
mg/mg C, the concentration of low molecular weight compounds would increase

after ozonation.

75

This section describes the results of the experiments that were conducted under
the same conditions as described in Sections 4.3 and 4.4 with the exception that ozone
dose was varied from 0 to 7 mg/mg C. Although ozone doses greater than 2 mg/mg C are
not practical from economic standpoint, this experimental design was used to verify
hypotheses described above.

As can be seen in Figures 4.13 and 4.14, the trends for TOC and UV-254 were
remarkably similar to those obtained earlier (see Figures 4.6 and 4.7). Ozonation could
potentially remove approximately 23% of organic carbon (see Fig. 4.13). As expected,
this also resulted in the oxidation of some potentially biodegradable organic matter. The
removal of TOC increased with ozone dose and reached a plateau at a dose of 1 mg/mg
C, which agreed with our previous ﬁndings. Further increases in ozone dose resulted in
an increased production of BDOC, which reached nearly 40 percent at an ozone dose of 7
mg/mg C.

Figure 4.14 shows that ozonation was able to remove up to 80 percent of UV254-
absorbing compounds at an ozone dose of approximately 2 mg/mg C. Further removal of
UV-254 was difﬁcult to achieve even at higher ozone doses.

Figure 4.15 shows AMW distribution after ozonation and FBT for an ozone dose
of 7 mg/mg C. As expected, ozonation resulted in a decrease in higher molecular weight
compounds and in a dramatic increase in the concentration of low molecular weight
organic matter. The FBT resulted in signiﬁcant removal of both low and high molecular

weight compounds.

76

70

 

 

TOC removal %
8 8
l
1
l

 

 

T ﬁr

0.00 1.00 200 3.00 4.00 5.00 6.00 7.00 8.00
Ozone dose. mg/mg C

 

(Es—Ozonation+FBT-e-Ozonation/FBT]

Figure 4.13. Effect of ozone dose on the removal of TOC after ozonation and FBT

(EBCT = 180 min, RT in ozone contactor =100 min, liquid ﬂow =10 mL/min, gas ﬂow
15 mL/min)

77

100 _ _

 

 

 

 

 

 

UV-254 cm 1
B 8 8 8 8

\
\

l

l

l

l

i

i

 

 

 

0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00
Ozone dose, mglmg C

 

L—o—Ozonation +FBT+OzonatioanBTj

 

Figure 4.14. Effect of ozone dose on the removal of UV-254 after ozonation and FBT

(EBCT = 180 min, RT in ozone contactor =100 min, liquid ﬂow =10 mL/min, gas ﬂow
15 mL/min)

78

 

 

 

 

 

 

Organic carbon. mglL

  
 
  

 
 

500-1000 1000-3000 3000-10000 10000-30000 >30000
AMW fractions

lRawLaterIOzoiniaEnitJOBnaﬁtioJ/FBTJ

Figure 4.15. AMW distribution after ozonation and F BT

(ozone dose -7 mg/mg C, EBCT — 180 min)

79

4.6. OZONATION/FBT WITH RECYCLE

The results of the study presented in the previous sections showed that, under
operational conditions studied, ozone was utilized inefficiently in the ozonation/FBT
system: ozone broke down higher molecular weight compounds and simultaneously
oxidized low molecular weight compounds that could, otherwise, be degraded
biologically.

This section describes preliminary evaluation of the recycle mode of operation in
which a portion of system efﬂuent was recycled back to the ozonation step. By recycling
a portion of system effluent, water alternately passed through ozonation and
biodegradation steps. During each ozonation step, only those compounds with highest
affmity to ozone were expected to react. The reaction products and nonreacted
compounds would pass to the biodegradation step, which would remove biodegradable
portion of organic matter. Nonbiodegradable organic materials would be recycled back to
the ozonation step for further ozonation. In this way, ozonation was expected to degrade
only those compounds that were not biodegradable. This approach was successfully
utilized for the treatment of bleaching efﬂuents in pulp and paper industry, which resulted
in increased biodegradation efﬁciency relative to the amount of ozone consumed

(Heinzle, F ahmy, and Kut, 1992).

The experimental results for the ozonation/FBT process with recycle at a recycle
ratio (recycle ﬂow to inﬂuent ﬂow ratio) and an ozone dose of 1 mg/mg C are presented
in Table 4.2. As was shown in the previous sections, at an ozone dose of 1 mg/mg C the
lowest level of BDOC concentration was achieved after ozonation , which is thought to

80

be because of the mineralization of organic carbon during ozonation. The results for

single-pass ozonation/FBT process using the same raw water are given for comparison.

Table 4.2. Comparison of recycle and single-pass modes of ozonation/FBT operation

 

 

 

 

 

Single-pass Ozonation/FBT
Parameter Raw water ozonation/FBT with recycle
TOC, mg/L 5.96:0.23 3.67:0.07 3.49i0.03
Turbidity, NTU 1.17:1:0. 1 3 0.621006 0.401007
UV-254, cm'I 0.163:I:0.002 0.049:i:0.001 0.048i0.001
THIVIFP, rig/L 425i15 120:34 75

 

 

 

 

 

 

The recycle mode of operation resulted in a slight increase in the removal of TOC
and UV-254 compared to the single-pass mode. The effect was more noticeable for
turbidity. Due to the malfunction of the gas chromatograph, only one data point was
obtained THMFP and it was not possible to determine if the difference was statistically
signiﬁcant.

Although the recycle mode of operation appears to result in a greater removal of
organic matter and THM precursors than did the single-pass process, the effect was less

than expected. This is thought to be because of slow biodegradation of organic matter.

81

4.7. SUMMARY

The study presented in this chapter conﬁrmed the hypothesis that the efficiency of
the combined ozonation/FBT process increased with ozone dose. The FBT efﬁciency
however, appeared to decrease when ozone dose increased from 0 to 1 mg/mg C. Further
increase in ozone dose did result in an increased F BT efﬁciency. This is thought to be
because along with breaking down high molecular weight compounds, ozone oxidized
some low molecular weight, potentially biodegradable, organic matter. Fractionation of
NOM by ultraﬁltration conﬁrmed this ﬁnding. The study also showed that FBT was able
to break down some high molecular weight compounds into smaller constituents.

The results of the experiments partially explained conﬂicting information reported
in different sources on the effect of ozone doses on production of biodegradable organic
matter and low molecular weight organic compounds. It appears that when ozonation is
conducted under conditions that encourage the mineralization of organic carbon, a
decrease in the concentration of BDOC and low molecular weight organic compounds
may occur. When no TOC removal is observed during ozonation, an increase in ozone
dose is likely to result in an increased production of biodegradable organic matter and
low molecular weight organic compounds. The operating conditions that affect the
mineralization of organic matter are investigated in Chapter 6.

The study presented in this chapter demonstrated a good potential of the
ozonation/FBT process to control THM precursors in drinking water. At an EBCT of 180
minutes and an ozone dose of up to 2 mg/mg C, the process was able to remove up to 50

percent of TOC and humic substances, up to 80 percent of UV-254, up to 70 percent of

82

THMFP, and up to 80 percent of turbidity. The pH and alkalinity did not signiﬁcantly
change during the treatment. The removal of TOC was comparable to that achieved at
Ann Arbor Water Treatment plant (Scadsen, 1997), which included softening,

ﬂocculation/sedimentation, ozonation, and activated carbon adsorption. Although the
recycle mode of operation resulted in increased process efficiency, the effect was less

than expected.

83

5. SURROGATE CHARACTERIZATION OF NOM

5.1. INTRODUCTION

The parameters that were used to evaluate the transformation of NOM during

ozonation and FBT included TOC, UV-254, humic substances and THMFP (see Chapter

4). Table 5.1 presents major characteristics of these parameters. (They were described in

more detail in Chapter 2.)

Table 5.1. Characteristics of surrogate parameters

 

 

 

 

Parameter Characteristics Ease of measurement
TOC Collective measure of NOM present in easy to measure;
water can be measured in-line
UV - 254 characterizes the aromatic portion of easy to measure;
NOM can be measured in—line
HS heterogeneous organic materials that difficult to measure

include most of the naturally occurring
organic matter in water;

precursors for THMs

THMFP potential of water to produce THMs on difficult to measure
chlorination;

any treatment process must be evaluated
on the basis of how effectively it can
reduce THMFP

 

 

 

 

 

 

The humic substance and THMF P analyses were laborious and time-consuming
and only four to six samples per week could be analyzed. This would limit the progress
of the study. Therefore, it was important, at that early stage of the project, to establish the

relationships between TOC, UV-254, HS, and THMFP, which could be used to substitute

84

one parameter for the others during further investigation of the ozonation and FBT

processes. The objectives of the study presented in this chapter were:

to test the hypothesis that there exists a correlation between TOC and UV-254;
the relationships between these parameters may, however, be different for
ozonation, FBT, or combined ozonation/FBT effluents;

to test the hypothesis that there exists a correlation between UV-254 and HS and
this correlation does not vary for waters at different treatment stages;

to test the hypothesis that there exists a correlation between TOC and THMF P
and, if so, to determine if this correlation varies for waters at different treatment
stages;

to test the hypothesis that there exists a correlation between UV-254 and THMFP
and, if so, to determine if this correlation varies for waters at different treatment
stages;

to determine if the relationships between surrogate parameters are ﬁrnctions of
ozone dose.

The experimental data used to develop the correlations between surrogate

parameters were generated from the pilot-scale ozonation/FBT system for ozone doses

ranging from 0 to 2 mg/mg C and an EBCT of 30 and 180 minutes. In the experiment

using FBT alone, the EBCT was varied ﬁom 15 to 360 min. The experiments were

conducted at room temperature. All experimental data generated during the period from

November 1996 to July 1997 were included in the analysis except for four data points

which appeared to be in error or because the paired datum was not obtained in the same

day.

85

5.2. UV-254 AND TOC

An important characteristic of the raw water quality observed during the period
when the data were generated was seasonal variations in TOC and UV-254. Figure 5.1
shows that TOC and UV-254 were relatively constant during the fall-winter period and
increased rapidly in the spring, which was probably the result of runoff from snowrnelt.
The peak of TOC was approximately 6.9 mg/L in May and then decreased in summer
period.

Figures 5.2-5.4 show the relationships between TOC concentration and UV-254
for ozonated water, water treated by F BT alone, and water treated by combined
ozonation and FBT, respectively. Equations representing the least-square linear ﬁt
through the data are included in each plot. Visual evaluation of the quality of ﬁt between
the data and the straight lines and the values of the correlation coefficients, r2, show that
there exist good correlations between TOC concentration and UV-254 for each treatment
stage. The slope of the correlation lines was, however, different, for different treatments.
The slope was the lowest for the ozonation and the greatest for the FBT, with the slope
for the ozonation/FBT being in the middle. This was not surprising considering the fact
that ozonation had greater effect on the removal of UV254-absorbing compounds rather
than on the removal of organic carbon, whereas biotreatrnent appears to be more effective
towards the removal of nonaromatic (nonUV-254 absorbing) portion of organic matter.
The results, thus, conﬁrmed the hypothesis that although there exists good linear

correlations between UV-254 and TOC concentration, the relationships were different for

86

waters at different treatment stages. Therefore, it does not appear that either of these

parameters can be substituted for the other when evaluating various treatment strategies.

 

 

 

 

 

8.00 , 0.3
7.00 ~~
1 0.25
6.00 1
~~ 02
5.00 1
§ 5
. 4.00 «- 015 3'
8 «i
'- B
3.00 -~
1 0.1
2.00 ..
«1 0.05
1.00 1
am i 1L + 1 1 1 *1 ‘1 0
Nov Dec Jan Feb Mar Apr May June July
Month

 

[+Toc ...... uv-254j

 

Figure 5.1. Seasonal variations of TOC and UV-254 in Huron River

87

 

    

 

 

 

71 e
O
61
S.
e 5*
S y=12.638x+3.8949
E4, .3 ’0 R2=0.8723
8 31
l-
21
1.
o . . . . ,
0 0.05 0.1 0.15 0.2 0.25 0.3

UV-254,1/cm

Figure 5.2. Correlation between TOC concentration and UV-254 for ozonated
water (ozone dose — 0-2 mg/mg C)

88

 

8.00

 

y = 44.257x- 1.8544
3.00« R2=0.9913

 

1.00 1

 

 

0.00

 

0 0.05 0.1 0.15 02 025
UV-254, 1/cm

Figure 5.3. Correlation between TOC concentration and UV-254 for biotreated water

(EBCT — 15-360 min)

89

 

 

 

  

y=18.333x+2.7144
R’=0.9423

 
 

 

 

 

0 0.05 0.1 0.15 02 025 0.3
UV- 2 5 4 . 1 [C m

Figure 5.4. Correlation between TOC concentration and UV-254 for ozonated and

biotreated water (ozone dose 1-2 mg/mg C, EBCT — 180 min)

90

An attempt was made determine if the relationships between TOC concentration
and UV-254 were functions of ozone dose. This was done by plotting average TOC
concentration versus average UV-254 for each ozone dose. Figure 5.5 shows the
correlations between average TOC concentrations and average UV-254 for ozonation and
ozonation/FBT effluents. The ozone-dose averaging technique produced good linear
correlations between TOC concentration and UV-254 despite the fact that the data had
relatively high standard deviations due to seasonal water quality variations.

From the plots in Figures 5.2, 5.4, and 5.5, ozone-dose averaging of the data did
not signiﬁcantly alter the relationships between TOC concentration and UV-254. The
advantage of the technique, however, is in that that it links the relationships to ozone
dose. For example, by observing the plots in Figure 5.5 one can conclude that most
signiﬁcant changes in TOC concentration and UV-254 occurred when ozone dose
increased ﬁ'om 0 to 1 mg/mg C. An increase in ozone dose to 2 mg/mg C did not result
in a signiﬁcant removal of either TOC concentration or UV-254.

It is worth to note that, an exponential relationship appears to better describe the
relationship between TOC and UV-254 (see Figure 5.6). Additional studies would be,
however, required to conﬁrm this. This was beyond the scope of this study, since it did
not appear that TOC-UV relationships could be used for the evaluation of the ozonation

and biodegradation processes.

91

 

 

  
 
  
 
  
 

 

 

 

74
y=13.398x+3.8229

6* R’=0.955
3 rawwater(Nov-July)
E 51
c‘
8
£4 0.5 mg OalmgC(Jan- Feb,June)
g 1mg Q,/mgC(Dec-Jan, Feb-Mar,July)
8 3‘ "T" 1
'- 2mg O3/mgC(Nov-Dec. Mar) ‘

21 y=20.065x+2.501

R2=0.9816
1.
0 . . .
0 0.05 0.1 0.15 02 025

UV-254,1/cm

l oozonation

I o 20 ha {lb n/_F BT51

 

 

Figure 5.5. Ozone-dose averaging linear correlation between TOC and UV-254

92

 

 

 

 

 

7 .
y=1.3707Ln(x) +8.5408
R2=0.992 _

6— /
_. rawwater(Dec-July)
Es. -
c’
.0
24+ 0.5 mg O3/mgC(Jan- Feb,June)
8
§ 1mg Q,/mgC(Dec-Jan. Feb-Mar,July)
8 31
'- 2 mg Oalmg C (Nov-Dec. Mar)

2 1 y=1.8958Ln(x)+9.2428

R2 =o.9917
11
o T T r r
0 0.05 0.1 0.15 02 0.25
UV-254.1/cm
f oozonation uozonation/FBT T

 

Figure 5.6. Exponential relationship between TOC and UV-254

93

5.3. HS AND UV-254

In order to establish the relationship between HS and UV-254 the data for raw
water, ozonated water, and water after ozonation and FBT, were plotted on the same
graph, shown in Figure 5.7. Visual observation of the plot suggests that the relationship
between HS and UV-254 appears to be the same for waters at different treatment stages.
A straight line ﬁtted well all HS and UV-254 data for raw and treated water.

The ozone-dose averaging technique was applied to HS and UV-254 data. Figure
5.8 shows excellent correlation between HS and UV-254 when the average data for each
ozone dose were plotted. The relationship was essentially the same as the one that was
established in Figure 5.7. The visual observation of the plot in Figure 5.8 shows that most
signiﬁcant changes of HS and UV-254 occurred when ozone dose increased from 0 to 1
mg/mg C. Further increase in ozone dose to 2 mg/mg C did not result in a signiﬁcant
removal of either HS concentration or UV-254.

The results of the study demonstrated that there exists a good UV-254 is a very
good linear correlation between HS and UV-254, which did not appear to vary for waters
at different treatment stages. The relationship did not appear to be dependent on ozone

doses either.

94

HS.mgCIL

 

 
  

3-51 y=13.145x+1.3065
R’=o.9586

2.5 1

1.5 1

0.5 1

 

0 r r 1 T 1 r 1 r T

 

 

0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18
U V- 2 5 4 , 1/c m

 

L erawwater Iozonation paozonation/FBT j

 

Figure 5.7. HS versus UV-254 (ozone dose — 0-2 mg/mg C, EBCT — 180 min)

95

0.2

4.50 1

4.“) 1

    
  
 

y= 14.442x+ 1.2041

3'50? R2=0.9961

raw water

3.00 ~

03 = 0.5 mglmgC

0;, =1 mglmg C

1.50 4 03=2mglmgC

1.“) -

0.50 1‘

 

0.00 . f w w T f r r —- —~w
O 0.02 0.04 0% 0.08 0.1 0.12 0.14 0.16 0.18 0.2

UV-254,1/cm

 

Figure 5.8. Correlation between average HS and UV-254 at each ozone dose

96

5.4. THMFP AND TOC

The relationship between THMFP and TOC was established using the data for
raw water, ozonated water, and water after ozonation and FBT, which were plotted on the
same graph. As can be seen in Figure 5.9, there exists an exponential relationship
between THMFP and TOC. The relationship appears to be the same for waters at
different treatment stages. The plot using average THMFP and UV-254 data categorized
by treatment and ozone doses (see Figure 5.10) also suggests that this regression line
could be applied for waters at different treatment stages. The averaging approach,
presented in Figure 5.10, could also be used to evaluate the effect of ozonation and FBT
on the removal of THMF P. For example, the removal of THMFP after ozonation with
doses of 1 and 2 mg/mg C was not statistically different from that obtained aﬁer
ozonation/FBT at an ozone dose of 0.5 mg/mg C. The removal of THMFP after
ozonation/FBT at an ozone dose of 1 mglmg C was greater than that obtained after
ozonation with a dose of 2 mglmg C (the difference was statistically signiﬁcant at 95%
conﬁdence interval). This suggests that the use of ozonation in combination with FBT for
the removal of THMFP was more efﬁcient than ozonation alone with respect to ozone

consumption.

97

 

450~ y=29.589e°-‘”“ .
R2=O.8543

1504

 

1001

 

o I f T T

 

 

—1

0 1 2 3 4 5 6
TOC, mg/L

 

 

 

L oRawwater '_-__Q£o_n_a~tio£m:f AOzonation/FBT J

 

 

Figure 5.9. THMFP versus TOC concentration

(ozone dose — 0-2 mg/mg C, EBCT = 180 min)

98

THMFP. ug/L

 

  
  
   

4501 y=29.17860.4476x
R280.9206
4004
raw water
3501

g
+1

1 ozone-0.5mgOB/rngc

 

 

 

 

 

250~
,1
2004 »-—< 4 ozone-1mg03lmgc
ozone-2mgO3/mgC l
150, ‘ ozone/FBT-O.5mgO$/mgc
ozone/FBT-1mgOS/mgc
1004
ozone/FBT-ngoalmgc
50.
o 7 r 1 . r
0 1 2 3 4 5 6 7
TOC. mg/L

Figure 5.10. Correlation between average THMFP and TOC concentrations

99

5.5. THMFP AND UV—254

The ozone-dose-averaging approach was used to establish correlations between
THMFP and UV-254. As can be seen in Figure 5.13, there exist excellent correlations
between THMFP and UV-254 for ozonated water and water after ozonated and FBT,
despite high standard deviations for THMFP. The relationships between THMFP and
UV-254 are, however, different for different treatment stages, and, therefore, their use

could be limited to a speciﬁc treatment train.

  
 

500]
450 y=135.926°““"
l R2 =o.9992
moi
l

03=0WOImQC

03‘2"“"90 oe=o.5mgmgc

R2=0.9901

 

 

f I T

0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
U V - 2 5 4 . 1 I c m

 

‘___ ° 02.018135 _ 4992931231 1

L ___J

 

Figure 5.11. Correlation between average THMFP and UV-254

(ozone dose — 0-2 mg/mg C, EBCT = 180 min)

100

5.6. SUMMARY

The study presented in this chapter conﬁrmed the hypothesis that there exists a
correlation between UV-254 and TOC and between UV-254 and THMFP. However,
these relationships were different for waters at different treatment stages. For this reason,
neither of these parameters could be substituted for the other one when evaluating the
ozonation and biodegradation processes.

This study showed that there exists a good linear correlation between HS and UV-
254 and between TOC and THMFP, which did not appear to be different for waters at
different treatment stages. When investigating ozonation and F BT processes, UV-254 and
TOC could be used as surrogates for humic substances and THMFP, respectively.

In addition, the averaging approach was proposed. According to this approach the
relationships between average data for each ozone dose were established. It was shown
that the relationships so established were useful for the evaluation of the effect of ozone

doses on the selected surrogate parameters.

101

6. OZONATION OF NOM

6.1 . INTRODUCTION

This chapter presents the results of the bench-scale study that investigated the

ozonation of NOM in Huron River water in more detail. The objectives of the study

presented in this chapter were:

To determine if the removal of organic carbon and UV-absorbing compounds during
ozonation is affected by gas ﬂow rate for the same ozone dose;

To assess the effect of hydraulic retention time on TOC and UV-254 removal;

To test the hypothesis that, when ozonation is conducted under conditions that do not
result in the oxidation of organic carbon, BDOC concentration increases with ozone
dose;

To determine if UV-254 absorbance can be used to measure the speciﬁc ozone
consumption rate;

To determine if the speciﬁc ozone consumption rate can be used for the evaluation of
the efficacy of the ozonation of NOM;

To determine if humic and nonhumic ﬁ'actions of NOM can be used to model the
kinetics of ozone reactions with NOM; and if so to develop a mathematical model
that describes the transformation of humic and nonhumic fractions during ozonation;
To test the hypothesis that the rate constant of the ozone reaction with humic

substances is a function of both temperature and the concentration of humic

102

substances and can be expressed in two independent terms: a temperature-dependent
term and a concentration-dependent term;

0 To test the hypothesis that the rate constant of the ozone reaction with nonhumic
substances is a function of temperature and the concentration of nonhumic substances
and can be expressed in two independent terms: a temperature-dependent term and a
concentration-dependent term;

0 To determine the kinetic parameters of the ozone reactions with humic and nonhumic
substances, including stoichiometric coefficients and activation energies;

0 To validate the developed model over a wide range of operational parameters,
including ozone dose, temperature, and hydraulic retention time.

It should be noted that since pH and alkalinity did not change signiﬁcantly during
ozonation of Huron River water, the effect of these parameters was not investigated in
this study. The study was conducted using the bench-scale ozonation system described in
section 3.2.1. The operational parameters that were investigated included gas ﬂow rate,
ozone dose, hydraulic retention time, and temperature. The major parameters that were
monitored in this study included TOC, UV-254, humic and nonhumic fractions, and
BDOC. Biotreated water was obtained from the bioﬁltration system that was described in
Section 3.2.5.

Section 6.2 investigates the effect of operational parameters, including ozone
dose, gas ﬂow rate, hydraulic retention and temperature, on the transformation of NOM
during ozonation. The kinetic study of the ozone reactions with NOM is presented in

Section 6.3. The summary of results is given in Section 6.4.

103

6.2. EFFECT OF OPERATIONAL PARAMETERS

6.2.1. Eﬁect of Gas Flow Rate on the Oxidation of NOM by Ozone

The goal of these experiments was to determine the appropriate gas ﬂow rate for
subsequent ozonation experiments. The experiments were conducted with gas ﬂow rates
of 15, 25, and 36 mL/min. The liquid ﬂow rate was 18 mL/min, which corresponded to a
hydraulic retention time of 20 min. The ozone dose was varied from O to 5 mg/mg C. All
experiments were conducted at room temperature.

Figure 6.1 shows the influent gas ozone concentrations that were required to
achieve a certain level of ozone dose for given gas ﬂow rates. As expected, the inﬂuent
gas ozone concentration requirements decreased with an increase in gas ﬂow rate. For
example, in order to achieve an ozone dose of 1 mg/mg C, the ozone concentration in the
inﬂuent gas was approximately 6 mg/L at a gas ﬂow rate of 36 mL/min. To achieve the
same ozone dose at a ﬂow rate of 15 mL/min, an inﬂuent gas ozone concentration of
approximately 14 mg/L was required. Similarly, an ozone dose of 2 mglmg C required
inﬂuent gas ozone concentrations of approximately 17 and 31 mg/L at gas ﬂow rates of
36 and 15 mL/min, respectively.

No dissolved ozone was detected at ozone doses less than 0.6 mglmg C (see
Figure 6.2). An increase in ozone dose (by increasing the concentration of ozone in the
inﬂuent gas) resulted in an increased dissolved ozone concentration. The gas ﬂow rate,
however, did not affect the dissolved ozone concentration.

Figure 6.3 shows that no TOC removal occurred at ozone doses less than 0.5

mglmg C. Further increase in ozone dose did result in partial TOC removal. It did not

104

appear, however, that the gas ﬂow rate affected TOC removal. This was expected
considering the fact that, as was shown earlier, the gas ﬂow rate did not affect dissolved
ozone concentration.

Unlike TOC, UV-absorbing compounds were effectively removed at low ozone
doses (less than 1 mglmg C). At ozone doses greater than 2 mglmg C, no signiﬁcant
removal of UV-254 was observed. The same effect was observed earlier in the pilot-scale
studies that were described in Section 4.3. The gas ﬂow rate did not effect the removal of
UV-254, because, as was discussed earlier, the gas ﬂow rate did not affect dissolved

ozone concentration (see Fig 6.2).

105

4.5

3.5

(a)

Ozonedose,mglmgc
.. z; N 3.’

.0
on

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

cg'". mg/L

 

[+15 mL/min +25 mL/min +36 mL/mirﬂ

 

Figure 6.1. cgin requirements for various gas ﬂow rates

(room temperature, RT = 20 minutes).

106

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6 .
l I l I l
I I l l
5 Mr , ‘~A4—~ : T '
l , I
l l
é“ L—P'T ”i ‘ I T
- l
. l ,
g, p d_CHLJ_wh zuwzhhmwm
l V. l ,
E I ’ ‘
g2 T} T '1. rﬁﬁ. _-
x ' I I
1 - off fmT' ”‘ r
: l l l
o 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Ozone dose,mg/mg C
L .15mL/min _ o25mL/_r_nin .36 mL/minaJ

 

 

Figure 6.2. Dissolved ozone concentrations for various gas ﬂow rates

(room temperature, RT = 20 minutes).

107

 

 

 

.4,__,.

I
o
o
.— é~————h

 

 

 

 

F" W.
1—4
h.
T

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

31mmi*—————rr——j :1” - 1
* I . ‘ l I

l l
. , 1 Si

N l l 1
1 a 1 +—¢l_if
. 1 - l ,1 1 i l 1
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5

Ozone dose, mglmg C

 

1015 mL/min A25 mL/min I36 mL/rniﬂn]

Figure 6.3. Effect of ozone dose on TOC concentration for various gas ﬂow rates

(room temperature, RT = 20 minutes).

108

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

4 “1 a H A 1
1 a h h l _ a
M _ _ I W _ .. _
1 1f1tn11~1 1e1n_,r 1+
. _ , _
h __ _ z N z
11 u1m1 F1 r1 4 1 +1 .\ _. -
_ _ a v. v A
ngm -+1+
z n _ . i if
._ . h g l __
. $1 1 $141» 41111141111“
U _ h . h +
Wh11444171;
“
. . _
A _ _
r 1 , IRT T1- 11 L11 _1
_ Y .
_ m . ..
1 11- 11 11L11 11.111111 11 1
_ h h 1 w
h h . lo
r 1 1111 11,11 j .1 #111 ﬂl 1. rial 1T1lw11 m
W .. J. _
. o m
1,1 1 1’rlfe11r1 .1
. J _
. ﬂ _
l 0 w t. _
m m m m m m m m m m.

ED: 6m”):

2.5 3.5 4.5

Ozone dose. mglmg C

1.5

0.5

in .25 mL/min .36 mggtﬂj

 

 

1.15 mL/m

Figure 6.4. Effect of ozone dose on UV-254 absorbance for various gas ﬂow rates

20 minutes).

(room temperature, RT

109

6.2.2. Effect of Hydraulic Retention Time

The goal of these experiments was to determine the effect of hydraulic retention
time (HRT) on TOC and UV-254 during ozonation. The study was conducted using raw
water. The HRT was varied from 10 to 60 min. The gas ﬂow rate was 36 mL/min. This
high ﬂow rate was selected in order to minimize the retention time of the effluent gas in
the reactor headspace. The ozone dose was varied from O to 6 mglmg C. All experiments
were conducted at room temperature.

Figure 6.5 shows the effect of HRT and ozone dose on the removal of TOC. At an
HRT of 10 minutes and ozone doses less than 0.5 mglmg C, no TOC removal was
observed. With an increase in ozone dose from 0.5 to 3 mglmg C, the removal of TOC
steadily increased. At HRTs of 20, 40 and 60 minutes, the TOC removal was observed at
at ozone doses less than 0.5 mglmg C and steadily increased with an increase in ozone
dose. As can be seen in Figure 6.5, the removal of TOC increased with HRT. For
example, at an ozone dose of l mglmg C and an HRT of 10 minutes, the removal of TOC
was not signiﬁcant at 95% conﬁdence interval. At the same ozone dose and HRTs of 40
and 60 minutes, TOC concentration decreased from 6.2 mg/L for raw water to
approximately 5.8 and 5.4 mg/L after ozonation, respectively.

Figure 6.6 once again showed (see also Figure 6.4) that UV-absorbing compounds
were effectively removed at low ozone doses (less than 1 mg/mg C). At ozone doses
greater than 2 mg/mg C, little removal of UV-254 was observed. The removal of UV-

absorbing compounds was not affected by HRT.

110

An attempt to explain the effect of HRT on the removal of TOC and UV-254 was

made. Figure 6.7 shows that the inﬂuent gas ozone concentration required to achieve a

given ozone dose decreased with an increase in HRT. As a result, the concentration of

dissolved ozone for a given ozone dose also decreased with an increase in HRT (see

Figure 6.8). The following hypothesis, therefore, could be formulated:

1) The removal of TOC decreases with a decrease in the concentration of dissolved
ozone; the HRT has, however, a greater effect on TOC removal than does
dissolved ozone concentration (as a result, for a given ozone dose, TOC removal
increases with an increase in HRT, even though the dissolved ozone concentration
decreases).

2) The removal of UV-254 is independent of HRT and dissolved ozone
concentration.

To verify these hypotheses, the efficacy of ozonation in the bench-scale system
was compared to that in the pilot-scale system. The operational conditions of the
experiments are shown in Table 6.1. In order to achieve similar HRT in both system, the

contact volume in the pilot-scale ozone contactor was reduced by half.

111

Table 6.1. Experimental conditions of bench- and pilot-scale ozonation systems

 

 

 

 

 

 

Parameter Bench-scale system Pilot-scale system
Water ﬂow rate, mL/min 18 20
Gas ﬂow rate, mL/min 15 15
Retention time, min 20 ca. 25
Ozone dose, mglmg C O - 2.7 O - 2.3
Temperature, 0C 20 20

 

 

 

 

 

The major difference in the performance of bench-scale and pilot-scale ozone
contactors (which actually made it possible to verify the hypothesis stated above) was in
the efficiency of ozone utilization, s, that was deﬁned as

in_ out

0
“lei-100%

In
cg

where cgin and cgout are the concentrations of ozone in inﬂuent and effluent gas, mg/L.

Figure 6.9 shows the ozone utilization efficiency in the bench-scale system. The
efﬁciency was approximately 60 percent at an ozone dose of 0.5 mg/mg C and decreased
with an increase in ozone dose. Only trace amounts of ozone were detected in the effluent
gas ﬁ'om the pilot-scale ozone contactor. So, it was assumed that the efficiency of ozone
utilization in the pilot-scale system was 100 percent.

Figure 6.10 shows the inﬂuent gas ozone concentrations in the bench- and pilot-
scale systems, which were required to achieve a given ozone dose. In order to achieve the

same ozone dose, the pilot system required lower ozone concentration in the effluent gas

112

than did the bench-scale system. As a result, under similar experimental conditions the
dissolved ozone concentration was lower in the pilot ozonation system than that in the
bench-scale ozonation system, as demonstrated in Figure 6.11.

As can be seen in Figure 6.12, the removal of TOC during ozonation was
essentially the same in both systems at ozone doses less than 0.6 mglmg C. The TOC
concentration in the pilot system appears to level off at an ozone dose of l mg/mg C,
whereas the TOC concentration in the bench system continued to decrease with an
increase in ozone dose (the difference in TOC concentrations for raw water and after
ozonation at a dose of 1 mglmg C was statistically signiﬁcant at 5 percent level). Since
the dissolved ozone concentration in the bench-scale system was several times greater
than that in the pilot-scale system and all other conditions were essentially the same for
both systems, it appears that the removal of TOC increases with dissolved ozone
concentration. This also suggests that HRT has greater effect on the removal of TOC than
does the concentration of dissolved ozone. This is because the removal of TOC increased
with HRT even though dissolved ozone concentration decreased.

Figure 6.13 shows that, for a given ozone dose, the removal of UV254-absorbing
substances was essentially the same in the bench- and pilot-scale systems. Since the
dissolved ozone concentration in the bench system was several times greater than that in
the pilot-scale system, it appears that the removal of UV254-absorbing compounds is
independent of ozone concentration. This also conﬁrmed the hypothesis that the removal

of UV-254 is independent of HRT for the range of HRTs studied.

113

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 1
6! kN Ar— _ #MJr _AJ,_#!- 12... -. -_.
\~%\N\J\l 1
5 \‘§:;\:T\\+=
\
a 1\\. \.
- 1
§. .1zw_z1-
2 _.H__ z
1
1 ,,_~__.___.__ *Hvi M _ ﬂ 1__ 1.__ H A_,
1 1
0 . 1 1 1
0 1 2 3 4 5 6 7

Ozone dose. mglmg C

 

[+RT = 10 min +RT = 20 min+RT =40 min+RT = 60 min]

 

Figure 6.5. Effect of ozone dose on TOC concentration for various HRTs.

114

W-254, 1lom

0.2

0.18

0.16

0.14

.o
a
N

.O
.s

,0
8

0.04

0.02

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 1 1
1 1. z, I; _ __ z
T 1 .' 1
1
#.m__ ﬂ-14A ;_ _-_ _- ____z -_
1 ___,_1_.~__ .____
1 ‘ L 1
1 ‘ 1 '
_i_1_ L + 1 l #2
‘ 1 1 I 1 1 1
_ - 1 7_w ‘wmﬁhw,__-___v,
1‘ 7L ‘ 1 1
l' - 1 _
r I‘ ‘
I 1
0' x .
O
1 1 1
1 1 1 l . 1
0 1 2 3 4 5 6

Ozone dose. mglmg C

 

ﬁRT=10min IRT=20min eRT=40min WxRT=60mirﬂ

 

Figure 6.6. Effect of ozone dose on UV-254 absorbance for various HRTs.

115

 

 

 

 

 

 

69‘", mg/L

 

 

 

lllL

 

 

 

    

 

 

 

 

"
h
7 6 5 4 3
o 9&9: .03.... 0:30

 

 

 

 

1.W

m

0

4

=

T. s
.H H
mm. H
w m
:nw. .m
P: V
.TH r.
E .m
.m e 6
m m 11..
o .u
2 u
= q
‘T m
R .m

+ cg
n 7.
.m 6.
o m.
= .1.
T F
R

1+1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

s m 3 2
.38 6882850 9.03 3293.0

Ozone dose,mg/mg C

40 min}

___. _______ ,,!___. _‘

 

 

git—RT: 10 min +RT=20 min+RT=30 min +RT

,-— ___

Figure 6.8. Dissolved ozone concentration for various HRTs

117

WI

 

60+

E

OumemMmmmﬁ%
8

10~

 

 

 

05 1 15 2 25
Ozonedose,mg/mgC

Figure 6.9. Efﬁciency of ozone utilization in the bench-scale ozone contactor

(HRT — 20 minutes)

118

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

c,‘".mglL

BBB—REF :p'ﬁ']

 

Figure 6.10. Inﬂuent gas ozone concentration requirements for bench- and pilot-scale

systems (see Table 6.1 for experimental conditions)

119

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Ozone dose, mglmgC

l
l

 

bench+pilot

 

i +

l

Figure 6.11. Dissolved ozone concentration for bench— and pilot-scale systems

120

 

 

0 0.5 1 1.5 2 2.5 3
Ozone dose, mglmgC

 

l-o—bench +pilolj

 

Figure 6.12. Removal of TOC during ozonation in the bench- and pilot-scale

systems (experimental conditions are given in Table 6.1)

121

02

' l
i 1
0.18 ﬂ,“ +~ #~ * I~——

 

 

 

 

 

 

lﬂzm#z_w

, __ __. _

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

~“~a
002+— L~‘_ﬁ..z.____4ﬂ__ T A
o l I I l 4' l
0 05 1 15 2 25 3 35 4

Ozone dose, mglmgc

 

t4—bench—tjﬂigﬂ

Figure 6.13. Removal of UV-254 during ozonation in the bench— and pilot-scale

systems (experimental conditions are given in Table 6.1)

122

6.2.3. Effect of Ozone Dose on BDOC

The results of the pilot-scale study described in Chapter 4 showed that an increase
in ozone dose from 0 to 1 mglmg C resulted in an increased removal of TOC. The
concentration of biodegradable organic matter appeared to decrease at the same time (see
Figures 4.13). The removal of TOC leveled off with further increase in ozone dose. This
was accompanied by an increase in the concentration of biodegradable organic matter.
The hypothesis formulated based on this ﬁnding was that when ozonation is conducted
under conditions that do not result in the oxidation of organic carbon, BDOC
concentration increases with an increase in ozone dose. To verify this hypothesis bench-
scale ozonation experiments were conducted. The ozonation efﬂuents were analyzed for
UV-254, TOC and BDOC. The BDOC was measured using the recirculating bioﬁltration
system that was described in Section 3.2.5. Ozone dose was varied from O to 1 mglmg C.
The results of the experiments described in the previous section showed that for this
range of ozone doses and a HRT of 10 minutes, the removal of TOC was not signiﬁcant
at 95% conﬁdence interval. However, to ensure that no TOC was removed during
ozonation, the experiments described in this section were conducted at a lower HRT of 7
minutes. As can be seen in Figure 6.14, the removal of UV-254 was the same at HRTs of
7 and 10 minutes.

Figure 6.15 shows the effect of ozone dose on BDOC concentration. As expected,
the removal of TOC aﬁer ozonation was not signiﬁcant at 95% conﬁdence interval.

BDOC concentration steadily increased with ozone dose and reached a concentration of

123

2.8 mg/L at an ozone dose of 1 mglmg C compared to a BDOC concentration of 1.1

mg/L in raw water.

 

 

(118 lL—M*——

 

 

016

 

c114 —

 

n12 ”"1,Jﬂ,_Az_ V~ ____

 

UV-254. 1/em
‘3
l
l
I
l
1

 

cos ———»—~ ﬂ~— -#—*t~‘r

 

 

 

nos

 

(104

_ _ , _._._.___- m _

 

om

 

 

 

0 (12 (14 Q6 Q8
Ozone dose, mglmg C

 

_ . _ {—1
LoHRT-len ﬂRT-g—mﬂ

 

1.2 1.4

Figure 6.14. Removal of UV-254 during ozonation at low HRTs

124

35

N
01

Organic carbon, mg/L
in N

05

 

 

 

 

 

 

 

 

 

 

 

 

 

__A
V
T’
T T 1 1

 

l

62 64 66 68 1 1.2
Ozone dose. mglmg C

 

r '—_“.

+TOC removal +BDOC

L__z

 

Figure 6.15. Effect of ozone dose on BDOC concentration

(HRT = 7 minutes)

125

6.3. OZONATION KINETICS

6.3.1. Introduction

This section describes the study of the kinetics of ozone reactions with NOM. The
kinetics of the ozonation reactions in natural waters can be described in terms of two key
process parameters: (1) speciﬁc ozone consumption rate of water, as proposed by
Staehelin and Hoigné (1985), and (2) apparent oxidation rate constant of water, proposed
by Yurteri and Gurol (1988). Although the Staehelin and Hoigné and Yurteri and Gurol
approaches are similar in nature, there is a difference in terms of the interpretation of
kinetic data. In natural waters, the speciﬁc rate of ozone consumption can be described as
an overall expression, which is ﬁrst-order with respect to ozone concentration (Staehelin

and Hoigné, 1985):

C110 3 ]
dt

 

= ‘k03 [03] (6-1)

where [03] is the concentration of dissolved ozone and ko3 is the speciﬁc ozone
consumption rate.
According to Yurteri and Gurol (1988), the speciﬁc rate of ozone consumption

can be expressed as

k03 = ZkiMi (6.2)
l

126

where Mi is the concentration of substrate i in water and k, is the apparent reaction rate
constant for Mi.
The apparent reaction rate constant, k,-, consists of two terms, which reﬂect direct

and radical reactions of ozone with organic matter, as follows (Yurteri and Gurol, 1988):

ki = (kdi +qui)Mi (63)

where kdi and k, are the rate constants for direct and radical reactions, respectively, and
M is the factor, which accounts for initiation and scavenging of radical reactions and can

be expressed as (Yurteri and Gurol, 1988):

. _ 2k1i[OH-]+ k21Mi
l stﬂsj]
J

 

(6.4)

where k“ and k2i are the rate constants of the initiation step in the decomposition of ozone
by OH' and the substrate [M], respectively; and Si and ks, are the concentration and rate
constants of scavengers, respectively.

Thus, it is mechanistically possible to obtain explicit relationships for the
reactions of ozone with NOM. These relationships, however, have limited practical use
because it is virtually impossible to account for all substrates and scavengers that
participate in the reactions with ozone. A simpler approach must, therefore, be identiﬁed,
which describes the reaction of ozone with NOM.

In this study, the kinetics of ozone reactions was expressed in terms of bulk

characterization of NOM, as follows:

127

d S
§=-th031[81 (6.5)

where, k1 is the total apparent rate constant of substrate [S]; and [S] is the bulk
characteristics of NOM.

It should be noted that the total apparent rate constant, k1 , represents both direct
and radical reactions and is dependent on the aqueous matrix, including such factors as
alkalinity. It is important to realize that k7 is the function of the concentration of the
solute [S], which is evident from the previous discussion, and, therefore, changes as the
ozone reaction proceeds.

The use of bulk characterization of NOM to describe ozone reactions in water is
not new (Bablon et al., 19913). The major challenge, however, was to select appropriate
bulk parameters, which would adequately describe the transformation of NOM in a
particular treatment train, e. g. ozonation and biotreatment.

The use of speciﬁc rate of ozone consumption (see Equation 6.1) is an example of
the bulk characterization approach. This parameter describes the extent of ozone
reactions in water and may be useful in evaluating the effect of operational parameters on
the ozonation of NOM.

This study also proposed the use of bulk characterization in terms of two
parameters, HS and nonHS. These parameters were selected for the following reasons:

0 HS and nonHS are representative of all NOM in water;

0 HS and nonHS are separate and distinct fractions of NOM;

128

HS and nonHS are categories of NOM, which are distinguished by its
biodegradability: HS are essentially nonbiodegradable, whereas nonHS is a
potentially biodegradable portion of NOM; and, ﬁnally,

HS and nonHS can be quantiﬁed.

The particular objectives of this study were:

To test the hypothesis that the speciﬁc rate of ozone consumption can be used to
the evaluate the efficacy of the ozonation of NOM;

To demonstrate the use of HS and nonHS to describe the kinetics of ozone
reactions in water;

To test the hypothesis that the ozonation of NOM follows the following pathways:
1. Ozone converts HS into nonHS

2. Oxidation of organic carbon occurs through the reaction of ozone with nonHS
3. Both reactions may proceed simultaneously.

To verify the ﬁrst order kinetics with respect to HS and nonHS;

To determine the stoichiometric coefﬁcients for the reactions of ozone with HS
and nonHS;

To determine the reaction rate coefﬁcients and activation energies for the reaction
of ozone with HS and nonHS; and

To develop and validate a kinetic model that describes the transformation of

NOM during ozonation.

129

6.3.2. Model Development

Two models describing the ozone reactions in water were developed. The models
were based on the bulk characterization of NOM expressed either in terms of the speciﬁc

rate of ozone consumption (Model I) or in terms of HS and nonHS (Model 11).

6.3.2.1. Model Based on the Speciﬁc Ozone Consumption Rate

According to the theory of absorption with reaction (Ansehni et al. 1985;
Dankwerts, 1970) and assuming that bench-scale ozone contactor could be considered as
a completely mixed reactor for both liquid and gas phase and that the resistance to
transport of ozone from gas phase to the interface was insigniﬁcant (Astarita, 1967), the
mass balance for ozone could be described by the following system of ordinary

differential equations:

ch

_dt_ : name‘s” —cL)V 4.03ch —QcL (6.6)
dc“) - ~
H.—.: mg»«tn-mmcr-cnv m

where V is the volume of the reactor, mL; Q and G are the liquid and gas ﬂow rates,
respectively, mL/min; CI, is the concentration of ozone in water, mg/L; CE0 is the

concentration of ozone in the inﬂuent gas, mg/L; cg“) is the interfacial concentration of

130

ozone in the gas phase, mg/L; ha is the mass transfer coefﬁcient, Marin; at is the ozone-
water equilibritnn constant; Hg is the gas hold-up.
For low ozone concentration, we could assume the validity of Henry’s law, which

could be written as (Roth and Sullivan, 1981):

Y = X (6.8)

where Y is the ozone mole fraction in gas phase; X' is ozone mole fraction in equilibrium
with the bulk mole fraction in the gas phase; H is the Henry’s law constant, atm/mole
fration: P is the gas pressure, atm.

Since the ozone concentration in water is low, X. could be expressed as (Sheffer

and Esterson, 1982):

t

6 CL MHZO _ 18

X = _ c‘ =3.75.10-7c‘ (6.9)
103pH20 MO 48-106 L L

 

3

where cL' is the ozone concentration in the equilibrium with the bulk concentration in gas
phase, mg/L; Mmo and M03 are the molecular weight of water and ozone, respectively,
mg/L; pmo is the density of water, g/L.

Similarly,

c M
3 02: 32 cg=4.66.10‘4cg (6.10)

—103p02 MO3 48143-103

 

 

Y

131

where c8 is the ozone concentration in gas phase, mg/L; M02 is the molecular weight of
oxygen, respectively, mg/L; p02 is the density of water, g/L.

Assuming that gas pressure in the reactor is 1 atm, combining the Equations (6.8)-

(6.10) gave

c1=2§cg (6.11)

which gave the relationship between a and H for ozone-water systems:

a=——— (6.12)

The Henry’s law constant was found to be satisfactorily modeled as a function of

pH and temperature over the range of conditions used in water treatment, as follows

(Roth and Sullivan, 1981):

2428
H = 3.84-107[OH']°'°35e T (6.13)

where, [OH'] is the concentration of hydroxide ions, g-mol/L; T is temperature, K.

132

6.3.2.2. Model Based on HS and nonHS.

By taking into account the hypotheses regarding the transformation of HS and
noan during ozonation and adding all the previous assumptions, the oxidation of NOM

by ozone could be described by the following system of ordinary differential equations:

dC i

V Ti" _ (kLa(ac(g) - cL ) — VHskHsCHsCL - VnonHSknonHSCnonHScL )V — QCL (6-14)
dens o

V—dt_=Q(cHS —CHS)_kHScHScLV (6-15)
dcnonHS 0

V T = Q(cnonHS — cnonHS) — knonHScnonHSch 1' kHSCHScLV (6-16)

dc“) . .
H8 .5:— = G(c(g°) — cg")— kLa(ac(g‘) — cL)V (6.17)

where kns and knonHS are the reaction rate coefﬁcients for HS and nonHS, respectively,
L/(mg min); VHS and VnonHS are stoichiometric coefﬁcients for HS and nonHS,

respectively; cHs" and cman° are the initial concentrations of HS and nonHS,

respectively, mg/L.

6.3.3. Rate of Ozone Consumption

Two methods for measuring the speciﬁc ozone consumption rate were

investigated in this study: (1) the method, based on measuring UV-254 reduction in

133

samples after ozonation, and (2) the method based on the model described by Equations

(6.6)-(6.7). The goal of the experiments presented in this section were:

0 To test the hypothesis that UV-254 absorbance can be used to measure the
concentration of dissolved ozone and the speciﬁc ozone consumption rate;

0 To determine if the model described by Equations (6.6)—(6.7) can be used to
determine the speciﬁc ozone consumption rate and to compare this method proposed
with the spectroscopic method;

0 To verify the ﬁrst-order kinetics for ozone consumption; and

0 To demonstrate the use of the speciﬁc rate of ozone consumption for the evaluation of

the ozonation of NOM.

6.3.3.1. Spectroscopic Method

The major difﬁculty in using UV-254 absorbance to measure the rate of ozone
depletion is that both ozone and NOM absorb UV light at a wavelength of 254 nm.
Preliminary experiments were conducted to determine the effect of residual ozone on the
UV-254 absorbance of NOM aﬁer ozonation. This was done by comparing (1) the UV-
254 of the ozone contactor efﬂuent after allowing the residual ozone to react with NOM
with (2) the UV-254 of the efﬂuent from the quenching tank (see Fig. 3.1) in which the
reaction was quenched by purging all residual ozone with helium. The quenching
procedure was described in Section 3.2.1. In addition, the samples were analyzed for
TOC in order to determine if the removal of organic carbon occurred through the

oxidation by ozone or through volatilization. The results are shown in Table 6.2.

134

Table 6.2 Effect of quenching on UV-254 and TOC after ozonation

 

 

 

 

 

 

 

 

 

 

Ozone dose, UV-254 absorbance, l/cm TOC, mg/L
mglmg C w/ quenching w/o quenching w/ quenching w/o quenching
0.8 0.102 0.101 6.4 6.37
1.5 0.068 0.068 5.91 5.87
3.4 0.051 0.050 5.53 5.58

 

 

As can be seen in Table 6.2, the residual ozone did not appear to affect either UV-
254 or TOC in ozonated samples. This suggested that the removal of TOC was due to the
oxidation reactions rather than volatilization. This also suggested that the base line
correction could be used to separate the UV absorbance of dissolved ozone from the UV
absorbance of NOM. To prove this hypothesis, the values of ozone concentration
measured using UV-254, as described below, were compared with those measured by the
Indigo method.

The UV-254 measurements of ozonated samples were done by diverting the
effluent to a head-space free 1-cm capped UV cell and measuring the reduction in UV-
254 over time. The measurements were usually started 30 seconds after sample collection
and, depending on the rate of UV-254 reduction, were taken every 15 to 60 seconds
afterwards. A typical kinetic curve for UV-254 in ozonated samples is shown in Figure

6.16.

135

The UV-254 decreased with time and reached a plateau aﬁer several minutes. The
UV-254 value at which the curve leveled off corresponded to the UV-254 absorbance of
ozone-free sample. The curve, shown in Figure 6.17, was obtained by substracting the
UV-254 of ozone-free water from the actual values of UV-254. This curve described the
kinetics of ozone depletion expressed in terms of UV-254. The tail of the curve was cut
off as the accuracy of measurements decreased when the values approached zero. Figure

6.17 shows that the data were ﬁtted very well by the exponential line:

UV254 = UV2540e""UV‘ (6.18)

where UV2540 is the concentration of the dissolved ozone in the ozone contactor
expressed in terms of UV absorbance, cm'l; kuv is the speciﬁc rate of UV254 depletion.
The dissolved ozone concentration in the ozone contactor was calculated using

the following equation:

h403-1000
CL = TW2540 (6.19)

where 0;, is the concentration of dissolved ozone in the ozone contactor, mg/L; 8 is the
extinction coefficient (a = 3000 L mole'l at 254 nm) (Masten, 1991); b is the cell width (b
= 1 cm).

The ozone concentrations that were determined using the spectroscopic method
were compared with those measured using the Indigo method over a wide range of

operational conditions, including:

136

- ozone dose - 0 to 5 mglmg C

0 gas ﬂow rate - 15 to 36 mL/min
o retention time - 7 to 40 minutes
0 temperature — 10 to 30°C.

For each ozonated sample, a kinetic curve of UV-254 depletion, similar to that
shown in Figure 6.17, was obtained and the ozone concentration was determined using
Equations (6.18) and (6.19). The-correlation coefﬁcient for the exponential regression
lines so obtained ranged from 0.992 to 1, with an average of 0.997. Figure 6.18 shows
that there exists a good agreement between the spectroscopic and Indigo methods (the
slope of the regression line was 0.9925). Thus, the study showed that the rate of ozone
consumption followed the ﬁrst order kinetics and the speciﬁc ozone consmnption rate

could be determined from the UV depletion curve as

1(03 = k UV (6.20)

The spectroscopic method for measuring the speciﬁc rate of ozone consumption is
simple and requires only the measurements of UV254 absorbance. It also allows for the
measurements of the dissolved ozone concentration. The limitations of the method are the
following:

o the method cannot be used for high ozone consumption rate when the residual
ozone is depleted quickly (for less than a minute); and

o the method should be validated for each water source under investigation.

137

UV-254. 11cm

0.18 ~

0.16 ~

0.14 1

0.12 <

p
.s

 

p
‘3

om. UV absorbance of ozone-free sample

0.04 4

0.02 -

 

 

 

 

Time, min

Figure 6.16. Kinetic curve of ozone depletion in water based on UV-254

138

0.12 .

 

 

 

0.1 1
y = 0.1094e'°“°7”‘
R2 = 0.9993

0.081
30.031
°.'

0.04 «

0.024

o I I ﬂ I T I ﬂ W l
0 0.5 1 1.5 2 25 3 3.5 4 4.5

Time. min

Figure 6.17. Kinetic curve of ozone depletion based on UV-254 using baseline correction

139

«b

q (UV254 method), mg/L
0

21

 

 

 

0 1 2 3 4 5 6
q(l ndigo method), mg/L

Figure 6.18. Correlation between dissolved ozone concentrations measured

spectrophotometrically and by Indigo method

140

6.3.3.2. Method Based on Ozone Consumption Model

The second method for measuring the speciﬁc ozone consumption rate pr0posed
in this study was based on the application of the ozone consumption model described by

Equations (6.6) and (6.7). For steady state, Equations (6.6) and (6.7) became:

0=kLa(acg> -cL)V—k03cLV—QcL (6.21)

0 = G(c‘g°’ —cg>) —- kLa(acg) -cL)V (6.22)

Because the resistance to transport of ozone from gas phase to the interface is
insigniﬁcant (Astarita, 1967), it was assumed that the interfacial concentration of ozone
in gas phase was equal to the bulk gas phase ozone concentration. Then, the mass transfer

coefﬁcient, kLa, could be calculated using Equation (6.22):

k a = G(cg —cg‘")
L V(0Lc‘g’In -cL)

 

(6.23)

and the speciﬁc rate of ozone consumption, ko3, could be calculated using Equation

(6.21):

141

_ kLa(0Lc;“t — cL )V — QcL

0 _
3 cLV

 

(6.24)

where cgout is the concentration of ozone in the efﬂuent gas, mg/L.

In order to calculate the speciﬁc rate of ozone consumption, km, the dissolved
ozone concentration and the concentration of ozone in the inﬂuent and efﬂuent gas were
measured. The ozone-water equilibrium coefﬁcient, a, was determined using Equations
(6.12) and (6.13). Since the ozone-water equilibrium coefficient is a function of pH and
temperature, these parameters were measured in each experiment. The calculations were
conducted for the same range of operational parameters that were described in the
previous section. The values of the speciﬁc rate of ozone consumption so obtained were
compared with those determined using the spectroscopic method. The results are shown
in Figure 6.19. There exists a good agreement between the values obtained by two
methods. The dispersion of the data was, however, greater than those for the dissolved
ozone concentration (see Fig. 6.18). This could be attributed to the fact that the speciﬁc
ozone consumption rate that was determined using the ozone consumption model (see
Equation 6.6) involved the measurements of several parameters (dissolved ozone
concentration, the concentration of ozone in the inﬂuent and efﬂuent gas, pH, and
temperature), each of which had an experimental error and could contribute to the total
error (see Equation 6.23).

The method based on the ozone consumption model allows for the determination
of both high and low speciﬁc rates of ozone consumption. The method, however, appears

to be less accurate than the spectroscopic method.

142

 

2.5

 

 

 

 

y = 0.9299x
R2: 0.8963
2.
1.5«
3n
.9
1_
0.5«
o I T 1 f
0 0.5 1 1.5 2 2.5

modd
1‘03

Fig 6.19. Correlation between the speciﬁc rates of ozone consumption determined by two

methods

143

6.3.3.3. The Use of Ozone Consumption Rate to Evaluate the Efﬁcacy of Ozonation

Hoigné (1994) identiﬁed the rate of ozone consumption as one of the parameters
necessary for the quantiﬁcation of physical and chemical properties of water with respect
to ozone reactions. The goal of the study presented in this section was to determine if the
speciﬁc ozone consumption rate could be used to evaluate the efﬁcacy of ozone reactions
in water. Particular objectives were to test the hypotheses that:

o the speciﬁc ozone consumption rate would decrease with the removal of NOM
during ozonation;

0 gas ﬂow rate would not affect the speciﬁc ozone consumption rate;

0 the speciﬁc ozone consumption rate would increase with retention time.

The experimental design was the same as described in Chapter 6.2. The speciﬁc
rate of ozone consumption was determined using the spectroscopic method (see Section
6.3.3).

Figures 6.20-6.22 show results of the experiments conducted at room temperature
and a retention time of 20 minutes. Ozone dose was varied from 1 to 5 mg/mg C. The gas
ﬂow rate was 15, 25, and 36 mL/min.

The ko3 decreased with an increase in ozone dose (see Figure 6.20). Two regions
could be identiﬁed. The ﬁrst region, which corresponded to ozone doses of
approximately 1.5 mg/mg C and less, was characterized by a sharp decrease in km when
ozone dose increased. In the second region, which corresponded to ozone doses of
approximately 1.5 mglmg C and greater, a slow decrease in km with an increase in ozone

dose’was observed. This correlated well with UV-254 data that exhibited a similar

144

behavior during ozonation (see Figure 6.4). The ﬂow rate did not appear to affect k03.
This is thought to be because neither UV-254 nor TOC removals were affected by gas
ﬂow rate (see Section 6.2).

The ko3 decreased exponentially with a decrease in the concentration of UV-
absorbing compounds during ozonation (see Fig. 6.21). The ko3 dropped ﬁ'om 1.6 to 0.4
min'1 while TOC concentration did not change signiﬁcantly (see Fig. 6.22). This decrease
could be attributed to the removal of UV254-absorbing compounds. Further ozonation
resulted in a decrease in TOC concentration from 5.8 to 5 mg/L, which was accompanied
by a decrease in km from 0.4 to 0.1 1/min. It was difﬁcult, however, to determine if this
decrease was due to the removal of TOC 0r UV-absorbing compounds or both. It did not
appear that gas ﬂow rate affected the correlations between TOC and ko3 or UV-254 and
k03-

Figures 6.23-6.25 show the plots of km versus ozone dose, UV-254 and TOC for
various retention times. The gas ﬂow rate was 36 mL/min in all experiments. Figure 6.23
shows that k03 decreased with an increase in ozone dose. It appears that the values of kg,
were higher for higher retention times. Similar phenomenon was observed for the
relationships between TOC and km or UV-254 and k0; (see Figures 6.24 and 6.25,
respectively). This affect could not be attributed to the removal UV-254 absorbing
compounds, since this removal during ozonation was not affected by retention time (see
Section 6.2). The extent of the reaction of ozone with TOC increased with retention time
(see Section 6.2), which could possibly be a contributing factor to higher ko3 values at

higher retention times.

145

The ﬁndings presented in this section suggest that the use kg; for the evaluation of
ozonation may be limited. This is because, without having additional information, it
would be difﬁcult to attribute the ko3 changes to either the reaction of ozone with UV-

254-absorbing compounds or TOC or both.

 

I: \1
1.2 ‘\\

 

 

 

 

;
1 _
l

 

 

___.z. .

 

-§

 

 

 

 

km 11mm
1/6

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0 2
\‘RKH
0 I.
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Ozone dose, mglmg C

 

[+15 mL/min +25 mL/min+36 mL/minj

 

Figure 6.20. Effect of ozone dose on 1(03 for various gas ﬂow rates

(HRT = 20 minutes)

146

hmpﬂmm

1B

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 i I, l
1 . l l A
16 _f_z_wr_uz_zm_nf_.-u__+___.-__r__h___
1
14 ' J = o 126e71.875x f ‘—
R21=0.9949 / 1?"!!03/"0‘3
1.2 J - '
I ‘f /
1
1+ 1 AJFTM—ﬂ 41-4-1 1 e/ ——
L ‘ J j 1 1.5m903ltrngC
0.8 -~—w~» 1‘ _ﬂz_- 9W —1’—w4~~~—— -———---—-—+ ————e
1 1 v . /
°‘6 1 l [I l /
1.9mgtbjlmgC
.-_n.l. , .
1 i a ‘ 214mngde
0.2 + ‘1 i , » ‘m, —
o . 4.4mgq/mgk’ﬁ5mgoumgc l !
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08

UV-254,1/cm

 

- 3 6 m LTrHiﬂn—Il

V_J

I .15 mL/min .25 mL/min

 

Figure 6.21. Correlation between UV-254 and km for various gas ﬂow rates

(HRT = 20 minutes)

147

koa. 1/min

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I 1 ‘
1 1 .
an?” ’ __-
1.2 9.6%, TE _-_1 z, |.____.5| ______ -
1 1-6 41__6
.. 1 1 HM“
, w
.

 

 

 

 

 

 

 

 

 

I
2 3 4 5 6 7
TOC.mg/L

 

*1

Lo15mL/min .25mL/min .36mL/minJ‘

 

Figure 6.22 Correlation between TOC and km for various gas ﬂow rates

(HRT = 20 minutes)

148

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

h z W
. _ I .
1111411 115111
2 1111_1111
A n ,9
1 19111111
.1111 119,111 1% 11111111-
111%.: m .
. 5111 .
1111,; “11-1, . 1-
. ._ . e .
1 11 11 4 11ﬁ1111
.1; -.
1%
z _ .

55: .85.

25 3.5 4.5
Ozone dose. mglmgC

1.5

0.5

30min_1

eRT

=20 min

IRT

10min

CRT

 

 

1
L

Figure 6.23 Effect of ozone dose on km for various retention times

149

kos. 1lmin

 

 

61.40

R2; .9990

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

UV-254,1/cm

 

1oRT = 10 min IRT= 20 min oRT =_'3_0 mirﬂ

 

Figure 6.24 Relationship between UV-254 and 1(03 for various retention times

150

O 0.01 0.02 0.03 0.04 0.05 015 0.07 0.08

 

0.09

 

__..____ _ __.. .- ..

 

 

 

 

 

 

 

2.5

TOC. mg/L

1.

oRT = 30 min

 

 

=20min

IRT

0RT= 10min

 

1

Figure 6.25. Relationship between TOC and km for various retention times

151

6.3.4. Reaction of Ozone with Humic Substances

This section presents the kinetic study of the ozone reaction with HS. The
objectives of the study were:

0 to test the hypothesis that, at an HRT of less than 10 minutes and ozone dose of 1
mglmg C and less, ozone is directed solely for the oxidation of humic substances,
which are converted into nonhumic materials;

0 to determine the stoichiometric coefﬁcient for the reaction of ozone with humic
substances;

0 to test the hypothesis that the rate constant of the ozone reaction with humic
substances is a function of both temperature and the concentration of hmnic
substances and can be expressed in two independent terms: a temperature-dependent
term and a concentration-dependent term;

0 to determine the kinetic parameters of the ozone reaction with humic substances.

6.3.4.1. Stoichiometric Coefficient

The results of the study presented in Section 6.2 showed that the removal of TOC
was not signiﬁcant at 95% conﬁdence interval at an HRT of 10 minutes and ozone doses
of 1 mglmg C or less. It was also shown that the removal of UV-254-absorbing
compounds increased with ozone dose and was independent of HRT (see Section 6.2)

Taking into consideration that there exists a good correlation between HS and UV-254

152

(see Section 5.3), it was expected that at an HRT of less than 10 minutes and ozone doses
of 1 mglmg C or less, ozone would oxidize only humic substances, which would result in
an increase in the concentration of nonhumic materials. Nonhumic substances were not
expected to be oxidized under such conditions.

Figure 6.26 shows the experimental results for ozonation conducted at room
temperature and a retention time of 7 minutes. The ozone dose was varied ﬁom 0 to 1
mg/mgC. It should be noted that on the plot in Figure 6.26, ozone dose was expressed in
mg/L, instead of mg/mg C. This was done in order to be able to calculate the
stoichiometric coefﬁcient. The TOC removal was essentially at 95 percent conﬁdence

interval. The removal of humic substances followed the linear equation:

HS = —0.193 1 - [ozone dose] + 3.4167 (6.25)

with R2 = 0.9947.

An increase in the concentration of nonhumic substances followed the linear

equation:

nonHS = 0.1925 - [ozone dose] + 2.395 (6.26)

with R2 = 0.8991.

As can be seen in Figure 6.26 and from Equations (6.25) and (6.26), ozone
oxidized humic substances, which were converted into nonhumic materials. The absolute
values of slopes of lines described by Eqs. (6.25) and (6.26) indicates that no oxidation of

nonhumic materials occurred.

153

The stoichiometric coefﬁcient for the ozone reaction with humic substances was

calculated from Equation (9.25) and was equal 5.17 mg O3/mg HS.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

7
6 : - _ ______ __ C ;
_. 5 —~ —— ____ ————
g y=0.1925x+2 395
:- R’=0.8991
4 _ fit _—__z ﬂ! _ ﬂ
.‘3
g | /’
:3 ‘\.k ‘ ___-— ,__.H
0 /\v\.
0
C
o l
02 m
y=-O.1931x+3.4167
R2=0.9947
1 4._.
o T 7 r T T T
0 1 2 3 4 5 6 7

Ozone dose. mg/L

 

r-o-TOC - HS . nonHS-‘1]

 

Figure 6.26. Transformation of organic matter during ozonation

(HRT = 7 minutes)

154

6.3.4.2. Reaction Rate Coefﬁcient

The study presented in Section 6.2.3 and the evidence from the literature (Bablon
et al., 1991b) showed that the rate coefﬁcient of the ozone reactions in water depends on
the concentration of NOM. It was thought that the rate coefﬁcient of the reaction of
ozone with humic substances is a function of temperature and the concentration of humic

substances and can be expressed in two independent terms, as follows:

k HS = kHS (t) ' kHS (CH3) (6-27)

where his is the reaction rate coefﬁcient; kHs(t) is the temperature-dependent term of the
reaction rate coefﬁcient; kHs(cHs) is the concentration-dependent term of the reaction rate
coefﬁcient.

The goals of the study presented in this section were to validate Equation (6.27)
and to determine the rate coefﬁcients for the ozone reaction with humic substances. The
experiments were conducted at ozone doses ranging from 0 to 5 mg/mg C and at
temperatures ranging from 10 to 30°C. The water ﬂow rate was 18 mL/min, which
corresponded to an HRT of 20 minutes. The reaction rate coefﬁcient, kHs, was

determined using Equation (6.15), which for steady state gave:

2___(°?+s “°Hs) (6.28)

kHS = V
CHSCL

155

Figure 6.27 shows the effect of temperature and the concentration of humic
substances on km. The kHs decreased exponentially with the removal of humic

substances, as follows:

k“, = aeb°HS (6.29)

As can be seen in Figure 6.27, the exponential factor appears to be independent of
temperature, whereas the preexponential factor increased with temperature. Thus,

Equation 6.29 can be rewritten as

kHS = k‘HSeEasmS (6.30)

where Ecus is the exponential factor of the concentration-dependent term; kHst is the
temperature-dependent term.

The kHst valuewas expected to vary with temperature by Arrhenius’ Law:

11s= Else “T (6.31)

where kHs° is the frequency factor; ET is the activation energy; R is the universal gas
constant; T is temperature, °K.
As can be seen in Figure 6.28, a plot of In kHst versus 1/T yielded a straight line,

where the slope was —ETHs/R and the intercept was In kHs°.

156

Combining Equations (6.30) and (6.31), the rate coefﬁcient of the ozone reaction

with humic substances could be expressed in two independent terms, as follows:

 

 

 

 

 

 

 

 

 

C
kHS = kgse RT e HS “S (6.32)
0.35
O
0.30 , ﬁrm- _- _ zzzzz __ z .
y=BE-06e3‘956’
J R’=0.9944
025 / 1_ _-
'Eozo~_———— e 5-7“ _--____ _,__z
:1
2 0.15 4» ~~4~eel—— “___. __ w.-- _ __ ,
x

y=3E-0563.9‘0«
R’=0.972

:7“ / //

j.__ _“ 7 .77 #. #_ -__.._ _ _ ;. . ¥ ._ ___ _ _—

/ //y=35-0694.0089!1

 

 

 

 

 

 

R’=0.9796
0.00 . .
0.00 0.50 1.00 1.50 2.00 2.50 3.00
HS. mglL
‘TuToc _‘ n=20c _’ 71500:,

 

 

Figure 6.27. Relationship between HS and km for various temperatures

(the data for 15°C are not included for ease of readability)

157

 

 

0.0035

 

 

0.00325 0.0033 0.00335 0.0034 0.00345 0.00355
.2.
4.
.5.
(I)
i
5 y=-9627.7x+21.298
43: R2= 0.9835
-101
-12. °
-14
HT

Figure 6.28. The plot of Arrhenius’ equation for the reaction of ozone with humic

substances

158

6.3.5. Reaction of Ozone with Nonhumic Substances

This section presents the kinetic study of the ozone reaction with nonhumic
substances. The objectives of the study were:

0 to determine the stoichiometric coefﬁcient for the reaction of ozone with
nonhumic substances;

0 to test the hypothesis that the rate constant of the ozone reaction with nonhumic
substances is a function of both temperature and the concentration of nonhumic
substances and can be expressed in two independent terms: a temperature-
dependent term and a concentration-dependent term;

0 to determine the kinetic parameters of the ozone reaction with nonhumic

substances.

6.3.5.1. Stoichiometric Coefﬁcient

The stoichiometric coefficient for the ozone reaction with nonhumic substances

was determined using the following equation:

(31c: —Cgm)""ns (c113 “0113»

 

Vnoan = (6.33)

0 0
(0.10an —cnonl-lS +9115 ”CHS)

159

where G and Q are the gas and water ﬂow rates, respectively, mL/min; cgin and cgout are
the concentration of ozone in inﬂuent and effluent, respectively, mg/L; cHs° and cnoan°
are the concentration of HS and nonHS in raw water, respectively, mg/L; cns and cman
are the concentration of HS and nonHS in the effluent, respectively, mg/L.

The experimental design was the same as described in the previous section with
the exception that ozone dose was varied from 1 to 5 mg/mg C. The gas and water ﬂow

rats were 36 amd 18 mL/min, respectively. The results of the experiments are presented

 

 

 

 

 

 

in Table 6.3.
Table 6.3. Experimental data for vnoan
Parameter Raw water Ozonated samples
cg‘“, mg/L 6.98 16.22 27.04
of“, mg/L 3.45 9.89 18.38
HS, mg/L 3.40 2.40 1.98 1.75
nonHS, mg/L 2.68 3.50 3.57 3.58
vnoan, mg Oglmg nonHS 10.1 10.0 11.7

 

 

 

 

 

 

 

The value of vnoan that was used in subsequent analysis was calculated as an

average of the values determined in these experiments and was equal 10.6 mg O3/mg

nonHS.

160

6.3.5.2. Reaction rate coefﬁcient

The reaction rate coefﬁcient, knonns, was determined by combining Equations

(6.15) and (6.16), which for steady state gave:

0 0
C - C + C — C
km S = 3 ( nonHS nonHS HS HS) (6.34)

 

C nonHSc L

The knonHS was determined using the data from the same series of experiments,
which were used to determine km with the exception that only data for an ozone dose of
1 mglmg C and greater were used.

The values of knonns for various ozone doses and temperatures are presented in
Table 6.4. As expected, knonl-lS increased with temperature, but did not appear to be
dependent of ozone dose. This is thought to be because the concentration of nonhumic
substances remained at the same level during ozonation (see Table 6.3). For the range of
operational conditions studied, the knonl-IS was assumed to independent of ozone dose and

had, therefore, only a temperature-dependent term and could be expressed as:

T
E nonHS

k nonHS : k gonHS 6 RT (63 5 )

A plot of 1n knoan versus l/T, shown in Figure 6.29, yielded a straight line,

where the slope was —ETnoan/R and the intercept was In knoan°.

161

Table 6.4 Values of knonns

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Ozone dose, knonns. L/mg min
mglmg C Temperature, °C
10 15 20 30
1 0.001
1.1 0.003
1.2 0.001
1.3 0.003
1.4 0.002 0.002
1.5 0.004
1.7 0.002
1.9 0.013
2.0 0.002
3 0.009
5 0.009

 

 

 

 

 

162

 

 

 

0.0)33

0.0034 0.00345

 

 

0.00325 0.00335 0.0035 0.00355

-1 -

.2.

.3.
g y=-9397.3x+26.392
.3 '41 R’=0.984
E

51

O

.6.

.7 .

.3

1/7

Figure 6.29. The plot of Arrhenius’ equation for the reaction of ozone with nonhumic

substances

163

6.3.6. Mass Transfer Coefﬁcient

The mass transfer coefﬁcient, kLa, was determined using Equations (6.23) and the
data ﬁ'om the same series of experiments, which were used to determine km and knoan-
The values of km for various ozone doses and temperatures are presented in Table 6.5.
The kLa did not appear to be affected by either ozone dose or temperature for the range of
operational conditions studied. The kLa used in the subsequent analysis was calculated as
an average of the data presented in Table 6.5.

Table 6.5. Values of kLa

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Ozone dose, kLa, l/min
mglmg C Temperature, °C
10 15 20 30

1 0.73 0.60

1.1 0.64

1.2 0.71

1.3 0.62

1.4 0.89 0.68 0.67

1.5 0.83

1.7 0.70

1.9 0.80

2.0 0.79
3 0.89
5 0.7

 

 

 

 

 

 

 

164

6.3.7. Ozonation Model

Resulting model, which described the ozonation kinetics of NOM, was obtained

by combining Equations (6.14)-(6.17), (6.32) and (6.35), as follows (with the kinetic

parameters are listed in Table 6.6):

 

EITis 5:0an
Vdc—L-(k a(0tc(’)-c )—v k° e-T‘TeEI‘SCHSc c —v k° e_ “T c c )V—Qc
dt - L g L HS HS HS L nonHS nonHS nonHS L L

 

 

(6.36)
chs 0 0 ’i—I‘TS EESCHs
V dt = Q(CHS _CHS)—kHSe e CHSCLV (6.37)
dc _EzonHS _% EC
V—flmtis- = Q(c:onHS " cnonHS ) — kgonHSe RT 0110anch 1' kgse RT 6 Hsc”$911591." (6'38)
c") . .
118 _d:_ = G(c;°> - cg") — kLa(acg’ — cL)V (6.39)

165

Table 6.6. Kinetic parameters for ozonation model

 

 

 

 

 

 

 

 

 

 

Parameter Dimension Value
VHS mg O3/mg HS 5,17
VnonHS mg O3/mg nonHS 10.6i0.04
k°Hs L/mg min 9628
E°Hs L/mg 3.97:1:0.02
ETHs/R °K 1.78E+O9
kononHS L/mg min 9397
ETman/R °K 2.89E+11
kLa 1/min O.7i0.1

 

 

 

166

 

6.3.8. Model Validation

The numerical calculations were performed using the ModelMaker computer
simulation program (Cherwel Scientiﬁc, Oxford, United Kingdom). The ozonation model
described by Equations (6.36)-(6.39) was validated for the following range of
experimental conditions:

0 Ozone dose — 0 to 4 mglmg C
0 Temperature — 10 to 25°C
0 Retention time — 7 and 20 minutes.

First, the speciﬁc rates of ozone consumption could be determined independently
using the spectroscopic method (see Section 6.3.3) and using the ozonation model
described by Equations 6.14, in which the speciﬁc rate of ozone consumption could be

described as follows:

1‘3; = VHSkHSCHS + VnonHSknonHScHS (6-40)

The values of the speciﬁc ozone utilization rate determined using the
spectroscopic method (ko3) were compared with those obtained using Equation (6.40)
over a range of ooperational parameters described above. Figure 6.30 shows that there
exists a good agreement between these parameters.

Figures 6.31-6.34 present the comparison of experimental and calculated data for

the ozonation of raw water conducted at room temperature and a retention time of 7

167

minutes. There was a good agreement between measured and calculated ozone
concentrations in the effluent gas (see Figure 6.31). The calculated dissolved ozone
concentrations were, however, lower than the experimental values (see Figure 6.32). The
model described very well the removal of HS and nonHS during ozonation (see Figures
6.33-6.34).

Figures 6.35-6.38 present the comparison of experimental and calculated data for
ozonation of raw water conducted at room temperature and a retention time of 20
minutes. As can be seen there was a good agreement between the model and
experimental data for all parameters.

Table 6.7 presents experimental and calculated data for ozonation at an ozone
dose of 1 mglmg C and a retention time of 7 minutes and temperatures ranging from 10
to 25°C. The calculated values of the inﬂuent gas ozone concentration were slightly lower
than the experimental ones. The difference, however, did not exceed 15 percent. A good
agreement between the calculated and experimental data was observed for the dissolved
ozone concentration except for the values obtained at 10°C. The difference between
calculated and experimental values for HS and nonHS did not exceed 5 percent.

Table 6.8 presents experimental and calculated data for ozonation at an ozone
dose of l mg/mg C and a retention time of 20 minutes and temperatures ranging from 10
to 20°C. Very good agreement between calculated and experimental data was observed
for all parameters except for dissolved ozone concentration, for which the calculated
values were approximately 30 percent lower than the experimental data.

Thus, the model and experiments produced agreeable results for all parameters

except for dissolved ozone concentration for which calculated values were lower than the

168

experimental values in two out of four series of experiments. Further studies are needed
to reconcile this discrepancy. The model adequately described the transformation of
humic and nonhumic substances. The relative difference between the experimental and

calculated values did not exceed 10 percent with an average of 6 percent.

 

35« y=09999x
R’=0.8997

2.5 1

1103'“, 1lmin

1.5 1

0.5 «

 

 

 

 

0.5 1 1.5 2 2.5 3 3.5 4

k0,,1/min

Figure 6.30. Speciﬁc ozone utilization rate determined spectrophotometrically and using

Equation (6.40).

169

cc“. mglL

12

101

 

 

 

 

0-2 0.4 0.6 0.8
Ozone dose, mglmg C

 

i—o—Expe rimental ------ Model_j

 

Figure 6.31. Experimental and calculated data for cg°ut

(t=20°C, RT=7 min)

170

1.2

.0
(D

 

Dissolved ozone concentration, mglL
.0 .0 .0 .0 .O .0 .0
N 0) #5 OI (D N a:

.0
_s

 

 

 

 

0.2 0.4 0.6 0.8 1 1.2
Ozone dose, mglmg C

 

L—e—Expe rimental ------ Modelj

 

Figure 6.32. Experimental and calculated data for cL

(t=20°C, RT=7 min)

171

 

 

11

0.5 «

 

 

 

0 0.2 0.4 0.6 0.8 1
Ozonedose,mghngC

 

E—e—Expe rimental ------ Model ]

 

Figure 6.33. Experimental and calculated data for HS

(t=20°C, RT=7 min)

172

1.2

nonHS. mglL

 

 

 

 

 

 

0.2 0.4 0.6 0.8
Ozone dose. mglmg C

:Expe rimental -:--Modelj

Figure 6.34. Experimental and calculated data for nonHS

(F20°C, RT=7 min)

173

 

181

16*

14~

121

10*

co”. mglL

 

 

 

 

T

0 0.5 1 1.5 2 25 3 3.5
Ozone dose. mglmg C

_4

 

L +Expe rimintal ------ Moder—1

___. ___—___. a

Figure 6.35. Experimental and calculated data for cg

(t=20°C, RT=20 min)

174

 

251

ImuummdomchnmxmmmbmrmyL
h

051

 

 

 

 

05 1 15 2 25 3 35
Ozonedose.mghngC

14* Experimental ------ Model

 

 

Figure 6.36. Experimental and calculated data for cl,

(t=20°C, RT=20 min)

175

HS. mglL

 

 

 

 

 

 

1.5 -

1 _

0.5 —
0 1 1 r 1 . .
0 0.5 1 1.5 2 2.5 3

Ozone dose,mg/mg C

 

1__ +Experimental ------ Model T

 

Figure 6.3 7. Experimental and calculated data for HS

(t=20°C, RT=20 min)

176

3.5

nonHS, mglL

N
01

N

.5
CI

0.5 ~

A

 

 

 

 

 

 

 

0.5 1 1.5 2 2.5 3 3.5
Ozone dose. mglmg C
Ll——o—Expe rimental ------ Modem

 

Figure 6.3 8. Experimental and calculated data for nonHS

(t=20°C, RT=20 min)

177

 

Table 6.7. Experimental and calculated data for ozonation of raw water
(ozone dose=1 mglmg C, HRT=7 minutes)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

t,°C cg, mglL cL, mg/L
Experim Model Rel. differ, % Experim Model Rel. differ, %
10 10.1 1 9.64 4.6 1.65 1.25 24.0
15 9.89 9.64 2.5 1.08 0.90 16.6
20 10.04 9.55 5.4 0.89 0.88 1.1
25 10.26 8.72 15.0 0.59 0.62 -5.0
t,°C cHS, mg/L* Cnoan, mg/L**
Experim Model Rel. differ, % Experim Model Rel. differ, %
10 2.34 2.43 -3.8 3.15 3.24 -2.8
15 2.39 2.38 0.4 3.15 3.28 -4.1
20 2.45 2.39 2.4 3.13 3.28 4.6
25 2.42 2.35 2.9 3.36 3.31 1.4

 

 

 

 

 

 

 

 

"‘ - HS concentration in raw water was 3.36 mg/L
" - nonHS concentration in raw water was 2.2 mg/L

178

 

 

 

Table 6.8. Experimental and calculated data for ozonation of raw water
(ozone dose=1 mglmg C, HRT=20 minutes)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

t,°C 08, mg/L 01,, mg/L
Experim Model Rel. differ, % Experim Model Rel. differ, %
10 4.17 4.14 0.7 0.98 0.69 29.6
15 4.1 4.04 1.4 0.67 0.49 26.8
20 3.95 3.97 -0.5 0.5 0.33 34.0
t,°C 035, mg/L* owns, mg/L**
Experim Model Rel. differ, % Experim Model Rel. differ, %

10 2.21 2.33 -5.4 3.13 3.15 -0.6

15 2.07 2.28 -10.1 3.37 3.19 5.3

20 2.05 2.25 -9.8 3.22 3.21 0.3

 

 

 

 

 

 

 

 

* - HS concentration in raw water was 3.4 mg/L
’1‘ - nonHS concentration in raw water was 2.3 mg/L

179

6.4. SUMMARY

The study presented in this Chapter investigated the effect of operational
parameters, including ozone dose and HRT on the transformation of NOM during
ozonation. Most of the UV-254-absorbing compounds in water were efficiently removed
during ozonation at doses less than 1.5 mglmg C. The HRT in the range from 10 to 60
minutes did not affect the removal of UV-254-absorbing compounds during ozonation. It
appears that the removal of UV-254 was independent of the concentration of dissolved
ozone. The removal of organic carbon increased with HRT and dissolved ozone
concentration. Under the experimental conditions studied, gas ﬂow rate did not affect the
removal of either organic carbon or UV-254-absorbing compounds. This study conﬁrmed
the hypothesis that when ozonation was conducted at an HRT of less than 10 minutes and
ozone doses of less than 1 mglmg C, the concentration of BDOC increased with ozone
dose.

This chapter presented the results of the kinetic study of the ozonation of NOM.
Two models were developed. The ﬁrst model was based on the speciﬁc ozone
consumption rate and the second model described the changes in the concentration of
humic and nonhumic substances during ozonation.

The speciﬁc rate of ozone consumption was measured using the spectroscopic
method, which was developed in this study, and the methods based on the ozone
consumption model described by Equations (6.6)-(6.7). It was shown that both methods
gave agreeable results. The spectroscopic method, however, appeared to be more accurate

than the method based on the ozone consumption model. The spectroscopic method was

180

also used to determine the concentration of dissolved ozone. It was shown that there
existed a very good correlation between the dissolved ozone concentrations determined
using the spectroscopic method and the Indigo method. The results of the experiments
conﬁrmed the ﬁrst-order kinetics with respect to the speciﬁc ozone consumption rate.
This study showed that the usefulness of the speciﬁc ozone consumption rate for the
evaluation of ozonation efficacy was limited. This is because it was difﬁcult to separate
the effect of ozonation on UV-254-absorbing compounds and TOC based solely on the
speciﬁc ozone consumption rate without additional information.

This study showed that humic and nonhumic fractions were good surrogate
parameters for modeling the kinetics of ozone reaction in water. The results of the
experiments conﬁrmed the hypothesis that ozone converts humic substances into
nonhumic materials and the removal of organic carbon appears to occur through the
oxidation of nonhumic substances. It was shown that both reactions could proceed
simultaneously. This study conﬁrmed the hypothesis that the rate constant of the ozone
reaction with humic substances was a function of temperature and the concentration of
hunric substances. This reaction rate constant could be represented as a product of two
independent terms: a temperature-dependent terms and a concentration-dependent term.
Under the operational conditions studied, the rate constant of the ozone reaction with
nonhumic substances appeared to be independent of the concentration of nonhumic
substances. The stoichiometric coefﬁcients for the reactions of ozone with humic and
nonhmnic substances were 5.2 and 10.6 milligram of ozone per milligram of humic and
nonhumic substances, respectively. This suggested that the amount of ozone required for

the oxidation of nonhumic materials was approximately two times greater than that

181

required for the oxidation of humic substances. As was stated earlier, a mathematical
model was developed that described the transformation of humic and nonhumic
substances during ozonation. The kinetic parameters were determined and the model was
veriﬁed over a wide range of ozone doses, temperatures, and HRTs. It was shown that the
model adequately described the transformation of humic and nonhumic substances. The
relative difference between the experimental and calculated data did not exceed 10

percent with an average of 6 percent.

182

7. BIODEGRADATION OF NOM

7.1 . INTRODUCTION

Microorganisms can metabolize low molecular weight fractions of NOM and high
molecular weight polymers can be hydrolyzed through the action of extracellular
enzymes, which appears to be a key process controlling the overall rate of the
degradation of NOM (Rittrnann and Woolschlager, 1996). Klevens and Collins (1996)
referred to these fractions as “fast” and “slow” BDOC. Monod kinetics adequately
describes the utilization of “fast” BDOC (Rittrnann and Woolschlager, 1996). However,
in order to determine the kinetic biodegradation parameters, measurements of biomass
concentration in the reactor are required. Considering low organic loading and low
concentration of biomass attached to the support medium in the FBT column, accurate
measurements of biomass concentration in the reactor were not possible. For this reason,
a different approach to describe the biodegradation of NOM was proposed was proposed
in this study, which is described in Section 7.2. The Monod kinetics was used only to
determine the minimum concentration of substrate that required to maintain a steady-
state bioﬁlm (see Section 8.6)

The objective of the study presented in this section were:

0 Using the concept of “fast” and “slow” BDOC, to determine biodegradation
characteristics that quantify the efﬁcacy of the biodegradation of NOM;
0 To compare the biodegradation efﬁciency of FBT to that of bioﬁltration;

- To determine the effect of temperature on the efﬁciency of FBT;

183

0 To test the hypothesis that when ozonation is conducted at an HRT of less than 10
minutes and ozone doses of 1 mg/mg C or less, the biodegradation efficiency
increases with ozone dose.

The study was conducted using the F BT system, the bioﬁltration system, and the
bench-scale ozonation system, which were described in Chapter 3. The experiments were
conducted using raw and ozonated Huron River water. The operational parameters that
were investigated included EBCT, ozone dose, and temperature. The parameters that
were monitored in this study included TOC and UV-254 as they were shown to be good

indicators of HS concentration and THMF P (see Chapter 5).

7.2. APPROACH

A typical curve of the removal of organic carbon in the bioﬁltration column or
F BT column is shown in Figure 7.1. The tangent line to the curve changed from a sharp
decrease after depletion of “fast” BDOC to a more graduate decrease during the
assimilation of slowly biodegradable compounds. Nonbiodegradable organic carbon was
estimated at the point where the curve leveled off. Klevens et al. (1996) operationally
deﬁned the distinction between “fast” and “slow” BDOC by bisecting the extreme
tangent lines of the organic curve, as shown in 7.1. This deﬁnition provides, however,
only graphical determination of “fast” and “slow” BDOC, but not the mathematical
description.

Several operationally deﬁned parameters were introduced in this study. As

outlined in Figure 7.1, the abscissa of the intersection of two extreme tangent lines

184

operationally deﬁned EBCTmin, which represented the EBCT that at least required to
remove “fast” BDOC. The ordinate of the intersection less the concentration of
nonbiodegradable organic carbon operationally deﬁned BDOC,.OW, which represented the
amount of BDOC that at least remained after biodegradation at EBCTmin. The maximum
rate of the biodegradation of fast BDOC, Rmax, was determined by the slope of the
tangent line to the curve at EBCT = 0. This tangent line was mathematically deﬁned as
the line that connected two adjacent points at the beginning of biodegradation, as shown
in Figure 7.1. The second extreme tangent line was mathematically deﬁned as the line

that connected two adjacent points at the end of biodegradation.

 

 

 

 

 

 

 

 

9
8 4
7 4
"fast" BDOC
"slow" BDOC
2 « E nonBDOC
14 ; :
EEBCTW,°,
0 “L’—‘*‘ 1 *\ f f r— ___-__ _— "1
T 50 ‘ 100 150 200 zjo
-1
EBCT, min

Figure 7.1. Biodegradable and nonbiodegradable fractions of NOM

185

7.3. FBT VS. BIOFILTRATION

The goal of the experiments presented in this section was to compare the
efficiency of F BT to that of bioﬁltration with respect to the removal of NOM from raw
water. It was expected that the biodegradation efﬁciency in the F BT column would be
greater than that in the bioﬁltration system. This is because the available surface area of
the ﬂuidized bed is greater than that of the bioﬁlter.

The study was conducted using the FBT column and the bioﬁltration column,
which were described in Chapter 3. The ﬂow rate to the FBT column was varied from 2
to 60 mL/min, which corresponded to EBCTs of 10 to 300 min. The ﬂow rate to the
bioﬁltration column was varied from 0.4 to 20 mL/min, which corresponded to EBCTs of
5 to 250 minutes. All experiments were conducted at room temperature. Most
experiments were conducted in summer 1997. Additional data were collected in
November 1997 to account for seasonal variations in the quality of Huron River water.

Changes in UV-254 of the raw water are shown in Figure 7.2. The removal of
UV-254 increased with EBCT and leveled off at an EBCT of approximately 160-200
minutes reaching a level of 16 percent. The difference in the removal of UV-254-
absorbing compounds by FBT and bioﬁltration was not signiﬁcant at 5 percent level.

Figure 7 .3 shows the removal of TOC by F BT and bioﬁltration. For both systems,
the removal of TOC increased with EBCT and appeared to level off at an EBCT of
approximately 150-200 minutes. It appears, however, that the FBT efficiency towards the

removal of TOC was greater than that of bioﬁltration.

186

18

161

14

12

UV254nmmwaL%l
8

 

 

 

 

 

 

 

CD

 

 

 

 

 

 

oBiothration

OFBT j

 

 

0 0
_~___,_ ._mﬂzht___ﬁz,__ 0 n
O
0 o o
0 A_g#__*~__*__w__«_____*._H_‘*_‘A‘vkﬂd_____
O
O
0 ° 7 __ETTTTHT
10 0 .‘#_n_zh,#____f_____.
0
-——o~—- ”W -___.__,._ ___“ 7~ WWW ___w #222...“
O
0
. z
0 50 100 150 26) 25) 3X) iﬁm
EBCT,mH1

Figure 7.2. Removal of UV-254 from Huron River water by FBT and bioﬁltration

187

TOC removal, %

25

 

 

 

 

 

 

 

 

O
0
201#———m_—ﬂ *# *#_—_——_—— -_
O O
0 o O
O
15 qu~ ~— —~ —— —— —————_—~— -
8 ° .
10 ”VWHHA+AA#44++#www—»A———
° .0
5————.——— ——~——-
0
o r r T ‘1 r r
0 50 100 150 200 250 300
EBCT.min
‘ oBiofiltration oFBT j

 

Figure 7.3. Removal of TOC ﬁ'om Huron River water by FBT and bioﬁltration

188

 

Using the approach, described in the previous section, an attempt was made to
quantify the biodegradation efﬁciency of FBT and bioﬁltration. Figure 7 .4 shows typical
curves of the removal of BDOC by FBT and bioﬁltration. For both systems, two kinetic
regions could be identiﬁed, which characterized the removal of “fast” and “slow” BDOC.
The removal rate of “fast” BDOC in the F BT column was greater than that in the
bioﬁltration column. Table 7.1 presents operationally deﬁned biodegradation parameters
for F BT and bioﬁltration. The Rm“ in the FBT system was two times greater than that in
the bioﬁltration system. Consequently, EBCTmin was almost two times greater in the
bioﬁltration system than that in the FBT system. The difference in BDOCslow for two
systems was not signiﬁcant at 5 percent level.

The results of the experiments showed that biodegradation efﬁciency in the FBT
system was greater than that in the bioﬁltration system. It should be noted, however, that
the systems were operated under different hydrodynamic conditions. In order to achieve
the same EBCT, the ﬂow rate to the FBT system was in the range from 2 to 60 mL/min,
whereas the ﬂow rate to the bioﬁltration was in the range from 0.4 to 20 mL/min. The
bioﬁltration system was operated under plug-ﬂow conditions, whereas the F BT system
was operated as a CSTR. Considering that a commercial F BT system will be operated
under conditions close to plug-ﬂow conditions, the FBT efﬁciency is expected to be

greater than that obtained in this study.

189

Table 7.1. Operationally deﬁned biodegradation parameters for FBT and bioﬁltration

 

 

 

 

 

 

 

 

 

 

 

Parameter
Treatment BDOCo Rm EBCTmin BDOC,.0w
mg/L mg/L min min mg/L
FBTI 1.2 :l: 0.03 0.034 1: 0.003 26 :t 2 0.35 10.06
Bioﬁltrationz 1.15 :l: 0.07 0.017 i 0.002 48 i 3 0.38 :1; 0.07
' two experiments
2 four experiments

 

 

 

 

 

 

 

«0.2

EBCT, min

 

 

+FBT :____-I—Biofiltration J

 

 

Figure 7.4. BDOC removal from Huron River water by FBT and bioﬁltration

190

7.4. EFFECT OF TEMPERATURE

The goal of the experiments presented in this section was to test the hypothesis
that the F BT efﬁciency would increase with temperature. The experiments were
conducted with raw water using the FBT system described in Section 3.4. Temperature
was varied from 10 to 250C. The EBCT was varied from 15 to 120 minutes.

Figures 7.5 and 7.6 show the effect of EBCT and temperature on the removal of
TOC and UV-254 absorbance, respectively. The removal of organic carbon and UV-254
absorbing compounds increased with EBCT and appeared to be greater for higher
temperatures. Figure 7.7 shows the removal of BDOC for various temperatures. For all
temperatures studied, two kinetic regions could be identiﬁed, which characterized the
removal of “fast” and “slow” BDOC. The removal rate of “fast” BDOC in the FBT
column was greater for higher temperatures. Table 7.2 shows biodegradation
characteristics for various temperatures. The Ram increased and EBCTmin decreased with
an increase in temperature. The difference in BDOCsiow does not appear to be signiﬁcant

at 5 percent level.

191

TOC removal. 96

20.0

18.0

16.0

_L
N
Q

d
p
O

.0“
o

4.0

2.0

0.0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

O

 

 

 

 

 

 

 

 

20 40 60 80 100 120 140
580 T. min

 

FT=10C+T=15C+T= 20C+T= 2st

 

Figure 7.5. Removal of TOC from Huron River water by FBT for various

temperatures

192

UV-254 removal. %

 

 

 

 

 

 

 

 

 

 

 

 

, 1
10.0 ._1_ #HE— ———+¥J _ ﬂ 1 #4.— -- 11H -1 1- r
1 1 1
1 . w 1 l 1
8.04'———— ——a L _,1fﬁ_f I‘ﬂiu_,1_-
1 ‘ 1
I 1 1
6.0 ——— ﬂ 1 __1. 114 #1... A 11 m 1,m
1 1 .' l
1 1 1 1‘ 1 1
401— A-%——~ﬁ1 ***+‘ A1 111 1 1 1 _-
1 1 j
1 1, 1
1 1 ' I
0.0 Ir 1 I .l 1 1
o 20 40 60 80 100 120 140
EBCT.min

31:10?sz e 20 c :4 =35—g

 

Figure 7.6. Removal of UV-254 from Huron River water by FBT for various

temperatures

193

1.40

1.20

0.40 ~

0.20 ~

0.00

 

 

 

40 60 80 100 120
EBCT. min

 

b—T=100+T=150+T=2001

 

Figure 7.7. Removal of BDOC by FBT for various temperatures

194

 

140

Table 7.2. Effect of temperature on the biodegradation parameters

 

 

 

 

 

 

 

 

 

 

Parameter
Temperature BDOCo Rmax EBCTmin BDOC,;ow
°C mg/L mg/L min min mg/L
10I 1.2 0.016 39 0.47
151 1.2 0.021 33 0.4
202 1.2 0.034 i 0.003 26 i 2 0.35 :1: 0.06

 

‘ one experiment
2 two experiments

7.5. BIODEGRADATION OF OZONATED WATER

The results of the experiments presented in Section 6.2.3 showed that during

ozonation at an HRT of 7 minutes and ozone doses ranging from 0 to 1 mglmg C, BDOC

concentration increased with ozone dose. This section presents the analysis of the effect

of ozonation on the biodegradation characteristics of raw water, which were deﬁned in

Section 7.2. It was expected that under experimental conditions stated above, the

biodegradation efﬁciency would increase with ozone dose.

The experimental design was described in Section 6.3.2. Brieﬂy, ozonation was

conducted in the bench-scale ozone contactor followed by biodegradation in the

bioﬁltration column. Experiments were conducted at room temperature. The HRT in the

ozone contactor was 7 minutes in all runs. Ozone dose was varied from O to 1 mg/L. No

TOC removal was observed during ozonation under these conditions (see Section 6.3.2).

195

 

Figure 7.8 shows the removal of BDOC during bioﬁltration of samples ozonated
at various ozone doses. As in previous cases, two kinetic regions could be identiﬁed in all
samples, which characterized the removal of “fast” and “slow” BDOC. From visual
observation of the graphs, ozonation resulted in an increase in the concentration of “fast”
BDOC. Table 7 .3 presents biodegradation characteristics of water aﬁer ozonation. For an
ozone doses of 0.6. 0.75 and 1 mglmg C, the Rm, was from 0.09 to 0.11mg/(L min)
compared to 0.017 mg/(L min) for raw water. This corresponded to an approximately
ﬁve-fold increase in Rmax after ozonation. The EBCTmin decreased from 48 minutes for
raw water to 17 minutes for the ozonated water. Considering that the BDOC
concentration in the ozonated sample was only two times greater than that in the raw
water and assuming that the BDOC removal follows the ﬁrst-order kinetics (Klevens and
Collins, 1996; Huck, Zhang, and Price, 1994), the result suggests that ozonation resulted
in an approximately two-fold increase in biodegradation efﬁciency. The BDOCslow,
however, increased after ozonation. This ﬁnding suggests that along with “fast” BDOC
ozonation also resulted in an increase in “slow” BDOC.

Table 7.3. Effect of ozonation on the biodegradation parameters

 

 

 

 

 

 

 

 

 

 

 

 

Parameter

Ozone dose BDOCo Rm“ EBCT.“n BDOC,-10w

mg/mg C mg/L mg/L min min mg/L
0I 1.15 i 0.07 0.017 i 0.002 48 i 3 0.38 1: 0.07

0.152 2.31 0.09 17 0.84

0.752 2.55 0.09 20 0.81

12 2.78 0.11 24 0.79

four experiments

one experiment

196

 

 

BDOC, mg/L
8

0.50 1

 

 

0.00 . 1 1 ‘ ’* ,. .. 1‘ _1
250

 

 

 

-0.5O
EBCT, min

 

--o--Raw water +0.6mg O3In‘gC —0—0.7Smg O3ln’gC +1019 03/"UC]

 

Figure 7.8. Biodegradation of ozonated samples

(ozonation effluent at an HRT of 7 minutes)

197

7.6. SUMMARY

Several biodegradation parameters were introduced in an attempt to quantify the
efﬁcacy of the biodegradation of NOM. These parameters included (1) the maximum rate
of the biodegradation of “fast” BDOC (Rm); (2) the minimum EBCT that required to
eliminate “fast” BDOC (EBCTmin); and (3) the minimum BDOC concentration that
remained after biodegradation at EBCTmin GBDOCslow). These biodegradation parameters
were used to quantitatively describe the effect of temperature and ozone doses on the
biodegradation of NOM and also to compare the efﬁciency of FBT to that of bioﬁltration.

This study showed that the removal of BDOC in the FBT system was greater than
that in the bioﬁltration system. The Rmax in the FBT column was two times greater than
that in the bioﬁltration column.

The results of the experiments showed that the removal of TOC and UV-254
increased with temperature. An increase in temperature resulted in an increase in Rmax
and in a decrease in EBCTmm. The removal of slowly biodegrading organic matter did not
appear to be affected by temperature.

This study showed that ozonation at an HRT of 7 minutes and ozone doses of 0.6
to l mglmg C signiﬁcantly increased the efﬁciency of biodegradation of NOM compared
to that of raw water. It also resulted in an increased the concentration of “fast” BDOC.
The BDOC,.ow increased after ozonation, which suggested that along with “fast” BDOC

concentration, ozonation also resulted in an increase in “slow” BDOC concentration.

198

8. EVALUATION OF COMBINED OZONATION AND FBT PROCESS

8.1. INTRODUCTION

The results of the experiments presented in Chapter 4 demonstrated a good
potential of the ozonation/FBT system to control THM precursors. The study presented in
Chapter 4, however, showed that along with breaking down high molecular weight
compounds ozone also oxidized lower molecular weight compounds (less than 1000
daltons) that could, otherwise, be degraded biologically. It was also shown that a portion
of ozone was consumed to mineralize organic carbon, which resulted in a decreased
production of biodegradable organic matter.

The subsequent studies presented in Chapters 6 and 7 attempted to explain these
ﬁndings. The bench-scale ozonation experiments described in Chapter 6 showed that the
mineralization of organic carbon during ozonation increased with an increase in HRT.
Since the ozone contactor in the ozonation/FBT system was operated at an HRT of
approximately 100 minutes, ozonation resulted in a mineralization of up to 25 percent of
organic carbon. From the bench-scale study (see Chapter 6), very little removal of
organic carbon was observed after ozonation at an HRT of less than 10 minutes and
ozone doses of up to 1 mglmg C. Under these conditions, the concentration of BDOC
and, hence the biodegradation efﬁciency, increased with ozone dose. The use of low HRT
may, however, be limited by considerations of mass transfer efﬁciency and/or the CT
requirements under SDWA D/DBP Rule if ozone is also used for disinfection. Another

problem may arise from the fact that biodegradable organic matter in Huron River water

199

consisted of rapidly and slowly biodegrading fractions (“fast” and “slow” BDOC) (see
Chapter 7). The “slow” BDOC accounted for nearly 30 percent of biodegradable organic
matter in raw water. The results of the experiments demonstrated that along with an
increase in “fast” BDOC concentration, ozonation also resulted in an increase in the
concentration of “slow” BDOC. The slowly biodegradable organic materials may pose
bacterial growth problems in the distribution system, if they are not removed in the
treatment system. Thus, the optimization of the combined ozonation and FBT process
should be directed towards:
1) increasing the removal of NOM;
2) reducing the ozone consumption by using ozone solely for the production of
BDOC rather than for the mineralization of organic carbon;
3) reducing the biodegradation time by converting “slow” BDOC into “fast”
BDOC.
The objectives of the study presented in this chapter was:
0 to use the biodegradation characteristics developed in Section 7.2 for the
evaluation of process efﬁciency;
0 to test the hypothesis that using FBT prior to ozonation increases the production
of low molecular weight compounds and biodegradable organic matter;
0 to test the hypothesis that using bioﬁltration after FBT/ozonation results in a
greater removal of NOM than does the ozonation/FBT process;
0 to determine if the recycle of a portion of the ozonation effluent back to the F BT
column increases the efﬁciency of the FBT/ozonation process combined with

bioﬁltration;

200

o to test the hypothesis that adding an easily biodegradable carbon source to the
FBT column increases the efficiency of the FBT/ozonation process combined
with bioﬁltration (the F BT/ozonation process using an additional carbon source
will be referred to as the stimulated F BT/ozonation process);

0 to test the hypothesis that a recycle of a portion of the ozonation effluent back to
the FBT column reduces the biodegradation time of the stimulated FBT/ozonation
process combined with.

It should be noted that the use of bioﬁltration in combination with FBT/ozonation
process does not represent a departure from the originally proposed ozonation/FBT
process. This is because the process design of either the ozonation/FBT or FBT/ozonation
process will require some sort of ﬁltration as the ﬁnal stage, which will be used to
remove turbidity remaining in the system efﬂuent and as a safeguard if any malfunctions
of FBT operation occur. Besides, most existing water treatment plants currently use sand
or GAC ﬁltration as the ﬁnal stage of the process. Therefore, if the proposed process is
used for retroﬁtting of an existing facility, bioﬁltration will not require additional capital
investments. The optimization of the combined ozonation and FBT process should be
directed towards the minimization of bioﬁltration time in the bioﬁlter rather than towards
the elimination of the bioﬁltration step.

It should also be noted that the goal of the study presented in this chapter was not
to determine the optimal conditions for the combined ozonation and FBT process, but to
evaluate approaches that were expected to increase the efficiency of the process. This is
because the F BT column in the existing pilot-scale system was operated as a C STR (see

Section 3.2.2). Commercial systems or a larger pilot-scale system (expected to be

20]

installed in the near ﬁtture) will be operated close to plug-ﬂow conditions, in which the
biodegradation efficiency is expected to be different from that in the existing system.
Section 8.2 presents the results of the study of the single-pass FBT/ozonation process.
The effect of the recycle on the performance of the FBT/ozonation is investigated in
Section 8.3. The study of single-pass stimulated FBT/ozonation and stimulated
FBT/ozonation with recycle are described in Sections 8.4 and 8.5, respectively. Section
8.6 provides additional information on the biodegradation kinetics in an attempt to
explain the ﬁndings obtained in previous sections. Section 8.7 presents the summary of

the results.

8.2. SINGLE PASS F BT/OZONATION PROCESS

As was stated earlier, the ozonation of Huron River water resulted in the oxidation
of low molecular weight compounds and potentially biodegradable organic matter (see
Chapter 4). The results of the experiments, presented in Chapters 7.3, showed that FBT
could potentially remove up to 20 percent of TOC from Huron River water. The rapidly
biodegradable fraction (“fast” BDOC) accounted for approximately 70 percent of the
biodegradable organic matter in the raw water, which would be eliminated if the FBT is
used prior to ozonation. In this case, ozone was expected to be used for the oxidation of
slowly biodegradable or refractory organic matter. The objectives of the study presented
in this section were:

0 to compare the effect of ozone dose on the removal of TOC and UV-254 from

the FBT effluent to that from raw water;

202

o to test the hypothesis that ozonation of the FBT efﬂuent results in an increase

in the concentration of low molecular weight compounds;

0 to evaluate the efficiency of bioﬁltration for the removal of biodegradable

organic matter produced from the F BT/ozonation efﬂuent

o to determine biodegradation characteristics of the FBT/ozonation effluent

using the approach developed in Section 7.2.

The study was conducted using the pilot-scale FBT/ozonation system described in
Section 3.2.3. As was shown in Section 7.3, the EBCT in the FBT column that required
to eliminate “fast” BDOC from raw water was approximately 30 minutes. As such, the
EBCT in the FBT column was reduced from 180 minutes to 30 minutes as described in
Section 3.2.3. The ﬂow rate of raw water entering the system was 20 mL/min, which
corresponded to an EBCT in the FBT column of 30 minutes and a retention time in the
ozone contactor of approximately 60 minutes. The experiments were conducted at room
temperature. Ozone dose was varied from 0 to 2 mg/mg C. The parameters that were
monitored in this study included TOC concentration, UV-254 absorbance, and AMW
distribution. The biodegradation characteristics, deﬁned in Section 7.2, were also
determined. A simpliﬁed schematic of the treatment train with sampling locations is

shown in Figure 8.1.

203

Raw water

 

 

 

V

Sample

 

FBT

 

 

 

 

 

11
Sample

 

Ozonation

Efﬂuent to bioﬁltration

 

 

 

 

l

7

L
7

Sample

Figure 8.1. Simpliﬁed schematic of the FBT/ozonation process

204

Figures 8.2 and 8.3 show the effect of ozone doses on the removal of UV-254-
absorbing compounds and TOC from raw water and from the FBT efﬂuent. The use of
FBT prior to ozonation resulted in a slight increase in the removal of UV-absorbing
compounds compared to that removed by ozonation alone. This was consistent with the
data obtained in Chapter 4 that showed that the removal of UV-254 absorbing
compounds occurred mostly through ozonation rather than biodegradation. The TOC data
were more remarkable. As expected, the removal of TOC after ozonation of raw water
increased with ozone dose and reached a level of approximately 23 percent at a dose of 2
mg/mg C. This was consistent with the results obtained in Chapter 4. The FBT resulted in
the removal of approximately 12 percent of organic carbon from raw water, which
accounted for approximately 60-65 percent of biodegradable organic matter in raw water.
The TOC removal during ozonation of the FBT effluent increased with ozone dose and
reached a plateau at a dose of approximately 0.4 mglmg C.

The study presented in Chapter 4 showed that when the removal of TOC during
ozonation leveled off, the concentration of low molecular weight compounds and
biodegradable organic matter increased with ozone dose. As such, ozonation of the FBT
efﬂuent at ozone doses greater than 0.4 mg/mg C was expected to result in an increased
production of low-molecular weight compounds and biodegradable organic matter.
Figure 8.4 shows the AMW distribution of organic matter in raw water, in the FBT
effluent and in water aﬁer F BT and ozonation at ozone doses of 0.5 and l mglmg C. The
FBT resulted in a partial removal of low-molecular weight compounds (less than 500
daltons). The F BT also broke down higher molecular weight compounds into smaller

constituents. This agreed with the previous ﬁndings (see Chapter 4). As expected, for

205

both ozone doses, ozonation of the FBT efﬂuent resulted in a decrease in the
concentration of high-molecular weight compounds and a simultaneous increase in the
concentration of organic carbon in low molecular weight fractions (less than 1000
daltons). This increase was not observed during the ozonation of raw water conducted
under similar conditions (see Section 4.3). This is thought to be because the removal of
organic carbon during ozonation of raw water in the pilot-scale system increased with an
increase in ozone dose of up to 1 mglmg C, whereas the removal of organic carbon
during ozonation of the F BT effluent leveled off at a dose of approximately 0.4 mglmg C.
The results of the experiments once again conﬁrmed the hypothesis (see Chapter 4) that
when ozonation does not result in the removal of organic carbon, the concentration of
low molecular weight compounds increases with ozone dose.

The removal of biodegradable organic matter produced from the ozonation of
F BT efﬂuent was accomplished by bioﬁltration. The removal of organic carbon by the
FBT/ozonation process followed by bioﬁltration was compared to that by the
ozonation/FBT process described in Chapter 4. The results are presented Table 8.1. The
ozonation/FBT process could potentially remove approximately 40 percent of organic
carbon, which was achieved at an EBCT of 180 minutes. The FBT/ozonation process
followed by bioﬁltration resulted in the same removal at a total EBCT of only 60
minutes. The process could potentially remove additional 7 percent of TOC, which
indicated that the removal efﬁciency of the FBT/ozonation process followed by
bioﬁltration was greater than that of the ozonation/FBT process. The removal of this

remaining biodegradable organic matter by bioﬁltration was, however, slow.

206

Figure 8.5 shows the kinetic biodegradation curve for the FBT/ozonation effluent
obtained at an EBCT in the FBT column of 30 minutes and an ozone dose of l mg/mg C.
This curve was used to determine biodegradation characteristics of the F BT/ozonation
effluent, which are presented in Table 8.2. As can be seen in Figure 8.5 and Table 8.2, the
FBT/ozonation efﬂuent contained at least 0.5 mg/L of “slow” BDOC. As a result,
prolonged bioﬁltration was required to completely remove all biodegradable organic

matter from the FBT/ozonation efﬂuent, which agrees with the results presented in Table

8.1.

207

UV254 removal, %

 

 

 

8

 

 

 

2?

1'?

 

 

 

 

 

 

 

2.5

 

 

 

 

 

 

 

 

 

 

 

0.00 0.50 1.00 1.50 200 250
Ozone dose, mglmg C

+FBT effluent +Rawwater 1

 

 

Figure 8.2. Removal of UV-254 during ozonation of raw water and FBT effluent

(F BT EBCT —30 minutes)

208

TOC removal. %

 

 

 

 

 

 

 

 

 

 

 

 

A f ; 3 A
20 f _ l
15 «—
1

10 /

5 4 —- ——

o 1 7 1 r

0.00 0.50 1.00 1.50 2.00

Ozone dose, mglmg C

 

‘L +FBT effluent

+Raw water

1

2.50

Figure 8.3. Removal of TOC during ozonation of raw water and FBT effluent

(FBT EBCT — 30 minutes)

209

TOC. m9"-

 

1.8 5

1.6 4 _—

1.4«

1.2 a

0.8 4

0.6 1

0.4 «

0.2 .

 

 

 

 

 

 

 

 

 

 

 

 

 

 

04 . T .
«5(1) 500-1000 1000-3000 3000-10000 10000-30000 >30000
AMWCutoffs

 

 

 

 

 

T

 

i Iraw water IF 8T '1
l UFBT/ozone=0.5mglmgc DFBT/ozone=1mglmgc 1

 

 

Figure 8.4. AMW of NOM after FBT and ozonation

210

Table 8.1. Efficiency of FBT/ozonation/bioﬁltration and ozonation/FBT

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Parameter F BT/ozonation/bioﬁltration Ozonation/
1 2 3 FBT‘
Ozone dose, mg/mg C 1.0 1.0 1.0 1.0
EBCT for FBT, min 30 3O 30 180
Bioﬁltration EBCT, min 30 70 150 n/a
Total EBCT, min 60 100 180 180
TOC removal, % 39.4i1.l 42.5il.1 46.5:l:0.2 39.7i2.1
1see Chapter 4
1.8
1.6 1
1.4 «
1.2 a
2f ‘ ‘
8 0.8
m
0.6 1
0.4
0.2
o f , , 1 4 1 1 .-
0 20 40 60 80 1(1) 120 140 160 180 200
E B c T. mi n

Figure 8.5. Biodegradation kinetics of the FBT/ozonation effluent

211

 

 

Table 8.2. Biodegradation characteristics of the F BT/ozonation effluent

 

 

 

 

 

Parameter Value
BDOCo, mg/L 1.59 :t 0.1
Rm, mg/L min 0.06 i 0.02
EBCTmin, min 18.2 i 0.6
BDOCslow, mg/L 0.46 i 0.16

 

 

 

 

8.3. FBT/OZONATION PROCESS WITH RECYCLE

The results of the study presented in the previous section showed that, under
similar experimental conditions, FBT/ozonation followed by bioﬁltration resulted in a
greater removal of organic matter and lower biodegradation time than did the
ozonation/FBT process. It was, however, shown that the F BT/ozonation effluent
contained at least 0.5 mg/L of slowly biodegradable organic matter, which was difficult
to remove by subsequent bioﬁltration. The goal of the study presented in this section was
to test the hypothesis that recycle of a portion of the ozonation effluent back to the FBT
column would result in an increase in the removal of organic matter by subsequent
bioﬁltration and in a decrease in biodegradation time. The effect of recycle on UV-254
and TOC concentration was also investigated.

The experimental design was the same as described in the previous section with
the exception that a portion of the ozonation effluent was recycled back to the FBT

column. The recycle ﬂow was varied from 10 to 40 lemin, which corresponded to

212

recycle ratios (recycle ﬂow to inﬂuent ﬂow) of 0.5 to 2. A simpliﬁed schematic of the

process with sampling locations is shown in Figure 8.6.

Raw water

 

 

 

 

 

FBT

 

 

 

 

Sample

 

Ozonation

 

Effluent to bioﬂltration

 

 

 

A
7

Sample

Figure 8.6. Simpliﬁed schematic of the FBT/ozonation process with recycle

213

Figure 8.7 shows the effect of recycle ratio on the removal of TOC after the FBT
and ozonation stages at an ozone dose of l mglmg C. With no recycle, FBT resulted in
the removal of approximately 12 percent of organic carbon and the following ozonation
eliminated approximately 9 percent of organic carbon. An increase in recycle ratio
resulted in a decrease in the removal of organic carbon by ozonation. This is thought to
be because of an increase in single-pass ﬂow rate through the ozone contactor and, hence,
a decrease in single-pass retention time. The removal of TOC after F BT and
FBT/ozonation increased with recycle ratio. At recycle ratios of 1 and greater, no TOC
removal during ozonation occurred and, as a result, the removal of TOC after
FBT/ozonation leveled off. As can be seen in Figure 8.8, for other ozone doses (up to 2
mglmg C) and a recycle ratio of 1, TOC removal during ozonation was not observed
either.

Figure 8.9 shows that recycle did not affect the removal of UV-254 aﬁer
FBT/ozonation. The UV-254 in the FBT effluent decreased with an increase in recycle
ratio. However, this decrease could probably be attributed to a dilution of system inﬂuent
with the recycle ﬂow.

Figure 8.10 shows TOC removal during bioﬁltration of effluent obtained from the
FBT/ozonation system at an ozone dose of l mglmg C and a recycle ratio of 1. The
results of bioﬁltration experiments using the efﬂuent obtained from the single-pass
FBT/ozonation system at an ozone dose of l mg/mg C are given for comparison. The
data points at EBCT = 0 represented the removal of organic carbon by FBT/ozonation. It
appears that using the recycle resulted in a slight increase in the removal of organic

matter by F BT/ozonation followed by bioﬁltration. Although this increase was

214

statistically signiﬁcant the effect was less than expected. As can be seen in Figure 8.10,

the biodegradation rate of “fast” BDOC in the effluent from the recirculating

FBT/ozonation system was lower than that from the single-pass FBT/ozonation system.

This is thought to be because, in the recirculating F BT/ozonation system, FBT resulted in

the removal of most easily biodegrading organic matter leaving slower biodegradable

materials in the system effluent. As a result, a decrease in EBCT in the bioﬁltration

system was not observed.

TOC removal. %

\

 

Ii

 

 

 

 

 

\

1
1
l
1

7

 

 

 

i

0 0.5 1 1.5 2 2.5
Recycle ratio

 

r—O-afterFBT +afterFBT/ozonation +in ozone contactorJ

Figure 8.7. Effect of recycle on the removal of TOC by FBT/ozonation

(EBCT =30 minutes, ozone dose = l mg/mg C)

215

 

 

 

 

 

 

 

 

 

 

 

 

3o-L_._m_ _._____-_-A_v__ﬂ E

25 / M _
32 /
a? 20 ~~ #1.“.W ”7’4
E
8 15 —— “-F ‘ v .1
1.—

10 #4

5 .2.

0 __m “__t.m ——1—-——— 1

0 05 1 15 2 25

Ozone dose. mglmgc

 

RFBTA+FBTIozonationl

|
1

Figure 8.8. Effect of ozone dose on the removal of TOC by FBT/ozonation with recycle

(EBCT =30 minutes, recycle ratio = l)

216

 

 

 

8

 

8

 

 

 

 

 

UV-254 removal. %
8

8

 

 

 

 

 

 

 

 

0 0.5 1 1.5

161

Recycle ratio

 

L +FBT fFBTIozonation 1

 

 

Figure 8.9. Effect of recycle on the removal of UV-254 by FBT/ozonation

(EBCT =30 minutes, ozone dose = 1 mglmg C)

217

2.5

TOC removal. %

 

 

 

 

 

 

 

 

 

 

 

 

w 5 W (m __._a
4
........i 4
ﬁw_.—4
20 15
10 _"n ~—~ ___ MNI-,_W___”_E
o r r 1 -
o 50 100 150 200 250

EBCT. min

 

7' ‘—1
.--D--FBTIozonation effluent -—e—FBT/ozonation effluent (recycle ratio-15

 

 

Figure 8.10. Bioﬁltration of the F BT/ozonation effluent

(FBT/ozonation: EBCT =30 minutes, ozone dose = 1 mglmg C; recycle ratio = l)

218

8.4. STIMULATED F BT/OZONATION PROCESS

The study presented in the previous section did not conﬁrm the hypothesis that
that recycle of a portion of the ozonation effluent back to the FBT column would result in
an increased bioﬁltration efﬁciency. This was thought to be because FBT was able to
remove only most easily biodegrading organic matter and slower biodegradable materials
were remained in the FBT/ozonation effluent.

This section describes the experiments in which the possibility of adding an easily
biodegradable carbon source in order to enhance the biodegradation of NOM was
explored. It was expected that adding an easily biodegradable material would result in an
increased growth of microorganisms that would be able to metabolize slow biodegradable
organic matter. Of course, there was a possibility that microorganisms would utilize only
easily biodegradable organic matter and leave slowly degrading materials.

The experimental design was the same as described in Section 8.2 with the
exception that an easily biodegradable carbon source was used in these experiments. A
solution of 500 mg/L of acetic acid in Milli-Q water was pumped into the bottom of the
F BT column at rates of 0.1 and 0.2 lemin, which corresponded to acetate dose of 2.5
and 5 mg/L. Ozone dose was 1 mglmg C in all experiments.

Figure 8.11 shows the effect of acetate dose on the removal of TOC from raw
water affer F BT (no ozonation). The TOC removal appears to increase with the addition
of acetate. However, the difference was statistically signiﬁcant only for data points
obtained at acetate doses of 0 and 5 mg/L. It was really surprising, however, that, when

acetate was added, no TOC removal was observed during ozonation of the F BT effluent.

219

The effect of bioﬁltration on the removal of organic matter from the stimulated
FBT/ozonation effluent is shown in Figure 8.12. The acetate dose was 2.5 mg/mg C. The
results of bioﬁltration of the efﬂuent from the FBT/ozonation system (no acetate) are
given for comparison. The data points at EBCT = 0 represented the removal of organic
carbon by FBT/ozonation. As can be seen, the rate at which “fast” BDOC was removed
from the stimulated FBT/ozonation effluent was greater than that from the
FBT/ozonation (no acetate) effluent. The potential removal of organic carbon by the
stimulated FBT/ozonation process followed by bioﬁltration was also greater than that by
the FBT/ozonation process (no acetate) followed by bioﬁltration.

Figure 8.13 shows the biodegradation kinetic curve for the stimulated
F BT/ozonation effluent obtained at an EBCT in the FBT column of 30 minutes, an ozone
dose of 1 mg/mg C, and an acetate dose of 2.5 mg/L. Table 8.3 presents biodegradation
characteristics of this efﬂuent. As can be seen in Figure 8.13 and Table 8.3, the
stimulated FBT/ozonation effluent had high concentration of “fast” BDOC, which was
rapidly removed by bioﬁltration. However, at least 0.5 mg/L of “slow” BDOC remained

in the effluent aﬁer stimulated FBT/ozonation.

220

25

 

 

 

 

 

"—" ‘1

 

TOC removal. %

 

 

 

 

 

Acetate dose, mglL

 

+

1 FBT +ozonatioanBTJ

Figure 8.11. Effect of acetate dose on the removal of TOC by FBT and FBT/ozonation

(ozone dose -— l mglmg C, F BT EBCT — 30 minutes)

221

60.0

 

 

 

 

TOC removal. %

 

 

 

 

 

 

 

0.0 * r * ’ '
0 20 40 so 80 100 120 140 160 180 200
EBCT, min
1 —+—Enhanced FBT/Ozonation --e-- FBT/Ozonation j

 

Figure 8.12. Bioﬁltration of efﬂuents from FBT/ozonation and stimulated FBT/ozonation

(FBT EBCT - 30 minutes, ozone dose -— 1 mglmg C, acetate dose 2.5 mg/L)

222

BDOC, mglL

3.0)

 

2501

2.00

1(0-

0.50 1

 

 

0.00 1 1 1 . T 1 1 1 c
O 20 40 60 80 100 120 140 160 180 200
E B C T. mi n

 

Figure 8.13. Biodegradation kinetics of stimulated FBT/ozonation effluent

(ozone dose — l mg/mg C, FBT EBCT — 30 minutes, acetate dose — 2.5 mg/L)

223

Table 8.3. Biodegradation characteristics of stimulated F BT/ozonation effluent

 

 

 

 

 

Parameter Value
BDOCo, mg/L 2.7 i 0.06
Rmmmg/Lmin 0.19:0.01
EBCTmin, min 12.3 i 0.8
BDOCslow, mg/L 0.47 i- 0.02

 

 

 

 

8.5. STIMULATED FBT/OZONATION PROCESS WITH RECYCLE

The results of the study presented in the previous section showed that stimulated
FBT/ozonation followed by bioﬁltration resulted in an increased removal of organic
matter. However, the stimulated FBT/ozonation effluent contained at least 0.5 mg/mg C
of slowly biodegrading organic carbon, which was difficult to remove by bioﬁltration.

The goal of the study presented in this section was to investigate the efficiency of
the stimulated FBT/ozonation process with recycle. Without adding an easily
biodegradable carbon source, the recycle operation was shown to be ineffective (see
Section 8.3). The result of experiments presented in Section 8.4 showed that adding 2.5
mg/L of acetate to the FBT column resulted in an increased production of “fast” BDOC
after FBT/ozonation, which was rapidly removed by bioﬁltration. This suggested that
stimulated FBT/ozonation was able to convert slowly biodegrading organic materials into
“fast” BDOC. It was, therefore, expected that using recycle in this case would result in

increased process efficiency.

224

The experimental design was the same as described in the previous section with
the exception that a portion of the ozonation effluent was recycled back to the FBT
column at recycle ratios of 0.5 and 1. Higher recycle ratios were not tested in this study.

Figure 8.14 shows the removal of organic carbon during bioﬁltration of the
efﬂuent from the recirculating stimulated FBT/ozonation system with recycle ratios
varying from 0 to 1. The data points at EBCT = 0 represented the removal of TOC after
stimulated FBT/ozonation with recycle ratios of 0, 0.5 and 1. As can be seen, the recycle
did not result in an increase in the maximum removal of organic carbon after stimulated
FBT/ozonation followed by bioﬁltration. However, bioﬁltration time appears to decrease
with an increase in recycle ratio. Figure 8.15 shows bioﬁltration kinetics for the efﬂuents
from the stimulated FBT/ozonation for various recycle ratios and Table 8.4 presents
biodegradation characteristics of these efﬂuents. As expected, BDOC concentration in the
F BT/ozonation efﬂuent (BDOCo) decreased with an increase in recycle ratio because of
an increased removal of organic carbon in the FBT/ozonation system (see data points at
EBCT = 0 in Figure 8.14). It appears that Rmax decreased proportionally to BDOCo. The
efﬂuents were characterized by low EBCTmin. The difference in EBCTmin for various
recycle ratios does not, however, appear to be statistically signiﬁcant. Finally, the
concentration of slowly biodegradable organic matter in the FBT/ozonation efﬂuent

decreased with an increase in recycle ratio.

225

60.0

 

 

 

 

 

 

 

 

 

 

 

 

#L a
50.0 4~—— —————~——,—— — i___._ _z_ _ _, ___ , f
40.0 PW mm
°\°
37
>
O
E30.0~ H—— ————ﬁ _ 77i7____!44_ _l
2 1
§
20.0 '/ __
l.
10.0-13~— _ — (Wmmr; 111 m1 ,1
0.0 T 1' Y
0 50 100 150 200 250
EBCT,min

T

L__-_0_—No recycle

 

+Recycle ratio-0.5 _ +Recyc|e ratio-1 _j

Figure 8.14. Bioﬁltration of effluents from stimulated FBT/ozonation with recycle

(F BT EBCT - 30 minutes, ozone dose — 1 mglmg C, acetate dose - 2.5 mg/L)

226

 

3.0)

2.501

2001

BDOC, mg/L
'8

1.00 5

05° ‘ \

0.11)

 

 

 

0 20 40 60 80 100 120 140 160 180 200
E BCT, min

 

Em recycle +Recycle ratio 0.5 +Recycle ratio 1 1

 

Figure 8.15. Biodegradation kinetics of effluents from stimulated FBT/ozonation with
recycle

(F BT EBCT - 30 minutes, ozone dose — l mglmg C, acetate dose - 2.5 mg/L)

227

Table 8.3. Biodegradation characteristics of recirculating stimulated F BT/ozonation

 

 

 

 

 

 

 

effluents
Parameter Recycle ratio
0 0.5 l
BDOCo, mg/L 2.7 i 0.06 2.18 1.64
Rmmmg/Lmin 0.19:0.01 0.15 0.13
EBCTmin, min 12.3 i 0.8 11.9 10.5
BDOCslow, mg/L 0.47 i 0.02 0.4 0.24

 

 

 

 

 

8.6. BIODEGRADATION KINETICS

Models describing the kinetics of substrate utilization by bioﬁlms generally
include terms for the mass transfer of the substrate from the bulk solution to the bioﬁhn
surface, as well as simultaneous diffusion and biotransformation within the bioﬁlm
(Criddle, Alvarez, and McCarthy, 1991). While the models are complex even for some
idealized situations (Rittrnann and McCarty, 1980), the models describing the
biodegradation in a recirculating FBT column for the treatment of drinking water could
be signiﬁcantly simpliﬁed. First, a recirculating FBT column could be considered as a
completely mixed reactor, since the recirculation rate was much greater than the rate of
substrate inﬂux. Secondly, since the liquid rate through the bed was high, the resistance
to mass transfer in the diffusion layer between the bulk liquid and the bioﬁlm could be
neglected.

For a CSTR, the rate of substrate utilization can be written as follows:

228

dsc
dt

 

=%(s0 —Sc)—aJ (8.1)

where So and Se are the substrate concentrations in the inﬂuent and effluent, respectively;
J is the ﬂux of the substrate into the bioﬁlm; a is the area of bioﬁlm per unit volume of
reactor; 1: is the hydraulic retention time.

At steady state, dSJdt = 0, and

380 - S.) = ad (82)

The ﬂux J is inﬂuenced by molecular diffusion of the substrate in the bioﬁlm and
by the rate of substrate utilization within the bioﬁlm. With low substrate concentrations,
the feed rate to the bioﬁlm is sufficiently low, so the active population of bacteria is
relatively low, and the bioﬁlm can be considered as fully penetrated (Criddle, Alvarez,
and MacCarthy, 1991). For the fully-penetrated bioﬁlm, the ﬂux is inﬂuenced only by
bioreaction kinetics. The kinetics of these reactions follows a Monod relationship

(Rittrnann and McCarty, 1981):

J = kmLfoSf

8.3
Sf + Ks ( )

229

where Sf is the substrate concentration within the bioﬁlm; k," is the maximum speciﬁc rate
of substrate utilization, time'l; K, is the half-velocity coefficient; Xfis the density of the
bioﬁlm; Lf is the bioﬁlm thickness.

Under CSTR conditions, for fully penetrated bioﬁlms and for low substrate
concentration, Sf = Se and S << K, (Criddle, Alvarez, and MacCarthy, 1991). Thus,
Equation (8.3) could be reduced to a ﬁrst-order relationship with respect to substrate

concentration, which agrees with the ﬁndings of Huck, Zhang, and Price (1994):

-:-(80 — 86) = use (8.4)

where X = aLfo is the concentration of biomass in the reactor, k' = km/Ks.

To verify the ﬁrst-order biodegradation kinetics with respect to substrate
concentration, the values of Rmax for all samples obtained as described in the previous
sections were plotted versus the concentration of BDOC. The plot in Figure 8.16 shows
that there existed a linear correlation between Rm.x and BDOC concentration. The results
agree with the ﬁndings by Huck, Zhang, and Price (1994), who obtained similar
relationships.

It should be noted that, considering the low organic load, the development of the
bioﬁlm on the support media was important. There exists a minimum substrate
concentration Smin below which a steady-state bioﬁlm cannot exist and the bioﬁlm will be
constantly losing its mass (Criddle, Alvarez, and MacCarthy, 1991). Huck, Zhang, and
Price (1994) identiﬁed Smin as the abscissa of the intersection of the regression line with

the abscissa line. From Figure 8.16, Smin was approximately 1 mg/L. This ﬁnding may

230

suggest that the existence of “slow” BDOC could not simply be attributed to the fact that
the concentration of BDOC was below Smin. Additional studies, however, are needed to
explain the efficient biodegradation of organic matter in the effluents from the stimulated

recirculating FBT/ozonation process at BDOC concentrations signiﬁcantly lower than

Smin-

 

0.25

0.2 1 e
y = 01103:: - 0.1148

R2 a 0.9598

Rm. mglL min

0.1 <

0.05 4

 

 

 

 

0 0.5 ‘l 1.5 2 2.5 3
BDOC. mglL

Figure 8.16. Relationship between BDOC concentration and biodegradation rate

231

8.7. SUMMARY

The study presented in this section showed that using F BT prior to ozonation
resulted in an increased production of low molecular weight compounds and
biodegradable organic matter. This was not observed in the ozonation/FBT process
conducted under similar conditions. The biodegradable organic matter produced from the
FBT/ozonation process was removed by bioﬁltration. It was shown that the
FBT/ozonation process followed by bioﬁltration was more efﬁcient than the
ozonation/FBT process in terms of the removal of organic matter and biodegradation
time. However, the effluent from the FBT/ozonation process contained at least 0.5 mg/L
of slowly biodegrading organic carbon, which was difﬁcult to remove by bioﬁltration.
The hypothesis that the recycle of a portion of ozonation effluent back to the FBT column
would increase the efﬁciency of the process was not conﬁrmed. This is thought to be
because of slow biodegradation of NOM. As a result, FBT resulted in the removal of
fastest biodegrading fractions of organic matter leaving slower biodegradable materials in
the system effluent. These materials were difficult to remove by following bioﬁltration.

The results of experiments conﬁrmed the hypothesis that adding an easily
biodegrading carbon source to the FBT column resulted in an increased removal of
organic matter. In addition, the stimulated FBT/ozonation process resulted in an
increased production of “fast” BDOC that was efficiently removed by bioﬁltration.
However, the efﬂuent from stimulated FBT/ozonation still contained at least 0.5 mg/L of
slowly biodegrading carbon, which was difﬁcult to remove by bioﬁltration. The recycle

mode of operation resulted in a decrease in the concentration of “slow” BDOC in the

232

effluent from the stimulated F BT/ozonation and, consequently, in low bioﬁltration time
required to remove remaining biodegradable organic matter. Thus, the study showed that
adding an easily biodegrading organic carbon in combination with the recycle mode
operation resulted in an increased process efficiency in terms of the removal of organic
matter and biodegradation time.

The study also showed that the effluent from the stimulated F BT/ozonation was
efﬁciently biodegraded even at the substrate concentration below Smin. Further studies are
needed to explain this phenomenon and to better understand the transformation of NOM
during stimulated biodegradation and to determine optimal operational parameters (dose
of a carbon source, recycle ﬂow, EBCT etc.). The measurements of assimilable organic

carbon (AOC) will also be needed to evaluate the regrth potential of treated water.

233

9. ECONOMIC ANALYSIS

In this chapter, the cost of the proposed combined ozonation and FBT process was
compared with the cost of conventional ﬂocculation-sedimentation-ﬁltration processes
with and without ozonation and granular activated adsorption for a design capacity of 1
MGD.

A simpliﬁed schematic of the combined ozonation and F BT process is shown in
Figure 9.1. The treatment train includes pumping water into an FBT column, from where
it ﬂows by gravity into an ozonation tank followed by sand ﬁltration and chlorination. A
portion of the ozonation effluent is recycled back to the F BT column. Values for design
and operating parameters used to estimate the cost of the facility are summarized in Table
9.1.

The annual cost analysis method was used to consolidate capital and operating
cost into a single estimate. The capital recovery factor was calculated using the
amortization period and interest rate given in Table 9.1. The salvage value cost was not
included in cost analysis. The conventional treatment costs were presented in Wiesner et
al. (1994) and were based on the “typical” values observed in practice rather than values
taken from speciﬁc operating facilities. Since these values were based on 1992-1993
prices, these costs were increased by 10 percent to reﬂect an increase in price for the last
5 years. This modest increase was assumed to yield a more cautious estimate of the cost-

effectiveness of the proposed process.

234

 

 

 

 

- FBT Ozonation Biofilratlon Chlorination _.

' Treated
Raw watef

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q“

Figure 9.1. Simpliﬁed schematic of the combined ozonation and FBT process

235

Table 9.1 Values for major cost, design and operating parameters

 

 

 

Parameter Value
Cost parameters

Amortization period, years 20
Annual interest rate, percent 10
Land cost, S/acres not included
Labor, S/hr 30
Electricity, $/kW-hr 0.07
Acetic acid, $/drum 465.60

Design and operating parameters
FBT media
FBT EBCT, min
Acetic acid dose, mg/L
Filter media
Bioﬁlter EBCT, min
Filtration rate, gpm/sq ft
Ozone dose, mg/L
HRT in ozone contactor, min
Design recycle ratio

Operating recycle ratio

 

activated carbon
30
2.5
sand
1 0
3
6'
1 0
2
1

 

Based on TOC concentration in raw water of 6 mglmg C and ozone dose of 1 mg/mg C

 

Compared to “typical” values observed in practice (Weisner et al. 1994), the cost
of ozonation treatment in the proposed process was increased by 60 percent. This was
done to reﬂect a greater anticipated consumption of ozone in the proposed process
compared to that in conventional ozonation facilities. (Typical ozone doses used in
conventional treatment are in the range from 3 to 5 mg/L (Scadsen, 1998; Bellamy et al.
1990).) Since the chlorination cost was not signiﬁcant compared to other costs, this cost
was roughly estimated using the six-tenths rule based on the cost of 50 MGD chlorination
facility (Water Treatment Plant Design, 1998).

The equipment and operating cost data for FBT were provided by EFX Systems,
Inc. (Rajan, 1998). Installation, piping, instrumentation and control, electrical, building,
and engineering and supervision costs were determined using designated percentage
factors to the equipment cost (Water Treatment Plant Design, 1998; Peters and
Tirnmerhaus, 1991). Considering that FBT systems are preengineered systems, the cost
factors for installation, piping, and engineering were reduced to 20, 20 and 10 percent,
respectively. The operating cost was calculated based on using activated carbon as an
FBT medium, rather than a signiﬁcantly cheaper inert material (e.g., sand), which was
used in pilot testing. This was done to yield a more cautious estimate of the cost-
effectiveness of the proposed process and to ensure the feasibility of future studies using
activated carbon.

The bioﬁlter was sized based on a ﬁltration rate of 3 gpm/sq ft (Water Treatment
Plant Design, 1998). It was assumed that a standby F BT pump would be used for ﬁlter

backwashing. The pumps were sized using the data presented in Robinson (1996) and in

237

Crane (1978). The equipment cost for the pump station and bioﬁltration were found in
Process Plant Construction Estimating Standards (1998). Installation, piping,
instrumentation and control, electrical, building, and engineering and supervision costs
were determined using designated percentage factors to the equipment cost (Water
Treatment Plant Design, 1998; Peters and Timmerhaus, 1991).

The calculations of the capital and operating cost of the proposed process are
presented in Appendix B. The data are summarized in Table 9.2. The most signiﬁcant
costs were associated with FBT and ozonation, which suggests that these are the areas in
which optimization efforts should be directed. The economical feasibility of increasing
recycle ratio is presented in Figure 9.2. The cost of treatment by conventional GAC is
given for comparison. When the recycle ratio increased from 2 to 3.5, the total cost of
treatment by the proposed process was estimated to increase by approximately 10
percent. At recycle ratios less than 7, the cost of treatment by FBT was estimated to be
lower than that by GAC. At recycle ratios greater than 6, GAC appears to be
economically more feasible.

Table 9.3 presents the cost of treatment by the combined ozonation and FBT
process (recycle ratio 2) and by conventional processes. The cost of treatment by the
ozonation and FBT process was estimated to be approximately 20 percent greater than
conventional ﬂocculation/sedimentation processes and 40 percent less than
ﬂocculation/sedimentation processes with ozonation and GAC. However, conventional
ﬂocculation/sedimentation processes are not technically feasible for the treatment of
waters having quality similar to that of Huron River water (Scadsen, 1998). The

ozonation and FBT process appears to be technically and economically feasible and

238

should be considered seriously when evaluating the construction of new or retroﬁtting

existing facilities, especially for small systems.

Table 9.2 Cost breakdown for the combined ozonation and F BT process

 

 

 

 

Unit operation Total cost, $/1,000 gal
Pump station 0.13
F BT 1.11
Ozonationl 0.72
Bioﬁltration 0.06
Chlorination 0.05
Total cost 2.07

 

IWiesner, 1994

Table 9.3. Cost of selected treatment systems

 

 

Treatment system Total cost, $/1,000 gal
F locculation/sedimentation + ﬁltration1 1.68
Flocculation/sedimentation + ozonation + GACl 3.55
FBT + ozonation + bioﬁltration 2.07

 

 

 

Wiesner, 1994

239

 

 

 

Cost, $11,000 gallons

 

 

 

 

GAC(no recycle)

 

 

 

 

0.5 ..

 

 

 

0 1 2 3 4 5 6 7 8
Recycle ratio

 

:40—FBT +Totai1

 

Figure 9.2. Effect of recycle on the cost of treatment by combined FBT/ozonation.

240

10. CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE RESEARCH

Section 10.] presents summary of the signiﬁcant results and conclusions from the
experimental results and data analysis. Recommendations for future research are given in

Section 10.2.

10.1. CONCLUSIONS

10.1.1. General Conclusions

The following conclusions highlight the major results of the study presented in
this thesis (particular results are discussed in subsequent sections):

1. A combination of factors affected the removal of organic carbon during ozonation of
Huron River water. These included ozone dose, hydraulic retention time (HRT), and
dissolved ozone concentration.

2. If ozonation is to be incorporated into an ozonation/biodegradation treatment train,
the ozonation should be conducted under conditions that do not result in the
destruction of organic carbon. In this way, ozonation is directed for the production of
biodegradable organic matter, which is removed by biotreatment.

3. An ozone-dose averaging approach for developing the correlations between water
quality surrogate parameters was proposed and shown to be useful in evaluating the

efﬁciency of ozonation and F BT processes.

241

4. The UV-254 absorbance of ozonated Huron River water could be used to determine
the dissolved ozone concentration and the speciﬁc ozone consumption rate.

5. Several biodegradation characteristics were developed and used to quantify the
efﬁciency of the biodegradation of NOM.

6. Combined ozonation and FBT treatment for the control THM precursors in drinking
water was shown to be technically and economically feasible.

7. Adding acetate to the F BT column in combination with the recycle mode of operation
provided a synergetic effect that resulted in an increase in the efﬁciency of the
combined ozonation and FBT process in terms of the ozone consumption and

biodegradation time.

10.1.2. Surrogate Characterization of NOM

The goal of the surrogate characterization study was to establish the relationships
between TOC, UV-254, HS, and THMFP and to determine if any of these parameters
could be used as surrogates for the others during the investigation of the combined
ozonation and F BT process. The following conclusions can be made upon the results of
this study:

1. There existed good correlations between TOC and UV-254 for raw water and
ozonated and biotreated waters. The relationships between these parameters were,

however, different for waters at different treatment stages. For this reason, these

242

parameters could not be substituted for each other when evaluating various treatment
alternatives.

There existed a good correlation between UV-254 and humic substance
concentrations, which did not change for waters at different treatment stages. As
such, the UV-254 could be used as a surrogate for humic substances.

There existed a good correlation between TOC concentration and THMFP, which did
not change for waters at different treatment stages. As such, the TOC concentration
could be used as a surrogate for THMF P.

There existed a good correlation between UV-254 and THMFP, which, however, was
different for waters at different treatment stages. For this reason, the UV-254 could

not be used as a surrogate for THMFP when evaluating various treatment alternatives.

10.1.3. Ozonation Studies

The following conclusions can be made upon the results of ozonation studies

using Huron River water:

1.

Generally, the removal of organic carbon increased with ozone dose, hydraulic
retention time (HRT), and dissolved ozone concentration. At low HRT (less than 10
minutes), ozonation at doses of up to 1 mglmg C resulted in very little destruction of
TOC. At a concentration of dissolved ozone of less than 1 mg/L (normally maintained
in commercial ozonation systems) and HRTs greater than 10 minutes, the removal of
organic carbon increased with ozone dose and leveled off at an ozone dose of

approximately 1-1.5 mglmg C.

243

. For the range of experimental conditions studied, the removal of UV—254-absorbing
compound increased with an increase in ozone dose (up to 5 mglmg C) and was not
affected by HRT (ranging from 7 to 60 minutes).

. Ozonation, conducted under conditions that resulted in the destruction of organic
carbon, did not result in an increased production of low-molecular weight compounds
(less than 1000 daltons) or biodegradable organic matter. When little removal of
organic carbon during ozonation was observed, the concentrations of low molecular
weight compounds and BDOC increased with an increase in ozone dose.

. The ozonation of NOM follows the following pathways: (1) ozone converts humic
substances into nonhumic materials; (2) the destruction of organic carbon occurs
through the reaction of ozone with nonhumic substances; (3) both reactions can

proceed simultaneously.

. The reaction rate constant of the ozone reactions with humic substances was a

function of temperature and the concentration of humic substances and was
represented as the product of two independent terms: temperature- and concentration-
dependent terms (see Equation 6.32). The reaction rate constant of the ozone
reactions with nonhumic substances was a function of temperature only.

. The amount of ozone required for the destruction of nonhumic substances was
approximately two times greater than that required for the oxidation of humic
substances. This suggested that ozonation should be directed towards the oxidation of
essentially nonbiodegradable humic substances rather than potentially biodegradable

nonhumic materials.

244

7. A mathematical model that described the transformation of NOM in terms of the

concentration of humic and nonhumic substances was developed. The kinetic
parameters, including stoichiometric coefficients and activation energies were
determined. The model adequately described the transformation of humic and
nonhumic ﬁ'actions during ozonation. The relative difference between experimental
and calculated data did not exceed 10 percent with an average of i6 percent.

The second model that was tested in this study was based on the speciﬁc ozone
consumption rate. Using the speciﬁc rate of ozone utilization for the evaluation of the
ozonation efficiency was, however, prone to possible misinterpretations. This is
because, using only the speciﬁc ozone consumption rate, it was difficult to separate
the effect of ozonation on the removal of organic carbon from that on the removal of
UV-254-absorbing compounds. For this reason, the model based on the speciﬁc rate
of ozone utilization was less useful than the model based on the concentrations of

humic and nonhumic substances.

10.1.4. Biodegradation Studies

The following conclusions can be made upon the biodegradation studies of raw

and ozonated waters:

1.

The biodegradable organic matter in Huron River water consisted of two fractions:
fast and slowly biodegrading fractions (“fast” and “slow” BDOC). Several
operationally deﬁned biodegradation parameters were introduced in an attempt to

quantify the efﬁcacy of the biodegradation of NOM. These parameters included (1)

245

the maximum rate of the biodegradation of “fast” BDOC (Rmax); (2) the minimum
EBCT that required to eliminate “fast” BDOC (EBCTmin); and (3) the BDOC
concentration that at least remained after biodegradation at EBCTmin (BDOC310w).

2. The removal of “fast” BDOC in the F BT system was greater than that in the
bioﬁltration system. The Rmx in the FBT column was two times greater than that in
the bioﬁltration column. The removal efficiency of “slow” BDOC was essentially the
same for both systems.

3. Increasing temperature resulted in an increase in Rmx and in a decrease in EBCTmin.
The removal of slowly biodegrading organic matter did not appear to be affected by
temperature.

4. Biodegradability of NOM in water increased after ozonation (under conditions when
little destruction of organic carbon occurred). However, ozonated water contained at
least 0.5 mg/L of slowly biodegradable organic carbon, which was difficult to remove
by biotreatment. These slowly biodegrading organic materials can pose the bacterial
regrowth problem in the distribution system if they are not removed in the treatment

system. (The approaches of reducing BDOC,.ow are summarized in Section 10.1.5.)

10.1.5. Pilot-Scale Study of the Combined Ozonation and FBT Process

The following conclusions are from the pilot-scale study of the combined
ozonation and FBT process:
1. At an EBCT of up to 180 minutes and ozone dose of up 2 mglmg C, the

ozonation/FBT process was able to remove up to 50 percent of TOC and humic

246

substances, up to 80 percent of UV-254, up to 70 percent of THMFP, and up to 80
percent of turbidity. The removal of organic matter was comparable to that achieved
at Ann Arbor Water Treatment plant, which included softening,
ﬂocculation/sedimentation, ozonation, and activated carbon adsorption.

. The use of recycle of a portion of the system effluent back to the ozone contactor in
order to increase the removal efﬁciency relative to ozone consumption was
investigated. Recycling at a recycle ratio of 1 resulted in an increase in the removal of
organic matter by 3 percent. Although the difference was statistically signiﬁcant, the
effect was less than expected.

. Using FBT prior to ozonation resulted in an increased production of low molecular
weight compounds and biodegradable organic matter, which was removed by
bioﬁltration. The FBT/ozonation process followed by bioﬁltration was more efficient
that the ozonation/FBT process in terms of the removal of organic matter and
biodegradation time. The effluent from the FBT/ozonation system contained at least
0.5 mg/L of “slow” BDOC, which was difficult to remove by bioﬁltration. Recycle of
a portion of the FBT/ozonation effluent back to the FBT column did not result in
signiﬁcant increase in the reduction of BDOC,.ow and total biodegradation time.

. Adding an easily biodegradable carbon source (acetate) to the FBT column resulted in
an increase in the removal of organic carbon by FBT. The stimulated FBT/ozonation
process resulted in an increased production of “fast” BDOC, which was efficiently
removed by bioﬁltration. The efficiency of the stimulated F BT/ozonation process
followed by bioﬁltration was greater than that of the F BT/ozonation process (no

acetate) followed by bioﬁltration in terms of the removal of organic matter relative to

247

ozone consumption. However, the stimulated FBT/ozonation efﬂuent still contained
at least 0.5 mg/L of slowly biodegrading organic carbon, which was difficult to
remove by bioﬁltration.

5. The recycle mode of operation resulted in a decrease in the concentration of “slow”
BDOC in the effluent from the stimulated FBT/ozonation process and, consequently,
in a decrease in bioﬁltration time.

6. For a design capacity of 1 MGD, the cost of treatment by the stimulated recirculating
FBT/ozonation process followed by bioﬁltration was estimated to be 40 percent less

than that by a ﬂocculation/sedimentation process with ozonation and GAC.

10.2. RECOMMENDATIONS FOR FUTURE RESEARCH

1. The study presented herein included one source water. The applicability of the results
to different sources of water supplies needs to be demonstrated.

2. The Huron River water contained low bromide concentration. As bromide may affect
the kinetics of ozone reaction in water and DBP speciation (Krasner et al., 1996, von
Gunten and Hoigné, 1996), the applicability of the results for water containing high
concentration of bromide needs to be demonstrated.

3. The pilot-scale FBT system was operated as a CSTR. The plug-ﬂow systems should
be investigated. It is expected that the efﬁciency of the efﬁciency of FBT would
increase when operated under plug-ﬂow conditions.

4. Detail study needs to be conducted to better understand the transformation of NOM

during stimulated biodegradation.

248

10.

This study used acetic acid as an easily biodegradable carbon source. Other sources of
easily biodegradable organic carbon need to be investigated for their applicability in
the ozonation and FBT systems. The effect of nutrients on the efﬁciency of stimulated
FBT should also be addressed.

The results of the studies 3 - 6 should serve as basis for optimizing the operational
parameters including the dose of carbon source and recycle ratio.

In this study, the FBT system used an inert material as a support medium. It needs to
be demonstrated if using activated carbon can increase process efﬁciency.

The analysis of assimilable organic carbon as an indicator of the efﬂuent regrowth
potential should be included in future studies. The water also needs to be analyzed for
haloacetic acid formation potential, as haloacetic acids are regulated by US EPA
(Oxenford, 1996).

The mathematical ozonation model and kinetic ozonation parameters, that were
developed in this study, needs to be veriﬁed for its applicability to biodegradation
effluents.

Studies are needed to determine if there exists a correlation between BDOC and the
concentration of nonhumic substances. If such correlation is established, a model that
describes the transformation of NOM during ozonation and FBT (including

recirculating and stimulated modes of operation) can be developed.

249

APPENDICES

250

Appendix A. List of formulas

251

Ozone dose (expressed as milligram of ozone per liter (mg/L»:

G(c:;n — cg")

Q

 

Ozone dose =

where, G and Q are gas and water ﬂow rates, respectively, mL/min; cgin and cg°“t

are the concentration of ozone in the inﬂuent and effluent gas, respectively, mg/L.

Ozone dose (expressed as milligram of ozone per milligram of TOC (mg/mg C)):

 

G cin _ cout
Ozone dose = ( g g )
Q-TOC

where, TOC is the concentration of total organic carbon in the inﬂuent, mg/L.

Gaseous ozone concentration using the spectroscopic method:

 

cg =48,000 A pm
8°bpatm +pgaugc

252

where, C8 is ozone concentration in gas phase, mg/L; A is absorbance; a is the
extinction coefﬁcient, L/mole cm; b is UV pathlength, cm; Palm is atmospheric

pressure, psi; Pgauge is pressure in the UV cell, psig.
EBCT in the F BT system (min):

2
EBCT = 7‘: H

 

where d is the column diameter, cm.; H is the bed height, cm; Q water ﬂow rate,

mL/min.

Bioﬁlter EBCT (min):

EBCT = ellapsed time bed volume

 

sample volume

253

Appendix B. Annual cost analysis

254

Amortization period
Annual interest rate
Capital recovery factor

Design capacity
Capital cost

Operating cost
Annualized capital cost

Annualized operating cost

Total annual cost
Cost per 1000 gallons

Capital cost

Pumps

Installation

Piping

Instrument and control
Electrical

Buildings

Engineering and Superv
Contingency

Total capital

Operating cost
Electricity

Labor
Total operating

Pumps

Pump speciﬁc speed
Pump efﬁciency
Motor efﬁciency
Capacity

Pressure head
Electricity

Work horsepower
BHP

Motor HP

Motor size

Kilowatts

Energy cost per year

W

Table B. 1. Pump station

% of equip

40
3O
20
10
20
10
15

20
0.1
0.1175
1 M60
$ 102,044.00
$ 33,435.79
$ 11,986.05
$ 33,435.79
$ 45,421.84
3 0.12
Quantity Cost
2 $ 19,400.00
Hrlyr Labor/hr
200 $ 30.00
Centrifugal
1800 rpm
0.8
0.80
1 M60
694 GPM
200 ft
$ 0.07 kW hr
35.11
43.89
55
60
44.742
$ 27,435.79

$ 38,800.00
$ 15,520.00
$ 1 1,640.00
$ 7,760.00
$ 3,880.00
$ 7,760.00
$ 3,880.00
$ 12,804.00
3 102,044.00
$ 27,435.79
$ 6,000.00
3 33,435.79

O

H

WH/(pump efﬁciency)

(Q'H'SG)/3956
BHP/(motor efﬁciency)

0.745‘motor HP

Robinson R.N.Chemical Engineering Reference Manual, Profess. Publ, Belmont, Ca, 1996
Crane, Flow of Fluids, Crane Co. 1978

255

Amortization period
Annual interest rate
Capital recovery factor

Design capacity
Capital cost

Operating cost
Annualized capital cost

Annualized operating cost

Total annual cost
Cost per 1000 gallons

Capital cost
Equipment (2 bioﬁlters)

Installation

Piping

Instrument and control
Electrical

Buildings

Contingency

Total capital

Operating cost
Labor

Total operating

Table B.2. Bioﬁltration

% of equip

40
20
20
10
20
15

“6993606999

Note: one of the FBR pumps can be used for backwashing

we
Capacity

EBCT

Filtration rate
Surface area required
Filter bed volume

Filter media depth
Filter diameter

References:

20
0.1
0.1175
1 MGD
74,250.00
15,000.00
8,721.38
15,000.00
23,721.38
0.06
Quantity Cost
2 $ 16,500.00
Hrlyr Labor/hr
500 $ 30.00
1 MGD
694 GPM
10 minutes
3 gpm/sq ft
231 sq ft
6944 gallons
928.4 cu. ft
4.01 ft
9.69 sq ft.

Water Treatment Plant Design. 3d edition. McGraw-Hills, NY, 1990
Process Plant Construction Estimating Standards, Richardson Engineering

Services, Arizona, 1997.

Peters and Timmerhaus. Plant Design and Economics, McGraw Hills,1991

256

“9

33,000.00
13,200.00
6,600.00
6,600.00
3,300.00
6,600.00
4,950.00
74,250.00

15,000.00
15,000.00

Table 3.3. Chlorine Basin

Amortization period
Annual interest rate
Capital recovery factor

Design capacity

Capital cost

Operating cost
Annualized capital cost
Annualized operating cost
Total annual cost

Cost per 1000 gallons

Capital cost % of equip
Chlorine basin

Installation n/a
Piping n/a
Instrument and control n/a
Electrical n/a
Buildings n/a
Engineering and Supervis n/a
Contingency n/a
Total capital

Qpergting cost

Chlorine

Labor
Total operating

W

”$696969“

20
0.1
0.1175

1 MGD

120,000.00
7,825.00
14,095.15
7,825.00
21,920.15
0.06

Quantity
1

Cost/kg
0.25
Hrlyr
200

Cost

$ 120,000.00 $

Quant, kg/day

$

20 $
Labor/hr

30.00 $

5

Water Treatment Plant Design. 3d edition. McGraw-Hills, NY, 1990
Peters and Timmerhaus. Plant Design and Economics, McGraw Hills,1991

257

120,000.00

120,000.00

1,825.00

6,000.00
7,825.00

Amortization period
Annual interest rate
Capital recovery factor

Design capacity

Recycle ratio

Treatment train

Bioﬁlm support medium
Capital cost

Operating cost
Annualized capital cost
Annualized operating cost
Total annual cost

Cost per 1000 gallons

Capital cost

Equipment

Installation

Piping

Instrumentation and control
Electrical

Buildings

Engineering and Supervision
Contingency

Total capital

Operating cost (no acetate)
Acetate cost
Total operating

Acetate

Capacity

Dose

Amount

Cost per kg

Annual cost

Cost per 1000 gallons

References:

Table B.4. FBT (recycle ratio — 2)

20
0.1
0.1175
1
2
dual
activated carbon
3 2,947,650.00
S 57,622.84
$ 346,229.86
$ 57,622.84
$ 403,852.71
5 1.11
% of equip
$ 1,290,000.00
20 $ 258,000.00
20 $ 258,000.00
20 $ 258,000.00
10 $ 129,000.00
20 $ 258,000.00
10 $ 129,000.00
15 $ 367,650.00
3 2,947,650.00
$ 50,000.00
$ 7,622.84
S 57,622.84
1 MGD
2.5 mg/L
9.45 kg/day
$ 2.21
$ 7,622.84
0.02

Rajan R.V. Equipment and Operating Cost for FBT Systems, 1998
Water Treatment Plant Design. 3d edition. McGraw-Hills, NY, 1990
Peters and Timmerhaus. Plant Design and Economics, McGraw Hills,1991

258

MGD

Table B.5. FBT (recycle ratio — 3.5)

Amortization period
Annual interest rate
Capital recovery factor

Design capacity

Recycle ratio

Treatment train

Bioﬁlm support medium
Capital cost

Operating cost
Annualized capital cost
Annualized operating cost
Total annual cost

Cost per 1000 gallons

Capital cost

Equipment

Installation

Piping

Instrumentation and control
Electrical

Buildings

Engineering and Supervision
Contingency

Total capital

Operating cost (no acetate)
Acetate cost
Total operating

Acetate

Capacity

Dose

Amount

Cost per kg

Annual cost

Cost per 1000 gallons

References:

20
0.1
0.1175

1
3.5
dual
activated carbon
3,427,500.00
70,588.35
402,592.86
70,588.35
473,181.21
1 .30

”9’99““

% of equip

20
20
20
10
20
10
1 5

$

3 1,500,000.00
$ 300,000.00
$ 300,000.00
$ 300,000.00
$ 150,000.00
$ 300,000.00
$ 150,000.00
$ 427,500.00
3 3,427,500.00

$

63,000.00
7,588.35
70,588.35

«69

1 MGD
2.5 mglL
9.45 kg/day

2.20

$ 7,588.35

0.02

Rajan R.V. Equipment and Operating Cost for FBT Systems, 1998
Water Treatment Plant Design. 3d edition. McGraw-Hills, NY, 1990
Peters and Timmerhaus. Plant Design and Economics, McGraw Hills,1991

MGD

Table B.6. FBT (recycle ratio — 5)

Amortization period 20
Annual interest rate 0.1
Capital recovery factor 0.1 175
Design capacity 1
Recycle ratio 5
Treatment train dual

Bioﬁlm support medium activated carbon
Capital cost $ 3,884,500.00
Operating cost $ 84,988.35
Annualized capital cost $ 456,271.91
Annualized operating cost $ 84,988.35
Total annual cost $ 541,260.26
Cost per 1000 gallons 3 1.48
Capital cost % of equip

Equipment $ 1,700,000.00
Installation 20 $ 340,000.00
Piping 20 $ 340,000.00
Instrumentation and control 20 $ 340,000.00
Electrical 10 $ 170,000.00
Buildings 20 $ 340,000.00
Engineering and Supervision 10 $ 170,000.00
Contingency 15 $ 484, 500.00
Total capital 3 3,884,500.00
Operating cost (no acetate) $ 77,400.00
Acetate cost $ 7,588.35
Total operating 3 84,988.35
Acetate

Capacity

Dose 1 MGD

Amount 2.5 mglL

Cost per kg 9.45 kglday

Annual cost $ 2.20

Cost per 1000 gallons 3 7,588.35

References:

Rajan R.V. Equipment and Operating Cost for FBT Systems, 1998
Water Treatment Plant Design. 3d edition. McGraw-Hills, NY, 1990
Peters and Timmerhaus. Plant Design and Economics, McGraw Hills,1991

260

MGD

Amortization period
Annual interest rate
Capital recovery factor

Design capacity

Recycle ratio

Treatment train

Bioﬂlm support medium
Capital cost

Operating cost
Annualized capital cost
Annualized operating cost
Total annual cost

Cost per 1000 gallons

Capital cost

Equipment

Installation

Piping

Instrumentation and control
Electrical

Buildings

Engineering and Supervision
Contingency

Total capital

Operating cost (no acetate)
Acetate cost
Total operating

Acetate

Capacity

Dose

Amount

Cost per kg

Annual cost

Cost per 1000 gallons

References:

Table 3.7. FBT (recycle ratio — 7)

dual
activat

20
0.1
0.1175

ed carbon

4,272,950.00

100,338.35

% of equip

20
20
20
10
20
10
15

”99699993

“$96899699’9’69

69

“69

$

1 MGD
2.5 mglL
9.45 kglday

2.20

$ 7,588.35

0.02

Rajan R.V. Equipment and Operating Cost for FBT Systems, 1998
Water Treatment Plant Design. 3d edition. McGraw-Hills, NY, 1990
Peters and Timmerhaus. Plant Design and Economics, McGraw Hills,1991

261

501,899.10
100,338.35
602,237.45

1 .65

1,870,000. 00

374,000. 00
374,000.00
374,000. 00
187,000. 00
374,000.00
187,000.00
532,950.00

4,272,950.00

92,750.00
7,588.35
100,338.35

MGD

Appendix C. Control charts

262

TOC, mg/L

0.3

 

 

 

 

 

 

 

 

O
0.2 1
O
0
O
0.1" ___————-—-———-—_———————
O
O O
O
O
O
0 "’— 1 . T . +——-——+r—. .*——-"——r . 4' r 1
3111197 4130197 6/1 9197 818m °°..912719: 11116197 115199 2124199 4115198 614199
0. 0 :0 .0 .
. L ° 21: ° .
O .0 O. O .
-0.1 e o e e o .0
O 00 .0 O
00 O O 00.0 0 O
00 00 O
00 0 O .0
O O 0
oz , ’
O
-0.3
-0.4
Date

 

[—Mean (l'o.97)— “11-2319. dev(0.1-7)]

 

Figure C.1. TOC standard — 0 mg/L

263

 

2.51

.1.5(

00mg"-

0.5 ~

 

 

0 e
————— L — — — — — — — — — - — — — * — — -
so 0
g’A 0“ 0 ° 0 o ...
a.“ ' . , 0" ‘1" e 03.: “‘00 i 0' I
——————— .— — — -.- —h.‘ — — — — — — — -

 

 

0

Ti

T 1

311/97 4130197 611997 8/8197 927/97 11I16/97 1W 2124M 415%

Date

 

 

 

Mean (2.5) J;- +I-2std acumen

L

 

Figure C.2. TOC standard - 2.5 mg/L

264

 

Sim

 

5.20

 

 

 

 

________L ______________
o e o
o o o o O
5.10 e
o
o e e
o
o o
o o e e
o
5113-4 . o .0 .0 o o. e
e o o
o o o
as
L o o - o ‘ e g
e r' ' 0' e e v
.0 o e. o O
-4911 o o o
e e e
o e
o o
o e o o o e
o o
o o. .
4.80— o
so
0
o
o
o e
_________________ .—
4.701
o
4.60 . v . T 1 . . .
3/11/97 413097 @1997 BMW 927/97 11/1897 115193 224/Q 411519
Date

 

F—Mean (4.94)— -+l-2std dev(0.21)i

|

Figure c.3. TOC standard — 5.0 mg/L

265

TOC,mgl

82

7.8 1

7.6 1

7.4 1

7.2 1

6.8

 

 

s
so
° es 0
o es
0
°, 0 e
‘ e
0
° 0
o . e
, o
’9 so
0 . e
-00
.0
s 0 °
’ o
O o
8‘ , .0
s so
,° 0‘
e
’0
e
s
e
o
e

 

 

T

311 1/97 413097 6/1997 8/8/97 927/97 11l16/97 115199 220% 41/1“

Date

 

E1177 7.133):- - 11;?myt9-Zél

Figure C.4. TOC standard —- 7.5 mg/L

266

 

Sim

Appendix D. Spreadsheet Formats

267

Table D. 1. Spreadsheet format for pilot-scale system data entry

 

 

Water source Huron River, Ann Arbor, MI DATEf J
Date collected

 

 

 

 

 

Mode of operation
Sequence
Recycle
Biostimulant

 

 

 

 

 

 

Operating conditions

FBR bed height Comments:

 

FBR diameter in

Bed volume mUmin
Water flow mL/min
Gas ﬂow mL/min
Recycle mUmin
Line pressure psig
Temperature C
Biostimulant ﬂow mUmin
Biostimulant dose mglL
lnﬂuent gas UV254 1lcm
Influent gas 03 mg/L
RT disolved 03 mg/L
(lnﬂuent TOC) mglL
Ozone dose mgO3/mgC
EBCT min
Water

Parameter RTIFBR

as CaCO3
T , NTU
TOC
UV254 1lcm
HS,

nonHS,
MW500,
CI2
THMFP
CHCI3,
CHC|2Br

 

268

Table D2. Spreadsheet format for bench-scale system data entry

 

 

 

 

 

 

 

 

 

 

 

 

5E EDE EDE 00
E as; 80 35.. *

 

 

 

 

 

35 35 35 35 35 35 50: .35 #435 .35 35 a 35 50: was 35 50:

_mzcoc— m: ooszwzoB 8.09 00» «8.5 a._< In 88 8.2.3213; 88 T9: 82.. 88 «m9:
_ as; 883 use. no so am no 5 no

 

 

 

 

 

 

 

 

 

 

I11 ‘a

 

 

 

 

 

 

 

 

 

 

 

 

50: _3353 39>:

269

 

35 as; :5 a 09.
4E oE:.o> 8.08m

 

 

0. 85>)

 

 

 

 

.p 881.00 oﬁo

 

 

 

 

 

”358800 8.58 35$

 

Table D.3. Spreadsheet format for HS data entry

ID
Date
Date

Treatment
Ozonation
Ret time, min
Ozone
T C
Biofiltration
EBCT min
FBR
EBCT min
T C
to XAD column
TOC
UV-254 1lcm
Column
TOC
NaOH
Column NaOH
DI water
Column DI
Column efﬂuent
Column eluent
HS concentration,
nonHS concentr,
TOC,
HS + nonHS
HS+nonHS
Comments

 

270

Table D4. Spreadsheet format for biodegradability data entry

 

 

 

 

 

 

 

 

 

 

 

 

Water source Start

Date collected Finish

Expen'mental setup Sample descripa’on

Column diameter System 1 efﬂuent

Bed height Mode: Flow/reel mL/min

Bed volume Dose: FBR EBCT;

Filter media Biostimularr Dose: mglL

Flow Inﬂuent TOI UV-254:

Mode of operation FBR effl F BT removal

Sample volume Efﬂuent TOC Ozonation removal

Flow rate Total removal

Single pass contct time Biodeg clmn Addnl biodegradble
Potential removal

 

 

 

Ellaps time EBCT UV-254 Comments
hr min 1lcm 96 removal

 

271

Table D.5. Spreadsheet format for MWD data entry

 

Water source

 

 

Date collected

 

 

System
Sequence
Recycle
Biostimulant

Operating conditions
Water flow

Gas ﬂow

Recycle

Line pressure
Biostimulant ﬂow
Biostimulant dose
Inﬂuent gas UV254
Inﬂuent gas 03

RT disolved O3
(Inﬂuent TOC)
Ozone dose

EBCT

Cumulative d’stn’buﬁon

‘ Mobcular w_eight cutoff

<500

 

1000

 

3000

 

10000

 

30000

 

Sample filtered

 

 

 

Sample unﬁltered

 

Fractions
MWCO
<500

 

 

500-1000

 

1000-3000

 

3000-10000

 

10000-30000

 

>30000

 

 

 

 

 

 

Sample collected

 

Sample processed

 

Sample analyzed

 

 

Comments

 

Samples:

Sample 1:
Sample 2:

Sample 3:

 

 

 

 

272

Table D6. Spreadsheet format for TOC data entry

DATE

 

273

REFERENCES

274

Ahmad, R. 1998. Biological Step Help Meets THM Regs. World Water and Environ.
Engineer. 2:18.

Amirtharajah, A. and O’Melia, CR. 1990. Coagulation Processes: Destabilization,
Mixing, and Flocculation. In Water Quality and Treatment. 4th edition. F .W. Pontius,
ed. McGraw-Hill, Inc., New York, pp. 269-366

Amy, G. L., Kuo, C., and Sierka, R. 1988. Ozonation of Humic Substances: Effect on
Molecular Weight Distribution of Dissolved Organic Carbon and Trihalomethane
Formation Potential. Ozone Sci. and Engnr, 10:39-54.

Amy, G.L., Collins, M.R., Kuo, Cl, and King, PH. 1987. Comparing Gel Permeation
Chromatography and Ultraﬁltration for the Molecular Weight Characterization of
Aquatic Organic Matter. J. A WWA., 79:1 :43-49.

Amy, G.L., Sierka, R.A., Bedessem, J ., Price, D., and Tan, L. 1992. Molecular Size
Distribution of Dissolved Organic Matter. J. A WWA, 84:6:67-75.

Anderson, L.J., Johnson, J .D., and Christrnan, RF. 1986. Extent of Ozone Reaction with
Isolated Aquatic Fulvic Acid. Environ. Sci. & Technol, 20:739-742

Anselmi, G., Lignola, P.G., Raitano, C., and Volpicelli, G. 1985. Ozone Absorption with
Reaction in Benzesulfonic Acid Aqueous Solutions. Chemical Engrg. Sci. 40:7:1033-
1040.

Arora, H., LeChevallier, M.W., and Dixon, KL. 1997. DBP Occurrence Survey. J.
AWWA, 89:6:60-68.

Astarita, G. 1967. Mass Transfer with Chemical Reactions. Elsevier Publishing Co., New
York.

Baboock, D. and Singer, P.1979. Chlorination and Coagulation of Humic and F ulvic Acids.
J. AWWA, 71:3:149-152.

Bablon, G. et al. 1991a. Fundamental Aspects . In Ozone in Water Treatment: Application
and Engineering. Ed. by B. Langlas, D.A. Reckhow and DR. Brink. Lewis Publisher,
Inc., Chelsea. pp. 11-132.

Bablon, G. et al. 1991b. Practical Application of Ozone. In Ozone in Water Treatment:

Application and Engineering. Ed. by B. Langlas, D.A. Reckhow and DR Brink. Lewis
Publisher, Inc., Chelsea.

275

Bellamy, W.D. et al. 1991. Economics of Ozone Systems. In Ozone in Water Treatment:
Application and Engineering. Ed. by B. Langlas, D.A. Reckhow and DR. Brink. Lewis
Publisher, Inc., Chelsea.

Botterro, J.Y. and Bersillon, J.L. 1989. Aluminum and Iron (III) Chemistry: Some
Implications for Organic Substance Removal. In: Aquatic Humic Substances. Inﬂuence
on Fate and Treatment of Pollutants. I.F. Suffet and P. MacCarthy, Eds. American
Chemical Society, Washington, DC. pp. 425-442.

Chadik, RA. and Amy, G.L. 1983. Removing Trihalomethanes Precursors From Various
Natural Waters by Metal Coagulants. J. AWWA, 75:10:532-535.

Cipparone, L.A., Diehl, A.C., and Speitel, GE. 1997. Ozonation and BDOC Removal:
Effect on Water Quality. J. A WWA, 89:2:84-97.

Cleasby, J. L. 1990. Filtration. In Water Quality and Treatment. 4th edition. F.W. Pontius,
ed. McGraw-Hill, Inc., New York, pp. 455-560

Cleasby, J .L. 1991. Source Water Quality and Pretreatment Options for Slow Sand
Filters. In Slow Sand Filtration. Eds. G.S. Logdson. New York: American Society of
Civil Engineers.

Cleasby, J .L., Hilmoe, D.J., Dimitracopoulos, CJ. 1984. Slow Sand and Direct In-line
Filtration of a Surface Water. J. A WWA, 76:12:44-54.

Collins, M.R., Amy, G.L. and Steelink, C. 1986. Molecular Weight Distribution,
Carboxylic Acidity, and Humic Substances Content of Aquatic Organic Matter:
Implications for Removal during Water Treatment. Environ. Sci. Technol,

20:10:1028-1032

Collins, MR. and Vaughan, CW. 1993. Assessing Bioﬁlter Treatability of Natural
Organic Matter. In Proc. 1993 Ann. AWWA Conf, Miami, Florida, pp. 1249-1273.

Crane, 197 8. Flow of Fluids. Crane Co.

Criddle, C.S, Alvarez, K., and MacCarthy, P.L. 1991.Microbial Processes in Porous
Media. In Transport Processes in Porous Media. Eds. J. Bears and M.Y.
Corapcioglu. NATO Advanced Studies Institute Series B: Applied Sciences, V. 202.
Kluwer Academic Publishing Co., Dordrecht, The Netherland, pp. 641-691.

Dankwerts, P.V. 1970. Gas-Liquid Reactions. McGraw-Hill, NewYork.

Dempsey, B.A., Ganho, R.M., and O’Melia, CR. 1984. The Coagulation of Humic
Substances by Means of Alum Salts. J. AWWA, 76:4:141-150.

276

Eaton, A. 1995. Measuring UV-absorbing Organics: a Standard Method. J. A WWA,
87:2:86-90

Edzwald, J .K., Becker, W.C., and Wattier, KL. 1985. Surrogate parameters for
Monitoring Organic Matter and THM Precursors. J. A WWA, 76:4: 120-132.

Flogstad, H and Odergaard, H. 1985. Treatment of Humic Waters by Ozone. Ozone Sci.
Eng, 7:2:121-136.

Fox, K.R., Miltner, G.S., Logsdon, G.S., Dicks, D.L., and Drolet, LP. 1984. Pilot-Plant
Studies of Slow-Rate Filtration. J. A WWA, 76:12:62-68.

Freeman, DH. and Cheung, LS. 1981. A Gel Partition Model for Organic Desorption from
a Pond Sediment. Science (Washington, D. C), 214:790-792.

Georgeson, D.L. and Karimi, AA. 1988. Water Quality Improvements with the Use of
Ozone at the Los Angeles Water Treatment Plant. Ozone Sci. Engrg., 10:3:255-276.

Glaze, W.H., Koga, M, and Cancilla, D. 1989. Ozonation By—Products. Improvement of
an Aqueous-Phase Derivitization Method for the detection of Formaldehyde and
Other Carbonyl Compounds Formed by the Ozonation of Drinking Water. Environ.
Sci. Technol., 23:7:838-847.

Glazer, HT. and Edzwald, J.K. 1979. Coagulation and Direct Filtration of Humic
Substances with Polyethylenirnine. Environ Sci. Technol., 13:3:299-307.

Goel, S., Hozalski, R.M., and Bouwer, E]. 1995. Biodegradation of NOM: Effect of NOM
source and Ozone Dose. J. AWWA, 87:1 :90-105.

Heinzle, E., Geiger, F ., Fahmy, M., and Kut, OM. 1992. Integrated Ozonation-Biotreatment
of Pulp Bleaching Eﬁluents Containing Chlorinated Phenolic Compounds. Biotechnol.
Prog, 8:67-77.

Hickey, RE, Wagner, D. and Mazewski, G. 1991. Treating Contaminated Groundwater
Using a F luidized-Bed Reactor. Remediation, Autumn, 447-459.

Hill, CG. 1977. An Introduction to Chemical Engineering Kinetics and Reactor Design.
New York: John Wiley and Sons, Inc. 1977 p. 36.

Hoigné, J. Characterization of water Quality Criteria for Ozonation Processes. Part 1:
Minimal Set of Analytical Data, 1994. Ozone Sci Engrg. 16:383-389.

Huck, P. M. 1990. Measurements of Biodegradable Organic Carbon and Bacterial
Growth Potential in Drinking Water. J. A WWA. 82:7:78-86.

277

Huck, P.M., Zhang, S., and Price, ML. 1994. BOM Removal During Biological
Treatment: a First-order Model. J. A WA, 83: 12:69-71.

Jackson, J .L., Hong, S., and Summers, RS. 1993. The Use of Ultraﬁltration to
Characterize GAC Breakthrough of Organic Matter in Molecular Size Fractions. In
Proc. 1993 Ann. A WWA Conf, Miami, Florida, pp. 93-107.

J oret, J .C. and Levy, Y. 1988. Rapid Method for Estimating Bioelirninable Organic
Carbon in Water. In Proc. 1988 A WWA Ann Conf, Orlando, Florida.

Kaastrup, E. and Halmo, T.M. 1989. Removal of Aquatic Humus by Ozonation and
Activated-Carbon Adsorption. In Aquatic Humic Substances. Inﬂuence on Fate and
Treatment of Pollutants. I.F. Suffet and P. MacCarthy, Ed. American Chemical Society,
Washington, DC. pp. 697-726.

Kasarabada, A.N. 1997. Use of THMFP, TOC and Chlorine Demand as Control
Parameters for Monitoring the Performance of an Ozonation/FBT System. MS.
Thesis. Michigan State University.

Killops, SD. 1986. Volatile Ozonation Products of Aqueous Humic Material. Water Res,
20:2:153-165.

Klevens, OM. and Collins, MR. 1996. Natural Organic Matter Characterization and
Treatability by Biological Activated Carbon Filtration. In Proc. 1996 ACS Symposium
— Disinfection By-Products and NOM Precursors. Washington, DC, pp. 211-246.

Kolb, B., Auer, M. and Pospisil, P. 1983. Methods for Quantitative Analysis of Volatile
Halocarbons from Aqueous Samples by Equilibrium Headspace Gas Chromatography
with Capillary Colums. J. Chromatography. 279:341-349.

Krasner, S.W., Croué, J-P., Buﬂle, J.. and Perdue, EM. 1996. Three Approaches for
Characterizing NOM. J. A WWA. 88:6:66—79.

Krasner, S.W., Sclirnenty, M.J., Chinn, R., Chowdhury, Z.K., and Owen, D.M. 1996. The
Impact of TOC and Bromide on Chlorination By-product Formation. In Disinfection
By-products in Water Treatment, Eds. R.A. Minear and G.L. Amy. Lewis Publisher,
pp. 59-90.

Krasner, S.W., Sclimenti, M.J., and Coffey, BM. 1993. Testing Biologically Active
Filters for Removing Aldehydes During Ozonation. J. A WWA, 85:5:62-71.

Kruithof, J .C., van der Gaag, M.A., and van der Kooy, D. 1989. Effect of Ozonation and
Chlorination on Humic Substances in Water. In Aquatic Humic Substances. Inﬂuence on
Fate and Treatment of Pollutants. I.F. Suﬂ‘et and P. MacCarthy, Eds. American Chemical
Society, Washington, DC. pp. 663-680.

278

Larose, J ., Grasmick, A., Blondeau, P., Elrnaleh, S., and Legeron, JP. 1982. Elimination of
Humic Materials. Ozone Sci. Eng, 4:2:79-89.

Lawrence, J ., Tosine, H, Onuska, RI, and Comba, ME. 1980. The Ozonation of Natural
Waters: Product Identiﬁcation. Ozone Sci. Eng, 2:1 :55-64.

Lee M.C., Snoeyink, V.L., and Crittenden, J.C. 1981. Activated Carbon Adsorption of
Humic Substances. J. AWWA, 73:8:440-446.

Legube B., Croué, J.P., de Laat, J., and Dore, M. 1989. Ozonation of an Extracted Aquatic
Fulvic Acid: Theoretical and Practical Aspects. Ozone Sci. Eng, 11:1:69-92.

Liao, W., Christrnan, RF, Johnson, J.D., and Millington, BS. 1982. Structural
Characterization of Aquatic Humic Material. Envir. Sci. Technol., 16:7:403-410.

Lykins, B.W. and Clark, RM. 1989. Trihalomethanes Precursor and Total Organic Carbon
Removal by Conventional Treatment and Carbon. In Aquatic Humic Substances.
Inﬂuence on Fate and Treatment of Pollutants. I.F. Suffet and P. MacCarthy, Eds.
American Chemical Society, Washington, DC., pp. 597-622.

MacCarthy, P. and Suffet, I.H. 1989. Introduction to Aquatic Humic Substances. Inﬂuence
on Fate and Treatment of Pollutants. I.F. Suffet and P. MacCarthy, Eds.. American
Chemical Society, Washington, DC. p. xvii-xxx.

Malevialle, J.P., Anselrne, C, and Marsigny, O. 1989. Effect of Humic Substances on
Membrane Process. In Aquatic Humic Substances. Inﬂuence on Fate and Treatment of
Pollutants. I.F. Suﬁ‘et and P. MacCarthy, Eds. American Chemical Society, Washington,
DC. pp. 749-768.

Malley, J.P., Eighmy, T.T., Collins, MR, Royce, J.A., and Morgan, DP. 1993. The
Performance and Microbiology of Ozone-Enhanced Biological Filtration. J. A WWA,
85:12:47-57.

Manahan, SE. 1991. Environmental Chemistry. Fiﬁh Ed., Lewis Publisher, Chelsea, MI.
pp. 83-92.

Manahan, SE. 1989. Interaction of Hazardous-Waste Chemicals with Humic Acids. In
Aquatic Humic Substances. Inﬂuence on Fate and Treatment of Pollutants. I.F. Suffet
and P. MacCarthy, Eds. American Chemical Society, Washington, DC, pp. 83-92.

Masten, SJ. 1991. Ozonation of VOC’s in the Presence of Humic Acid and Soils. Ozone:
Sci.Engrg, 13:3: 287-312.

McCreary, J .J . and Snoeyink, V.L. 1980. Characterization and Activated Carbon Adsorption
of Several Humic Substances. Water Res, 14:151-160.

279

Meijers, AP. 1976. Quality Aspect of Ozonation. Water Res, 10:647-656.

Mellema, J. 1998. The use of AMWD for the Evaluation of Ozonation and Biotreatment
Processes. MS. Thesis. Michigan State University.

Mogren, E.M., Scarpino, P., and Summers, RS. 1990. Measurements of Biodegradable
Dissolved Organic Carbon in Drinking Water In Proc. 1990 A WWA Ann Conf,
Cincinnati, Ohio.

Moore, J .W. and Pearson, RG. 1981. Kinetics and Mechanisms: A Study of Homogeneous
Chemical Reactions, 3rd ed. New York: Wiley-Interscience, pp. 68-70.

Morris, J .K. and Knocke, W.R. 1984. Temperature Effects on the Use of Metal Ion
Coagulants for Water Treatment. J. A WWA, 76:3:74-79.

Najm, I.N., Patania, N.L., Jacangelo, J .G., and Krasner, SW. 1994. Evaluating
Surrogates for Disinfection By-products. Jour. A WWA, 86:6:98-106.

National Research Council, 1997. Safe Water from Every Tap. Improving Water Service
to Small Communities. National Academy Press, Washington, DC.

Oxenford, J. L. 1996. Disinfection By-products: Current Practice and Future Directions.
In Disaﬂection By-products in Water Treatment, Eds. R.A. Minear and G.L. Amy.
Lewis Publisher, pp. 3-16.

Peters, MS. and Tirnmerhaus, KD. 1991. Plant Design and Economics for Chemical
Engineers. McGraw-Hill. New York.

Pirbazary, M., Ravindran, V., Wong, S-P., and Stevens, MR 1989. Adsorption of
Micropollutants on Activated Carbon. In Aquatic Humic Substances. Inﬂuence on Fate
and Treatment of Pollutants. I.F. Suffet and P. MacCarthy, Eds. American Chemical
Society, Washington, DC. pp. 549-578.

Pontius, W.F. 1995. An Update of the Federal Drinking Water Regs. J. AWWA, 87:2:48-58.

Price, M.L., Bailey, RW., Enos, A.K., Hook, M., and Hermanowicz, W. 1993. Evaluation
of Ozonation/Biological Treatment for Disinfection Byproducts Control and Biologically
Stable Water. Ozone Sci. Engr., 15:95-130.

Process Plant Construction Estimating Standards. 1998. Richardson Engineering Services,
Mesa, Arizona.

Rachwal, A.J., Foster, D.M., and Holmes, M. 1992. Combining Ozone/Advanced

Oxidation and Biological Filtration Processes for Organics Removal from Water. In
Proc. 1992 A WWA Confer., Toronto, Ontario, 543-559.

280

Rajan, R.V. 1998. Capital and Operating Costs of FBT Systems. EFX Systems, Inc.

Reckhow, DA. and Singer, RC. 1990. Chlorination By-Products in Drinking Waters:
From Formation Potentials to Finished Water Concentrations. Jour. A WWA,
82:4:173-180.

Reckhow, D.A., Xie, Y., McEnroe, R., Byrnes, P., Tobiason, J .E., and Switzenbaum,
MS. 1993. The Use of Chemical Surrogates for Assimilable Organic Carbon. Proc.
A WWA Annual Conference, Miami, Florida.

Reckhow, D.A., Legube, B., and Singer, RC. 1986. The Ozonation of Organic Halide
Precursor: Effect of Bicarbonate. Water Res, 20:8:987-998.

Rittmann, BE. and McCarthy, PL. 1980. Model of Steady-State Bioﬁlm Kinetics. Biotech.
and Bioengrg , 22:2343-2357.

Rittrnann, BE. and J .E. Woolschlager, III, 1996. Fundamental Components for Modeling
BOM and Regrowth. In Proc. 4'" International BOM Conference, Waterloo, Canada.
p. 66-72.

Robinson, NR. 1996. Chemical Engineering Reference Manual.Professional
Publications. Belmont, CA.

Roth, J.A. and Sullivan, DE. 1981. Solubility of Ozone in Water. Ind. Eng. Chem.
Fundam, 20:2:140-144.

Scadsen, J. 1997. 1996 Laboratory Report. Environmental Health Laboratories.

Scadsen, J. 1998. Personal Communication.

Schnitzer, M. and Khan, S.U. 1972. Humic Substances in the Environment, New York:
Marcel Dekker, Inc.

Semmens, M.J. and Field, T.K. 1980. Coagulation: Experience in Organic Removal. J.
A WWA, 72:8:476-482.

Servais, P., Billen, G., and Hascoet, M. 1987. Determination of Biodegradable Fraction
of Dissolved Organic Matter in Waters. Water Res, 21 :4z445-450.

Sheffer, S. and Esterson, G.L. 1982. Mass Transfer and reaction Kinetics in the
Ozone/Tap Water System. Water Res, 16:383-3 89.

Shukairy, HM. and Summers, RS. 1992. The Impact of Preozonation and Biodegradation
on Disinfection By-product Formation. Water Res, 26:9: 1217-1227.

281

Shukairy, HM. and Summers, RS. 1996. DBP speciation and Kinetics as Affected by
Ozonation and Biotreatrnent. In Disinfection By-products in Water Treatment, Eds.
RA. Minear and G.L. Amy. Lewis Publisher, pp. 311-338.

Singer, RC. 1989. Impact of Ozonation on the Removal of Particles, TOC, and THM
Precursors. A WWA Research Foundation Report.

Singer, RC. and Chang, SD. 1989. Correlation Between Trihalomethane and Total
Organic Halides Formed During Water Treatment. J. A WA, 81 :61-69.

Sontheimer, H. and Hubele, C. 1987. The Use of Ozone and Granular Activated Carbon
in Drinking Water Treatment. In Treatment of Drinking Water for Organic
Contaminants. Eds. P.M. Huck and P. Toft. Pergamon Press, New York, pp. 45-66.

Speece, R.E., Madrid, M., and Needhan, K. Downﬂow Bubble Ozone Aeration. 1971. J.
Sanitary Engineering Division. Proc. Of the American Society of Civil Engineers, V.
97, No. SA4, 433-441.

Speitel, G.E., Symons, J .M., Diehl, A.C., Sorensen, H.W., and Cipparone, LA. 1993.
Effect of Ozone Dosage and Subsequent Biodegradation on Removal of DBP
Precursors. J. AWWA, 85:5:86-95.

Staehelin, J. and Hoigné, J. 1985. Decomposition of Ozone in Water in the Presence of
Organic Solutes Acting as Promoters and Inhibitors of Radical Chain Reactions.
Environ. Sci. Technol., 19: 12: 1206-121 3.

Standard Methods for the Examination of Water and Wastewater. 18th ed. 1992. Greenberg,
A.E., Clesceri, LS, and Eaton, A.D., Eds, APHA, AWWA, WEF.

Stevens, A.A., Moore, L.A., Slocum, C.J., Smith, B.L., Seeger, DR. and Ireland, J .C. 1989.
Chlorinated Humic Acid Mixtures: Criteria for Detection of Disinfection Byproducts in
Drinking Water. In Aquatic Humic Substances. Inﬂuence on Fate and Treatment of
Pollutants. I.F. Suffet and P. MacCarthy, Eds. American Chemical Society, Washington,
DC. pp. 681-695.

Thurman, EM. 1985. Humic Substances in Groundwater. Humic Substances in Soil,
Sediment and Water. G.R Aiken, D.M. McKnight, RB. Wershaw, and P. MacCarthy,
Eds. New York: John Wiley and Sons. pp. 87-103.

Thurman, EM. and Malcolm, R.L. 1981. Preparative Isolation of Aquatic Humic
Substances. Envir. Sci. Technol., 15:4:463-466.

Trussel, RR. and Umphres, MD. 1978. The formation of Trihalomethanes. J. AWWA.
70:11: 604-612.

US. EPA Drinking Water Regulations. 1997. Ozone News, 25:4:10-19.

282

US. EPA, 1993. Technical and Economic Capacity of States and Public Water Systems
to Implement Drinking Water Regulations: Report to Congress. EPA 810/R-93/001,
Sept. 1993.

van der Kooij, D. 1992. Assimilable Organic Carbon as an Indicator of Bacterial
Regrowth. J. AWWA, 84:2:57-65.

van der Kooij, D., Visser, A. and Hijnen, W.A.M. 1982. Determining the Concentration
of Easily Assimilable Organic Carbon in Drinking Water. Jour. A WWA, 74:10:540-
545.

Ventresque, C., Bablon, G., and Jadas-Hécart, A. 1990. Ozone: A Means of Stimulating
Biological Activated Carbon Reactors. Ozone Sci. Engnr., 13:91-107.

Veenstra, J.N., Barber, J.B., and Khan, P.A.1983. Ozonation: Its Effect on the Apparent
Molecular Weight of Naturally Occurring Organics and Trihalomethane Production.
Ozone Sci. Eng, 5:4:225-244.

Vik, EA. and Eikebrokk, B. 1989. Coagulation Process for Removal of Humic Substances
from Drinking Water. In Aquatic Humic Substances. Inﬂuence on Fate and Treatment of
Pollutants. I.F. Suffet and P. MacCarthy, Eds. American Chemical Society, Washington,
DC, pp. 385-408.

Voice, T.C., Pak, D., Zhao, X., Shi, J ., and Hickey, RF. 1992. Biological Activated
Carbon in Fluidized Bed Reactors for the Treatment of Groundwater Contaminated
with Volatile Aromatic Hydrocarbons. Water Res, 26:10:1389—1401.

Volk, C., Roche, P., Renner, C., Paiullard, H., and Joret, J.C. 1993. Effects of Ozone-
Hydrogen Peroxide Combination on the Formation of Biodegradable Organic Carbon.
Ozone Sci Engrg. 15:405-418.

von Gunten, U. and Hoigné, J. 196. Ozonation of Bromide-Containing Waters: Bromate
Formation through Ozone and Hydroxyl Radicals. In Disinfection By-products in
Water Treatment, Eds. R.A. Minear and G.L. Amy. Lewis Publisher, pp. 187-206.

Water Treatment Plant Design, 1990. McGraw-Hill. New York

Watts, CD. 1985. Organic By-products of Ozonation of Humic and Fulvic Acid Proc. Intl.
Conf, Role of Ozone in Water and Wastewater Treatment. R Perry and RE. McIntyre,
Eds. Selper Ltd., London, England.

Weisner, M.R., Hackney, J ., Sethi, S., Jacangelo, J .G., and Lainé, J .-M., 1994. Cost

Estimates for Membrane Filtration and Conventional Treatment. J. A WWA, 86:12:33-
41.

283

Yurteri, C. and Gurol, MD. 1988. Evaluation of Kinetic Parameters for the Ozonation of
Organic Micropollutants. Water Res, 22:465-476.

284

 

"111111111111111“