&EPA
United States
Environmental Protection
Agency
Municipal Environmental Research EPA 600 2-80-1 30a
Laboratory August 1980
Cincinnati OH 45268
Research and Development
Removing Potential
Organic Carcinogens
and Precursors from
Drinking Water
Volume I
and
Appendix A
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-80-130a
August 1980
REMOVING POTENTIAL ORGANIC CARCINOGENS AND
PRECURSORS FROM DRINKING WATER
Volume I and Appendix A
by
Paul R. Wood - Principal Investigator
Daniel F. Jackson
Drinking Water Quality Research Center
Florida International University
Miami, Florida 33199
James A. Gervers
Doris H. Waddell
Miami-Dade Water and Sewer Authority
Louis Kaplan
Dade County Department of Public Health
Miami, Florida 33125
Grant No. R804521-01
Project Officer
Jack DeMarco
Drinking Water Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental
Research Laboratory, U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the U.S.
Environmental Protection Agency, nor does mention of trade names
or commercial products constitute endorsement or recommendation
for use.
ii
-------
FOREWORD
The U.S. Environmental Protection Agency was created be-
cause of increasing public and government concern about the
dangers of pollution to the health and welfare of the American
people. Noxious air, foul water, and spoiled land are tragic
testimonies to the deterioration of our natural environment.
The complexity of that environment and the interplay of its
components require a concentrated and integrated attack on the
problem.
Research and development is that necessary first step in
problem solution; it involves defining the problem, measuring
its impact, and searching for solutions. The Municipal Envir-
onmental Research Laboratory develops new and improved tech-
nology and systems to prevent, treat, and manage wastewater and
solid and hazardous waste pollutant discharges from municipal
and community sources, to preserve and treat public drinking
water supplies, and to minimize the adverse economic, social,
health, and aesthetic effects of pollution. This publication
is one of the products of that research and provides a most
vital communications,link between the researcher and the user
community.
To protect the consumer of public drinking water, this
study was undertaken to develop feasible and economical metho-
dology for reducing the amount of specific organic contaminants
in drinking water.
Francis T. Mayo, Director
Municipal Environmental Research
Laboratory
111
-------
ABSTRACT
The principle objective of the two-year Research Project
was to devise feasible and economical methodology for removing
existing organic contaminants from and preventing development of
potential carcinogens in the public water supplies in Bade
County, Florida. Specifically, development of methodology to
reduce the amount of four trihalomethanes (chloroform, bromo-
dichloromethane, chlorodibromomethane, and bromoform) present in
drinking water was the prime objective.
A four-phase study was designed to evaluate the efficiency
of three adsorbents in removing 19 individual halogenated
organics and trihalomethane precursors. These adsorbents were
XE-34Q—a carbonized polymeric macroreticular resin; IRA-904—
a strong base cationic resin designed to remove large molecular
weight substances such as precursors from water; and granular
activated carbon (GAC). Adsorbent columns were placed at
various stages in the water processing system; i.e., the raw
water stage, the lime softened stage at the up-flow Hydrotreator
effluent and the finished water stage.
Four GAC Filtrasorb 400 columns, each 0.76 meters (2.5 feet)
deep, arranged in series on the finished water line were most
effective in reducing the level of the trihalomethanes present
in the finished water that would be consumed by the public.
The Polanyi-Manes Theory of adsorption was applied and
found helpful in interpreting results. Preliminary studies were
made of the bacterial profile of the Preston Water Treatment
Plant, raw and finished water, and effluent from four GAC
columns. Distribution system samples were also analyzed.
This report was submitted in fulfillment of Grant No.
R804521-01 by Dade County Department of Public Health under the
Sponsorship of the U.S. Environmental Protection Agency. This
report covers a period from June 1976 to June 1980, and work was
„ completed as of May 1980.
iv
-------
CONTENTS
Disclaimer ii
Foreword iii
Abstract iv
Figures viii
Tables xviii
List of Abbreviations xx
Acknowledgments xxi
Section I. Introduction 1
Section II. Conclusions 3
Section III. Recommendations 9
Section IV. Plant and equipment description 11
Preston Plant water source 11
Preston Plant site 11
Bench scale adsorption test unit 13
Section V. Methods and procedures 19
Operation of bench scale adsorption unit 19
GC analytical method 19
TOC analysis. 21
THM, Terminal THM and THM FP 21
Data analysis 24
Section VI. Experimental Plan 26
EDI 26
v
-------
ED1R 28
ED2 28
EDS 28
ED4 28
Section VII. Results and discussion 32
Full slcale plant studies 32
Specific halogenated organics 32
Raw water source
Hydrotreator effluent source 36
Finished water source 38
TOC and THM FP organics 40
Raw water source 40
H.T. water source 40
Finished water source 42
Other parameters 48
Rainfall and chlorides 48
Rainfall and TOC 48
pH - 48
Turbidity 49
Color 49
Bench scale studies 50
Specific halogenated organics 50
Raw water source 50
H.T. water source 76
THM 76
Finished water source 105
Adsorption by XE-340 187
Adsorptive capacity and competitive
adsorption 188
TOC and THM FP organics 196
Raw water source 208
H.T. water source 220
Finished water source 220
Other parameters 246
Chlorine 246
Turbidity 253
^ Color 253
pH 253
vx
-------
Comparison of laboratory and distribution
system aging 253
Total THM growth in adsorbent
colunn effluents 256
Bed life criteria in deep GAG columns 262
Relationship of TOC and THM FP data 268
Leaching study on XE-340 resin column 270
Biological activated carbon (BAG) 281
Polanyi-Manes Adsorption Theory 283
Theory development 283
Theory application 290
GC/MS HOC confirmation data 309
References 311
Appendix A. Frances Parsons
Part I. Microbial Flora of Granulated Activated
Carbon Columns Used in Water Treatment 314
Part II. Chlorination of Granulated Activated
Carbon (GAG) Column Effluent to Control
Bacteria 349
VI1
-------
FIGURES
Number Page
1 Flow diagram of John E. Preston Water
Treatment Plant -1-2
2 Bench Scale Column Adsorption Unit 16
3 Plumbing for adsorption column 17
4 Detailed view of column fittings 18
5 Backwash system for columns 20
6 Typical chromatogram of halogenated organics 22
7 Flow diagram of Bench Scale Adsorption Unit for
GAG and XE-340 study in EDI 27
8 Flow diagram of Bench Scale Adsorption Unit for
leaching study in ED2 29
9 Flow diagram of Bench Scale Adsorption Unit for GAC
and IRA-904 resin study in EDS 30
10 Flow diagram of Bench Scale Adsorption Unit for deep
bed study in ED4 31
11^13 TOG in raw, Hydrotreater and finished water 44-46
14 THM FP in raw water and removal by lime softening.. 47
15-18 cis 1,2-Dichloroethene in raw water and removal
by 0.76 meter (2.5 feet) of GAC and 0.76
meters (2.5 feet) of XE-340 52-55
19 Vinyl chloride in raw water and removal by 0.76
meter (2.5 feet) GAC and 0.76 meters
(2.5 feet) XE--340 57
20 Vinyl chloride in raw water and removal by 0.76
and 1.52 meters (2.5 and 5 feet) of IRA-904
resin 58
Vlll
-------
Number
Page
21 trans 1,2-Dichloroethene in raw water and removal
by 0.76 meter (2.5 feet) of GAC 60
22 trans 1,2-Dichloroethene in raw water and removal
by 0.76 meter (2.5 feet) of XE-340 61
23 1,1-Dichloroethane in raw water and removal by
0.76 meter (2.5 feet) of GAC and 0.76 meters
(2.5 feet) of XE-340 63
24 1,1,1-Trichloroethane, 1,2-dichloroethane and
carbon tetrachloride in raw water and removal
by 0.76 meter (2.5 feet) of GAC 66
25 1,1,1-Trichloroethane, 1,2-dichloroethane and
carbon tetrachloride in raw water and removal
by 0.76 meter (2.5 feet) of XE-340 67
26 Trichloroethylene in raw water and removal by
0.76 meter (2.5 feet) of GAC 69
27 Trichloroethylene in raw water and removal by
0.76 meter (2.5 feet) of XE-340 70
28 Tetrachloroethylene in raw water and removal by
0.76 meter (2.5 feet) of GAC and 0.76
meter (2.5 feet) of XE-340 71
29 Chlorobenzene in raw water and removal by 0.76
meter (2.5 feet) of GAC and 0.76 meter
(2.5 feet) of XE-340 73
30 p-Chlorotoluene in raw water and removal by 0.76
meter (2.5 feet) of GAC and 0.76 meter
(2.5 feet) of XE-340 74
31 o, m and p-Dichlorobenzene in raw water and removal
by 0.76 (2.5 feet) of GAC and 0.76 meter
(2.5 feet) of XE-340 75
32-33 .Chloroform in H.T. water and removal by 0.76
meter (2.5 feet) XE-340 79-80
34-35 Bromodichloromethane in H.T. water and removal
by 0.76 meter (2.5 feet) XE-340 81-82
36-37 Chlorodibromomethane in H.T. water and removal*
by 0.76 meter (2.5 feet) XE-340 83-84
38-39 Bromoform in H.T. water and removal by 0.76
meter (2.5 feet) XE-340 85-86
ix
-------
Number
40-41 cis 1,2-Dichloroethene in H.T. water and
removal by 0.76 meter (2.5 feet) XE-340 ..... 88-89
42 cis 1,2-Dichloroethene in H.T. water and
removal by 0.76 meter (2.5 feet) IRA-904
on
resin .......................................... au
43 Vinyl chloride in H.T. water and removal by
0.76 meter (2. 5 feet) of XE-340 ............... 92
44 trans 1,2-Dichloroethene in H.T. water and
removal by 0.76 meter (2.5 feet) of XE-340 ---- 93
45 1,1-Dichloroe thane in H.T. water and removal by
0.76 meter (2. 5 feet) XE-340 .................. 95
46 1,1,1-Trichloroethane, 1,2-dichloroethane and
carbon tetrachloride in H.T. water and
removal by 0.76 meter (2.5 feet) XE-340 ...... 97
47 Trichloroethylene in H.T. water and removal by
0.76 meter (2. 5 feet) of XE-340 ............... 99
48 Tetrachloroethylene in H.T. water and removal
by 0.76 meter (2.5 feet) of XE-340 ............ IOC
49 Chlorobenzene in H.T. water and removal by 0.76
meter (2.5 feet) of XE-340 .................... 101
50 p-Chlorotoluene in H.T. water and removal by
0.76 meter (2.5 feet) of XE-340 ............... 102
51 o, m and p-Dichlorobenzene in H.T. water and
removal by 0.76 meter (2.5 feet) of XE-340.... 104
52-54 Chloroform in finished water and removal by
0.76 meter (2.5 feet) of XE-340 ........... 107-109
55 Chloroform in finished water and finished water thru
0.76 meter (2.5 feet) of IRA-904 resin and
0.76 meter (2.5 feet) of GAC .................. 109
56 Chloroform in finished water and in the effluent from
0.76, 1.52, 2.29 and 3.05 meters (2.5, 5, 7.5
and 10 feet) of GAC ............................ 11C
57-59 cis 1,2-Dichloroethene in finished water and
removal by 0.76 meter (2.5 feet) of XE-340. 115-117
x
-------
60 cis 1,2-Dichloroethene in finished water and
removal by 0.76 meter (2.5 feet) of GAG
and 0.76 (2.5 feet) of IRA-904 resin 118
61 cis 1,2-Dichloroethene in finished water and
finished water thru 0.76, 1.52, 2.29 and
3.05 meters (2.5, 5, 7.5 and 10 feet) of GAG... 119
62-64 Bromodichloromethane in finished water and
removal by 0.76 meter (2.5 feet) of XE-340.121-123
65 Bromodichloromethane in finished water and
removal by 0.76 meter (2.5 feet) of GAG and
0.76 meter (2.5 feet) of IRA-904 resin 124
66 Bromodichloromethane in finished water and removal
by 0.76, 1.52, 2.29 and 3.05 meters (2.5, 5,
7.5 and 10 feet) of GAG 125
67-69 Chlorodibromomethane in finished water and removal
by 0.76 meter (2.5 feet) of XE-340 128-130
70 Chlorodibromomethane in finished water and removal
by 0.76 meter (2.5 feet) of IRA-904 resin
and 0.76 meter (2.5 feet) of GAG 131
71 Chlorodibromomethane in finished water and removal by
0.76, 1.52, 2.29 and 3.05 meters (2.5, 5,
7.5 and 10 feet) of GAG 132
72-74 Bromoform in finished water and removal by 0.76
meter (2.5 feet) of XE-340 134-136
75 Bromoform in finished water and removal by 0.76
meter (2.5 feet) of GAC and 0.76 meters
(2.5 feet) of IRA-904 resin 137
76 Bromoform in finished water and removal by 0.76,
1.52, 2.29 and 3.05 meters (2.5, 5, 7.5 and
10 feet) of GAC 138
;
77-78 Vinyl chloride in finished water and removal by
0.76 meter (2.'5 feet) of XE-340 140-141
79 Vinyl chloride in finished water and removal by
0.76 meter (2.5 feet) of IRA-904 resin and
0.76 meter (2.5 feet) of GAC 142
80 Vinyl chloride in finished water and removal by
0.76 and 1.52 meters (2.5 and 5 feet) of GAC.. 143
xi
-------
Number Page
81 Vinyl chloride in finished water and removal by
2.29 and 3.05 meters (7.5 and 10 feet) of
GAG 144
82-83 trans 1,2^-Dichloroethene in finished water and
removal by 0.76 meter (2.5 feet) of XE-340.147-148
84 trans 1,2-Dichloroethene in finished water and
removal by 0.76 meter (2.5 feet) of GAG
and 0.76 (2.5 feet) of IRA-904 resin 149
85 trans 1,2-Dichloroethene in finished water and
removal by 0.76, 1.52, 2.29 and 3.05 meters
(2.5, 5, 7.5 and 10 feet) of GAG 150
86-87 1,1-Dichloroethane in finished water and removal
by 0.76 meter (2.5 feet) of XE-340 153-154
88 1,1-Dichloroethane in finished water and removal
by 0.76 meter (2.5 feet) of GAG and 0.76
meter (2.5 feet) IRA-904 resin 155
89 1,1-Dichloroethane in finished water and removal
by 0.76 and 1.52 meters (2.5 and 5 feet) of
GAG 156
90 1,1-Dichloroethane in finished water and removal
by 2.29 and 3.05 meters (7.5 and 10 feet) GAG.. 157
91-92 1,1,1-Trichloroethane, 1,2-dichloroethane and carbon
tetrachloride in finished water and removal
by 0.76 meter (2.5 feet) of XE-340 159-160
93 1,1,1-Trichloroethane, 1,2-dichloroethane and
carbon tetrachloride in finished water and
removal by 0.76 meter (2.5 feet) of GAG and
0.76 meter (2.5 feet) of IRA-904 resin 161
94 1,1,1-Trichloroethane, 1-2-dichloroethane and
carbon tetrachloride in finished water and
removal by 0.76, 1.52, 2.29 and 3.05 meters
(2.5, 5, 7.5 an.d 10 feet) of GAG 162
95-96 Trichloroethylene in finished water and removal
by 0.76 meter (2.5 feet) of XE-340 165-166
97 Trichloroethylene in finished water and removal
by 0.76 meter (2.5 feet) of IRA-904 resin.... 167
XI1
-------
Number Page
98-99 Trichloroethylene in finished water and removal
by 0.76 meter (2.5 feet) of GAC 168-169
100 Trichloroethylene in finished water and removal
by 1.52 meters (5 feet) of GAC 170
101 Trichloroethylene in finished water and removal
by 2.29 meters (7.5 feet) of GAC 171
102 Trichloroethylene in finished water and removal
by 3.05 meters (10 feet) of GAC 172
103 Tetrachloroethylene in finished water and removal
by 0.76 meter (2.5 feet) of XE-340 174
104-105 Chlorobenzene in finished water and removal by
0.76 meter (2.5 feet) of XE-340 176-177
106 Chlorobenzene in finished water and removal by 0.76
meter, (2.5 feet) of GAC and 0.76 meter
(2.5 feet) of IRA-904 resin 178
107 Chlorobenzene in finished water and removal by
0.76, 1.52, 2.29 and 3.05 meters (2.5, 5,
7.5 and 10 feet) of GAC 179
108-109 o, m and p-Dichlorobenzene in finished water and
.removal by 0.76 meter (2.5 feet) of XE-340.182-183
110 o, m and p-Dichlorobenzene in finished water and
removal by 0.76 meter (2.5 feet) of IRA-904
resin 184
111 p-Dichlorobenzene in finished water and removal
by 0.76, 1.52, 2.29 and 3.05 meters (2.5, 5,
7.5 and 10 feet) of GAC 185
112 o-Dichlorobenzene in finished water and removal
by 0.76, 1.52, 2.29 and 3.05 meters (2.5, 5,
7.5 and 10 feet) of GAC 186
113 Cubic centimeters adsorbed by each GAC column
for all halogenated compounds added together... 189
114 Cubic centimeters of total HOC entering and
adsorbed by GAC column #1, 2, 3 and 4 in
122 days 190
115 Cubic centimeters of total HOC adsorbed per
100 grams of GAC in GAC columns ttl, 2, 3 and
4 in 122 days 191
xiii
-------
Number Page
116 Adsorption wave front defined by breakthrough
and saturation time for HOC and Type II and
Type III substances thru 0.76 meter
(2.5 feet) of GAG 193
117 THM FP in finished water and removal by 0.76, 1.52,
2.29 and 3.05 meters (2.5, 5, 7.5 and 10
feet) of GAG 198
118 TOG in finished water and removal by 0.76, 1.52, 2.29
and 3.05 meters (2.5, 5, 7.5 and 10 feet) of
GAG 199
119 THM FP in finished water and removal by 0.76
meter (2.5 feet) of GAG 200
120 THM FP in finished water and removal by 1.52
meters ( 5 feet) of GAG 201
121 THM FP in finished water and removal by 2.29
meters (7.5 feet) of GAG 202
122 THM FP in finished water and removal by 3.05
meters (10 feet) of GAG 203
123 THM FP in raw water and removal by 0.76 and
1.52 meters (2.5 and 5 feet) of IRA-904 resin. 204
124 Test extention data for THM FP removal by 0.76,
1.52, 2.29 and 3.05 meters (2.5, 5, 7.5 and
10 feet) of GAG 207
125-126 THM FP in raw water and removal by 0.76 meter
(2.5 feet) of GAG and 0.76 meter (2.5
feet) of XE-340 211-212
127 THM FP in raw water and removal by 0.76 and 1.52
meters (2.5 and 5 feet) of IRA-904 resin 213
128 TOG in raw water and removal by 0.76 meter
(2.5 feet) of GAG and 0.76 meter (2.5 feet)
of XE-340 218
129 TOG in raw water and removal by 0.76 and 1.52
meters (2.5 and 5 feet) of IRA-904 resin 219
130 THM FP in raw water and removal by lime softening
and by 0.76 meter (2.5 feet) of XE-340 on
H.T. water 223
xiv
-------
Number
131 THM FP in raw water and removal by lime
softening and by 0.76 meter (2.5 feet)
XE-340 on H.T. water 224
132 THM FP in raw water and removal by lime
softening and by 0.76 meter (2.5 feet)
of IRA-904 resin on H.T. water 225
133 TOC in H.T. water and removal by 0.76 meter
(2.5 feet) of XE-340 228
134 TOC in H.T. water and removal by 0.76 meter
(2.5 feet) of IRA-904 resin 229
135-136 THM FP in finished water and removal by 0.76
meter (2.5 feet) of XE-340 232-233
137 THM FP in finished water and removal by 0.76
meter (2.5 feet) of GAG and 0.76 meter
(2.5 feet) of IRA-904 resin 234
138 THM FP in finished water and removal by 0.76, 1.52
2.29 and 3.05 meters (2.5, 5, 7.5 and
10 feet) of GAG 235
139 THM FP in finished water and removal by 0.76
meter (2.5 feet) of GAG 236
140 THM FP in finished water and removal by 1.52
meters (5 feet) of GAG 237
141 THM FP in finished water and removal by 2.29
meters (7.5 feet) of GAG 238
142 THM FP in finished water and removal by 3.05
meters (10 feet) of GAG 239
143 THM FP substances in grams entering and adsorbed
by each GAG column in 115 days 241
144 THM FP adsorption by GAG, XE-340 and IRA-904
resin per column 0.76 meter deep (2.5 feet)
at 49 days 243
145 THM FP adsorption by GAG, XE-340 and IRA-904 resin
per 100 grams of adsorbent at 49 days 245
146 TOC in finished water and removal by 0.76
meter (2.5 feet) of XE-340 249
xv
-------
Number
147 TOG in finished water and removal by 0.76
meter (2.5 feet) of IRA-904 resin and
by 0.76 meter (2.5 feet) of GAG 250
148 TOG adsorption by GAG, XE-340 and IRA-904
resin per column 0.76 meter (2.5 feet)
deep at 49 days 251
149 TOG adsorption by GAG, XE-340 and IRA-904
resin per 100 grams of adsorbent at 49 days....252
150 Comparison of laboratory bottle aged and
distribution system THM growth 255
151 Total THM growth in rechlorinated - 2 day
aged IRA-904 resin column effluent 258
152 Total THM growth in IRA-904 resin column effluent
due to THM FP conversion 259
153 Total THM growth in rechlorinated - 2 day aged
GAG column effluent 260
154 Total THM growth in GAG column effluent due to
THM FP conversion 261
155 Total THM growth in rechlorinated - 2 day aged
2,29 and 3.05 meters (7.5 and 10 feet) GAG 263
156 THM breakthrough and THM FP conversion components
of Total-THM in two day aged effluent from
2.29 meters (7.5 feet) deep GAG column 265
157 THM breakthrough and THM FP conversion components
of total-THM in 2 day aged effluent from
3.05 meters (10 feet) deep GAG column 266
158 Inst. THM in finished water and finished water
through 2.29 and 3.05 meters (7.5 and 10
feet) of GAG 267
159 Relationship of TOG and THM FP data in raw, H.T.
and finished water 269
160 TOG and THM FP in finished water thru 3.05
meters (10 feet) of GAG 271
161 Level of chlorodibromomethane entering and
leaving the partially exhausted 0.76
meter (2.5 feet) deep XE-340 column 272
xvi
-------
162 Adsorption and leaching of chlorodibromomethane
on a 0.76 meter (2.5 feet) deep XE-340 column274
163 Level of Bromodichloromethane entering and
leaving the partially exhausted 0.76 meter
(2.5 feet) deep XE-340 column 275
164 Adsorption and leaching of bromodichloromethane
on a 0.76 meter (2.5 feet) deep XE-340
column 276
165 Level of chloroform entering and leaving the
partially exhausted 0.76 meter (2.5 feet)
deep XE-340 column 277
166 Adsorption and leaching of chloroform on a 0.76
meter (2.5 feet) deep XE-340 column 278
167 Level of cis 1,2-Dichloroethene entering and
leaving the partially exhausted 0.76 meter
(2.5 feet) deep XE-340 column 279
168 Adsorption and leaching of cis 1,2-dichloroethene
on a 0.76 meter (2.5 feet) deep XE-340
column 280
169 Chloroform adsorption by 0.76 meter (2.5
feet) of GAC 286
170 Butane adsorption curves for F-400 GAC and
XE-340 resin 287
171-173 Chloroform adsorption by 0.76 meter (2.5 feet)
of XE-340 293-294, 296
174 cis ±f 2-rDichloroethene adsorption by 0.76 meter
f2.S feet) of XE-340 297
175 cis 1,2-Dichloroethene adsorption by 0.76 meter
(2.5 feet) of GAC 299
176 Bromodichloromethane adsorption by 0.76 meter
(2.5 feet) of XE-340 302
177 Bromodichloromethane adsorption by 0.76 meter
(2.5 feet) of GAC 303
178 Chlorodibromomethane adsorption by 0.76 meter
(2.5 feet) of XE-340 305
179 Chlorodibromomethane adsorption from 0.76
meter (2.5 feet) of GAC 306
xvii
-------
TABLES
Number Page
1 Typical partial analyses, John E. Preston
Water Treatment Plant 14
2 Average chemical application to raw water,
John E. Preston Water Treatment Plant. ... 15
3 Experimental design number and starting
and ending dates 26
4-6 Average concentration of specific halogenated
organics in raw, H.T. and finished water. . . 33-35
7-8 Percent removal or increase factor for
specific HOC in H.T. and finished water. . . 37, 39
9 Average total Inst THM and percent of individual
THM in each experimental design 40
10 TOC and THM FP removal by lime softening in
full scale plant 41
11 TOC, terminal THM and THM FP reduction by chlori-
nation, contact basin and/or sand filtration. . 43
12 Average pH and free chlorine (ppm) values in
each experimental design 48
13-18 Specific HOC adsorption data, raw water. . . betw. 51-72
19-24 Specific HOC adsorption data, H.T. water. . .betw. 77-103
25-36 Specific HOC adsorption data? finished water.betw.105-181
37 Observed adsorptive capacity of 100 grams of GAG
for'five HOC from finished water compared to
the Polanyi-Manes predicted value for each
compound from pure water (adsorbed by 0.76
meter [2.5 feet] of GAG) 194
38-39 THM FP and TOC adsorption data, raw water. . . . 209, 217
40-41 THM FP and TOC adsorption data, H.T. water. . . ,221, 227
XVlll
-------
Number page
42 THM FP adsorption data from finished water 230
43 Effect of THM FP influent concentration in raw
water on adsorption data interpretation 244
44 TOC adsorption data from finished water 247
45 Sampling procedure to compare laboratory and
distribution system aging 254
46 Total THM growth in adsorbent column effluents 257
47 Physical data for Polanyi-Manes calculations 289
48 Chloroform adsorption data from finished water 291
49 cis 1,2-Dichloroethene adsorption data from raw,
H.T. , and finished water 300
50 Bromodichloromethane adsorption data from H.T.
and finished water 301
51 Chlorodibromomethane adsorption data from H.T.
and finished water 304
52 Summation of adsorption parameters by GAC 308
53 GC/MS HOC confirmation data 310
xix
-------
GAC
H.T.
NORS
GC
GC/MS
TOC
THM
HOC
THM FP
Total THM
Inst. THM
Terminal THM
BAG
UV
m3/s
mgd
ED
L/h
GPH
MTZ
DOM
EBCT
LIST OF ABBREVIATIONS
- granular activated carbon
- Hydrotreator (up-flow lime softening unit)
- National Organics Reconnaissance Survey
- gas chromatograph
- gas chromatograph/mass spectrograph
- total organic carbon
- trihalomethane(s)
- halogenated organic compound(s)
- trihalomethane formation potential
• total trihalomethane(s)
• instantaneous trihalomethane(s)
• terminal trihalomethane(s)
• biologically activated carbon
- ultraviolet
• cubic meters/second
• million gallons per day
• experimental design
• liters/hour
gallons per hour
mass transfer zone
dissolved organic matter
empty bed contact time
xx
-------
ACKNOWLEDGMENTS
Unquestionably, the most valuable help, leadership and
guidance for the Project was provided by the Project Officer,
Mr. Jack DeMarco. His unselfish and tireless effort in getting
the study organized and keeping it moving in the proper direc-
tion was one of the major reasons for its success.
The support and cooperation of President Harold Crosby,
Florida International University; Mr. Garrett Sloan, Director
of the Miami-Dade Water and Sewer Authority, and Dr. Richard
Morgan, Director of the Dade County Department of Public Health,
are greatly appreciated.
The advice of the Technical Advisory Board was an important
factor in evaluating accumulating results of the work. Their
comments on projected goals are appreciated.
Advisory Board members were Mr. Anthony Clemente, Dade
County Department of Environmental Resources Management;
Mr. Glenn Dykes, State Department of Environmental Regulation;
Mr. Sidney Berkowitz, Consultant; Dr. John Davies and Dr. Henry
Enos from the University of Miami, as well as Dr. Morgan and
Mr. Sloan.
The secretarial services of Ms. Phyllis Engles, Ms. Barbara
Weil, Ms. Linda Rountree, Ms. Sarah Bostwick, Ms. Bess Simon,
and Ms. Marlene Blosucci were well performed and appreciated.
The technical staff of Florida International University
and the Metropolitan Dade County Water and Sewer Authority
contributed much to the success of the program. These include
Mr. Hunt Harween, Mr. Besteiro Palomeque, Mr. Kenneth Kirkman,
Mr. Russell Lang, Mr. Cesar Ordaz, Mr. William Booth, and
Ms. Laurie Miller.
Appreciation must be expressed also for the cooperation of
the personnel in the business office and accounting department
of the Dade County Department of Public Health, Florida Inter-
national University, and the Metropolitan Dade County Water and
Sewer Authority.
The technical assistance of Dr. Milton Manes, Chemistry
Department, Kent State University, Kent, Ohio and Dr. Michael
Rosene, Calgon Corporation, Calgon Center, Pittsburgh, Penn-
sylvania is also acknowledged.
xxi
-------
SECTION I
INTRODUCTION
In 1975, the U.S. Environmental Protection Agency announced
by their release of a report on "National Organics Reconnais-
sance Survey for Halogenated Organics in Drinking Water" (1)
(NORS) that the drinking water of Dade County, Florida contained
over 300 ppb chloroform and nearly 6 ppb vinyl chloride. Person-
nel at the Dade County Health Department, Miami-Dade Water and
Sewer Authority, and the Drinking Water Quality Research Center
of Florida International University in cooperation with the U.S.
Environmental Protection Agency in Cincinnati, Ohio developed a
research project to 1) develop an effective and economical
method for removing, or substantially reducing, the chemicals of
concern and 2) remove organic solutes (precursors) from the
water to prevent regrowth of these chemicals in the distribution
system where free chlorine is present. On June 22, 1976 a one-
year study was approved by EPA and later extended to September
5, 1978.
The scope of the Project included 1) a study of the effec-
tiveness of various adsorbents in removing potential carcinogens
already present in raw water and those generated in the water
treatment process, 2) removal of precursor substances which form
halogenated organics upon reaction with chlorine in the treat-
ment plant and distribution system, and 3) effect of the stage
of treatment process on efficiency for removing contaminants.
Early in the Research Project the high levels of halogen-
ated organics reported by the NORS (1) study were verified both
in concentration and identification by GC/MS. These results
were reported at once to local and state authorities.
This report describes the results of a study of two adsorb-
ent resins and granular activated carbon for their effectiveness
and efficiency in removing trihalomethane precursors, halogen-
ated organic compounds and total organic carbon from three
locations in the treatment plant, raw, lime softened and fin-
ished water. The resins were Ambersorb XE-340, a carbonized
polymeric macroreticular resin, and IRA-904, a strong base
cationic resin for anion exchange, both manufactured by Rohm and
Haas Company, Philadelphia, Pa. The Granular Activated Carbon
was Filtrasorb 400, 12 x 40 mesh, manufactured by Calgon Cor-
poration, Pittsburg, Pa.
-------
The Miami-Dade Water and Sewer Authority furnishes water
for over one million people through three major water plants.
The John E. Preston Water Treatment Plant in Hialeah, Dade
County, Florida, which operates at 2.63 m3/s (60 mgd), draws
water from the Biscayne Aquifer from seven wells located on the
plant site. The wells are approximately 27.4 meters (90 feet)
deep. The raw ground water, which contains an average of 10
mg/L of Total Organic Carbon, is treated by lime softening in an
up-flow Hydrotreator, breakpoint chlorination, and sand filtra-
tion. Before the water leaves the plant the free chlorine level
is adjusted to 2.5 ppm.
The strata overlying the recharge area of the Biscayne
Aquifer are predominately muck, which accounts for the rela-
tively high color present in the source water. Thus organics of
a natural origin comprise one problem that cannot be easily pre-
vented by attempting to change sources of drinking water in
this area. Other organic substances that are a result of man's
activities are also present and pose another facet of the prob-
lem presented in using the ground water in the area. Thus
initial studies of practical methods of removing organics were
directed at attempting to find a broad based organic removal
method.
The three adsorbents were studied in four experimental
designs developed by Jack DeMarco, EPA Project Supervisor.
Glass columns 2.54 cm (one inch) in diameter were connected
directly to raw, lime softened (Hydrotreator effluent) and fin-
ished water lines from the Preston Plant. Adsorbent bed depths
studied were 0.76, 1.52, 2.29 and 3.05 meters (2.5, 5, 7.5 and
10 feet) for Granular Activated Carbon, 0.76 and 1.52 meters
(2.5 and 5 feet) for IRA-904 resin, and 0.76 meter (2.5 feet)
for XE-340. A flow rate of 122L/min./m2 (3 gal./min./ft.2) was
maintained by rotometers through each column. Thus empty bed
contact times were always 6.2 minutes whenever a 0.76 meter
(2.5 feet) bed depth of adsorbent was used, 12.4, minutes for a
1.52 meter (5 feet) bed, 18.6 minutes for a 2.29 meter (7.5
feet) bed and 24.8 minutes for a 3.05 meter (10 feet) bed.
The 3.05 meter (10 feet) bed depth of Granular Activated
Carbon (24.8 minutes Empty Bed Contact Time) showed the most
promise for achieving broad based removal of organics. Water in
the normal distribution system was evaluated as a comparison
with experimental results obtained by using adsorbents.
-------
SECTION II
CONCLUSIONS
Full Scale Plant Performance
1. Over the two-year study, the high levels of halogenated
organic compounds reported by the national Organics Recon-
naissance Survey (1) study were verified both in concen-
tration and identification by Gas Chromatograph/Mass
Spectrograph.
2. The average concentration of halogenated organic compounds
present in raw water was approximately 15 percent less after
lime softening (Hydrotreator effluent).
3. The average level of trihalomethanes in finished water
leaving the plant over the two-year study was 67, 43, 28 and
2 yg/L respectively for chloroform, bromodichloromethane,
chlorodibromomethane and bromoform. The average total
trihalomethane level leaving the plant was 140 yg/L. In the
distribution system, this level can double in less than two
days. The level would rise higher, but the free chlorine is
exhausted in one to two days. When, additional free chlorine
is added at booster stations in the distribution system,
trihalomethane levels reach their maximum.
4. The level of some non-trihalomethane halogenated organic
compounds increased during the plant treatment process. A
consistent increase was found for the summed concentration
of 1,1,1-trichloroethane, 1,2-dichloroethane and carbon
tetrachloride. Increases of this summed concentration
ranged from 1.2 to 77 times the level in raw water. Since
the three compounds were summed due to overlapping gas
chromatograph peaks we do not know if the increase was due
to one or more of the three substances. Some other non-
trihalomethane halogenated organic compounds may have shown
intermittent increases.
5. In general, non-trihalomethane halogenated organic sub-
stances were not consistently well removed by the existing
full scale treatment plant.
6. Lime softening removed an average of 28 percent of the
trihalomethane formation potential and little additional
-------
actual removal was achieved by the sand filtration process.
Calcium carbonate floe was believed to provide the mechanism
for trihalomethane formation potential removal in the full
scale plant.
7. In the Preston Plant, the amount of precursor removal by
conversion to trihalomethanes by the combined chlorination-
sand filtration process averaged 23 percent of the Hydro-
treator effluent level.
Specific Organic Removal by Adsorbents
8. In both raw and finished water, XE-340 has more adsorptive
capacity in weight of organic substance adsorbed per unit
weight or volume of adsorbent, for individual halogenated
organic compounds than granular activated carbon. While
the values are different for each halogenated organic com-
pound and different in raw and finished water, in general,
XE-340 has approximately three times the adsorptive capacity
of granular activated carbon. .
9. The adsorptive capacity in weight of organic substances
adsorbed per unit weight or volume of adsorbent of XE-340
for halogenated organic compounds is only slightly greater
when treating raw water than when treating Hydrotreator
water. The lower total organic compound concentration in
the Hydrotreator water (approximately 30 percent lower) did
not enhance the ability of the adsorbents for halogenated
organic compound removal.
10. The adsorptive capacity of both XE-340 and granular acti-
vated carbon for cis 1,2-dichloroethene is less in finished
water than raw water despite a reduction of 34 percent total
organic carbon. The percent of cis 1,2-dichloroethene of
total halogenated organic compounds in raw and finished
water is 86.5 and 10 percent respectively. We attribute the
30 percent reduction in adsorptive capacity for cis 1,2-
dichloroethene in finished water to increased competitive
adsorption from the additional halogenated organics present
in the finished water.
11. On raw and Hydrotreator water, IRA-904 resin showed no
removal of any of the halogenated organic compounds present.
12. On finished water, IRA-904 resin appears to enhance the
reaction of free chlorine with precursors to form halo-
genated organic compounds. The effluent of a 0.76 meter
(2.5 feet) deep bed (empty bed contact time of 6.2 minutes)
contained 1.75 and 1.13 times the influent concentration of
-------
chloroform and bromodichloromethane respectively. Increases
in concentration occurred in some of the non-trihalomethane
halogenated organic compounds, but the majority showed no
increase nor decrease in concentration as a result of the
IRA-904 resin as observed in raw and Hydrotreator water.
13. A 3.05 meter (10 feet) deep bed of granular activated carbon
with an empty bed contact time of 24.8 minutes was ineffec-
tive for vinyl chloride removal.
Total Organic Carbon and Trihalomethane Formation Potential
Removal by Adsorbents
14. Total organic carbon data did not consistently correlate
with trihalomethane formation potential data. Also, one
cannot be converted into the other by a single conversion
factor since total organic carbon analysis measures some
substances that are not trihalomethane precursors. As
expected, total organic carbon analysis is not a precise
useful indicator of trihalomethane precursors. However,
general trends might be noted at a given site.
15. In this report, the shape of the adsorption breakthrough
curves for total organic carbon and trihalomethane formation
potential removal by adsorbents are similar to specific
halogenated organic compound removal curves.
16. A system was devised that could be used on total organic
carbon and trihalomethane formation potential substances to
allow a more complete understanding and comparison of
adsorbent performance.
17. IRA-904 resin removed trihalomethane formation potential
more efficiently from raw water than did the other two
adsorbents tested based on percent removal and based on
weight adsorbed per unit volume of adsorbent. However,
granular activated carbon removed trihalomethane formation
potential more efficiently than the other adsorbents based
on the weight of trihalomethane formation potential adsorbed
per unit weight of adsorbent.
18. Although IRA-904 resin was more effective for trihalomethane
formation potential removal than granular activated carbon
on a volume basis, and only 20 percent less effective on a
weight basis, it is important to note that the resin allowed
about 100 ug/L of trihalomethane formation potential to pass
through the bed at start-up even with a 1.52 (5 feet) bed
depth. The 0.76 meter (2.5 feet) granular activated carbon
system was able to produce an effluent with about 12 yg/L of
trihalomethane formation potential at start-up. Thus, a
single measure of effectiveness cannot be applied without
knowledge of performance required of the adsorbent. The
breakthrough curve is important in determining the adsorbent
-------
performance for a specific effluent criteria.
19. A direct comparison of 0.76 meter (215 feet) beds (6.2
minutes empty bed contact time) of granular activated carbon
and XE-340 shows that carbon removed more trihalomethane
formation potential on both an equal adsorbent weight and
volume basis when receiving raw water. Furthermore, the
breakthrough plot indicates that carbon maintained a lower
effluent concentration than XE-340 for about 17 days.
During this time period neither adsorbent was very effective
for trihalomethane formation potential removal at the con-
ditions tested.
20. Lime softening removed an average of 28 percent of trihalo-
methane formation potential precursors from raw water. This
compares with 29 and 24 percent removal by 0.76 meter (2.5
feet), 6.2 minutes empty bed contact time, of granular
activated carbon and XE-340 over a 119-day test and 46 per-
cent removal by a bed of IRA-904 resin after 49 days of
operation. If all three adsorbents are compared after 49
days of operation time, removals affected are 26, 24 and 46
percent for carbon, XE-340 and IRA-904 resin respectively.
A bed of IRA-904 resin 1.52 meters (5 feet) deep, 12.4
minutes empty bed contact time, removed 55 percent after 49
days of operation.
21. On a weight basis, calcium carbonate floe removed one-third
as much trihalomethane precursor from raw water as carbon.
22. The XE-340 bed removed an average of four percent trihalo-
methane formation potential from Hydrotreator water as
compared with 24 percent removed from raw water.
23. The IRA-904 resin bed removed an average of 32 percent
trihalomethane formation potential from Hydrotreator
effluent as compared to 46 percent from raw water.
24. An XE-340 column 0.76 meter (2.5 feet) deep removed no
trihalomethane formation potential from finished water.
25. A 0.76 meter (2.5 feet) deep bed of IRA-904 resin removed
13 percent trihalomethane formation potential from finished
water. At no time during the test period was the effluent
trihalomethane formation potential concentration from this
column below 180 yg/L.
26. Granular activated carbon was more efficient in removing
trihalomethane formation potential precursors from finished
water than the other two adsorbents. For example, in two
separate runs 0.76 meter (2.5 feet) of carbon removed 18
and 22 percent, and it was the only adsorbent tested that
removed enough precursor to keep the trihalomethane form-
-------
ation potential level below 100 yg/L.
Finished Water
27. Granular activated carbon was chosen for deep bed studies
because it was the best broad based adsorbent for removal
of organics in our system. The deep bed studies were car-
ried out on finished water which had the lowest level of
total organic carbon of any location in the plant, and did
not suffer from execessive calcium carbonate precipitation.
In our system, at a flow rate of 122L/min./m2 (3 gpm/ft.2),
an empty bed contact time of approximately 24.8 minutes in
a 3.05 meter (10 feet) deep granular activated carbon bed
was necessary to achieve a bed life of 81 days.
28. Free chlorine residuals were completely removed by 0.76
meter (2.5 feet) of granular activated carbon and IRA-904
resin throughout their respective test periods, whereas the
XE-340 completely removed the free chlorine residual for
about 17 days.
29. Combined chlorine residuals penetrated all adsorbents
tested.
30. Laboratory bottle aging of finished water as a means of
predicting trihalomethane growth in the distribution system
produced comparable results.
31. An XE-340 adsorbent column, partially saturated with halo-
genated organics in finished water, was treated with halo-
genated organic-free water to test for halogenated organic
leaching (desorption). Desorption of cis 1,2-dichloro*
ethene, chloroform, bromodichloromethane, and chlorodi-
bromomethane appeared to follow a curve that was the
reverse of the adsorption curve.
General
32. The Polanyi-Manes Theory of adsorption was useful in inter-
preting and explaining our data.
33. A bacterial profile study was made on raw and finished
water at the Preston Plant, and the effluent from the
granular activated carbon columns. Two reports of this
work are appended to this report. Clearly, we had a
Biological Activated Carbon system. As no additional
oxygen was added to the water (as in European practice), we
call our system a partial biological activated carbon
system. We cannot speculate on the results of the bacte-
rial growth were not present. We do feel, however, that
despite the massive bacterial growth that eventually pre-
vented back washing of the columns, adsorptive capacity
-------
of the granular activated carbon for halogenated organic
carbon was not decreased. Initial breakthrough and
saturation time for each halogenated organic compound
through each column were too consistent to suggest blocking
of active sites by the bacteria. Bacteria develop large
populations in granular activated carbon columns which
slough off into the column effluents in large numbers,
necessitating disinfection before release into the distri-
bution system.
-------
SECTION III
RECOMMENDATIONS
1. As finished water leaves a 3.05 meter (10 foot) deep granu-
lar activated carbon bed, it has nil free chlorine and a
high population of bacteria. It would have to be rechlori-
nated to again achieve disinfection and according to pre-
sent practice in Florida it would have to contain approxi-
mately 2.5 ppm of free chlorine to provide residual disin-
fectant before it could enter the distribution system. As
some precursors are still present, trihalomethane regrowth
in the distribution system will occur. We therefore define
granular activated carbon exhaustion or bed life as the
point where trihalomethane regrowth in a sample of the 3.05
meter (10 foot) granular activated carbon bed effluent
(after rechlorination to 2.5 ppm of free chlorine and aging
for two days) reaches the proposed Minimum Concentration
Level of trihalomethane or 0.1 mg/L (100 ppb). This level
was reached in 81 days. Failure at 81 days, however, was
not due solely to trihalomethane growth from precursors.
The column had become saturated to chloroform. If the
influent water had not contained such a high concentration
of trihalomethanes (140 yg/L), the bed life would have been
somewhat longer. We can only guess that bed life would have
been extended another two weeks. If ozone replaced break-
point chlorination after the Hydrotreator water, the high
trihalomethane load would be eliminated, and, according to
European reports, precursors would also be greatly reduced
by subsequent granular activated carbon-biological activated
carbon treatment. Bed life then would be greatly prolonged.
This remains to be confirmed in our system. We therefore,
recommend a pilot plant research project located at the
Preston Water Treatment Plant to study the possibility of
prolonging granular activated carbon bed life in our system
by the use of ozone followed by biologically activated
carbon.
2. Since a granular activated carbon column effluent contains
a high population of bacteria and since such a column would
probably be used at the end of a conventional treatment
plant we recommend that a bacterial study be made before
widespread use of such a system is adopted. Specifically,
to determine if the disinfection with 2.5 ppm of free
chlorine at the end of the treatment process is adequate
-------
with the associated treatment plant contact time which will
range widely from a few minutes to 24 hours or more before
discharge into the distribution system. Our work indicates
that the standard plate count method is apparently inade-
quate in assessing numbers and types of bacteria that are
found in a granular activated carbon effluent. Conditions
for an optimal bacterial method apparently are still being
worked out. We recommend that a committee of bacteriolo-
gists should be formed to adopt interim bacteria test
methods and set up research projects for further study to
determine optimal methods.
3. Considerable work is being done in the field of adsorption
kinetics. Perhaps more should be done with the Polanyi-
Manes adsorption theory. This might include the individual
adsorption of most of the specific halogenated organic com-
pounds found in raw and treated water from purified water
and water containing known amounts of total organic carbon
and added amounts of other specific halogenated organic
carbons to study competitive effects.
4. There are several questions about the source of halogenated
organic carbon in raw water that should be answered, such
as;
a. Source of our high level of cis 1,2-dichloroethene
b- Source of vinyl chloride
c. Lack of or very low level of trihalomethanes when other
volatile halogenated organic carbons are present
d. Effect of untraviolet and bacterial enzyme action on
specific halogenated organic carbon.
10
-------
SECTION IV
PLANT AND EQUIPMENT DESCRIPTION
PRESTON PLANT WATER SOURCE
The terrain in the vicinity of Miami and its neighboring
municipalities consists of an outcropping of soft, porous,
Oolite limestone rock. Water enters this porous rock from local
rainfall, runoff, and from the extensive canal system in south-
ern Florida. The approximate annual rainfall in the area is
152.4 cm (60 inches) per year. The porous water-bearing rock is
the aquifer from which raw water is drawn by the water treatment
plants in the area.
Seven wells have been drilled into the surface rock on or
near the Preston Plant site. This groundwater has high color
and contains dissolved iron. The color is attributed in large
part to the organic matter leached from decaying vegetation
through which the groundwater percolates. The water is slightly
basic with an average pH of 7.2.
PRESTON PLANT SITE
The Miami-Dade Water and Sewer Authority, through its three
major water plants, furnished water, either directly or indi-
rectly for over one million people. One of these three major
water plants is the John E. Preston Water Treatment Plant (Flow
Diagram in Figure 1), located at 1100 West Second Avenue,
Hialeah, Florida. The Preston Plant has been in operation
approximately seven years.
At present the Preston Plant is rated at 2.63 m3/s (60
mgd), and in general is operated near or at maximum capacity.
The wells supplying the Preston Plant are approximately 27.4 m
(90 feet) deep. Each well can produce over 34065 m3 (nine
million gallons) per day. Raw water from these seven wells is
fed to a combination of three upflow Hydrotreator (H.T.) soft-
eners, each rated at 0.88 m3/s (20 mgd). Silica, activated with
chlorine, is added to the raw water just prior to its entrance
into the upflow softener. The upflow softener effluent is
channeled into a recarbonation flume. Sodium silica fluoride is
added at this point. Chlorine is then added just before the
water enters the chlorine contact basin. After an average
retention time of 1.25 hours in the chlorine contact basin, the
11
-------
3 O
»z
CHLORINE DIFFUSION
IB.O ±PPM
0.5 ± PPM FLUORIDE
I 2
5 MOD RAPIO
SAND FILTERS
' 60 M.CD
CLEAR
WELL
O.6 MOD
VENTURI
METER
RESERVOIR
».0 H6
wf
/
Q_5) CAPAC1TT OF PUMPS IN MOO
O WELL
-O«6
Figure 1. Flow diagram of John E. Preston Water
Treatment Plant.
12
-------
water flows into rapid sand filters 0.22 m3/s - 122L/min./m2
(5 MGD - 3 GPM/ft.2), then to a 34065 m3 (nine million gallon)
reservior. From the reservoir the water is diverted to the
Hialeah Plant for pumping, or pumped directly to high pressure
distribution lines. Tables 1 and 2 contain chemical data
related to the Preston Water Treatment Plant.
BENCH SCALE ADSORPTION TEST UNIT
The Bench Scale Adsorption unit is located in the second-
floor laboratory of the Preston Plant. Three sampling lines
enter the laboratory from the plant raw water, Hydrotreator,
and clear well composite lines. The water is constantly moni-
tored by pH chart recorder, and samples are readily available
at this one location. Routine testing is done hourly around-
the-clock. A flow diagram of the Bench Scale Adsorption unit is
shown in Figure 2. Each glass column is 1.52 meters (five feet)
long by 2.54 cm (one inch) in diameter.
A flow rate of 122L/min./m2 (3 gal./min./ft.2) was main-
tained by rotometers.
/
The pumps used on the three sample lines from the plant to
the laboratory were three-quarter housepower, water lubricated,
with Teflon seals. All lines were copper pipe. Construction
details of the Bench Scale Adsorption unit are shown in Figures
3 and 4.
13
-------
TABLE 1. TYPICAL PARTIAL ANALYSES,
JOHN -E. PRESTON WATER' TREATMENT PLANT
Alkalinity (CaCO,)
Phenolphthalein
Methyl Orange
Hardness (CaC03)
Non-Carbonate
Total
Carbon Dioxide, Free (CO,)
Chlorine Residual (Cl,)
at plant •
Chlorides (Cl)
Fluorides (F)
Sulfates (SO4)
Calcium (Ca)
Iron (Fe)
Magnesium (Mg)
Sodium & Potassium (as Na)
Silica (Si02J
turbidity
Total Solids
Electrical Conductivity
(EC x 10 @ 25°C)
Treated Water
After Softening, Entering
Before Distribution
Well Water Chlorination System
0.
230.
20.
250.
25.
0.
40.
0.2
24.
88.
0.8
7.0
29.
8.0
Nil
350.
16,
32.
21.
53.
4.
40.
35.
75.
0. 0.
0.05(Combined) 2.0 (Free)
41.
0.1
0.1
Excess 50 units
55.
0.7
24.
23.
0.0
4.2
32.
9..0
Nil
205.
580. 305.
Color 50. 25. 6.
pH 7.3 10.0'- 10.3 8.8
lAll units 'expressed as mg/Lr (ppm), .where applicable)
14
-------
TABLE 2. AVERAGE CHEMICAL APPLICATION TO
RAW WATER, JOHN E. PRESTON WATER TREATMENT PLANT
Raw water influent to softening units:
1.0 to 2.0 ppm chlorine-activated silica —
chlorine 0.3 to 0.7 ppm.
Softening units:
160 to 180 ppm CaO as slaked Ca(OH)_
Carbon dioxide addition after softening unit, as
needed to achieve desired degree of stabilization.
Sodium silica fluoride addition, ±0.5 ppm to bring
fluoride level to 0.7 ppm in treated water.
Fifteen to 17 ppm chlorine dosage to achieve a chlorine
residual in treated water leaving plant of 1.5 to
3.0 ppm free chlorine. Chlorine contact basin
average retention time 1.25 hours. Average minimum
free chlorine residual at far points in distribution
system 0.5 ppm.
15
-------
^s
1.52 meters
( 5 feet)
Figure 2. Bench Scale Column Adsorption Unit
16
-------
Figure 3. Plumbing for adsorption column.
17
-------
00
Figure 4. Detailed viev? of column fittings.
-------
SECTION V
METHODS AND PROCEDURES
OPERATION OP BENCH SCALE ADSORPTION UNIT
A flow rate of 122L/min./m2 (3 gal./min.ft.2) was maintained
for this study and is equal to approximately 3.78L per hour
(one gallon per hour), or 89.3 liters per day. Flow through the
column was maintained at 3.78L/hr. (1.0 GPH) by adjusting the
rotometer. This was checked periodically. The pressure was
adjusted with the regulating valve in order to provide enough
head to maintain the desired flow. Whenever the increase in
pressure in the column was approximately 7 or 8 psi, the column
was backwashed. The H.T. effluent pump, lines, and column were
backwashed every day because of the rapid build-up of calcium
carbonate. The backwash system for the columns is shown in
Figure 5. Backwash water was prepared by passing tap water
through a Barnstead Still, an ion-X-changer, and an activated
carbon filter. This was followed by all glass redistillation.
This water was then boiled and purged with zero-grade helium to
remove volatile halogenated organics.
Specific data for each Experimental Design not specified
below, such as backwash dates and time, pH, turbidity, color,
and chlorine are presented in Appendix B of this report.
GC ANALYTICAL METHOD
Purgeable Halogenated Organic Compounds were analyzed
according to the purge and trap method of Bellar and Lichtenberg
(2) with modifications by Dressman and McFarren (3) for
analysis of vinyl chloride.
Pure helium was bubbled through a water sample (0.5 to 7 mL)
at a rate of 20 mL/min. for 11 minutes. Volatile halogenated
organics were retained in a one-eight inch O.D. by eight inch
stainless still trap. The first two-thirds of the trap con-
tained Tenax-GC and"the upper one-third contained Silica Gel-15.
Tenax-GC efficiently adsorbes most of the compounds, but Silica
Gel-15 is more efficient in adsorbing the lower molecular weight
highly volatile compounds such as vinyl chloride.
The trap was backflushed for six minutes at 220?C with
helium (20 ml/min.) onto the GC column (1/8 in. x 7 ft. stain-
19
-------
2.3
Figure 5. Backwash system for columns.
-------
less steel packed with Tenax-GC). Helium carrier gas (40 mL/
mm.) was then turned on and the oven set for an 18-minute
isothermal hold at 95°C. The balance of the full 52-minute run
was programmed at 4°C/min. to 220°C. The Hall Electrolytic
Conductivity Detector reduces all halogenated compounds to
halogen acids, which are then detected.
A typical chromatogram of a standard sample is shown in
Figure 6. These are the 19 specific halogenated organic com-
pounds routinely monitored. Each peak is identified by the
chemical name and number and concentration in micrograms in the
standard. No significant amount of methyl iodide was found
during this study. Values for compounds No. 7, 8, and 9 were
summed because the three peaks overlapped and made separate
analysis impractical. In most of this report values,for the
three isomers of dichlorobenzene were summed. In ED4 they were
reported separately, because improved chromatographic technique
separated the three isomers.
TOG ANALYSIS
TOC values were obtained by the EPA Laboratory in Cincinnati,
Ohio on a Dohrmann-Envirotech Organic Analyzer with an Ultra Low
Organics Module.
TRIHALOMETHANES, TERMINAL TRIHALOMETHANES AND TRIHALOMETHANE
FORMATION POTENTIAL
Four general individual THM compounds were qualified and
quantified in the study as a part of the HOC analytic program.
They were chloroform, bromodichloromethane, dibromochloro-
methane, and bromoform. In order to facilitate the investiga-
tion of THM's and their control, other parameters were also
utilized. These parameters are discussed elsewhere (4) in more
detail and are defined here as they applied to this project.
1. Total trihalomethane (total THM) concentration is the sum-
mation of the concentrations of the individual THMs in a
sample.
2. Instantaneous THM (inst. THM) is the concentration of TTHM
in the water at the time the sample is collected.
3. Terminal THM (term. THM) is the sum of TTHM present in the
water at the moment of sampling and TTHM subsequently formed
during additional reaction time under defined conditions.
This value may be used as a general estimate of THM con-
centrations that the consumer would receive if the water
from the location sampled were subjected to the pH, temper-
ature, free chlorine residual and storage time conditions
that were used for the sample. During the project, the
21
-------
,— --•!-- i "••
C.T.; r_
L...U -i.
Figure-6. Typical chromatogram of halogenatett organics.
22
-------
reaction was driven toward completion by adding chlorine to
exhaust the precursor. Samples were routinely stored at
the finished water pH of 9.0 and a temperature of 22°C for
six days, i.e., beyond the normal detention time in the
distribution system of the utility, with sufficient free
chlorine added to satisfy demand. After six days under
storage conditions, a concentration was reached that was
assumed to represent a maximum reaction for Preston plant
distribution water. Modification of these conditions were
made on additional samples in ED3 and ED4 as described
later. This value may be used as a general estimate of THM
concentrations that the consumer, would receive if the water
from the location sampled were subjected to the pH, tempera-
ture, free chlorine residual and storage time conditions
that were used for the sample.
Trihalomethane formation potential (THM FP)is the difference
between the terminal TTHM and the instantaneous TTHM (term.
TTHM - inst. TTHM = THM FP), an indirect measure of the
unreacted precursor in the water sampled. It is the
increase in the TTHM concentration that occurred during the
storage period for the determination of the terminal TTHM
concentration. The unreacted precursor has the potential to
further increase TTHM concentrations in the presence of free
chlorine.
Handling procedures for HOC including Inst. THM's and Term.
Trihalomethane Samples. All HOC samples for raw water and
adsorbent column effluents on the raw water line were taken
in septum bottles filled to the top and sealed with no air
entrapment. No reagents were added. These are referred to
as the odd number samples 1, 3 and 5. All HOC samples for
the plant Hydrotreator and clear well effluent, as well as
the adsorbent columns receiving these waters were sampled in
the same way except the septum bottles contained one drop of
10 percent sodium thiosulfate. The sodium thiosulfate was
added to quench the presence of any free chlorine residual
in the sampled water. These samples are referred to as odd
numbered samples 7, 9, 11, 13, 15, 17, 19, 21, 23 and 25.
All term. THM samples were taken at the same time as the HOC
samples at each location. The septum bottles were filled to
about the 3/4 level. An appropriate amount of buffer solu-
tion and free chlorine solution was added and the septum
bottle then-was quickly filled to the top with water sample.
Care was taken to allow no overflow, but yet to avoid air
entrapment in the sample bottle prior to sealing. The
samples were delivered to FIU for six-day storage at 22°C.
These samples are designated as even numbered samples 2
through 16, 20, 22, 24 and 26.
23
-------
Additional samples were taken during ED3 and ED4 to compare
the value, of the laboratory stored samples for estimating
the THM concentrations that occurred in the actual distri-
bution system. Clear well effluent samples (designated as
11+2) were taken in empty septum bottles, filled to the
top and sealed without the addit-ion of any reducing agent or
additional chlorine. These samples were stored for two days
at 22°C and then delivered to FIU for analysis. Free
chlorine residuals and pH were determined on these samples
at the time of analysis. The samples were collected for
comparison with inst. THM concentrations found at the Red
Road distribution system sampling station which is two days
water travel time from the plant. Thus, a clear well water
sample was stored in a septum bottle for two days and com-
pared to a sample taken in the distribution system two days
later.
Additional samples from the Red Road distribution system
location were also taken and buffer plus chlorine solution
were added to these samples prior to bottle storage at pH 9
and 22°C for four days. These samples are designated as
17+4 and represent the THM concentration that was present
in a water that was in the actual distribution system for
two days with the pH and chlorine residuals normally present
and four additional days in bottle storage with a pH of 9.0,
temperature of 22°C and presence of free chlorine residual.
The THM concentration of this six-day stored sample, two
distribution days plus four bottle storage days (sample 17 +
4) was compared to the normally obtained six-day bottle
stored term. THM concentration of the clear well water
(sample 12).
Along with the sample of the clear well stored for two days
in a bottle (sample 11 + 2), samples were taken from the
effluent of the adsorbent columns. The column effluent
samples were adjusted as required to assure that a free
chlorine residual was present prior to sealing and two-day
storage. The adsorbent effluent samples were used to esti-
mate the THM concentration that might be received by a con-
sumer two days from the plant, if a specific adsorbent were
a part of the normal treatment system. The concentrations
of these adsorbent effluent samples (i.e. samples designated
13+2, 15+2, 23+2 and 25 +2) were compared with clear
well water samples stored for two days in a bottle as well
as the inst. THM samples collected at the Red Road sampling
station two days water travel time from the plant.
DATA ANALYSIS
The format in this study was to plot all individual data
points to form the adsorption or breakthrough curve. A typical
plot of actual adsorption data for bromodichloromethane is shown
24
-------
in Figure 66, page 125. Initial breakthrough times for each of
the four columns in series is shown. Saturation times for the
first two columns are shown. An extrapolated saturation time
for the third column is shown and the saturation time for the
fourth column cannot be extrapolated because of insufficient
data to establish the slope. The average influent concentration
in ug/L is determined and each adsorption or breakthrough curve
is integrated to determine the amount of substance entering,
passing and adsorbed by the column at breakthrough, saturation
and at the end of the test period or at some other time period
in common with data in another ED. The breakthrough point was
sometimes difficult to ascertain from the actual data and plot-
ted curves. For each HOC, adsorbent type and bed depth studied
there was sometimes intermittent low level leakage before the
time we picked as the breakthrough point. In general, we define
the breakthrough point as the time required to reach 2 yg/L on
the breakthrough curve. Actual leakage values during the period
before breakthrough are too low to plot on the yg/L scale used
for the complete breakthrough curves . In some cases the actual
low level of breakthrough is shown above the plotted data point.
Throughout the study, 76.2 cm (30-inch) bed depths were used
for each column. In all calculations an average weight value
was used for GAC, XE-340 and IRA-904 resin per column. These
values were, respectively, 176, 215, and 275 grams. The flow
rate through each column was 122L/min./m2 (3 gal./min./ft.) ,
resulting in a flow of 89.3 L/day and approximately 3.785 L/hr.
(1 gal./hr.) .
Interpretation of results includes consideration of Mass
Transfer Zone (MT?) . The Mass Transfer Zone for a specific sub-
stance is the minimum bed depth at a given flow rate necessary
to prevent column breakthrough after initial flow. We used the
following equation:
MT = Ts " Tb
s
T = saturation time
s
T, = breakthrough time
D
The raw water feed for the Preston plant averaged 10 mg/L
of TOC. However TOC is merely a representation of the carbon
fraction present. The concentration of Dissolved Organic Matter
(DOM) is some higher amount depending on the relative percent of
carbon in the total molecular weights of organics present. We
have extimated that the organics present contain an average of
60 percent carbon. We calculate approximate DOM values by
dividing TOC values by 0.6. Our raw water feed thus contained
approximately 17 mg/L of DOM. The DOM value is helpful when
discussing competitive adsorption.
25
-------
SECTION VI
EXPERIMENTAL PLAN
The two-year study contained four Experimental Designs (ED).
These designs are listed below in Table 3 with their starting
and ending dates. During the conduct of the experimental
designs for the bench scale studies, the samples that were taken
described the influent water to the bench scale experiments are
also designed to describe the operation of the full scale
Preston plant. Thus a long term comparison of the raw water,
Hydrotreater effluent and finished water for the full scale
treatment plant was a designed phase of the project and desig-
nated as Full Scale Plant Studies. The full scale plant study
data can also be used with each bench scale study conducted.
For example, during EDI the samples on Figure 7 designated as
1, 2, 7, 8, 11 and 12 describe both the influent to the bench
scale columns as well as the full scale treatment plant water at
the locations noted.
TABLE 3. EXPERIMENTAL DESIGN NUMBER
AND STARTING AND ENDING DATES
ED
Dates
1
1R
2
3
4
Aug. 13 - Dec. 7, 1976
Jan. 18 - May 20, 1977
Jun. 3 - Aug. 5, 1977
Aug. 26 - Oct. 18, 1977
Nov. 1, 1977 - Mar. 3, 1978
The general purpose for each ED was as follows:
EDI
The flow diagram for EDI is shown in Figure 7. Two
26
-------
NJ
PRESTON PLANT
Hydrotreator
Raw Water from Wells
en f
M I I
§ « 3
. G-
•A-
.c
'.E
'.3.
>
•V.
J I Sample Point and Number
2
I'
Chlorine
Contact
Basin
Bench Scple Adsorption Unit
8
. E
'."3 •'
*
*
Sand
Filter
Clear
Well
Figure 7. Flow diagram of Bench Scale Adsorption Unit for GAG and XE-340 study
in EDI.
-------
adsorbents were studied. Filtrasorb 400, 12 x 40 mesh obtained
from Calgon Corporation, Pittsburgh, Pennsylvania, was chosen as
the GAC adsorbent. The second adsorbent was Ambersorb XE-340
from Rohm and Haas Company, Philadelphia, Pennsylvania. Amber-
sorb XE-340 is a polymeric carbonaceous adsorbent tailored for
removal of low molecular weight organics from water.
As shown in Figure 7, both adsorbents were placed in the raw
water line. This enabled a comparison of their abilities to re-
move TOC, HOC and precursor substances from raw water. Pre-
cursor removal was measured by the THM FP method. Ambersorb
XE-340 columns were also placed in H.T. and finished water lines
to study HOC, TOC, and THM FP removal with the respective influ-
ent waters.
ED1R
As EDI work progressed, changes in methodology were made and
the desired complete data base from initial start-up was not
obtained. This work was thus considered mainly as a shake-down
phase and EDI was repeated as originally planned.
ED2
The Flow Diagram for ED2 is shown in Figure 8. At the end
of ED1R, the partially exhausted XE-340 column on the finished
water line was selected for a leaching study because this type
of data had not been previously collected. As shown in Figure 8,
a fresh XE-340 column was placed on the finished water line
ahead of the partially exhausted column. For a period of time,
essentially all halogenated organics would be removed by the
fresh column. The halogenated organic leaching rate for the
second column was determined by analyzing the effluent sample
13A.
ED3
The Flow Diagram for ED3 is shown in Figure 9. Rohm and
Haas indicated that IRA-904 resin, avstrong base cationic ad-
sroption resin, was one of the better polymeric adsorbers for
precursor type substances. We did not select it for halogenated
organic adsorption. To study the effect of bed depth, two IRA-
904 resin columns in series were placed on the raw water line.
One IRA-904 resin column was placed on the H.T. line. To com-
pare the effectiveness of IRA-904 resin with GAC Filtrasorb 400,
one column of each was placed on the finished water line.
ED4
The Flow Diagram for ED4 is shown in Figure 10. Four GAC
Filtrasorb 400 columns in series were placed in the finished
water line to study the effect of bed depth and contact time on
halogenated organic and precursor removal.
28
-------
PRESTON PLANT
Raw Water
from Wells
Hydro-
treator
o
i
Chlorine
Contact
Basin
Sand
Filter
Clear
Well
N)
U>
Bench Scale Adsorption Unit
In —T-
H 19
-5) *O 3
a S3
O r-l
It r-t
•f-
:«-:
!o-:
.E'
•;-4
Partially
Exhausted Column
9n Clear Well Water
(Experimental
Design No.1-Repeat)
["] Sample Points and Numbers
Figure 8. Flow diagram of Bench Scale Adsorption Unit for
leaching study in ED2.
-------
PRESTON PLANT
CO
O
Raw Water from Wells
y-
'
H
rr
t
_ 1 1,^.
1 r
rn
^•M!
*.9-."
,
."
.
•"
0.'
4;;
. .
."•;
m - - V-
. I. > 1 '
• i
— y
i
'G.'
'.A'.'
•t;' '
* ."
" * '."
s
I I Sample
Point and Number
Figure 9. Flow diagram of Bench Scale Adsorption Unit for GAG and IRA-904 resin
study in EDS.
-------
PRESTON PLANT
Hydro-
Raw Water tre
from Wells /"""
j 1 j
^
PI
| 2 j 1 7 1
ator
>
n
|_8 J
Chlorine
[ll+ll
Clear
Well
Benol1
(Uj
m—f~*
IT3
1—3
1
••
a T.
I "8 1
Scale Adsorption Unit
EH
*'.'•
C.
00
1
E
..'A'.
'•- 9-
.* .* "-"
L
0 EE3
i
J£
'•*'.'
i
D I241
i
JT
.' "G".
"*•
OIE
1 23+21 [25+21
Q Sample Point and Number
Figure 10. Flow diagram of Bench Scale Adsorption Unit for* deep bed study in
ED4.
-------
SECTION VII
RESULTS AND DISCUSSION
Section VII, Results and Discussion, is divided into two
parts; Full Scale Plant Studies and Bench Scale Studies. Full
Scale Plant Studies refers to data obtained at sample points
from the Preston Water Treatment Plant during the conduct of
each bench scale experiment. These Full Scale Plant sample
points include the raw water feed to the plant, effluent from
the lime softening unit (H.T.), and finished water from the
clear well. Bench Scale Studies refers to data obtained at
sample points from the Bench Scale Column Adsorption Unit.
FULL SCALE PLANT STUDIES
Specific Halogenated Organics
>
Nineteen specific HOC were studied. The specific compounds
with their chemical identification number for this report are
given in Table 4. They are presented in their order of elution
from the gas chromatograph using a Tenax GC column (Figure 6).
The average concentration in the full scale raw, H.T. and fin-
ished water during the conduct of each bench scale ED are shown
in Tables 5 and 6 respectively.
Raw Water Source—
Vinyl chloride levels were 0.8, 6.9, and 12.8 yg/L respec-
tively in ED1R, EDS, and ED4. This is an insufficient data base
to indicate that vinyl chloride levels might be increasing, but
the possibility should be studied further.
The level of methylene chloride also increased, 0.08, 0.1,
and 0.45 yg/L. The level of trans 1,2-dichloroethene varied
erratically from 1.3 to 2.0 yg/L. 1,1-Dichloroethane varied
from 0.3 to 0.6 yg/L. The compound cis 1,2-dichloroethene was
the highest level HOC in raw water and the four ED varied in no
set pattern from 21.0 to 29.0 yg/L. The four THM present,
chloroform, bromodichloromethane, chlorodibromoethane, and
bromoform essentially averaged nil concentration in all ED
except ED1R which showed levels of 0.16, 0.11, 0.04, and 0.02
yg/L respectively. The summed value of 1,1,1-trichloroethane,
1,2-dichloroethane and carbon tetrachloride varied from 0.1 to
0.2 yg/L in ED1R, EDS and ED4. Trichloroethylene varied from
0.13 to 0.4 yg/L. Chlorobenzene varied from 0.19 to 1.3 yg/L.
32
-------
TABLE 4). AVERAGE CONCENTRATION OF SPECIFIC HALOGENATED
ORGANICS IN RAW WATER
Average concentration in ]_ig/L
Chem.
I.D.#
20
1
3
4
5
6
. 7
8
9
10
11
12
13
14
15
16
: 17
18
19
Chemical Name
Vinyl Chloride
Methylene Chloride
Trans 1 , 2-Dichloroethene
1 , 1-Dichloroethane
Cis 1 , 2-Dichloroethene
Chloroform
1,1,1-Trichloroethane \
1 , 2-Dichloroethane )
Carbon Tetrachloride /
Trichloroethylene
Bromodichlorome thane
Tetrachloroethylene
Chlorodibromomethane
Chlorobenzene
Bromoform
. p-chlorotoluene
m-dichlorobenzene
p-dichlorobenzene
o-dichlorobenzene-
Raw Water
EDI
21
N
N
N
N
ED1R
.8
.08
1.3
.3
29
.16
.11
.13
.11
.06
.04
.19
.02
.11
\
-
/
EDS
6.9
.1
2
.6
26.9
N'
.2
.4
N
N
N
1.3
N
.2
\
)l.O
/
ED4
12.8
.45
1.5
.58
25.6
.08
.1
.34
N
.003
N
1.1
N
.02
.51
.17
Total
N = Nil
ND = Not Determined
33.51
39.6 43.25
33
-------
TABLE 5. AVERAGE CONCENTRATION OF SPECIFIC HALOGENATED
ORGANICS IN H.T. WATER
Average concentration in yg/L
Chem.
I.D.# Chemical Name
20
1
3
4
5
6
7
8
9
10
11
12
13
14
15
16
; 17
18
19 .
Vinyl Chloride
Methylene Chloride
Trans 1 , 2-Dichloroethene
1 , 1-Dichloroe thane
Cis 1 , 2-Dichloroethene
Chloroform
1,1/1-Trichloroethane \
1, 2-Dichloroethane /
Carbon Tetrachloride /
Trichloroethylene
Bromodichlorome thane
Tetrachloroethylene
Chlorodibromomethane
Chlorobenzene
Bromoform
p-chlorotoluene
m-dichlorobenzene
p-dichlorobenzene
o-dichlorobenzene
H.T. Water
EDI
20
4
1.7
.62
.09
ED1R
.7
ND
.4
.13
25.4
1.1
.2
.07
.26
.003
.24
.03
N
.03
\
).39
/
ED3
6
ND
1.8
.89
24.1
1.43
.09
.44
.6
N
.46
.84
.013
.16
\
) .56
/
ED4
9.7
ND
.95
.45
22.3
1.2
.12
.33
.35
.004
.13
.72
.007
.03
N
.28
.16
Total
28.95
28.75 37.38 36.73
N = Nil
ND = Not Determined
34
-------
TABLE § . AVERAGE CONCENTRATION OF SPECIFIC HALOGENATED
ORGANICS IN FINISHED WATER
Average concentration pg/L
Chem.
I.D.#
20
1
3
4
5
6
7
8
9
10
11
12
13
14
15
16
; 17
18
19
Chemical Name
Vinyl Chloride
Methylene Chloride
Trans 1 , 2-Dichloroethene
1 , 1-Dichloroethane
Cis 1 , 2-Dichloroethene
Chloroform
1,1, 1-Trichloroethane \
1 , 2-Dichloroethane )
Carbon Tetrachloride /
Trichloroethylene
Bromodichlorome thane
Tetrachloroethylene
Chlorodibromomethane
Chlorobenzene
Bromoform
p-chlorotoluene
m-dichlorobenzene
p-dichlorobenzene
o-dichlorobenzene
Finished Water
EDI
10.9
80.2
37.1
12
.13
ED1R
.6
ND
.18
.2
17.2
71.4
.66
.2
42.7
.02
24.5
.1
1.9
N
\
) .63
ED2
.6
ND
.86
.18
18.4
64
1.47
.57
42.4
N
26.7
.08
1.91
N
\
)"
/
ED3
5.4
ND
1
.3
18.3
57
5.3
.1
39
N
27
.8
2.5
.2
\
•'
/
ED4
6.2
ND
.77
.4
19.9
67.3
7.7
.68
47
.003
33.6
.86
2.5
.1
N
..21
.14
Total
N = Nil
ND = Not Determined
160.29 159.27 157.2 187.36
35
-------
p-Chlorotoluene varied from 0.02 to 0.2 yg/L. The summed value
of m, p, and o-dichlorobenzene varied from 1.0 to 1.1 yg/L in
ED1R and ED3. In ED4, the three isomers of dichlorobenzene were
reported separately with values of nil, 0.51 and 0.17 yg/L
respectively.
Thus, in general, the raw water contaminants were fairly
consistent during the project.
v
Hydrotreator Effluent Source—
Three main factors in the lime softening stage of the plant
probably contribute to changed levels of the specific HOC origi-
nally present in raw water. These factors are, volatile loss,
adsorption on precipitated calcium carbonate (most of which is
removed as sludge) and THM generation by a small amount of
chlorine which is added before the lime to activate the sodium
silicate used as a coagulating aid.
The percent removal or increase factor (the symbol "X" used
as "times") based on raw water for the 19 specific HOC are shown
in Table 7. Vinyl chloride was reduced by 12, 13, and 24 per-
cent in ED1R, EDS, and ED4 respectively. Methylene chloride was
not determined on H.T. or finished water since it was used as an
internal standard for each GC determination. The average con-
centration of trans 1,2-dichloroethene was reduced by 69, 10,
and 37 percent. The level of 1,1-dichloroethane was reduced by
57 and 22 percent in ED1R and ED4 respectively. In ED3, an
increase factor of 1.5X was observed for 1,1-dichloroethane.
This factor is determined by dividing the average concentration
of the compound in the H.T. effluent water by the average con-
centration in raw water. There were 16 data points in EDS and
only 5 showed increase factors. Unless most of the data points
show a consistent increase factor we should probably not put too
much weight on the increase factor as determined.
The H.T. reduced levels of cis 1-2-dichloroethene by 5, 12,
10, and 13 percent in the four ED. The four THM, chloroform,
bromodichloromethane, chlorodibromomethane and bromoform
increased from almost nil levels in raw water to average values
of 1.9, 0.7, 0.4, and 0.03 yg/L in the H.T. effluent water. The
summed value of 1,1,1-trichloroethane, 1,2-dichloroethane and
carbon tetrachloride in EDI and ED4 show increase factors of
1.8X and 1.2X respectively. The summed value was reduced 55
percent in ED3. Again, the actual data points do not show a
consistant increase and partial removal is the usual case.
Trichloroethylene, tetrachloroethylene and p-chlorotoluene
exhibit a similar pattern. The H.T. reduced levels of chloro-
benzene by 84, 35, and 34 percent in ED1R, ED3, and ED4 respec-
tively. The isomers of dichlorobenzene are reduced by 65, 44,
and 51 percent.
36
-------
TABLE 7.
PERCENT REMOVAL OR INCREASE FACTOR FOR
SPECIFIC HALOGENATED ORGANICS IN H.T. WATER
Percent removal or increased
factor based on raw water
Chem.
I.D.t Chemical Name
20
-1
3
4
5
6
7
8
9
10
11
12
13
14
15
16
; 17
18
19
Vinyl Chloride
Methylene Chloride
Trans 1 , 2-Dichloroethene
1 , 1-Dichloroe thane
Cis 1 , 2-Dichloroethene
Chloroform
1,1, 1-Tr ichloroethane \
1 , 2-Dichloroethane /
Carbon Tetrachloride /
Trichloroethylene
Bromodichloromethane
Tetrachloroethylene
Chlorodibromomethane
Chlorobenzene
Bromoform
p-chlorotoluene
m-dichlorobenzene
p-dichlorobenzene
o-«dichlorobenzene
H.T. Water
EDI
5
ED1R
12
69
57
12
1.8X
46
95
84
73
\
65
1
ED3
13
10
1.5X
10
55
1.1X
N
35
10
\
>44
/
ED4
24
37
22
13
1.2X
3
1.3X
34
1.5X
N
45
6
N = Nil
ND = Not Determined
X = Times Factor
37
-------
Generally, the full scale plant H.T. process did not achieve
significant 'reductions in the concentrations of the specific
organics routinely monitored.
Finished Water Source—
Four main factors probably contribute to changed levels of
HOC in the finished water of the full scale treatment plant.
Volatile loss, removal of calcium carbonate (turbidity) by sand
filtration which may contain adsorbed HOC, and oxidation of HOC
by chlorine contribute to the overall HOC reduction. Break-
point chlorination will greatly increase THM levels.
In Table 8 the reduction of vinyl chloride in finished
water, based on raw water levels, is 25, 22, and 52 percent in
ED1R, ED3, and ED4 respectively. The reduction in vinyl
chloride is about the same in the H.T. portion of the plant and
the breakpoint chlorination—chlorine contact basin—sand
filtration stage of the plant. A pattern of further reduction
in finished water compared to reduction in H.T. effluent was
observed with trans 1,2-dichloroethene, 1,1-dichloroethane,
cis 1,2-dichloroethene, tetrachloroethylene, chlorobenzene,
p-chlorotoluene and the isomers of dichlorobenzene.
In Table 8 increase factors of 1.5X and 2X are shown for
trichloroethylene in ED1R and ED4. The individual data points
in both these ED show quite a consistent pattern of increase
suggesting that this compound may indeed be increasing in
finished water. The data in Table 8 for chlorobenzene suggest
that in ED1R and ED4, less of the compound is removed on a
percentage basis from finished water than H.T. water based on
the original amount present in raw water. An explanation might
be that chlorobenzene is actually increasing between the H.T.
and finished water stages, but not enough to indicate an over-
all increase factor. Of the non-THM HOC compounds it appears
that one or more of the summed group consisting of 1,1,1-tri-
chloroethane, 1,2-dichloroethane and carbon tetrachloride
increases in the finished water. The actual individual data
points clearly show that the finished water concentrations were
higher than raw and H.T. water in most of the samples in each
design phase. Trichloroethylene and chlorobenzene may also show
some increase. These increases may be formed by the reaction of
chlorine with precursors or may be introduced with chlorine, or
both.
The increase in inst. THM's through the treatment plant
(raw water versus finished water) are clearly shown by comparing
the chloroform data in Table 4 with the chloroform data in
Table 6. The average summation of the concentrations of the
four inst. THM species are shown for each ED in Table 9 (i.e.
129.4, etc.). Also the average percent of the total inst. THM's
are shown for each species (i.e. chloroform was 62 percent of
the 129.4 yg/L concentration for total inst. THM's for EDI).
38
-------
TABLE; 8. PERCENT REMOVAL OR INCREASE FACTOR FOR SPECIFIC
HALOGENATED ORGANICS IN FINISHED WATER
Chem.
I.D.#
Percent removal or increase factor
based on raw water
Chemical Name
Finished Water
EDI
ED1R
ED3
ED4
20
Vinyl Chloride
25
22
52
Methylene Chloride
Trans 1,2-Dichloroethene
86
50
49
1,1-Dichloroethane
33
50
31
Cis 1,2-Dichloroethene
48
41
29
Chloroform
1,1/1-Trichloroethane
\
1,2-Dichloroethane
I
6X
26.5X
77X
Carbon Tetrachloride
Z
10
Trichloroethylene
1.5X
75
2X
11
Bromodichloromethane
12
13
Tetrachloroeth;
Chlorodibromomethane
m-dichlorobenzene
p-dichlorobenzene
o-dxchlorobenzene
N « Nil
ND = Not Determined
X = Times Factor
39
-------
TABLE 9. AVERAGE TOTAL INST. THM AND PERCENT OF
INDIVIDUAL THM IN EACH EXPERIMENTAL DESIGN
EDI ED1R ED2 ED3 ED4
Total Inst. THM (yg/L) 129.4 140.5 135.0 125.5 150.4
TOC (mg/L) 9.8 8.6 8.3
Percent of Individual
Inst. THM
chloroform 62 50.8 47.4 45 44.8
bromodichloromethane 28.6 30.4 31.4 31 . 31.2
chlorodibromomethane 9.3 17.4 19.8 22 22.3
bromoform 0.1 1.4 1.4 2 1.7
In Table 9 the data show that the average total inst. THM
varied from 125.5 yg/L to 150.4 yg/L. TOC values in mg/L for
ED1R, ED3, and ED4 are also shown in Table 9 and they do not
correlate with the average total inst. THM values. There
appears to be a consistent trend in the data for the ratio of
bromine compounds to increase from EDl through ED4. The percent
of individual inst. THM data indicates that there was a shift in
the composition of the total inst. THM's, whereas chloroform
comprised 62 percent of the total inst. THM's during EDl, it
comprised only 44.8 percent during ED4. Other species increased
accordingly. Although the reason is unknown, the possibility of
slight salt water intrusion could exist.
TOC and THM FP Organics
Raw Water Source—
Average THM FP and TOC levels of raw water for EDl, ED1R,
ED3 and ED4 are shown in Table 10. There appears to be no
direct relationship between the concentrations of TOC and THM
FP. Comparisons of data in Table 10 show that the highest
average concentration of TOC was 9.8 mg/L in ED1R with a
corresponding average THM FP concentration of 659 yg/L and that
the lower average TOC concentration of 8.3 mg/L in ED4 was not
accompanied by a corresponding lower THM FP concentration.
H.T. Water Source—
Average THM FP levels of H.T. water for EDl, EDlR, ED3 and
ED4 are shown in the upper half of Table 10. The percent of THM
FP removed from raw water by lime softening is also shown.
40
-------
TABLE 10. TOO AND THM FP REMOVAL BY
LIME SOFTENING IN FULL SCALE
PLANT
ED
1
1R
3
4
ED
1
1R
3
4
Ave. THM FP
in raw water
>ug/L
816
659
591
662
Ave. TOC in
raw water
mg/L
-
9.8
8.6
8.3
Ave. THM FP
in H.T. water ~
yg/L
573
471
389
531
Ave. TOC in
H.T. water
mg/L
-
6.8
6.0
5.8
Ave . Percent
Removal
%
30
28
34
20
Ave. Percent
Removal
%
-
31
30
31
41
-------
In 404 days of testing over the two-year study, the weighed
average removed by lime softening was 27 percent.
Average TOC levels of H.T. water for ED1R, ED3 and ED4
appear in the lower half of Table 10 with percent removal data
from raw water by lime softening. The weighed average removal
was 31 percent. TOC and THM FP removals by lime softening
appear to correlate quite well with values of 31 percent and
27 percent respectively.
Raw water entering the Preston Plant had an average total
hardness of 245 ppm. Lime softening reduced the hardness to
about 85 ppm. Non-carbonate hardness averaged 6 ppm. A
decrease in carbonate hardness of 154 ppm is equal to 308 mg
of calcium carbonate floe per liter. In all four ED the average
THM FP removed fro'm raw water by lime softening was 205 yg/L.
This corresponds to 0.07 gram of THM FP adsorbed per 100 grams
of calcium carbonate floe.
Finished water Source—
TOC and THM FP removal data resulting from the combined
effects of breakpoint chlorination, residence in the chlorine
contact basin and sand filtration are shown in Table 11. TOC in
finished water (lower half of Table 11) is removed an average of
6 percent and the average THM FP removal is approximately 24
percent. However, the THM FP removal was in actuality simply a
conversion of a part of the THM FP in the H.T. water to a com-
bination of actual THM plus remaining THM FP (the sum of inst.
THM and THM FP is terminal THM) in finished water. A comparison
of the terminal THM values shown in parentheses in Table 11
shows this result. For example, in Table 11 the terminal THM
concentration for the H.T. effluent during EDI was 586 yg/L and
the terminal THM concentration for the finished water was 580
yg/L. No practical difference exists between these two average
valuers. Thus the H.T. water contained a THM FP concentration of
573'yg/L plus an inst. THM concentration of 13 yg/L while the
finished water contained a THM FP concentration of 448 yg/L and
an inst. THM concentration of 132 yg/L. Thus, whereas the com-
bination of THM FP and inst. THM concentrations for the two
locations were about equal, the THM's in the finished water
increased by about 120 yg/L while the THM FP decreased by about
the same amount. Thus the chlorination—contact basin—sand
filtration step really achieved no removal of precursor but
merely a conversion. The results are somewhat in line with the
low TOC removal.
TOC data for raw, H.T. and finished water are plotted for
ED1R, ED3 and ED4 in Figures 11, 12, and 13. THM FP data for
raw and H.T. water for ED4 are plotted in Figure 14. These
plots are presented at this point, mainly to show the variation
in values of these parameters from sample date to sample date.
Similar plots for the other ED appears later in the report with
42
-------
TABLE 11. TOG, TERMINAL THM AND THM FP
REDUCTION BY CHLORINATION, CONTACT
BASIN AND/OR SAND FILTRATION
^HVBHHBIWBB^MB4^HIHHMHallH
ED
1
1R
3
4
ED
1
1R
3
4
^••••^••••'••••••••••••i™ m ^•.^•••i i !• —•.man I ii.iin
Ave. THM FP
in H.T. water
yg/L
573 (586)*
471 (476)*
389 (397)*
531 (533)*
Ave . TOC
in H.T. water
mg/L
-
6.8
6.0
5.8
Ave. THM FP
in finished water
yg/L
448 (580)*
349 (495)*
274 (400)*
434 (584)*
Ave. TOC
in finished water
mg/L
-
6.1
5.9
5.4
HMHH^^HMm^MklMMmMMIIMIIIIIIiaHmHVBVBHII^HM^^B
Percent
Removal
%
22 (0)**
26 (0)**
30 (0)**
18 (0)**
Percent
Removal
%
-
10
2
6
* Terminal THM figures in parentheses
** Percent removal based on Terminal THM values
43
-------
11-
10-!
8—
n
-------
en
13-.
12-
11-
10-
9-
5 8-1
7-
u
o
5 -
4 -
v
v
1 -
Days
0
Av. 8.6
30% reduction
32% reduction
0 Raw water
—D— Hydrotreator water
— A— Finished water
04 11 18 25 32 39 46 53
Figure 12. TOC in raw, Hydrotreator and finished water (EDS).
-------
30% reduction
O
SAv. 5.8
35% reduction
Raw water
O— H.T. water
Finished water
Days 0
59 63 70 77 84 91
Figure .13. TOG in raw, Hydrotreator and finished water (ED4).
-------
900
800
Raw water
0— Hydrotreator water
20% reduction
Days
03 7 10
1721 2k 28 3135 38 421+5 4952 5659 6366 70 77 8084 8791 9498 101
Figure 14. THM FP in raw water and removal by lime softening (ED4) .
-------
adsorbent plots.
Other Parameters
I
As an aid in interpreting data for each ED, plant profile
information such as color, pH and turbidity of raw, H.T.
effluent, and finished water, as well as free and combined
chlorine for H.T. effluent and finished water were collected
and the raw data is available in Appendix B. Rainfall data and
chlorine levels of raw water are also available in Appendix B.
Rainfall and Chlorides—
The chloride concentration in raw water generally fe-llows
a cycle in response to the seasonal rainfall. Chloride levels
rise when rainfall is low and decrease when rainfall is high.
In southern Florida the wet season extends from May through
October with the heaviest rains usually occurring in May. The
dry season occurs between November and April. March and April
are usually very dry. Chlorides reach a low concentration in
September when rainfall is heavy and remain low until about
February as the ground water level subsides, then gradually
climb to a maximum in June.
Rainfall and TOC—
TOC levels in raw water may also be influenced by rainfall.
TOC data were collected only in ED1R, ED3, and ED4. The long-
est period of least rainfall in the two-year study period
occurred during most of ED1R which averaged the highest TOC
level of 9.8 mg/L. EDS and ED4 with more rainfall averaged 8.6
and 8.3 mg/L.
pH~
The pH of raw water remains constant at 7.2 ± .01 through-
out the year. The average values of the pH in H.T. effluent
and finished water, and the level of free chlorine in finished
water is shown in Table 12.
TABLE 12 . AVERAGE pH AND FREE CHLORINE (PPM) VALUES IN EACH
EXPERIMENTAL DESIGN
EDI
H.T. Fin. H.T. Fin. H.T. Fin. H.T. Fin.
PH
free chlorine
(ppm)
9.80
9.20 9.92 9.16
2.49
9.80 9.06 9.90
1.91
9.11
2.06
48
-------
There appears to be no apparent relationship between Total
THM values in Table 9 with THM FP values in Table 10, nor with
rainfall, chlorides, TOG, pH and chlorine data. Perhaps these
two parameters which are controlled by the chemistry of the
reaction of free chlorine with precursors is influenced by
subtle changes in pH, time and temperature which are beyond the
scope of our data.
Turbidity—
A record of H.T. effluent turbidity may be important
because with higher levels, additional amounts of precursors may
be carried over to the breakpoint chlorination step. However,
questions about the turbidity data prevented assessing the
relationship between the turbidity and THM FP. Tables of H.T.
turbidity results are included in Appendix A. In EDI and ED1R
the turbidity of the H.T. effluent fluctuated widely from day
to day. In ED3 and ED4 turbidity levels appear more uniform.
The average turbidity during EDI and ED1R was 10.2 NTU and 9.8
NTU respectively. In EDS and ED4 it was 3.4 and 4.6 respec-
tively. EDS and ED4 data may be misleading because during these
last two phases sampling was done only when the organic samples
were taken, whereas it was done daily during EDI and ED4. A
check of the operators' daily turbidity records during ED4
showed a high of 25 NTUs, and an average of 9.3 NTUs. These
values are more comparable to values reported for the first
phases. Thus, it is possible that the H.T. effluent turbidity
did not change substantially during the project.
Turbidity increased in the distribution system. Values
increased from an average of 0.32 NTU in finished water to an
average of 1.1 NTU in the distribution system sample.
Color—
Lime softening removed an average of 55 percent of the
color from raw water. Another 10 PCU is removed by chlorination
and sand filtration.
49
-------
BENCH SCALE STUDIES
This portion of the report presents data on the effects of
the three adsorbents evaluated on removal of specific HOC and
other organics as measured by THM FP and TOG in raw, H.T.
effluent and finished water. The pilot column configuration for
each ED was previously presented in Section VI.
Specific Halogenated Organics
Raw Water Source —
The effect's of adsorbents on raw water were studied in
EDI, ED1R, and EDS. Although the levels of the four THM were nil
or too low to be evaluated, the adsorption results for the
•specific compounds discussed below show that the XE-340 was more
efficient than GAG.. IRA-904 resin removal, as expected, was
poor and removal data for the raw source is only presented for
cis 1,2-dichloroethene and vinyl chloride to show typical
results with IRA-904 resin. Appendix A contains additional raw
data tables for all substances if further data are required.
cis 1,2-Dichloroethene — The HOC occurring in highest con-
centration in raw water was cis 1,2-dichloroethene. It will be
discussed first. Its general pattern will aid in interpreting
the data from some of the substances present in low concentra-
tions. Adsorption data appears in Table 13.
The adsorption data in Table 13 were obtained by integra-
ting the actual breakthrough curves which appear in Figures 15,
16, 17, and 18. The breakthrough point (B) and saturation point
(S) are shown on the curves. Using Table 13, a comparison of
the effectiveness of GAC versus XE-340 can be made on an equal
volume and equal weight basis at column saturation. At equal
volumes of adsorbent, XE-340 had 3.8 times and 3.4 times the
adsorptive capacity of GAC in EDI and ED1R respectively. At
equal weights of adsorbent, XE-340 had 3.2 times and 2.8 times
the capacity of GAC. Column breakthrough on GAC occurred at 21
and 16 days and column saturation at 69 and 73 days respectively.
Breakthrough occurred at 61 and 58 days for XE-340. Extra-
polated column saturation values of 280 and 242 days were
obtained for XE-340.
The MTZ for XE-340, 24 and 23 inches, is slightly more
than for GAC, 21 and 23 inches. It is apparent that GAC
(Figure 15) and XE-340 (Figure 17) , both allow low level passage
of cis 1,2-dichloroethene long before the value we have recorded
as the breakthrough point. The actual value in yg/L, which are
too low to plot on the "Y" axis scale, appear above the data
point. No number above a data point means nil concentration.
Consideration of this low level passage as breakthrough would,
of course, greatly change the recorded MT_ values.
50
-------
TABLE 13. cis 1,2-DICHLOROETHENE ADSORPTION DATA FROM RAW WATER
ED
1
1
1R
1R
3
3
•P -P
^
fl
»
•0
Feet
2.5
2.5
2.5
2.5
2.5
5
•P
Adsorbe
GAG
XE-
340
GAC
XE-
340
904
904
.p
Average
Influen
ng/i
21
21
29
29
26.9
26.9
2.5
5
A
Di
3
0
H
Column
Break th
Days
21
61
16
58
no
no
feet
feet
e
o
•H
Column
Saturat
Days
69
280
73
242
adsorj
adsorj
= o.:
= l.E
MT
z
Inch
21
24
23
23
tion
tion
6 met
2 met
C!
Test
Duratio
Days
117
117
122
122
53
53
er
ers
Gi
•H
n a -P
0) G m
•P 3 o>
c -i e*
H O
CJ o*
id J3 -H
•P o H
O n) 3
Grams
.219
.219
.316
.316
.127
.127
•O c *J
0) B H)
.Q P 0)
o o
(fl U i -P
f xi id
Grams
.084
.212
.115
.285
0
0
u
w
K^
O rj
4J 0
T3 Id -H
01 -P
x> c
%
65
61
61
62
C
X
l> 1| tji
0) O M
Di in C
O U
VI O
-o o -P
•a; H m
Grams
.048
.099
.065
.135
0
0
fj
01
^i ^
-------
en
to
90
80
70
60
50
40
30
20
10
Dayr
Raw water
— O— Raw water thru 2.5 ft. GAC (0.76 meter )
Raw water thru 2.5 ft. XE-340 (0.76 meter )
Av.21
Figure 15. cis 1,2-Dichloroethene in ITAW water and removal by p.70 netor
(2.5 feet) of GAC and 0.76 meter (2.5 feet) of ::E-340 (EDI).
-------
Raw water
)__Raw water thru 2.5 feet GAC (0.76 meter )
40
en
94 98
105
03 7 10 14 17 2124 2831 35 38 4245 4952 56 63 6670 73 77 84
Figure 16 . cis 1,2-Oichloroethene in itaw water and removal by 0.76 meter (2.5 feet)
of GAC (ED1R).
112
122
-------
Raw water
Raw water thru 2.5 feet XE-340 (0.76 meter )
Days
94 98
105 112
122
Figure 17. cis 1/2-Dichloroethene in raw water and removal by 0.76 meter (2.5 feet)
Of XE-340 (ED1R).
-------
en
ui
Av. 26.9
Raw water thru 2.5 feet (01:76 meter ) IRA-904
Raw water thru 5 feet (1.52 meters) IRA-904
7 11 Ik 18 21 25 28 32 35 39 42 46 49 53
Figure 18. cis 1,2-D.ichloroethene in »aw water and removal by 0.76 meter
(2.5 feet) and 1.52 meters (5 feet) of IRA-904 resin.
-------
In ED3, 0.76 (2.5 feet) and 1.52 (5.0 feet) meters of the
IRA-904 resin adsorbed no cis 1,2-dichloroethene. Discussion of
the other HOC will follow their order of presentation in Table 8.
Vinyl chloride— Analysis for vinyl chloride began on Test
Day 94 of ED1R after modifications were made to the present
equipment. Since vinyl chloride analysis did not begin until
toward the end of the test period, we do not have breakthrough,
saturation or MTZ information. The adsorption data obtained are
plotted in Figure 19. From Test Day 94 to 122, the average level
of vinyl chloride in the raw water was 0.80 yg/L and the average
level through GAC and XE-340 was 0.72 yg/L and 0.77 yg/L respec-
tively. Therefore, if the differences in the average are con-
sidered significant, from day 94 to 122 there was 10 percent and
4 percent removal respectively.
Results on 0.76 (2.5 feet) and 1.52 (5.0 feet) meters of
IRA-904 resin on raw water appear in Figure 20. Throughout the
entire two-year study, IRA-904 resin did not adsorb other HOC
from raw, H.T. or finished water. Therefore we read the
individual curves and averages in Figure 20 as indicating no
removal of vinyl chloride.
trans 1,2-Dichloroethene—The results of adsorbents for
removal of trans 1,2-dichloroethene from raw water are shown in
Table 14 and the breakthrough curves in Figures 21 and 22.
The HOC, trans 1,2-dichloroethene did not break through the
XE-340 column (Figure 22) during the 122-day test period, there-
fore, we cannot compare GAC and XE-340 at column saturation.
However, it is obvious that XE-340 has greater adsorptive capac-
ity for trans 1,2-dichloroethene than GAC, both on an equal
volume or equal weight basis. The GAC column allowed low level
passage (Figure 21) long before the time designated as break-
through. At the end of the test period, the GAC column had
adsorbed 82 percent of the entering trans 1,2-dichloroethene and
73 percent at extrapolated saturation. XE-340 had adsorbed 100
percent at the end of the test period.
1,1-Dichloroethane—Removal results for 1,1-dichloroethane
by adsorbents from raw water appear in Table 15 and Figure 23
respectively.
Table 15 data shows that breakthrough occurred at 21 days
and 94 days for GAC and XE-340 respectively. Figure 23 shows
that the GAC reached saturation at 94 days and that saturation
did not occur in the XE-340 column. However, again from the
Table 15 and Figure 23 data it is obvious that XE-340, both on
an equal volume and equal weight basis has greater adsorptive
capacity for 1,1-dichloroethane than GAC.
56
-------
i.o-i
VI
a
Ol
-J
.7-
.6-
.5 —
Raw water
— O— Raw water thru 2.5 feet (0.76 meter) GAC
•••O-" Raw water thru 2.5 feet (0.76 meter) XE-340
Days
Figure 19.
94 98 105 112
Vinyl chloride in raw water and removal by 0.76 meter (2.5 feet) GAC
and 0.76 meter (2.5 feet) XE-340 (ED1R).
122
-------
20-
15.
Ul
00
0)
4J
-H
tr>
10-
5-
Raw water
— O Raw water thru 2.5 feet IRA-904 resin (0.76 meter )
D Raw water thru 5.0 feet IRA-904 resin (1.52 meters)
0
Days 0
T
7
T
11
18
21 25 28
Figure 20. Vinyl chloride in raw water and removal by 0.76 and 1.52 meters
(2.5 and 5 feet) of IRA-904 resin (EDS).
-------
TABLE 14. trans 1,2-DICHLOROETHENE ADSORPTION DATA FROM RAW WATER
10
ED
1R
1R
5
ft
S
TJ
0)
CO
Feet
2.5
2.5
Adsorbent
GAC
XE-
340
Jj Average
j£». Influent
1.3
1.3
2.5
PI Column
to Breakthrough
66
none
feet =
I? Column
*eJ Saturation
142
= 0.7(
MT
z
Inch
16
raete
S? Test
"n Duration
122
122
r
o Total Entering
g Each Column
» During Test
.0142
.0142
" Total Adsorbed
§ by Each Column
01 at End of Test
.0017
.0142
£5 Adsorbed by Each
1 Column at
-------
4.9
o\
o
Raw water
Raw water thru 2.5 feet GAG (0.76 meter )
Days
03 7 10 Ik 17 21 2k 2831 35 38 k2 kS 49 52 56
122
Figure 21. trans 1,2-Dichloroethene in raw water and removal by 0.76 meter (2.5 feet)
of GAG (ED1R) .
-------
4.9
4-
a\
Raw water thru 2.5 feet XE-340
(0.76 meter )
Av.
o 4*
o
Days
7 10 14 17 2124 2831 3538 k2 k5 %9 52 56 6366 70 73 77 84 94 98 105 112 122
Figure 2.2.. trans 1,2-Dichloroethene in raw water and removal by 0.76 meter (2.5 feet)
6* XE-340 (ED1R).
-------
TABLE 15. 1,1-DICHLOROETHANE ADSORPTION DATA FROM RAW WATER
ED
1R
1R
5
1
•0
Column
w Breakthrough
21
94
= 0.
a? Column
"M Saturation
94
76 me
MT
z
Inch
23
ter
% Test
"M Duration
122
122
o Total Entering
g Each Column
w During Test
.0037
.0037
•O C 4J
v § to
xi 3 «j
M rH EH
0 0
in u
-------
(U
4J
•H
tn
3.
.5
Days
• 0ii Raw water
— O— Raw water thru 2.5 feet GAG (0.76 meter )
Raw water thru 2.5 feet XE-340 (0.76 meter )
3 7 10 14 17 21 24 28 31 35 38 42 45 49 52 56
Figure 23
63 66 7073 77
94 98
105 112
122
1,1-Dichloroethane in raw water and removal by 0.76 meter (2.5 feet)
of GAC and"0.76 meter (2.5 feet) of XE-340 (ED1R).
-------
l,l/l-Trichloroethane/ 1,2-dichloroethane, carbon tetra-
chloride—The removal results for the summed value of these
three HOC by adsorbents from raw water appear in Table 16 and
the breakthrough curves appear in Figures 24 and 25.
With the low average influent concentration of 0.104 yg/L,
and the spread in individual data points, estimation of break-
through and saturation times is difficult. Table 16, and Figures
24 and 25 show the estimated time of breakthrough for GAC and
XE-340 to be about 21 and 77 days respectively. Reported sat-
uration times are questionable. However, since the breakthrough
times for XE-340 is greater than for GAC, we would expect XE-340
to again have a greater adsorptive capacity than GAC.
Trichloroethylene—Adsorption data appear in Table 17 and
breakthrough curves in Figures 26 and 27.
Breakthrough occurred in 77 days and 96 days for the GAC
and XE-340 respectively. The saturation times cannot be extrap-
olated because of insufficient data points after breakthrough
to establish the slope of the curve. However, it is again
apparent that since XE-340 breakthrough occurred after GAC break-
through, XE-340 will have a higher adsorptive capacity than GAC.
Tetrachloroethylene—Adsorption of tetrachloroethylene by
adsorbents from raw water was studied only in ED1R. The influent
concentration to GAC and XE-340 columns was very low and erratic,
0.072 yg/L average for the first 31 days of the test and nil to
traces for the balance of the 122-day test. Influent level and
adsorption curves are shown in Figure 28. It is interesting to
note that even at this low concentration, both adsorbents do
adsorb a high percentage of the compound. No other conclusions
are drawn.
Chlorobenzene—Adsorption data appear in Table 18 and
adsorption curves in Figure 29. Influent concentration, plot-
ted in Figure 29, was very erratic. XE-340 removed all of the
compound for the entire test period. GAC removed essentially
all of the compound except for the three test dates shown in
Figure 29 when trace amounts passed.
p-Chlorotoluene—p-Chlorotoluene was studied in ED1R. The
erratic level of influent concentration is shown in Figure 30.
The average influent concentration was 0.38 yg/L. Both GAC and
XE-340 removed all of the compound throughout the test period.
o, ra and p-Dichlorobenzene—Adsorption from raw water by
adsorbents of the summed value of the three isomers of dichloro-
benzene was studied in ED1R. The influent concentration curve
appears in Figure 31. The average concentration was 1.1 yg/L.
Both GAC and XE-340 removed all of the compounds throughout the
test period.
64
-------
TABLE 16. 1,1,1-TRICHLOROETHANE, 1,2-DICHLOROETHANE,
CARBON TETRACHLORIDE ADSORPTION DATA FROM RAW WATER
en
ED
1R
1R
S
8-
Q
•O
&
Feet
2.5
2.5
Adsorbent
GAG
XE-
340
2.5
Jj Average
t* Influent
104
104
feet
Hi Column
n Breakthrough
21
77
= 0.
g Column
"« Saturation
98
98?
'6 me1
MT
z
Inch
24
6?
.er
g Test
"M Duration
122
122
Q Total Entering
§ Each Column
w During Test
.00113
.00113
•o g p
0) | in
ja P i V
f J3 <0
Grams
.00055
.0008
n Adsorbed by Each
| Column at'
in Saturation
.00055
.0008
%Adsorbed at
End of Test
51
71
^ % Adsorbed at
Saturation
56
88
n Adsorption per
g 100 gms. Adsorbent
01 at End of Test
.0003
.00037
S?
| Adsorption per
100 gms. Adsorbent
.0003
.00037
at Saturation
o
o
-------
.8 -
.7 -
.6 _
Raw water
•- Raw water thru 2.5 feet GAG
(0.76 meter )
Days
0 3
7 10 14 17 212+ 28 31 3538 4245 49 52 56 63 66 70 73 77 84 94 98 105 11
Figure 24, 1,1,1-Trichloroethane, 1,2-dichloroethane and carbon tetrachloride in
raw water, and removal by 0.76 meter
122
(2.5 feet) of GAC (ED1R).
-------
O Raw water thru 2.5 feet XE-340
(0.76 meter )
Days
17 21 24 28 31 35 38 42 45 49 52 56 63 66 70 73 77 84 94 98 105
Figure 25. 1,1,1-Trichloroethane, 1,2-dichloroethane and carbon tetrachloride in
raw water and removal by 0.76 meter (2.5 feet) of XE-340 (ED1R).
122
-------
TABLE 17. TRICHLOROETHYLENE ADSORPTION DATA FROM RAW WATER
oo
ED
1R
1R
5
ft
&
•a
£
Feet
2.5
2.5
Adsorbent
GAC
XE-
340
2.5
5 Average
"t* Influent
.14
.14
feet
m Column
w Breakthrough 1
77
96
= 0.7
p? Column
"a Saturation
i met*
MT
Inch
r
g Test
"a Duration
122
122
o Total Entering
w Each Column
01 During Test
.00153
.00153
£ Total Adsorbed
| by Each Column
01 at End of Test
.00137
.00148
-------
vo
Days
Raw water
• O— Raw water thru 2.5 feet GAG (0.76 meter )
63 66 70 73 77
112
122
0 3 7 10 14 17 21 24 28 31 35 38 42 45 49 52 56
Figure 26. Trichloroethylene in raw water and removal by 0.76 meter (2.5 feet), of GAG (ED1R)
-------
Raw water
•a--- Raw water thru 2.5 feet XE-340 (0.76 meter )
-j
o
CP
3.
Dav 0 3 7 10 14 17 21 24 28 31 3538 4245 49 52 56 63 66 70 73 77 84 94 98 105 112 122
Figure 27. Trichloroethylene.in raw water and removal by 0.76 meter (2.5 feet) of XE-340 (ED1R)
-------
.04 —
.03-
M
s
•H
.02-
.01-
0
Days
.3 .06 .1
tlltllt
.05
Average level
first 31 days
.072
Raw water
O — Raw water thru 2.5 feet GAC (0.76 meter )
Q.'.. Raw water thru 2.5 feet XE-340 (0.76 meter )
Av.
.023
0 '3 7 10 Ik 17 2121* 28 31 35 38 ^2 i*5 4952 56 63 66 70 73 77 &t 9^ 98 105 112 122
Figure 28. Tetrachloroethylene in raw water and removal by -°-76 meter (2.5 feet) of GAC
and 0.76 meter (2.5 feet) of XE-340 (ED1R).
-------
TABLE 18. CHLOROBENZENE ADSORPTION DATA FROM RAW WATER
NJ
ED
1R
1R
5
Ot
&
•O
01
m
Feet
2.5
2.5
.
Adsorbent
GAC
XE-
340
2.1
S Average
£t Influent
.19
.19
feel
CD Column
co Breakthrough
none
none
= 0.
I? Column
"M Saturation
76 me
MT
Inch
ter
§? Test
"a Duration
122
122
O Total Entering
| Each Column
w During Test
.0021
.0021
£ Total Adsorbed
H by Each Column
w at End of Test
.0021
.0021
n Adsorbed by Each
g Column at
ca Saturation
% Adsorbed at
End of Test
100
100
^% Adsorbed at
Saturation
0 Adsorption per
r\
| 100 gms. Adsorbent
w at End of Test
.001
.001
o
n
1 Adsorption per
100 gms. Adsorbent
at Saturation
8
-------
1.53
U)
Raw water
— 0—Raw water thru 2.5 feet GAC (0.76 meter )
Raw water thru 2.5 feet XE-340 (0.76 meter )
(all data points nil)
0
Days
0 3 "7 10 14 1721 24 2831 3538 4245 4952 56 63 66 7073 77 84 94 98 105 112
Figure 29. Chlorobenzene in raw water and removal by 0.76 meter (2.5 feet) of GAC and
0.76 meter (2.5 feet) of XE-340 (ED1R).
122
-------
1.
Raw water
Raw water thru
Raw water thru
all data points nil
(0.76 meter )
0
0 3 7 10 14 17 21 24 28 31 35 38 4245 49 52 56 63 66 70 73 77 84
Days Figure 30. p-Chlorotoluene in raw water and removal by 0.76 meter
0.76 meter (2.5 feet) of XE-340 (ED1R).
94 98 105 108 112 US 119 122
(2.5 feet) of GAG and
-------
ui
Raw water
Raw water thru 2.5 feet GAC
Raw water thru 2.5 feet XE-3401
all data points nil
(0.76 meter )
Days
31 ?5 38
£2 56
Figure 31.
18 an 35 38 H2 « k$ & 36 63 66 ?0 ^3 /7 * g s 105
o, m and p-Dichlorobenzene in raw water and removal by 0.76 meter
of GAC and 0.76 meter (2.5 feet) of XE-340 (ED1R) .
l'l9'l22
(2.5 feet)
-------
H.T. Water Source—
Low levels of the four THM, chloroform, bromodichloro-
methane, chlorodibromomethane and bromoform were present in H.T.
effluent water and results on these compounds will be discussed
first in this section. The synthetic organic removal results
for IRA-904 resin again were poor, as expected, and only cis
1,2-dichloroethene data is presented to illustrate the poor
adsorption. Additional synthetic organic data for the IRA-904
resin experiment (EDS) is available in Appendix B, Raw Data
Tables.
THM
Tables 19 and 20 are presented for chloroform and chloro-
dibromomethane adsorption by XE-340 from H.T. water. Figures 32
through 39 are presented to show the influent and XE-340
effluent concentrations for the four THM substances measured.
The variation in influent and effluent concentrations at these
low concentrations makes conclusions relative to breakthrough
times and saturation times difficult. In general the data show
that these substances appear in lower concentrations after
treatment of H.T. effluent water using XE-340 columns. Data
for the THM removal by XE-340 and GAC are presented later based
on a finished water influent.
cis 1,2-Dichloroethene—Adsorption data appear in Table 21.
Breakthrough curves appear in Figures 40, 41, and 42.
XE-340 was studied in EDI and ED1R. The breakthrough and
saturation performance of XE-340 on H.T. water in both ED
follows the pattern on raw water (Table 13). The adsorptive
capacity for XE-340 at 21 yg/L influent was calculated from the
H.T. water tests by using a log-log plot to compare directly
with raw water data at this influent concentration. Raw water
and calculated H.T. water values were 0.32 and 0.3 gram per
column and 0.117 and 0.108 cc per 100 grams respectively. From
Table 4 and 5, the total HOC in EDlR for raw and H.T. water
were 33.51 and 28.95 yg/L respectively. The percent of cis 1,2-
dichloroethene of the total HOC were 86.5 percent and 87.7 per-
cent respectively. Since the cis 1,2-dichloroethene to HOC
ratio is quite similar, we would expect little change in
adsorptive capacity due to competitive HOC. The closeness of
the raw and H.T. water results calculated at a common influent
level of 21 yg/L (0.32 and 0.3 gram per column) supports this
assumption.
The effect of competitive adsorption by TOC on the adsorp-
tive capacity for cis 1,2-dichloroethene can also be compared
using TOC data from the raw water and H.T. locations. The
average TOC concentration in the raw water was 9.8 mg/L and in
the H.T. effluent the concentration was 6.8 mg/L. Since there
was more TOC present in the raw water one might expect that the
adsorptive capacity for cis 1,2-dichloroethene would be
76
-------
TABLE 19. CHLOROFORM ADSORPTION DATA FROM H.T. WATER
ED
1R
1R
•p -P
s
t
•o
stion
152?
= 0.7
MT
Inch
ible
28.6
5 met
Test
Duration
Days
117
122
;r
o*
.3
M C -P
01 § W
4J 3 OJ
(3 H EH
32i
•P U M
o io 3
E-i W Q
Grams
.0355
.013
*o g p
xi i s
o "o *
CO U MH
O -O
rH 10 C
rt W W
O >i -P
EH A «
Grams
.0081
0
10
M
£ c
Adsorbed
Column at
Saturatio
Grains
.0085
•P 4J
a 01
Adsorbed
End of Te
<*>
*
62
*J
10 C
Adsorbed
Saturatio
<*>
%
52
c
A
M H -P
« o w
Qi W 0)
CSH
Adsbrptio
100 gms.
at End of
Grams
.00377
c
a
M M
0) O C
a w o
•0 -H
O U3
01 O
-g o 4J
Grams
. 00395
CC
.0027
-------
TABLE 20. CHLORODIBROMOMETHANE ADSORPTION DATA FROM H.T. WATER
oo
ED
1
1R
Q<
a
•o
s
Feet
2.5
2.5
Adsorbent
XE-
340
XE-
340
-
•p -P
Average
Influent
ug/L
1.0
.25
2.5
X
Oi
3
Column
Breakthro
Days
83?
?
:eet
c
Column
Saturatio
Days
none
none
= 0.7(
MT
z
Inch
mete
Test
Duration
Days
117
122
r
Oi
K
•H
•M C -P
fl) S in
4-1 3 0)
C H t<
W8 *
rH C
Id JG -H
•P O M
o « 3
fri W Q
Grams
.0047
.0027
•a c 4J
H
•H • O
J-l (0
ft S "O
H & C
O H
CO O
•d O 4J
< rH Id
Grams
.0022
.0013
C
5
fc M
(uoc
Qi 0) O
•0 -rt
c <: -p
Adsorptio
100 gms.
Grams
at Satura
CC
-------
\o
9 —
8 —
7 —
6 —
5-
„
a. 4 —
3-
10.3
H.T. water
. O— H.T. water thru 2.5 feet XE-340 (0.76 meter )
t
13.4
Days
7 9 13 20 32 6164 6971 7573 8385 9092 97 104 117
Figure 32. Chloroform in H.T. water and removal by 0.76 meter (2.5 feet) XE-340 (EDI)
-------
4-
00
O
0)
+J
•i-l
H
\
D"
2 -
1 -
H.T. water
— O — H.T. water thru 2.5 feet XE-340 (0.76 meter.)
03 7 10
17 21 2it 28 31 35 38 42 45 49 52 56 63 66 70 73 77 84
94 98
105 112
ays Figure 33. Chloroform in H.T. water and removal by 0.76 meter (2.5 feet) XE-340 (ED1R) .
122
-------
6.5
oo
Days
Average from 61
to 117 days=2.42
—O— H.T. water thru 2.5 feet XE-340 (0.76 meter )
Figure 34.
20 32 6164 6971 7678 8385 9092 97 104 117
Bromodichloromethane in H.T. water and removal by 0.76 meter (2.5 feet) XE-340 (EDI)
-------
00
H.T. water thru 2.5 feet
XE-340 (0.76 meter )
0
Days
63 66 70 73 77
84
3 7 10 14 17 21 24 28 31 35 38 42 45 49 52 56
Figure 35. Bromodichloromethane in H.T. water and'removal by 0.76 meter
(2.5 feet) XE-340 (ED1R).
94 98
105 112
122
-------
H.T. water
-------
0)
-p
•H
00
Days
7 10 14 17 21 24 2831 35 38 42451*9 52 56 6366 7073 77 84 91 98 1Q5
Figure 37. Chlorodibromomethane in H.T. water and removal by 0.76 meter
(2.5 feet) XE-340 (ED1R).
112
122
-------
oo
Cn
H.T. water
—-O— H.T. water thru 2.5 feet XE-340 (0.76 meter )
61 64 6971 7678 8385 9092 97
10«t
117
«-. U / _/ A vJ *-V v**- "* W-r w_f / 4. * w ^ i_* w»y w-^ ^u _/*_ J/ 1 U *t 11
y Figure 38 . Bromoform in H.T.Water and removal by 0.76 meter (2.5 feet) XE-340 (EDI).
-------
00
H.T. water
H.T. water thru 2.5
feet XE-340
(0.76 meter )
122
Figure 39. Bromoform in H.T. water and removal by 0.76 meter (2.5 feet) XE-340 (ED1R).
-------
TABLE 21. cis 1,2-DICHLOROETHENE ADSORPTION DATA FROM H.T, WATER
oo
-o
ED
1
1R
3
S
Ci
s
1
n
Feet
2.5
2.5
2.5
Adsorbent
XE-
340
XE-
340
904
2.5
g Average
t> Influent
20
25.4
24.1
feet
pi Column
in Breakthrough
32
28
no a
= 0.
$ Column
'Jn Saturation
274
270
Isorpt
'6 mei
MT
z
Inch
27
27
.on
.er
!? Test
"M Duration
3117
122
53
o Total Entering
1 Each Column
<" During Test
.209
< .277
.114
•Q G 4>
a) 1 oi
.0 3 4)
fc i-4 EH
o o
Bl O
-------
00
00
0)
4J
•H
Days
H.T. water
H.T. water thru 2.5 feet
XE-340 (0.76 meter )
7678 8385 9092 97
117
Figure 40. cis 1,2-Dichloroethene in H.T. water and removal by 0.76 meter
(2.5 feet) of XE-340 (EDI).
-------
40 -
H.T. water
H.T. water thru 2.5
feet XE-340 (0.76 meter
30 -
M
0)
4J
•H
00
VO
20
10
Av.
25.4
Days
03 7 10 I1* 17 21 24 28 31 35 38 42 45 49 52 56
63 66 70 73 77
94 98
105 112
119
Figure 41 . cis 1,2-Dichloroethene in H.T. water and removal by 0.76 meter
(2.5 feet) of XE-340 (ED1R).
-------
40
vo
o
30
CP
p.
20
10
0
Days
H.T. water
H.T. water thru 2.5 feet (0.76 meter )
Av. 24.1
0 4 7 11 lit 1821 2528 32 35 3942 4649 53
Figure 42. cis 1,2-Dichloroethene in H.T. water and removal by 0.76 meter (2.5 feet)
of IRA-904 resin (ED3).
-------
decreased in the raw water if TOC competes with the cis 1-2-
dichloroethene and HOC's adsorption sites. Since the adsorp-
tive capacity for cis 1,2-dichloroethene by XE-340 receiving
the raw water (0.32 gram per column) was not appreciably
different from the adsorptive capacity for cis 1,2-dichloro-
ethene by XE-340 receiving H.T. water (0.30 gram per column)
one might tentatively conclude that TOC does not compete with
cis 1,2-dichloroethene for the XE-340. The general lack of
removal of TOC by the XE-340 also may support this observation.
Vinyl chloride—Influent and effluent curves appear in
Figure 43. The average influent was 0.69 yg/L and the average
effluent was 0.5 yg/L. Since there was essentially no adsorp-
tion of vinyl chloride by XE-340 from raw water in the same
time period of 94 to 122 days, these values may not indicate
adsorption. If they do represent adsorption, 27 percent was
removed.
trans 1,2-Dichloroethene—Breakthrough curves appear in
Figure 44.
The influent concentration curves in Figure 44 illustrates
a condition which makes data interpretation difficult. For the
first 84 days of the test, the influent concentration averaged
0.11 yg/L. From day 84 to 112 the average was 1.2 yg/L.
Breakthrough occurred at the same time the influent concentra-
tion increased approximately tenfold. Research needs to be
done on adsorption of a single substance from pure water as well
as a mixture of many substances from pure and actual plant water.
Such research may indicate that breakthough under a given
condition may occur at approximately the same time for a wide
range of concentrations for a specific substance. If such is
the case, breakthrough might occur at approximately the same
time, but the level of breakthrough would be much less if the
influent concentration remained at the 0.11 yg/L level through-
out the test. In Figure 22, when the average raw water influent
concentration of trans 1,2-dichloroethene was 1.3 yg/L, the
same bed depth of XE-340 exhibited no breakthrough throughout
the 122-day test. Without the additional research mentioned
above, it is probably best not to attempt to explain such
differences obtained on such a complex system. We can conclude
that XE-340 removes all compound from H.T. water for a period
of 84 days even when the concentration averaged only 0.11 yg/L.
1,1-Dichloroethane—Adsorption data appear in Table 22.
The breakthrough curve appears in Figure 45.
The influent concentration, plotted in Figure 45, appears
quite erratic, averaging 0.13 yg/L. Estimation of the break-
through point from this curve is not too difficult, but the
extrapolated saturation point is questionable and was based on
adsorption data of this compound through the whole two-year
91
-------
1.0.
0)
4J
•rl
.9 -
.8
"7 IL1LJ
• /
.6-
.5
.4-
.3
.2-
Av.0.69
Av. 0.5
H.T. water
H.T. water thru 2.5 feet XE-340 (0.76 meter
.1
Days
-n -
Figure 43.
Ifl
1'05
Vinyl chloride in H.T. water and removal
by 0.76 meter (2.5 feet) of XE-340 (EDlR)
92
-------
vo
2 _
1.9-
1.8_
1.7-
1.6-
-P 1.4-
•H
1.2_
1.1 —
1 _
.9-
.8-
.6 —
.5-
.4-
.3-
.2-
H.T. water
H.T. water thru 2.5 feet XE-340 (0.76 meter )
n 0 3 7 10 14 17 21 ZH 28 31 35 38 HZ tb <+9 bZ bb bi bb /U /d // B4 yH y«
ys Figure 44 . trans. 1,2-Dichloroethene in H.T. water and removal by 0.76 meter
(2.5 feet) of XE-340 (ED1R).
98 ,105 112 122
-------
TABLE 22 . 1,1-DICHLOROETHANE ADSORPTION DATA FROM H.T. WATER
vo
ED
1R
Q.
Q
^D
Q)
fft
Feet
2.5
2.5
Adsorbent
XE-
340
eet
g Average
^i Influent
.13
= 0.7
PI Column
oi Breakthrough
94
6 met
p? Column
"M Saturation
131
ar
MT
z
Inch
9
^ Test
"ra Duration
122
o Total Entering
|j Each Column
w During Test
,00142
gj Total Adsorbed
• § by Each Column
01 at End of Test
.00129
n Adsorbed by Each
g Column at
n Saturation
.00131
^ %Adsorbed at
End of Test
91
^ % Adsorbed at
Saturation
92
o Adsorption per
w 100 gms . Adsorbent
a at End of Test
.0006
O
H
| Adsorption per
100 gms. Adsorbent
.0006
at Saturation
8
.00052
-------
Ul
H.T. water
— O—' H.T. water thru 2.5 feet XE-340 (0.76 meter )
0 3 7 10
Oays Figure.
14 17 21 24 28 31 35 38 42 45 49 52 56 63 66 70 73 77 84 94
5. 1,1-Dichloroethane in H.T. water and removal by 0.76 meter
(2.5 feet) XE-340 (ED1R).
-------
study. Such estimations are intended to aid overall data inter-
pretation and are not intended to be considered factual. We
can conclude that XE-340 on H.T. water as on raw water removes
all the compound even when present at the average low level of
0.13 yg/L for a period of time up to approximately 94 days.
1,1,1-Trichloroethane, 1,2-dichloroethane, carbon tetra-
chloride (summed concentration)—The average influent concentra-
tion, plotted in Figure 46, was only 0.022 yg/L, which is
probably too low to attempt conclusions. Periodically through-
out the test, the XE-340 column appears to allow some of the
material to pass while most is adsorbed.
Trichloroethylene—Adsorption data appear in Table 23. The
breakthrough curve appears in Figure 47.
Along with a very low influent concentration averaging 0.066
yg/L, we again have a much lower average of only 0.020 yg/L
entering for the first 84 days with the remaining number of
days averaging 0.240 yg/L. XE-340 removed all the compound for
the first 84 days except for trace amounts on two sample dates.
Tetrachloroethylene—The average influent concentration was
only 0.0025 yg/L. The influent and adsorption curves are shown
in Figure 48. The average concentration was higher during the
first part of the test and the XE-340 column allowed some pas-
sage even on initial start-up.
Chlorobenzene—The average influent concentration in ED1R
was 0.048 yg/L and individual data points are plotted in Figure
49. For the first 84 days of the test the average was 0.007
yg/L which was all removed by the XE-340 column. From day 94 to
the end of the test, the average entering was 0.08 yg/L, of
which a high percentage was removed.
p-Chlorotoluene—Seven samples of the 30 sampled during the
test period had low levels of p-chlorotoluene ranging from .008
yg/L to .540 yg/L. The influent data point curve appears in
Figure 50. All data points for samples through the XE-340
column showed nil concentration except for the sample on day
108, at 1.2 yg/L. No conclusions are made.
o, m and p-Dichlorobenzene—Adsorption data appear in
Table 24. Influent and adsorption curves appear in Figure 51.
Except for three sample points showing low levels of the
summed values of the compounds, XE-340 removed all the
compounds from H.T. water. Removal was essentially complete.
96
-------
06
—O— H.T. water thru 2.5 feet XE-340
(0.76 meter }
#
0 . ^ ^ / \l
Day^5 3 7 10 1417 2124 2831 35 3842 45 4952 56 63 66 7073 77 84 94 98 105
Figure 46. 1,1,1-Trichloroethane, 1,2-dichloroethane and carbon tetrachloride
in H.T. water and removal by 0.76 meter (2.5 feet) of XE-340 (EDlR).
112
122
-------
TABLE 23 . TRICHLOROETHYLENE ADSORPTION DATA FROM H.T. WATER
10
CO
ED
1R
P.
5T
Q
*O
Column
« Breakthrough
88
••f
= 0.'
° Column
"m Saturation
6 me1
MT
z
Inch
er
^ Test
"m Duration
122
o Total Entering
* Each Column
" During Test
.0007
•d c -P
o g 01
ja 3 o
8-3H
1^°
-^-g^
rH (!) C
«J W M
o s -P
EH 43 H)
Grams
.00057
o Adsorbed by Each
g Column at
in Saturation
^ %Adsorbed at
End of Test
81
^ %Adsorbed at
Saturation
o Adsorption per
| 100 gms. Adsorbent
01 at End of Test
.00027
CD
H
% Adsorption per
100 gms. Adsorbent
at Saturation
8
-------
IO
VD
M
(U
+J
•H
3-
.2-
.1-
H.T. water
H.T. water thru 2.5 feet XE-340 (0.76 meter.)
45 49 52 56
63 66 70 73 77
84 94 98 105 112
Days Figure 47. Trichloroethylene in H.T. water and removal by 0.76 meter (2.5 feet)
of XE-340 (ED1R).
122
-------
O
O
H.T. water
H.T. water thru 2.5 feet XE-340 (Q.76 meter .)
Days 03 7 10 Ik 17 2124 2831 35 38 42 4549 52 56 6366 7073 77
94 98
105 112
122
Figure 48. Tetrachloroethylene in H.T. water and removal by 0.76 meter (2.5 feet)
of XE-340 (ED1R).
-------
.09.
.08 —
.07 —
.06-
.05,
.04-
.03-
.02-
.01 -
Days
H.T. water
H.T. water thru 2.5 feet XE-340
j(0.76 meter )
03 7 10 1'ft 1721 2^ 2831 36 38 42^5 «t9 52 56 6366 7073 77
94 98
105 112
122
Figure 49. Chlorobenzene in H.T. water and removal by 0.76 meter (2.5 feet) of XE-340 (EDI)
-------
.4 -
.3 -
O
to
a)
•P
en
p.
.1 —
• H.T. water
O- —t-H.T. water thru 2.5 feet XE-340 (0.76 meter^
.54
Days
3 7 10 I«tl7 21 21+28 31 3538 4245 4^ 52 56 6*366 70 73 77 8% 9"+ 98 ids 108 U2. l'l5 l'l9122
Figure 50. p-Chlorotoluene in H.T. water and removal by 0.76 metex (2.5 feet) of XE-340 (ED1R)
-------
TABLE 24. O, m, AND p-DICHLOROBENZENE ADSORPTION DATA FROM H.T. WATER
o
OJ
ED
1R
3
8-
Q.
•8
«
Feet
2.5
Adsorbent
XE-
340
2.5
5 Average
f* Influent
.39
feet
(u Column
"5 Breakthrough
none
= 0."
jj? Column
*g Saturation
6 met
MT
z
Inch
sr
g Test
>oj Duration
122
o Total Entering
g Each Column •
<» During Test
.00425
•o e 4J
M
13 O
o w
1-4 m c
n) W W
O >i 4J
EH f> US
Grams
.00425
n Adsorbed by Each
g Column at
to Saturation
^ *Adsorbed at
End of Test
100
# % Adsorbed at
Saturation
n Adsorption per
g 100 gms. Adsorbent
01 at End of Test
.002
O
H
g Adsorption per
100 gms. Adsorbent
at Saturation
8
-------
2 —
1.5-
1 _
3
•H
.5 —
H.T. water
H.T. water thru 2.5 feet XE-340 (0.76 meter.)
2.57
2.41
Days 83 7 10 14 17 21 2k 28 31 35 38 42 45 49 52 56 63 66 70 73 77 84 94 98 105
Figure 51. o, m and p-Dichlorobenzene in H.T. water and removal by 0.76 meter
(2.5 feet) of XE-340 (ED1R).
112
122
-------
Finished Water Source—
Chlorination in the plant process between the H.T. effluent
sample point and the finished water sample point resulted in
large increases of the THM. Chloroform data will be discussed
first in this section. A discussion of cis 1,2-dichloroethane
follows chloroform so the two compounds occurring most fre-
quently and at the highest concentrations can be compared.
Chloroform—Adsorption data appear in Table 25. Influent
and adsorption curves appear in Figure 52, 53, 54, 55, and 56.
ED4, is shown in Table 25. Although the average influent level
varied from 57 for EDS to 67.3 yg/L for ED4, initial break-
through and saturation times were quite uniform at approximately
7 and 22 days and 8 and 23.5 days respectively. The MTZ also
were quite uniform at 21 and 20 inches respectively. Thus, The
variation in influent water conditions were such that they did
not greatly affect these parameters. This is more likely to
occur in ground water sources which are generally subject to
lesser variations in quality than river water sources. Also
maintenance of consistent contact times are more easily accomp-
lished in pilot systems as compared to full scale plants.
For these two study periods, a comparison of the grams of
chloroform adsorbed per 100 grams of adsorbent at saturation
(last column of figures in Table 25) with their respective
average influent level, shows that the adsorptive capacity of
GAC for chloroform increased (0.028 cc and 0.0358 cc) as the
influent concentration increased (57 to 67.3 yg/L), as predicted
by the Polanyi-Manes Theory, which is discussed beginning on
page 283. . In ED4, the adsorptive capacity in cc's per 100 grams
of GAC at saturation appears to have increased slightly with
increasing bed depth, 0.0449, 0.048, and 0.049 cc for 1.52
(5.0 feet), 2.29 (7.5 feet), and 3.05 (10 feet) meters of bed
depth respectively. This is to be expected because more
strongly adsorbed substances, both HOC and non-HOC, are removed
in the shallower bed depths thus reducing competitive adsorptive
effects. This point will be discussed later in more detail.
Chloroform adsorption from finished water by 0.76 meter
(2.5 feet) of XE-340 was studied during EDI, ED1R and ED2 (Table
25). Although the average influent concentration varied from
80.2, 69.3, and 64 yg/L, initial breakthrough and saturation
times for each ED were quite uniform at approximately 3 and 150
days respectively. The MTZ also were quite uniform at nearly
the entire column length of 76.2 cm (30 inches) which was
greater than that found for GAC. A comparison of the grams of
chloroform adsorbed per 100 grams of adsorbent at saturation
with their respective average influent level shows that the
adsorptive capacity of XE-340 for chloroform decreased (0.177,
0.148", and 0.134 cc) as the influent concentration decreased.
105
-------
TABLE 25. CHLOROFORM ADSORPTION DATA FROM FINISHED WATER
ED
1
1R
2
3
3
4
4
4
4
2.!
5
a
Feet
2.5
2.5
2.5
2.5
2.5
2.5
5
7.5
10
feet
Adsorbent
XE-
340
XE-
340
XE--
340
GAC
904
GAC
GAC
GAC
GAC
= 0
JJ 4J
Average
Influent
Vg/L
80.2
69.3
64
57
57
67.3
67.3
67.3
67.3
.76 It
CP
Column
Breakthrou
Days
3
3
0
7
8
29
49
72
eter
Column
Saturation
Days
156
150
150
22
23.!
49
76
98
5
MT
Inch
29
29
30
21
20
24.5
32
31.8
feet '
Test
Duration
Days
117
122
63
53
53
122
122
122
122
= 1.5:
tr»
c
•H
M d JJ
SI "
3 0)
C r-l E-«
3-s-S
o n> 9
tt W Q
Grams
.838
.736
.36
.27
avera<
.733
.733
.733
.733
meters
T3 C 4J
(U n W)
•S Jj £
0 0
in U . 4J
0" JQ it)
Grams
.497
.456
.247
.074
e incree
.094
.235
.376
.511
7.5 f(
u
M
is
Adsorbed b
Column at
Saturation
Grams
.569
.473
.429
.074
se of 1."
.094
.235
.376
.511
Set = 2.
i) 1 1
%Adsorbed a
End of Tes
*
59
62
69
27
5X
13
32
51
70
29 m>
.u
^Adsorbed a
Saturation
*
51
52
50
66
67
80
82
87
sters
C
-------
•H
iH
90_
80-
70_
60-
Finished water
thru 2.5 feet
XE-340
(0.76 meter )
Figure 52. Chloroform in finished water and removal by 0.76 meter (2.5 feet) of XE-340
IED1).
-------
o
00
90-
80-
70 —
60-
$50
•H
40 -I
30 -
20-
10
0
150 days
0 - Finished water
Finished water thru 2.5 feet XE-340 (0.76 meter*)
Days 03 7 10 14 17 2124 28 31 3538 4245 49 52 56 63 66 7073 77 84 94 98 l"05 " 112' ' 12~2
Figure 53. Chloroform in finished water and .removal by 0.76 meter (2.5 feet) of XE-340 (ED1R)
-------
O
VO
90
80
70
60
50
40
30
20
10
Finished water
Finished water thru 2.5 feet XE-340
(0.76 meter )
0 «t 7 11 I1* 25 32
5^ 5*6 S~
Days
Figure 54. Chloroform in finished water and removal by Q.76 meter f2.5feet) of XE-340 (ED2) .
-------
M
0)
O
200
175
150
125
100
Finished water
Finished water thru 2.5 feet IRA-904' resin (0.76 meter )
Finished water thru 2.5 feet GAC (0.76 meter )
Av. 100
25
0 4 7 11 14 1821 2528 3235 3942 46 49 53
Days Figure 55 . Chloroform in finished water and finished water thru 0.76 meter
2.5 feet) of IRA-904 resin and Q.76 meter (2.5 feet) of GAC (EDS)
-------
120
110
100
3.27
Finished water
O- 2.5 feet GAG (0.76 meter )
O" 5 feet GAG (1.52 meters)
-.7.5 feet GAG (2.29 meters)
10 feet GAG
35 38 1*21*5
10
• • a^^M VMMHMiri^^^^v
03 7 10
Days Figure 56
i t r
52 5659 63 66 7073 7780 8487 91 94 98101
112
122
Chloroform in finished water and in the effluent from 0.76, 1.52, 2.29 and
3.05 meters (2.5, 5, 7.5 and 10 feet) of GAG (ED4).
-------
Because the adsorptive capacity for GAG and XE-340 is
dependent on influent concentration, the influent concentrations
should be equal to compare the two adsorbents. If the three
XE-340 data points for adsorptive capacity and influent con-
centration are plotted on a log-log scale, the resulting plot
is a straight line. This also applies to our GAC data and to
other HOC on both adsorbents. This straight line applies to
the concentration range of interest in the particular water
tested, and only when the total HOC profile varies in intensity,
but not when there is a large change in the ratio of specific
HOC. It appears that the raw, H.T. effluent, and finished water
individually meet these requirements. However, the finished
water location could not be compared to the H.T. or raw because
the HOC ratios are not the same. Therefore, we can predict the
adsorptive capacity at saturation for a specific HOC from the
straight line plot for a given water. In this way, two adsor-
bents, or the same adsorbent run at times of different influent
concentrations can be compared at the same concentration. An
explanation of why these data points form a straight line on a
log-log scale plot and why predictions can be made is presented
in the section on the Polanyi Theory and Manes modifications,
page 283.
Using the log-log straight line plot for XE-340, we can
calculate the capacity at 67.3 yg/L and compare it directly with
the value for GAC at 67.3 yg/L in ED4. For XE-340 at 67.3 yg/L,
the grams adsorbed per column is 0.458 and the cc's adsorbed per
100 grams is 0.144. The GAC data (ED4) in Table 25 shows that
for 0.76 meter (2.5 feet) of GAC that the grams adsorbed per
column is 0.094 and the cc's adsorbed per 100 grams is 0.0358.
Therefore, XE-340 has 4.9 times (0.458 divided by 0.094) the
adsorptive capacity for chloroform in our finished water as GAC
per column where the volume of the two adsorbents are the same.
XE-340 had 4 times (0.144 divided by 0.0358) the capacity of GAC,
calculated on an equal weight basis of 100 grams of adsorbent.
This information, the data in Table 25, and the individual
breakthrough curves give a comprehensive view of chloroform
adsorption in our system.
As shown in Table 25, the effect of IRA-904 resin on fin-
ished water was studied in ED3. The level of chloroform leav-
ing the 0.76 meter (2.5 feet) deep column was 1.75 times the
level entering. A possible explanation is that the resin was
acting as a phase-transfer catalyst, accelerating the reaction
of free chlorine with precursors to form HOC in the empty bed
contact time of only 6.2 minutes. A review of phase-transfer
catalysts is available from Aldrich Chemical Company (5).
During ED1R on H.T. water, the influent concentration to
the XE-340 column for chloroform was 1.2 yg/L and the adsorptive
capacity at saturation of XE-340 was 0.0027 cc per 100 grams
(Table 19). The log-log straight line plot of XE-340 influent
112
-------
concentration and adsorptive capacity for finished water pre-
dicts 0.004 cc per 100 grains for finished water. When extra-
polated to 1.2 yg/L using the log-log plot of finished water
data. Based on the data in Table 5, during ED1R, chloroform
was 3.8 percent of the total HOC in H.T. water and 45 percent
of the total in finished water,(Table 6). Because of the higher
ratio of competing HOC in H.T. water we would expect less
chloroform adsorptive capacity in H.T. water. Thus, the observed
0.0027 cc value obtained on H.T. water is, as expected, less than
the finished water predicted value of 0.004 cc. As previously
mentioned, the ratio of specific HOC is an important factor
in predicting performance under various conditions.
cis 1,2-Dichloroethene—Adsorption data appear in Table 26.
Influent and adsorption from finished water curves appear in
Figures 57, 58, 59, 60 and 61.
XE-340, 0.76 meter (2.5 feet) deep, was studied in EDI,
ED1R and ED2. Column breakthrough and saturation time in ED1R
and ED2 are almost identical. The column breakthrough reported
in EDI is questionable since no data points were taken between
day 32 and 61 (Figure 57).
GAG, 0.76 meter (2.5 feet) deep, was studied in ED3 and
ED4. Breakthrough and saturation times are very close in both
ED, as shown in Table 26 and Figures 60 and 61. Using the log-
log plot to compare GAG and XE-340 at the same influent concen-
tration, at column saturation, XE-340 has an average of 3.1
times the capacity of GAG at equal volumes of adsorbent and an
average of 2.5 times per 100 grams. Calculated at the same
influent level of 21 ug/L, the adsorptive capacity of both GAC
and XE-340 is approximately 30 percent less in finished water
than it is in raw water. The TOG values for raw and finished
water average approximately 8.5 and 5.7 mg/L respectively (Table
4). Again, one might expect somewhat increased capacity as the
TOG values decrease. During ED1R the total HOC level in fin-
ished water was 159.6 yg/L and cis 1,2-dichloroethene was only
10 percent of the total (Table 6) as compared to the previously
stated 86.5 percent of the total HOC in the raw water for ED1R.
The 30 percent reduction in capacity in finished water probably
was due to the competitive adsorption of other HOC.
As was seen with chloroform, the adsorptive capacity of GAC
for cis 1,2-dichloroethene appeared to increase slightly per 100
grams of GAC as the bed depth increased. The IRA-904 resin
removed no cis 1,2-dichloroethene from H.T. or finished water
and none was generated.
Bromodichloromethane—Finished water adsorption data appear
in Table 27. Curves appear in Figures 62, 63, 64, 65 and 66.
113
-------
TABLE 26 . cis 1,2-DICHLOROETHENE ADSORPTION DATA FROM FINISHED WATER
ED
1
1R
2
3
3
4
4
4
4
2.5
J"|
+J
CM
s
•a
ID
m
Feet
2.5
2.5
2.5
2.5
2.5
2.5
5
7.5
10
feet
4J
q
3
ti
o
XE-
340
XE-
340
XE-
340
GAC
904
GAC
GAC
GAC
GAC
=0.7(
4J
Averag
Influe
W3/L
10.9
19.4
18.4
18.3
18.3
19.9
19.9
19.9
19.9
met*
£
o<
P
0
JE*
H 0)
8M
m
Days
61?
38
38
18
no c
17
59
101
none
r E
c
Q
•H
4J
Column
Satura
Days
128
191
190
53
dsorpt
66
119
171
feet
MT
z
Inch
16
24
24
20
ion
22
30
37
=1.52
B
O
Test
Durati
Days
117
122
63
53
53
122
122
122
122
mete
tji
a
•H
^1 f* ^J
m § m
•P P 0)
C. H EH
H O
Grams
.114
.206
.1035
,0867
.09
.217
,217
.217
.217
rs 7.5
•O C 4J
IV 1 V)
.0 9 o>
V< rH IH
O O
T3 O
rt fC
O t>
•H Id C
rt W W
O S 4J
EH ft a
Grams
.091
.166
.099
.058
0
.073
,159
.212
,217
Ceet=2 .
•g
IB
W
N
A G
•P O
•a n) -H
a» .p
,Q C fd
SIS
in .H -P
t< o m
rt U w
Grams
.092
.2
.187
.058
.073
.159
.242
29 metei
1 1 [ t
m ui
0)
tj t1*
a)
M O
O
ui -d
•ti C
rt u
%
80
81
96
67
0
34
73
98
100
S 1
*)
m e
*O *H
(U 4J
Adsorb
Satura
*
74
60
60
66
63
75
79
3 fe
£•
JS
V4 H -P
(U O Ul
ft U) 0)
'O EH
c rt
O
-------
Ul
M
+
DayS Figure 57. cis 1,2-bichloroethene in finished water and removal by 0.76 meter
(2.5 feet) of XE-340 (EDI).
117
158
-------
I-1
a\
0
Days
0 Finished water
— O — Finished water thru 2.5 feet XE-340 (0.76 meter )
b
-W-
-9-
9 99 <
/°N ^0-
i 0--0--0 °
| /
M 9 9*+ 28 31 35 38 42 45 49 52 56 63 66 70 73 77 84 94 98 105
Figure 58. cis 1,2-Dichloroethene in finished water and removal by 0.76 meter
(2.5 feet) of XE-340 (ED1R).
112
119
-------
40 -
35
30 -
VI
-------
Finished water
00
Finished water thru 2.5 feet
(0.76 meter ) IRA-904 resin
Finished water thru 2.5 feet GAC
(0.76 meter .)
Av. 19
25 28 32 35 39 42 46 49 53
Figure 60. cis 1,2-Dichloroethene in finished water and removal by 0.76 meter
(2.5 feet) of GAC and 0.76 (2.5 feet) of IRA-904 resin (ED3).
-------
Fin. water
— O— Fin. water thru 2.5' GAC
Fin. water thru 5' GAC
Fin. water thru 7.5' GAC
all data points zero - Fin.
water thru 10'
(0.76 meter )
(1.52 meters)
29 meters)
(3.05 meters)
Days
0 3 7 10 141721242831 353842454932 565963 667073 77 80 B>87 9194 981OJ105 112115 122
Figure 61 . cis 1,2-Dichloroethene in finished water and finished water thru 0.76, 1.52, 2.29
and 3.05 meters (2.5, 5, 7.5 and 10 feet) of GAC (ED4).
-------
TABLE 27 . BROMODICHLOROMETHANE ADSORPTION DATA FROM FINISHED WATER
N)
O
ED
1
1R
2
3
3
4
4
4
4
2.5
1
Feet
2.5
2.5
2.5
2.5
2.5
2.5
5
7.5
10
feet
Adsorbent
XE-
340
XE-
340
XE—
340
GAC
904
GAC
GAC
GAC
GAC
=0.7(
Average
Influent
WfL
37.1
42.7
42.4
39
39
47
47
47
47
met
j3
£
Column
Break thro
Days
20
20
22
16
14
42
73
105
sr !
a
Column
Saturatio
Days
216
210
can't
sxtrap
53
56
98
139
can ' t
extrap
feet
MT
z
Inch
27
27
21
22.5
34.3
42.7
=1.52
Test
Duration
Days
117
122
63
53
53
122
122
122
122
mete
S1
to a 41
O 8 111
sJU
M8*
-A**
JJ O H
O «S 5
E-i W Q
Grams
.388.
.465
.239
.185
averac
.512
.512
.512
.512
CS 7.5.
•d g 4>
41 g W
A 3 «
o"^
$r°
o *o
rH « C
td W w
Grams
.301
.36
.225
.121
e increa
.147
.293
.444
.51
Eeet=2 .
•g
(8
H
Z C
Adsorbed
Column at
Saturatio
Grams
.391
.439
.121
se of 1.1
.147
.293
.444
29 metei
JJ 4J
IB 10
Adsorbed
End of Te
df
%
78
77
94
65
3X
29
57
87
99
S 1
4J
at a
Adsorbed
Saturatio
Of
*
55
54
65
63
71
76
0 fe
G
Jl
H tl *J
(DOW
d w a>
dS^
O C
O H
0! O
-d o 4J
rtl ^1 (8
O W
01 O
•o o v
rtj f— (
-------
70 -i
60
ro
Finished water
Finished water thru 2.5 feet XE--340 (0.76 meter )
Days
0 7 9 13 20 32 61 6^ 6971 76 78 8385 9092 97 107 117
Figure 62. Bromodichloromethane in finished water and removal by 0.76 meter (2.5 feet)
of XE-340 (EDI).
-------
to
to
40
30
20
10
0
Days
Finished water
Finished water thru 2.5 feet XE-340 (0.76 meter )
i i 'i i—i i i—IT
03 7 10 14 17 21 24 2831 35 38 4245 49 52 55 63 66 7073 77
Figure 63.. Bromodichloromethane in finished water and removal by 0.76 meter
of XE-340 (ED1R).
(2.5 feet)
-------
52
to
to
40-
30-
r-l
P.
20 _
10-
t
Av. 42.4
Finished water
Finished water thru 2.5 feet XE-340
(0.76 meter )
B
Figure 64. Bromodichloromethane in finished water and removal by 0.76 meter (2.5 feet)
of XE-340 (ED2).
-------
to
Finished water
O— Finished water thru 2.51 IRA-904 resin
(0.76 meter )
......Q.™.. Finished water thru 2.5' GAC
(0.76 meter )
18 21 25 28 32 35 39 42 46 49 53
Figure 65. Bromodichloromethane in finished water and removal by 0.76 meter
of GAC and 0.76 meter (2.5 feet) of IRA-904 resin (ED3).
(2.5 feet)
-------
tO
cn
10
0
Finished water
— Finished water thru 2.5 feet GAC (0.76 meter ,)
Finished water thru 5.0 feet GAC (1.52 meters)
Finished water thru 7.5 feet GAC (2.29 meters)
Finished water thru 10.0 feet GAC (3.05 meters)
03 7 10 14 17 2124 2831 3538 42 45 49 52 56 59 6366 70 73 77 80 84 87 91 94 98 101 112
Days Figure 66. Bromodichloromethane in finished water and removal by0.76, 1.52, 2.29 and
3.05 meters (2.5, 5, 7.5 and 10 feet) of GAC (ED4).
122
-------
On finished water, calculated at the same influent concen-
tration of 39 yg/L, XE-340 had 3.4 times the capacity of GAG
per column and 2.8 times per 100 grams. It appears that for
both GAG and XE-340 the tests with higher influent concentra-
tions yielded higher amounts of bromodichloromethane adsorbed
than the tests with lower influent concentrations as expected
and demonstrated throughout for all substances. The adsorptive
capacity per 100 grams of GAG did not appear to change as the
GAG bed depth increased. Initial breakthrough of bromodichloro-
methane occurred in all runs on GAG and XE-340, but saturation
was not always reached. The IRA-904 resin caused an increase of
1.13 times the influent level.
Chlorodibromomethane—Adsorption data appear in Table 28.
Curves appear in Figures 67, 68, 69, 70, and 71.
On finished water at calculated equal influent concentra-
tions, XE-340 had 3.0 times the capacity of GAG per column and
2.5 times per 100 grams. There appeared to be no change in
capacity of GAG at 0.76 meter (2.5 feet) and 1.52 meters (5.0
feet) of depth. The IRA-904 resin did not reduce or increase
the level of chlorodibromomethane in finished water.
Bromoform—Adsorption data appear in Table 29. Curves
appear in Figures 72, 73, 74, 75, and 76.
Saturation by bromoform from finished water was not reached
consistently on XE-340. Saturation was reached in 0.76 meters
(2.5 feet) of GAG, but not in deeper GAG beds. There was no
removal and no increase of bromoform by the IRA-904 resin, in
finished water.
Vinyl chloride—Adsorption data appear in Table 30. Curves
appear in Figures 77, 78, 79, 80, and 81.
The curves for ED1R appear in Figure 77. The influent
average from day 94 to day 122 was 0.55 yg/L. The average
effluent from the 0.76 meter (2.5 feet) deep XE-340 column was
0.4 yg/L. As discussed on H.T. water, it is questionable
whether these figures represent adsorption. If adsorption did
occur, 27 percent was removed.
In ED2, the curves in Figure 78 show that column break-
through occurred very early. Since samples were taken at 0 days
and 4 days, we show a figure of 2 days for breakthrough.
In ED3, the curves in Figure 79 show that the vinyl chloride
effluent from a 0.76 meter (2.5 feet) deep IRA-904 resin column
averaged 3.9 yg/L over the test period. The influent average
was 5.4 yg/L. Based on our other IRA-904 resin data with other
HOC, we do not believe that this represents any adsorption. In
Figure 79, the average level of effluent from a 0.76 meter
126
-------
TABLE 28. CHLORODIBROMOMETHANE ADSORPTION DATA FROM FINISHED WATER
K)
-4
P P
ED
1
1R
2
3
3
4
4
4
4
2.5
j3
Di
S
•o
(1)
03
Feet
2.5
2.5
2.5
2.5
2.5
2.5
5
7.5
10
feet
4*
(U
J)
in
S
XE-
340
XE-
340
XE-
340
GAC
904
GAC
GAC
GAC
GAC
=0.7€
P
4) C
0> 0)
Id 3
to •-!
a
ft H
ug/L
12
24.5
26.7
27
27
33.6
33.6
33.6
33.6
met
tf*
&
3
o
il
H 01
Sto
m
Days
47
45
25
21
no <
21
59
98
none
:r S
c
o
•H
4J
9 3
rH +J
8 id
' W
Days
260
260
can't
extraj
76
dsorpt
101
174
can't
extrai
feet
MT
Inch
25
25
21.7
ion
23.8
39.7
=1.52
§
•H
P id
in to
1
»Q fl
P O
•0 id M
a) P
01 H P
•O O id
< CJ w
Grams
.165
.334
.117
.183
.349
can't
extrap
29 metei
P P
It) VI
a)
t» EH
•Q m
M O
O
W -O
•O fi
< w
%
87
89
99
82
0
50
95
99
100
S 1
£j
id c
o
•O -H
ID P
•P 1s
M b
O 3
W P
*O id
< U]
%
59
59
64
60
67
D fe<
C
ti t 1 Ij
ID o w
ft w
-------
-H
NJ
00
25-,
20-
Days
Finished water
Finished water thru 2.5 feet XE-340 (0.76 meter .)
61 6*t 6971 7678 8385 90 92 97
104
117
Figure 67. Chlorodibromomethane in finished water and removal by 0.76 meter (2.5 feet)
of XE-340 (EDI).
-------
to
\£>
10
03 7 10 14 172124 28 31 3538 4245 4952 56
63 66 70 73 77
94 98 105
ft 5
Figure 68. Chlorodibromomethane in finished water and removal by 0.76 meter (2.5 feet)
of XE-340 (ED1R).
-------
Finished water
Finished water thru 2.5 feet XE-340 (0.76 meter )
U
O
3.
35 —
30 _
25 J
20 -
15 -
10 _
5 -
B.
Figure 69. Chlorodibromomethane in finished water and removal by 0.76 meter (2.5 feet)
of XE-340 (ED2).
-------
(U
+J
•H
3.
50
20
10
Finished water
Finished water thru 2.5 feet IRA-904 resin (0.76 meter )
Finished water thru 2.5 feet GAG (0.76 meter )
0 47 11 14 1821 25 28 3235 3942 4649 53
76
ayS Figure 70. Chlorodibromomethane in finished water and removal by 0.76 meter (2.5 feet)
of IRA-904 resin and 0.76 meter (2.5 feet) of GAC (EDS).
-------
40
30
U)
NJ
M
fl
-H
Cn
3.
20
10
Finished water
Finished water thru 2.5 feet GAG
(0.76 meterJ
Finished water thru 5.0 feet GAG
(1.52 meters)
Finished water thru 7.5 feet GAG
mftt-.ftr.si
Finished water thru 10 feet
GAG - all data points nil
(3.05 meters)
0 _ ^ ^ ^ ^ __
T'T' TTo T4 17 21 T4 28 31 35 ^SjT 42 45 4952 5659 63 66 7073 7780 8487
Days • Figure 71. Chlorodibromomethane in finished water and removal by
3.05 meters (2.5, 5, 7.5 and 10 feet) of GAG (ED4).
91 94 98 101 112
0.76, 1.52-, 2.29 and
122
-------
TABLE 29. BROMOFORM ADSORPTION DATA FROM FINISHED WATER
co
co
ED
1
1R
2
3
3
4
4
4
4
2.
,rj
£
•d
0)
Feet
2.5
2.5
2.5
2.5
2.5
2.5
5
7.5
10
> fee
4.)
Adsorben
XE-
340
XE-
340
XE-
340
GAG
904
GAG
GAG
GAG
GAG
.=0.'
Average
Influent
U9/L
.1?
1.9
1.91
2.5
2.5
2.5
2.5
2.5
2.5
6 met
•
,£*
tn
3
O
Column
Break thr
Days
?
63
none
42
91
none
none
er
G
O
Column
Saturati
Days
can't
extra]
94
can't
extra^
5 fee
MT
Inch
16.6
'
t= 1.
Test
Duration
Days
117
122
63
53
53
122
122
122
122
>2 me
O*
G
•H
feet=
•g
fO
w
^1
n fl
4J O
M r-l 4J
rt U to
Grams
.013
2.29 met
JJ 4J
Id (A
(U
Adsorbed
End of T
*
%
100
100
0
56
98
100
100
ers
4J
n) C
O
Adsorbed
Saturati
*
*
72
10 J
C
M M P
u o n
Oi U) HI
TJ EH
G <
O MH
•H • O
P in
& §>*§
o w
(0 O
SO -P
in id
Grams
.005
.007
.0071
.0075
.005
.004
eet=3 .
(5
-------
.9 -.
.8. -
.7-
.6_
Finished water
O Finished water thru 2.5 feet XE-340 (0.76 meter,)
.5-
3
•H
.4-
average from day 92
to day 117 =0.47
Days
7 g 13 20 32 61 64 6971 7678 8385 9092
Figure 72. Bromoform in finished water and removal by 0.76 meter
XE-340 (EDI).
97
(2.5 feat) of
-------
-------
U)
M
-------
U)
0 Finished water
—-O-- Finished water thru 2.5 feet IRA-904 resin (0.76 meter )
—A— Finished water thru 2.5 feet GAC (0.76 meter )
all data points nil
Days
» Av. 2.7
— -O Av. 2.5
3235 39
0 if 7 11 lit IB 21 25 28 32 3S 39 42 US ?9 5~3
Figure 75. Bromoform in finished water and removal by 0.76 meter (2.5 feet)
of GAC and 0.76 meter (2.5 feet) of IRA-904 resin (ED3).
-------
00
Finished water
_O — Finished water thru 2.5 feet GAC (0.76 meter )
-- Finished water thru 5.0 feet GAC (1.52 meters)
Finished water thru 7.5 feet GAC - all data points nil (2.29 meters)
Finished water thru 10.0 feet GAC - all data points nil (3.05 meters)
17 21 24 28 31 35 38 42 45 49 52 56 59 6366 70 73 77 80 84 87 91 94 98 101
112
03 7 10
Figure 76. Bromoform in finished water and removal by 0.76, 1.52, 2.29 and 3.05 meters
(2.5, 5, 7.5 and 10 feet) of GAC (ED4) .
122
-------
TABLE 30. VINYL CHLORIDE ADSORPTION DATA FROM FINISHED WATER
u>
VD
ED
1R
2
3
3
4
4
4
4
2.5
,£•
ft
g
•a
«
Feet
2.5
2.5
2.5
2.5
2.5
5
7.5
10
feet
.P
Adsorben
XE-
340
XE-
340
GAC
904
GAC
GAC
GAC
GAC
•-o.it
Average
Influent
pg/L
.55
.6
5.7
5.7
8.4
8.4
8.4
8.4
mete
x
Cn
p
o
Column
Break thr
Days
2
no
3
10
17
35
r
a
0
Column
Saturati
Days
7
emovai
10
21
45
87
feet
MT
z
Inch
21
21
31
56
72
=1.52
Test
Duration
Days
122
63
53
53
122
122
122
122
mete:
Cn
c
•rl
0) 6 W
4J 9 0)
C H H
"is JB -S
•POM
o id P
EH W 5
Grams
;.0014
.0034
.027
.027
.092
.092
.092
.092
:s 7.5
•o g -P
» a in
V4 r-i EH
0 0
in U MH
id H M
O ^t -P
EH .Q id
Grams
0
.0049
.0117
.019
.045
:eet=2 .
A
o
id
w
^i
XI C
-P 0
•0 fl) -H
(U -P
^3 C «
si^
in H 4J
•O o id
< U 03
Grams
.00025
.0049
.0117
.019
.045
9 meter
•P -P
id co
IV
Adsorbed
End of T
<*>
%
0
5
13
21
49
S 1
•P
Id d
o
Adsorbed
Saturati
dP
%
34
65
74
56
69
fe<
•C
f^
rl M -P
-------
0)
-M
•H
tn
3.
.8 -i
.7 -
.6 -
.5 -
.4 -
.3 -
.2 -
.1 -
0 .
Days
Finished water
—O-— Finished water thru 2.5 feet XE-340 (0.76 meter.)
Av. .55
\ *• i j i i i i i r
9*t gS 105 112 122
Figure 77. Vinyl chloride in finished water and removal by 0.76 meter (2.5 feet)
of XE-340 (ED1R).
-------
0)
-p
-H
01
0
Days
Av. .6
Finished water
Finished water thru
2.5 feet XE-340
(0.76 meter )
0 k 7 11 1\ 25 32 53 56
Figure 78- Vinyl Chloride in finished water and removal by 0.76 meter (2.5 feet)
of XE-340 (ED2).
-------
—•— Finished water
-O— Finished water thru 2.5 feet IRA-904 resin (0.76 meter )
...p... Finished water thru 2.5 feet GAC (0.76 meter )
10 -
14.2
14.1
Av. 5.7
Av. 3.9
Av. 2.2
0
off 11 ll 1*8 \\ 25 Ye h ^5 & «fe 4*6 *f9 5$
Figure 79.. Vinyl chloride in finished water end removal by
0.76 meter (2.5 feet) of IRA- 904 resin and
0.76 meter (2.5 feet) of GAC (ED3).
142
-------
34.7
CO
25
n. 0 Finished water
— Q — Finished water thru 2.5 feet GAC (0.76 meter )
. Finished water thru 5.0 feet GAC (1.52 meters)
Days
3 7 10 l
-------
0)
-p
•H
Cn
3.
Finished water
— D— Finished water thru 7.5 feet GAC (2.29 meters)
A Finished water thru 10.0 feet GAC (3.05 meters)
15
10
Days
o 1 17 21 3t 2831 35~38 4245 4952 5659 63 66 7073 7780 8487 91 9498101 112
Figure 81. Vinyl chloride in finished water and removal by 2.29 and 3.05 meters
(7.5 and 10 feet) of GAC (ED4).
122
-------
(2.5 feet) deep GAG column was 2.2 yg/L. Based on our other
work with GAG, with other HOC, this probably represents some
adsorption, but it is probably incorrect to divide the GAG
effluent by the influent to get a percentage figure of 59 per-
cent removal. With vinyl chloride data, many more data points
over the test period would have to be taken to determine the
exact nature of adsorption and possible desorption (roll-over)
that may be occurring with this substance.
The curves in Figures 80 and 81 show the influent and
effluent concentration for four bed depths of GAG for ED4.
Breakthrough and saturation times for the four bed depths (as
seen from the curves and reported in Table 30) show a steady
increase with bed depth indicating that adsorption does appear
to be taking place. It is possible that after initial saturation
is reached on each column that roll-over occurs. That is, some
of the previously adsorbed vinyl chloride is desorbed. After
roll-over it appears that another period of adsorption may occur.
Considering the average influent and effluent from each bed depth
over the 122 day test period has questionable merit but may show
a trend. The average influent was 8.4 yg/L while the effluent
was 4.8 yg/L, 3.1 yg/L, 2.9 yg/L, and 2.8 yg/L respectively for
0.76 (2.5 feet), 1.52 (5.0 feet), 2.29' (7.5 feet), and 3.05
(10 feet) meters of GAG bed. This represents removal of 43 per-
cent, 63 percent, 64 percent and 67 percent.
trans 1,2-Dichloroethene—Adsorption data appear in Table
31. Curves appear in Figures 82, 83, 84, and 85.
Columns 0.76 meter (2.5 feet) deep of XE-340 were studied
in ED1R and ED2. In Figure 82, breakthrough was reported at 84
days and extrapolated saturation at 134 days. The same bed
depth of XE-340 was studied in ED2. In Figure 83, no break-
through occurred, which is as expected since the test duration
for ED2 was only 63 days, which was considerably less than the
breakthrough time of 84 days in ED1R.
In ED3, the 0.76 meter (2.5 feet) deep IRA-904 resin
column removed none of the compound (Figure 82). GAG columns,
0.76 meter (2.5 feet) deep, were studied in ED3 and ED4 (curves
in Figures 84 and 85) . The adsorption curve for GAG in Figure
84 except for two data points, indicates essentially complete
removal for the 53-day test period. No breakthrough is recorded
for EDS. We are probably justified in ignoring the two low
level passages at days 14 and 35 in Figure 84 since the adsorp-
tion curve for 0.76 meter (2.5 feet) of GAG in ED4 (Figure 85)
shows no breakthrough at all (all sample points nil) up to day 56.
Comparing XE-340 and GAG, both in 0.76 meter (2.5 feet)
deep columns, at the same influent concentration (using log-log
method) XE-340 had 2.0 times the adsorptive capacity of GAG at
equal volumes of adsorbent, and 1.3 times at equal weights.
145
-------
TABLE 31. trans 1,2-DICHLOROETHENE ADSORPTION DATA FROM FINISHED WATER
ED
1R
2
3
3
4
4
4
4
2.5
ti
O,
s
•d
0!
Feet
2.5
2.5
2.5
2.5
2.5
5
7.5
10
feet=
Adsorbent
XE-
340
XE-
340
GAC
904
GAC
GAC
GAC
GAC
0.7€
Average
Influent
W9/L
.54
.86
1.04
1.04
.77
.77
.77
.77
met«
x:
c
3
Column
Breakthro
Days
84
none
none
no
52
none
none
none
r
B
Column
Saturatio
Days
134
dsorpl
63?
feet
| i _^j
MT
z
Inch
11
ion
5
=1.52
Test
Duration
Days
122
63
53
53
122
122
122
122
mete
O1
•rl
H fi *J
01 G 01
•p 3 oi
B H EH
W O
id X! -H
Grams
.0059
.00484
.0049
.0049
.0084
.0084
.0084
.0084
:s 7.5
tJ fi 4->
O B B)
xi 9 > 4J
B X» i«
Grams
.0052
.00484
.0049
0
.004
.0084
.0084
.0084
:eet=2 .
•s
tf
W
rN
•Q a
Adsorbed
Column at
Saturatio
Grains
.0053
.004
29 metei
i) i)
id 01
^Adsorbed
End of Te
%
88
100
100
0
48
100
100
100
S 1
^j
id B
fe Adsorbed
Saturatio
%
90
93
) fe<
B
Ij ^J J t
von
t3 EH
B ri!
O UH
•rH « O
4-1 I/I
O W
01 O
t) 0 4J
ie£ rH Id
Grams
.003
.0023
.0028
0
.0023
.0034
.0016
.0012
!t=3.05
c
JS
M U
01 O B
a 01 o
•a -H
B < -P
O id
•H • to
4J 10 3
&&S
O ui
01 O
•a o .P
Grams
.003
.0023
meters
CC
.00238
.00183
-------
Finished water thru 2.5
feet XE-340 (0.76 meter )
99 90
7 10 li+ 17 21 24 28 31 3T5 38 42 45 49 52 56
63 66 70 73 77
94 98 10S
112
122-
DaySFigure 82 . trans 1,2-Qichloroethene in finished water and removal by 0.76 meter
(2.5 feet) of XE-340 (EDlR).
-------
Finished water
Finished water thru 2.5 feet XE-340 (0.76 meter )
1.5-
1 _
CO
0)
-P
•H
i-l
.5-,
Av. .86
? Y ll It °5 8 ?3 & 63
Y Figure 83. trans 1,2-BLchloroethene in finished water and removal by 0.76 meter
t
(2.5 feet) of XE-340 (ED2).
-------
4 -.
o
-P
-H
•H
tn
vo
Finished water
—"O— Finished water thru 2.5 feet IRA-904 resin (0.76 meter )
Finished water thru 2.5 feet GAG (0.76 meter )
Av. 1.04
Days
lilt 1821 25 28 3235 3942 4649 53
Figure 84. trans 1,2-Dichloroethene in finished water and removal by 0.76 meter
(2.5 feet) of GAG and 0.76 meter (2.5 feet) of IRA-904 resin (EDS).
-------
Ul
o
Finished water
Finished water thru 2.5 feet GAC
(0.76 meter )
Finished water thru 5.0 feet GAC, all data points nil
(1.52 meters)
Finished water thru 2.29 and -3.05 metprs
(7.5 and 10 feet) GAC, all data points nil
03 7 10 1<4 17 2124 28 31 3538 4245 4952 5659 63 66 7073 7780 84 8791 94 98101 112 122
Days Figure 85. trans 1,2-Dichloroethene in finished water and removal by 0.76, 1.52, 2.29 and
3.05 meters (2.5, 5, 7.5 and 10 feet) of GAC (ED4).
-------
The adsorption curve for the 1.52 meters (5.0 feet) deep
GAG column in ED4, Figure 85, shows three low level passages of
the compound at three widely separated sample dates (days 28, 91,
and 122). Since all the other data points showed nil concen-
tration, these three points are not considered as breakthrough.
All data points for the 2.29 (7.5 feet) and 3.05 (10 feet) meters
deep GAC columns were nil. Adsorption was complete over the full
test range for these two bed depths.
1,1-Dichloroethane—Adsorption data appear in Table 32.
Curves appear in Figures 86, 87, 88, 89, and 90.
XE-340, 0.76 meter (2.5 feet) deep, was studied in ED1R
and ED2. Taking both adsorption curves into account, Figures 86
and 87, we record breakthrough at 84 days in ED1R and 63 days in
ED2. In Figure 86 we also could consider the extremely low
level passage (.002 yg/L) on days 49, 63, and 66 as part of the
breakthrough curve, thus setting the breakthrough point at 49
days which is still in fair agreement with the 63 days in ED2.
The adsorption curve in Figure 88 for the IRA-904 resin, 0.76
meter (2.5 feet) deep indicates no removal of the compound in
EDS. GAC, 0.76 meter (2.5 feet) deep, was studied in ED3 and
ED4^ Considering the data as a whole, the breakthrough and
saturation times recorded in Table 32 are fairly close for ED3
and ED4. In ED4, Table 32, we observe a steady increase in
breakthrough and saturation time as GAC bed depth increases
despite the spread of data points in the adsorption curves in
Figures 89 and 90.
1,1,1-Trichloroethane, 1,2-dichloroethane, carbon tetra-
chlorlde (summed value)—Adsorption data appear in Table 33.
Curves appear in Figures 91, 92, 93, and 94.
XE-340, 0.76 meter (2.5 feet) deep, was studied in ED1R
and ED2. In Table 33, breakthrough at 98 days was reported for
ED1R. The adsorption curve in Figure 91 indicates complete
removal up to day 49, at which time a very low passage occurred
up to day 98 when passage increased sharply. On raw water,
Figure 25, a similar low level passage occurred from day 3 to
day 84, at which time the passage increased sharply. On H.T.
water, Figure 46, the low level passage started on day 7. It
appears that the trend for XE-340 is to allow some low level
passage of this summed group of substances from very early after
initial flow has begun, then to reach a period of increased
breakthrough varying from 98 days, 84 days and 66 days respec-
tively for ED1R, ED2 and ED4.
IRA-904 resin, 0.76 meter (2.5 feet) deep, was studied in
ED3, and the adsorption curve in Figure 93 indicates no removal.
GAC, 0.76 meter (2.5 feet) deep, was studied in EDS and ED4.
If the data point at day 3 (Figure 94) for the 0.76 meter
(2.5 feet) deep GAC column is not considered, both 0.76 meter
151
-------
TABLE 32. 1,1-DICHLOROETHANE ADSORPTION DATA FROM FINISHED WATER
Ul
K)
ED
1R
2
3
2
4
4
4
4
2.5
I
Q
Feet
2.5
2.5
2.5
2.5
2.5
5
."
10
feet
Adsorbent
XE-
340
XE-
340
GAC
904
GAC
GAC
GAC
GAC
=0.7*
Average
Influent
yg/L
.18
.8
.32
.32
.4
.4
.4
.4
mete
Column
Breakthrough
Days
84
none
16
no
28
42
52
77
•C I
Column
Saturation
Days
105
21
idsorp
49
56
59
87
feet
MT
z
Inch
6
7
:ion
13
15
11
14
=1.52
Test
Duration
Days
122
63
53
53
122
122
122
122
mete:
S1
Total Enteri
Each Column
During Test
Grains
.002
.001
.0015
.0015
.0044
.0044
.0044
.0044
rs 7.5
•a g -P
v i ia
A 9 <»
83*
01 U
%
90
88
79
88
94
94
) fe<
c
J3
Adsorption p
100 gms. Ads
at End of Te
Grams
.0009
.00047
.00025
0
.0008
.0005
.0004
.0004
!t=3.05
c
4)
A
01 .3 .3
d rtl JJ
0 n)
•H • M
4J 01 3
a E 4->
vj a> m
o ui
in o
•O O 4J
< I-H n)
Grams
.0009
.00025
.0008
.0005
.0004
.0004
meters
cc
.00077
.00021
.00068
.00042
.00034
.00034
-------
1.1
Ul
w
Finished water
Finished water thru 2.5 feet XE-340 (0.76 meter )
I
11
I I
I I
I I
I I
Days
03 7 10
17 21 24 28 31 35 38 42 45 49 52
94 98
105 112
122
Figure 86. 1,1-Dichloroethane in finished water and removal by 0.76 meter (2.5 feet)
of XE-340 (ED1R).
-------
tn
Finished water
Finished water thru 2.5 feet XE-340
(0.76 meter )
Figure 87. 1,1-Dichloroethane in finished water and removal by 0.76 meter (2.5 feet)
of XE-340 (ED2).
-------
tn
ui
11; 14 10 21 25 28 32 33 39
Finished water
— Q _ Finished water thru 2.5'
IRA-904 resin (0.76 meter )
...Q... Finished water thru 2.5' GAG
(0.76 meter )
Days
Figure 88. 1,1-Dichloroethane in finished water and removal by 0.76 meter (2.5 feet)
of GAC and 0.76 meter (2.5 feet) IRA-904 resin (ED3).
-------
2.0
1.5
a)
•P
•H
Ul
(Tl
1.0
.5
• Finished water
O —Finished water thru 2.5 feet GAC (0.76 meter )
...Finished water thru 5.0 feet GAC (1.52 meters)
Days
9194 98101 112
Figure 89. 1,1-Dichloroethane in finished water and removal by 0.76 and 1.52 meters
(2.5 and 5 feet) of GAC (ED4).
122
-------
Finished water
Finished water thru 7.5 feet GAG
(2.29 meters)
Finished water thru 10 feet GAC
(3.05 meters)
Days
28 U2 56 73 9H 108
Figure 90. 1,1-Dichloroethane in finished water and removal by 2.29 and 3.05 meters
(7.5 and 10 feet) of GAC (ED4).
122
-------
TABLE 33. 1,1,1-TRICHLOROETHANE, 1,2-DICHLOROETHANE AND
CARBON TETRACHLORIDE ADSORPTION DATA FROM FINISHED WATER
oo
4J 4-*
ED
1R
2
3
3
4
4
4
4
2.5
jCj
^3
Q<
S
•O
0)
n
Feet
2.5
2.5
2.5
2.5
2.5
5
7.5
10
feet
4J
fj
dsorbe
rij
XE-
340
XE-
340
GAC
904
GAC
GAC
GAC
GAC
=0.71
4>
verage
nfluen
< H
ug/L
.66
1.47
5.3
5.3
7.7
7.7
7.7
7.7
3 met
.a
Cn
3
olumn
reakth
u «
Days
98
none
28
38
87
none
none
sr 1
a
Q
•rH
C n)
o *$
o w
Days
58
73
199
i feel
MT
z
Inch
16
14
34
.=1.52
c
0
4-> n)
0) H
3
EH Q
Days
122
63
53
53
122
122
122
122
mete
&*
C
•rl
o S w
4J 3 fl)
f* f^ C>|
O «3 3
Grams
.0072
.0083
.025
average
.084
.084
.084
.084
rs 7.5
*O C 4-*
0) B O
•Q 3 (U
M •-< EH
0 0
*o o
•H S fl
n) W W
O >i 4J
EH ^J id
Grams
.0067
.0083
.02
increas
.038
.078
.084
.084
f eet=2 .
g*
o
16
W
s-
ja c
4J O
13 tj *H
0) 4J
XI g 10
(0 r-l 4J
^ O ifl
< U UJ
Grams
.0204
i of 1.5X
.038
.098
29 meteJ
-M -P
ID !/>
QJ
r^ g^
XI
%
93
100
80
45
93
100
100
•s ]
•u
%
74
76
72
0 fe
c
V
.q
H tl 4J
(!) O Ul
Q, w a)
e <
O (0 3
&&S
o w
in o
T3 O 4J
rtl rH nj
Grams
.0116
.022
.0278
meters
cc
-------
Ul
10
Days
Finished water
— O —• Finished water thru 2.5 feet XE-340 (0.76 meter )
3538, 42 45 49 52 56
94 98
105
112
122
Figure 91.
1,1,1-Trichloroethane, 1,2-dichloroethane and carbon tetrachloride in finished water
and removal by 0.76 meter (2.5 feet) of XE-340 (ED1R).
-------
Finished water
Finished water thru 2.5 feet XE-340 (0.76 meter )
•H
H
\
3-
Av. 1.47
Days
Figure 92. 1,1,1-Trichloroethane, 1,2-dichloroethane and carbon tetrachloride in finished
water and removal by 0.76 meter (2.5 feet) of XE-340 (ED2).
-------
A
11.7
. Finished water
— O — Finished water thru 2.5' IRA-904 resin
(0.76 meter. )
Finished water thru 2.5' GAC
(0.76 meters)
Av. 7.9
Av. 5.3
0
Days
58
Figure 93. 1,1,1-Trichloroethane, 1,2-dichloroethane and carbon
tetrachloride in finished water and removal by
0.76 meter (2.5 feet) of GAC and 0.76 meter
(2.5 feet)of IRA-904 resin (EDS).
161
-------
Finished water
O—-Fin. water thru 2.5' GAC
(0.76 meter )
-Fin. water thru 5.0' GAC
(1.52 meters)
Fin. water thru 7.5' GAC -
all data points nil
(2.29
Fin. water thru 10.0' - all data
points nil
(3.05 meters)
03 7 10 14 17 2124 28 31 35 38 42 45 4952 56 59 63 66 70 73 7780 84 87 91 9498 101
Days Figure 94.
112
122
1,1,1-Trichloroethane, 1-2-dichloroethane and carbon tetrachloride in
finished water and removal by 0.76, 1.52, 2.29 and 3.05 me-ters (2.5,
5, 7.5 and 10 feet) of GAC (ED4).
-------
(2.5 feet) deep GAG curves (Figures 93 and 94) are similar
because breakthrough times are 28 days and 38 days and satura-
tion times are 58 days and 73 days respectively. As expected,
as the GAG bed depth increases, the time to breakthrough
increases (Figure 94).
Trichloroethylene—Adsorption data appear in Table 34.
Curves appear in Figures 95, 96, 97, 98, 99, 100, 101 and 102.
XE-340, 0.76 meter (2.5 feet) deep, was studied in ED1R
and ED2. A breakthrough time of 94 days was determined from the
adsorption curve in Figure 95. This compares closely with a
reported breakthrough on raw water of 96 days and 88 days for
H.T. water. Saturation was not reached and at the end of the
test, 122 days, the adsorbent removed 96 percent of the compound
entering. The adsorption curve in Figure 96 for ED2 appears to
be the first contradictory data of the whole project. Tri-
chloroethylene breakthrough occurred on initial start-up and
XE-340 effluent from test day 14 to day 32 was higher than the
influent. All other results on trichloroethylene adsorption
from raw, H.T. and finished water in the two-year study were in
line and as expected with related substances and concentrations.
Results with a 0.76 meter (2.5 feet) deep column of IRA-
904 resin in ED3 are shown in Figure 97. The IRA-904 resin
effluent average concentration over the test period was 10 times
the influent concentration. It appears that trichloroethylene
is being generated by the column. The adsorption curve for
0.76 meter (2.5 feet) deep of GAG in ED3 is shown in Figure 98.
The average influent concentration was very low, 0.075 ug/L.
The GAG column effluent contained trichloroethylene on two
sample dates. From test day 28 to the end of the test at 53
days, the influent and effluent concentration was nil. Because
of the very low average influent, establishment of a break-
through or saturation time was not considered.
In ED4, the average influent concentration was 0.68 yg/L
and adsorption curves for 0.76 '(2.5 feet), 1.52 (5.0 feet),
2.29 (7.5 feet) and 3.05 (10 feet) meters of GAG are shown in
Figures 99, 100, 101 and 102 respectively. The reported break-
through and saturation times are open to question, but when con-
sidering all four bed depths, overall removals were 39 percent,
80 percent, essentially 100 percent, and essentially 100 percent
respectively in the 0.76 (2.5 feet),1.52 (5.0 feet), 2.29 (7.5
feet) and 3.05 (10 feet) meters deep columns. The respective
adsorption curves clearly show increased adsorption with
increasing bed depth. The effluent concentration for all data
points fox the 2.29 (7.5 feet) and 3.05 (10 feet) meters deep
columns was nil except for trace passage for two points and
points and three points respectively.
163
-------
K
H
I
Q
W
W
W
H
&
H
I
IS
o
H
EH
O
W
^
!H
W
EH
H
O
K
O
U
H
8
uoT^BJm^es 4«
^uaqjospv 'sui6 QOT
xad uoT^dJospv |
h
o
}ssi jo pus ^« a
HiaqjospY *sui6 QOT 3
jsd uoT^clzosptf u
uoT^Bjn^pg
^B psqaospv% *
3ssj 30 pug
5B paq;tosp\f% *
UOT^Bjn^BS 01
^B UUJtlXOD 1
qoeg Aq peqjospv u
^sai 50 pua ^B w
UBintoo tpea ^q §
psqaospv T^OI g
^ssi BUT ana
CM
CM
O
O
•
ro
CM
O
o
*
CM
CM
i-l
can't
extrat
•<*
en
rH
CM
•
1 0
as
in
CM
«
H
CM
en
O
O
m
VD
9)
jj
can't
alcula
0
r«
in
1 0
%%
in
*
CM
CM
0)
4J
14
^
(d
S
s
r-l
9
•ation t<
concent:
co
in
m
r-
o
•
%
in
.
CM
n
X
O
H
M
o
iM
co
0)
4J
1
CTI
CN
01
11
4J
0)
0)
-------
en
0
Days
63 66 7073 77
9+ 98
IDS
03 7 10 14 17 21 24 28 31 35 38 42 45 49 52 56
Figure 95. Trichloroethylene in finished water and removal by 0.76 meter (2.5 feet>
of XE-340 (ED1R).
112
122
-------
2.CU
cn
0 Finished water
O — Finished water thru 2.5 feet XE-340
(0.76 meter )
Av. .57
ays Figure 96. Trichloroethylene in finished water and removal'by 0.75 meter (2.5 feet)
of XE-340 (ED2).
-------
Finished water
- Finished water thru 2.5 feet IRA-904 resin (0.76 meter )
2 -
1 -
Days
0 4 7 1114 18 21 2528 3235 39 42 46 49 53
Figure 97 . Trichloroethylene in finished water and removal by 0.76 meter (2.5 feet)
of IRA-904 resin (ED3).
-------
0)
4J.
CTl
00
i.o—
Finished water
Finished water through 2.5 feet GAG (0.76 meter )
Av. 0.075
^Hft£HHM^B^^HH&*^fl^^^RM«HB^^ftM^^B^H^^^^^^^^BB«M^^
ww vv
-------
0 . Finished water
—-O- - Finished water through 2.5 feet GAC (0.76 meter )
Av. .68
-^Q 99
5 59 63 66 70 73 77 ffO
91
/\
/ \
9^ i b i iTsi'oei? 21*15 1*1
122
Days
Figure 99. Trichloroethylene in finished water and removal by 0.76 meter (2.5 feet) of GAC (ED4).
-------
-------
1 -
S-l
0)
+J
Finished water
Finished water through 7.5 feet GAG (2.29 meters)
Av.
-• .68
oo
o
0
Days
• + f y 99 99 99 99 » r"*N> ? O f ? » P 9 f> *» r^f^f » f » » »
7 IT) 1U 17 212«t2831 35384245 4952 5659 6366 7073 77 ffO 8487 91 9*t 98 TOlltTSlOai 12115119122
Figure 101. Trichloroethylene in finished water and removal by 2.29 meters (7.5 feet) of GAG (ED4).
-------
4J
-H
to
1 —
0 Finished water
•O — Finished water thru 10 feet GAG (3.05 meters)
Y-
CM
O
9999 99 P 0
03 7 10 14 17 21 2i+ 2831 35 38 W 45 4952 5659 6266 69 73 7780 8487 9194 98 101 105 108 112 115119"l22
Days Figure 102.Trichloroethylene in finished water and removal by 3.05 meters (10 feet) of GAG (ED4).
-------
Tetrachloroethylene—The influent concentration in all ED
was very low. The average in ED1R was 0.016 yg/L. The concen-
tration in ED2 and ED3 was nil for all test points. In ED4, all
points were nil to test day 108 and from day 108 to the end of
the test at 122 days, the average concentration was only 0.02
pg/L. The influent concentration and adsorption curves for
XE-340, 0.76 meter .(2.5 feet) deep, in ED1R are plotted in
Figure 103. The column allowed trace passage at test day 10,
14, 17 and 98. In EDS, while the influent concentration was nil
for all test points, the effluent from the IRA-904 resin column,
0.76 meter (2.5 feet) deep, had low levels on six of the six-
teen test points. In ED3, all effluent test points showed nil
concentration through a parallel GAG column 0.76 meter (2.5
feet) deep. In ED4, all effluent test points showed nil con-
centration through all four bed depths of GAG.
Chlorobenzene—Finished water adsorption data appear in
Table 35. Curves appear in Figures 104, 105, 106, and 107.
The influent concentration to the 0.76 meter (2.5 feet)
deep XE-340 column in ED1R was erratic as shown by the curve in
Figure 104. The influent concentration was nil on more than
half of the test days. The column effluent contained no chloro-
benzene during any of the test dates. In ED2, the average
influent concentration was lower, Figure 105. The XE^340 column
removed all chlorobenzene up to day 56, when a trace level was
noted in the effluent occurring after a peak in the influent
concentration. On test day 63, the column again had no chloro-
benzene in the effluent. In ED3, Figure 106, the IRA-904 resin
column had more chlorobenzene in the effluent than in the
influent, the average increase being 1.4 times. GAG, 0.76
meter (2.5 feet) deep, was studied in EDS and ED4. In EDS,
Figure 106, except for one data point, test day 4, all the
chlorobenzene was removed for the entire test period of 53 days.
When repeated in ED4, Figure 107, all the chlorobenzene was
removed to the breakthrough time of 84 days. It is questionable
whether saturation occurred as indicated in Figure 107, or if
the data point at day 115 was merely a spike in the effluent.
In deeper GAG columns, 1.52(5.0 feet), 2.29 (7.5 feet), and
3.05 (10 feet) meters, no trace of chlorobenzene was found in
the column effluents throughout the test period of 122 days.
p-Chlorotoluene—The influent concentration in ED1R, ED2,
EDS, and ED4 was essentially nil and the effluent from all
adsorbent columns was essentially nil.
o, m and p-Dichlorobenzene—Adsorption data appear in Table
36. Curves appear in Figures 108, 109, 110, 111 and 112.
The influent and adsorption curves for the summed value of
the three isomers in ED1R are shown in Figure 108. The XE-340
column, 0.76 meter (2.5 feet) deep, removed the isomers to
173
-------
.05
t
.0
Average first 7 days
of test .072
.Finished water
O —Finished water thru 2.5 feet
XE-340 (0.76 meter )
I \
I I
/ I
/ \
I
\
Days
3 7 10 14 17 21 2'8 31 35 38 4!> 4"5 49 52 56 63 6& 70 73 77 84 95 §8 1
Figure 103. Tetrachloroethylene in finished water and removal by 0.76 meter (2.5 feet)
of XE-340 (ED1R)
1T2
-------
TABLE 35. CHLOROBENZENE ADSORPTION DATA FROM FINISHED WATER
en
ED
1R
2
3
3
4
4
4
4
2.E
f .
•o
a
< H
Ug/L
, .14
.08
.8
.8
.86
.86
.86
.86
j met
t*
o>
3
lumn
eakthro
u m
Days
none
none
none
84
none
none
none
5r
c
lumn
turatio
O id
O' w
Days
115?
5 fee
MT
Inch
8
XL. 5:
§
•H
4i
P 10
W fc
o to
Q, in d)
C ^
O M-<
•rl • O
P B)
0 W
in o
TJ O -P
< rH id
Grams
.0007
.00021
.0022
.0043
.0027
.0018
.0013
3t=3'.OI
a
(V
0! O C
a w o
•O -H
a < -P
o id
P (IJ 3
& § -P
o w
ra o
•a o -P
ft >-t 10
Grams
.0043
meters
CC
.0039
-------
Finished water
Finished water thru 2.5 feet XE-340
(0.86 meter )
- all data points nil
0 3 7 10 14 17 21 24 28 31 35
Days Figure 104: Chlorobenzene in finished water and removal by 0.76 meter
94
98 105 112 122
(2.5 feet)of XE-340 (ED1R).
-------
Finished water
Finished water through 2.5 feet XE-340
(0.76 meter )
Av. .08
53 56 3
Figure lOS.Chlorobenzene in finished water and removal by 0.76 meter (2.5 feet) XE-340 (ED2)
-------
00
Finished water
Finished water through 2.5 feet IRA-904 resin (0.76 meter )
Finished water through 2.5 feet GAC (0.76 meter )
0 h 7 1114 1821 2528 3235 39 42 4649 53
DayS Figure 106..Chlorobenzene in finished water and removal by 0.76 meter (2.5 feet) of
GAC and 0.76 meter (2.5 feet) of IRA-904 resin (ED3).
-------
2.5
vo
Finished water
Finished water thru 2.5 feet GAC
(0.76 meter )
Finished water thru 5 feet GAC
(1.52 meters)
Finished water thru 7.5 feet GAC
(2.29 meters)
Finished water thru 10 feet GAC
(3.05 meters)
all points nil
0.5
03 7 10 14 17 21 24 28 31 3538 4245 49 52 5659 5355 70 73 77 80 8487 91 94 98 101
Days Figure 107. Chlorobenzene in finished water and removal by 0.76, 1.52, 2.29 and 3.05
meters (2.5, 5, 7.5, and 10 faet) of GAC (ED4).
-------
TABLE 36. o, m, AND p-DICHLOROBENZENE ADISORPTION
DATA FROM FINISHED WATER
oo
o
ED
1R
2
3
4
4
4
4
5
S1
Q
•a
4)
Feet
THRt
2.5
2.5
2.5
m-Dl
2.5
5
7.5
10
Adsorbent
E ISO
XE-
340
YT?—
340
904
CHLOF
GAC
GAC
GAC
GAC
|
I
i
I
A
\ ^^
I
fi 0 M
< H 1 CJ ffl
pg/L Days
1ERS SUMMED
.63 1 none
I
2.1 1 none
.3 1 inc
OBENZBNF.
nii conce
during er
nii conce
duifing er
nil cone*
during er
nil conce
during er
1
Column
Saturation
Days
reasec
ntrati
tire t
ntrati
tire t
ntrati
tire t
ntrati
tire t
KT
z
Inch
2.7X
on
est
on
est
on
est
on
est
Test
Duration
Days
122
63
122
122
122
122
0>
•H
^J f^ ^3
O S 01
4J 3 01
fi H EH
w o
Grams
.0069
.0068
«a g 4J
OEM
f> 3 i -P
EH J3 rt
Grams
.0069
.0068
J3
O
n)
H
^H
Adsorbed b
Column at
Saturation
Grams
-U 4-1
^Adsorbed a
End of Tes
%
100
100
Jj
^Adsorbed a
Saturation
%
-M
q)
o
M M 4->
4) O in
O, in
-------
TABLE 36. (CONT.)
CO
4J 4J
ED
4
4
4
4
4
4
4
4
5
I
4)
0)
Feet
P-D3
2.5
5
7.5
10
o-Dl
2.5
5
7.5
10
orbent
<
CHLOF
GAC
GAC
GAC
GAC
CHLOI
GAC
GAC
GAC
GAC
if
M rH
& C
< H
yg/L
3BENZE
.24
.24
.24
.24
DBENZE
.14
.14
.14
.14
JS
0>
umn
akthrou
•H
•rl
M C JJ
(D B I/I
JJ 3 0>
"8*
id tC *H
-POM
O it) 3
EH W Q
Grams
.0026
.0026
.0026
.0026
.0015
.0015
.0015
.0015
*O C JJ
•838
rl H EH
O O
W U MH
*o o
rt J3
O TJ
rH Id C
id w w
JJ
O S JJ
EH J3 id
Grams
.0026
.0026
.0026
.0026
.0015
.0015
.0015
.0015
A
o
n>
H
^i
13 id -H
a) JJ
X) d id
10 rH JJ
•o o as
< O W
Grams
JJ JJ
id 10
•8s
!°
KC H
OP
%
100
100
100
100
100
100
100
100
JJ
orbed a
uration
01 JJ
•O id
< CO
dP
%
a
0)
f%
M M JJ
ao to
n a)
O M-l
•H • 0
JJ Ul
O W
m o
•o o JJ
rtl rH Id
Grams
.0015
.00074
.00049
.00037
.00085
.00043
.00028
.00021
c
lt\
t^ ^J
2.8S
•a -n
•rl • rl
JJ W 3
a & JJ
o w
0) O
•O O JJ
Grams
CC
.0012
.00065
2.5 feet=0.76 meter 5 feet=1.52 meters 7.5 feet=2.29 meters 10 feet=3.05 meters
-------
00
to
Finished water
— O — Finished water thru 2.5 feet XE-340 (0.76 meter )
Days
03 7 10 14 17 21 2H 2831 35 38 42,^5
94- 98
105
115
122
Fiqure 108 o, m and p-Dichlorobenzene in finished water and r.emoval by 0.76 meter
(2.5 feet) of XE-340 (EDlR).
-------
00
CO
Av. 2.1
Finished water
Finished water thru 2.5 feet
(0.76 meter ) XE-340
Figure ]_09 °' m ant^ p-Dichlorobenzene in finished water and removal by 0.76 meter
(2.5 feet) of XE-340 (ED2).
-------
00
*>.
-
4-
3-
^H
2 -
••
1 -
«•
Days
^
\
Lx4
4
1 1 7 l'l
P-i m-iv*^
? 4'8
1
1
1
1
! 0 Finished water
1
1 __Q-_ Finished water thru 2.5 f eet IRA-904 resin
1 (0.76 meter )
1
1
I
I
1
1
1
1
1
1
1
1
! A
! 0 \ o AV. .8
I / \
VAy ^V^L, * v
m 18 21 25 28, 32 35 39 42 k& 49 52
Tin n^ m and r>— ni nVil nvnhpnT'.pne in f'inishfid wati<=r anrl i-oTn/->Tr=> 1 K™ n 7ft moi-eiT
(2.5 feet) of IRA-904 resin (ED3).
-------
CO
en
Finished water
(0.76 meter )
Finished water thru 2.5 feet GAG
Finished water thru 5 feet GAC
(1.52 meters)
Finished water thru 7.5 feet GAC
(2.29 meters)
Finished water thru 10 feet GAC
(3.05 meters)
all points nil
0
Days 03 71
Figure 111.
2«t 2831 35 38 "»2iT5 49 52 56 59 63 66 7077 80 8>4 87 91 94 98 101 112 122
p-Dichlorobenzene in finished water and removal by 0.76, 1.52, 2.29 and 3.05 meters
(2.5, 5, 7.5 and 10 feet) of GAC (ED4).
-------
00
Finished water
Finished water thru 2.5 feet GAG (0.76 meter )
Finished water thru 5 feet GAG
(1.52 meters)
Finished water thru 7.5 feet GAC
(2.29 meters)
Finished water thru 10 feet GAC
(3.05 meters)
all poxnts ni
"03 7 10 It 17 21 24 28 31 3538 42 45 4952 56 5963 66 7073 7780 8487 91 9498 101 112 122
Figure112. o-Dichlorobenzene in finished Water and removal by -0.76, 1'.52, 2.29 and 3..05 meters
(2.5, 5, 7.5 and 10 feet) of GAC (ED4).
-------
test day 122 when a low level of passage occurred. A large
spike in the influent concentration on test day 122 was noted.
XE-340, 0.76 meter (2.5 feet) deep, gave similar results in
ED2, as indicated by the curves in Figure 109. In ED3, Figure
110, the IRA-904 resin, 0.76 meter (2.5 feet) deep, appears to
have a higher level of isomers in the column effluent (2.7 times)
than in the influent.
In ED4, each isomer was reported separately. The influent
concentration and effluent concentration of m-dichlorobenzene
was nil for all GAC bed depths. The average influent concen-
tration, Figure 111, for p-dichlorobenzene was 0.24 yg/L. The
effluent concentration for all bed depths was nil for all test
points. The average influent concentration, Figure 112, for
o-dichlorobenzene was 0.14 yg/L. The effluent concentration for
all bed depths was nil for all test points.
Adsorption by XE-340—
We have presented much data showing that XE-340 resin has
approximately three times the adsorptive capacity for individual
HOC as GAC. The adsorptive capacity of adsorbents is usually
compared by measuring their capacity to adsorb butane gas. The
adsorbent which adsorbs the most butane is usually considered to
have more adsorptive capacity for substances like HOC than an
adsorbent which adsorbs less butane. Butane gas phase adsorp-
tion data for the GAC and XE-340 used in this study appear in
Figure 170. The method of obtaining these curves and their
interpretations are discussed in the section of the report in
which they appear. For the moment we will say only that the
curves show that based on butane adsorption data GAC should have
about seven times the adsorptive capacity of XE-340. This
figure is arrived at by projecting a vertical line, from 20 for
example on the "X" axis to the Butane Gas Phase (XE-340) and
Butane Gas Phase (GAC) curves, and then at the intersect points,
reading the corresponding cc adsorbed per 100 grams of adsorbent
on the "Y" axis. Values of 0.7 cc and 0.1 cc are obtained for
GAC and XE-340 respectively. Thus GAC adsorbs seven times as
much butane as an equal weight of XE-340. It has been shown by
Neely (6) that XE-340 adsorption does not follow the usual pat-
tern of physical adsorption on GAC because in addition to
adsorption in micropores, substances like HOC are taken into the
polymer matrix of the resin. The resin matrix swells as a
result of this incorporation. Thus, in our tests, XE-340
exhibits approximately three times the adsorptive capacity for
HOC as GAC because of adsorption into the polymer matrix. We
will show later in the report, that for substances which are not
so readily taken up by the polymer matrix, GAC has more adsorp-
tive capacity than XE-340, as predicted by the order of butane
adsorption data. These are the fulvic acid degradation sub-
stances which make up the bulk of substances measured by TOC and
THM FP analysis, generally known as precursors.
187
-------
Adsorptive Capacity and Competitive Adsorption—
Figure 113 is a plot of the total volume in cc's of all the
purgeable HOC in the finished water entering and leaving each
of the four GAG beds in the 122-day test, ED4. Integration of
these curves produces the three curves in Figure 114. Curve I
indicates the total volume in cc's of HOC entering each column.
Curve II indicates the total volume in cc's of HOC adsorbed by
each column and the cc's adsorbed per 100 grams of GAG. Curve
III indicates the total cc's adsorbed by 0.76 (2.5 feet), 1.52
(5.0 feet), 2.29 (7.5 feet) and 3.05 (10 feet) meters of GAC.
In 122 days, the finished water entering column 1 contained
1.235 cc of HOC, the entire first column adsorbed 0.382 cc or
0.217 cc per 100 grams of GAC. The next three columns each
received less and adsorbed less HOC than the preceding column.
When Curve II is projected to the "Y" axis the total adsorptive
capacity in cc's of HOC, 0.27 cc per 100 grams of GAC, for our
particular system is indicated. Adsorption per column decreased
as the concentration of adsorbate decreased. Therefore, for
maximum adsorbent use, the greatest amount of adsorbent possible
should be in contact with the highest possible concentration of
adsorbate. These data support the generally accepted view on
carbon use. The most efficient adsorbent usage would be a
continuous in-out GAC system. According to ,our data such a
system would have approximately 43 percent greater adsorptive
capacity than a single 3.05 meters (10 feet) deep bed. Multiple
beds in series also offer advantages in better carbon usage.
General practice is to operate the series system until the last
bed in series reaches the effluent criteria. The first bed is
then replaced with regenerated carbon and becomes the last bed in
the series arrangement. The new lead bed is the one that was
previously second in the series. In this way, four columns,
each 0.76 meter (2.5 feet) deep, arranged in series would have
about 31 percent more adsorptive capacity than one single column
3.05 meters (10 feet) deep. However, the actual design con-
figuration must take into consideration the higher capital costs
of these systems, as well as, the reduced operating costs. The
lowest total cost system is the one desired. Data described in
this study are important in achieving the most practical design
(7).
The curves in Figure 115 were obtained by integrating the
curves in Figure 113 at various sampling dates and more clearly
show the rates of saturation. It is evident from the top curve
in Figure 115 that column 1 will probably reach saturation at
the 0.27 cc value discussed above.
We have already presented data suggesting that the adsorp-
tive capacity of GAC and XE-340 for cis 1,2-dichloroethene is
30 percent less in finished water than in raw water, probably
due to competitive HOC adsorption. To aid the discussion of
competitive HOC adsorption, Figure 116 shows the adsorption wave
188
-------
125
oo
vo
Finished water enter-
ing column 1
— O — Column 1 effluent
•••D"" Column 2 effluent
A Column 3 effluent
— •—Column 4 effluent
Days
03 7 10 14 17 21 24 2831 35 38 4245 49 52 56 59 63 66 7073 7780 84 87 91 94 93 J 105 112
Figure 113. Cubic centimeters adsorbed by each GAC column for all halogenated compounds
added together (ED4).
122
-------
I-1
vo
o
w
85
K +J
rH
-------
217
•a
o
0)
"0
u
u
GAG Column #1
GAC Column #2
GAC Column #3
GAC Column #4
3 7 10 It 17 2124 2831 3538 4245 4952 5659 6366 70 737780 84 87 9194 98101 112 122
Figure 115. Cubic centimeters of total HOC adsorbed per 100 grams of GAC in GAC columns
#1, 2, 3 and 4 in 122 days (ED4).
-------
fronts defined by breakthrough and saturation time for several
HOC and Type II and III substances on a 0.76 meter (2.5 feet)
deep bed of GAG. The vertical height at the end of each HOC
curve represents the concentration in yg/L of each HOC. The '
concentration is read on the "Y" axis scale. The "X" intercept
for each curve is the days until breakthrough occurs and the
vertical projection of the end of each line to the "X" axis is
the time until saturation for that substance. The first number
in the parenthesis also gives the breakthrough time in days and
the second number gives the time to saturation. The Type II and
III substances shown in Figure 116 are defined in the discussion
of precursor removal that follows on page 192. Finished water
entering the first GAG column contains 7.8 mg/L Dissolved
Organic Matter (DOM) of Type II substances and 0.58 mg/L of Type
III, which corresponds to 4.69 and 0.36 mg/L respectively of TOG.
The curve for the Type II substances in Figure 116 merely repre-
sents the initial breakthrough and saturation time of 0 and 16
days respectively. The arrow at the end of the curve indicates
that the mg/L concentration cannot be shown on the yg/L scale
on the "Y" axis. The dash-line curve representing the Type III
substances in Figure 116 merely indicates that these strongly
adsorbed substances have an initial breakthrough and saturation
time of unknown values which are much beyond the breakthrough
and saturation time for all the HOC studied in this work. We
have calculated an average MTz for Type III substances of about
three inches. Type III substances, which are adsorbed by the
top portion of the GAG column, do not offer competition to
adsorption of all the HOC throughout most of the column. On the
other hand, Type II substances compete with HOC throughout the
column.
Of the five HOC shown in Figure 116, chloroform encounters
the least competitive HOG adsorption and bromoform encounters
the most competition. The results of this competitive adsorp-
tion are shown in Table 37. These values show the predicted
adsorptive capacity of each of five HOC from pure water at
saturation for the GAG used in this study. These values are
based on calculations using the Polanyi Theory and modifications
by Manes and Hofer which are described beginning on page 283.
Chloroform at the concentration in our finished water is
adsorbed 5 percent of its predicted capacity from pure water.
Bromodichloromethane, chlorodibromomethane and bromoform which
are present in progressively decreasing concentration, never-
theless, have higher predicted capacities than chloroform. How-
ever, the percent of predicted values steadily decreases (3.6,
1.8 and 0.12 percent) because of increasing competitive HOC
adsorption as indicated by the order of their wave front curves
in Figure 116. Cis 1,2-dichloroethene is present at a lower
concentration than chloroform and has a predicted adsorptive
capacity, as shown in Table 37, less than chloroform (0.46 com-
pared to 0.68). The observed adsorptive capacity is 6.5 percent
192
-------
DOM - Dissolved Organic Matter
TOC - Total Organic Carbon
Time in days to breakthrough
Time in days to saturation
64
80
96
Figure 116. Adsorption wave front defined by breakthrough and saturation time for HOC and
Type II and Type III substances thru 0.76 meter (2.5 feet) of GAC (ED4).
-------
of the predicted value. These data appear to present no problem
until the wave front curve for cis 1,2-dichloroethene is con-
sidered. The Manes-Hofer scale factor calculated from the
refractive index of a compound (described on page 284 ) in general
predicts the order of elution of the HOC compounds in our water
both from the GC Tenax column used for their analysis and from
the bench scale GAG columns on the water lines. On the GC Tenax
analysis column, cis 1,2-dichloroethene elutes before chloroform.
On the GAG columns on the water lines, cis 1,2-dichloroethene
elutes after bromodichloromethane (Figure 116). Because of its
wave front position in Figure 116, cis 1,2-dichloroethene
encounters more competitive HOC adsorption than chloroform, yet
in Table 37 the percent of predicted adsorption is higher than
for chloroform. Apparently the refractive index scale factor
does not predict the stronger than predicted adsorption shown by
this compound. Manes (private communication) has indicated that
carbon tetrachloride exhibits less adsorption than predicted by
its scale factor. Molecular geometry and other physical chemical
properties such as dipole moment can result in divergence from
the scale factor predicted value. Carbon tetrachloride has a
dipole moment of zero and due to its molecular structure presents
a small surface area to the carbon surface compared to its molar
volume. This results in less adsorption than predicted by the
Manes scale factor based on refractive index. Perhaps the d6uble
bond in cis 1,2-dichloroethene has greater affinity for the
carbon surface than a single bond resulting in greater adsorp-
tion than predicted. Our data also indicate that trans 1,2-
dichloroethene and trichloroethylene exhibit more adsorption
than predicted and a shift in the wave front as found with
cis 1,2-dichloroethene.
TABLE 37.' OBSERVED ADSORPTIVE CAPACITY OF 100 GRAMS OF GAC FOR
FIVE HOC FROM FINISHED WATER COMPARED TO THE POLANYI-
MANES PREDICTED VALUE FOR EACH COMPOUND FROM PURE
WATER (adsorbed by 0.76 meter [2.5 feet] of GAC)
Polanyi-Manes Observed
Predicted Capacity Percent
Capacity from from of
Pure Water Finished Water Predicted
cc cc
cis 1,2-Dichloroethene
Chloroform
Bromodichloromethane
Chlorodibromome thane
Bromoform
0.46
0.68
1.14
2.2
1.7
0.029
0.032
0.04
0.04
0.002
6.5
5.0
3.6
1.8
0.12
194
-------
The percent of predicted values in Table 37 and their
relationship to the wave fronts shown in Figure 116 apply only
for our finished water as it appeared during this study. Any
treatment plant modification that changed the ratios of the
component HOC would change all values in Table 37. We have
indicated that the maximum adsorption capacity of 0.76 meter
(2.5 feet) of Filtrasorb 400 GAG for total HOC in our finished
water was approximately 0.27 cc .per 100 grams (Figure 114). We
expect that if the concentration of the four THM were reversed,
the maximum adsorptive capacity would rise considerably above
the present 0.27 cc capacity. All observed capacities and per-
cent of predicted values would change in Table 37.
In Table 37, the observed adsorptive capacity of chloroform
per 100 grams of GAC is 0.032 cc, only 5 percent of the pre-
dicted value for pure water. It would be interesting to know
how much of this reduction is due to the competitive adsorption
of DOM ( or TOC) and how much is due to other HOC. We can
probably determine this from our existing data. Chloroform made
up approximately 98 percent of the entire volume of all the HOC
entering GAC column 4 during ED4 for up to 98 days. We there-
fore can conclude that this column was receiving only chloroform
from finished water. Integration of the 3rd and 4th GAC column
curves in Figure 56 shows that the 4th GAC column adsorbed
0.093 cc of chloroform at saturation (98 days), which corres-
ponds to 0.053 cc per 100 grams of GAC. This value is 8 percent
of the value that the Polanyi-Manes Theory predicts should be
adsorbed from pure water. Therefore, DOM (or TOC) substances
accounted for 92 percent of the reduction in adsorptive capacity
of chloroform from finished water compared with its capacity
from pure water. The competitive effect of other HOC in fin-
ished water further reduces the capacity to 95.3 percent of the
capacity for pure water (Table 37). GAC column 1, had an adsorp-
tive capacity for chloroform of 0.0358 cc/100 grams of GAC.
Column 4 had a capacity of 0.053 cc/100 grams of GAC, which is
32 percent higher than column 1. This percentage value is very
close to the 30 percent increased capacity found for cis 1,2-
dichloroethene in raw water compared with finished water, which
we feel is due to the absence of competitive adsorption of HOC.
As shown in Table 25, the four bed depths of GAC in ED4
show increasing values of adsorptive capacity for chloroform per
100 grams of GAC; 0.0358, 0.0449, 0.048, and 0.049 cc respec-
tively for each column. This was because the adsorptive capac-
ity for chloroform increased in each consecutive column as the
HOC competitive adsorption decreased.
The percent of predicted adsorptive capacity decreased as
the number of bromine atoms in the adsorbed molecular increased
(Table 37). The adsorptive capacity of these bromine-containing
THM on GAC from pure water has not been determined experimentally.
The Polanyi-Manes predicted values given in Table 37 for these
195
-------
three compounds are thus not confirmed values. Therefore, we do
not know if the decrease in percent of predicted value for these
three compounds as shown in Table 37 is caused by only increas-
ing competitive HOC adsorption or if part is due to an error in
the predicted value itself. For example, it is possible that the
predicted values are too high because of steric exclusion result-
ing from the greater bulk of bromine atoms compared to chlorine
atoms. Further work is needed to determine the adsorptive capac-
ity of these molecules from pure water.
The amount of total cubic centimeters of HOC adsorbed in a
given column was always less than in a preceding column (Figure
114). It might first appear that the explanation for this is
simply that, as expected with a single HOC in water, the adsorp-
tive capacity decreases as the concentration in the column
influent decreases. It is not that simple. We have already
shown that if a single HOC in our system reached saturation in
all four columns (as chloroform did), the adsorptive capacity
actually increased from column 1 to column 4. To explain the
results shown in Figure 114, one must consider the contribution
of each individual HOC. Column 1 adsorbed more HOC than the
other columns because it was receiving more of certain individual
HOC that had higher adsorptive capacities (Table 37).
Throughout this study, we did not observe much roll-over
(displacement of an adsorbed substance by a more strongly
adsorbed substance). There are several possible explanations
that may apply either alone or collectively. Unknowns such as
this make prediction of adsorptive capacity difficult if the
ratios of HOC in our system were to change greatly. We are not
without predictive capabilities, but we also recognize the
limitations of present methods and the direction further research
should take to improve this capability.
TOC and THM FP Organics
Precursors in this study were measured by TOC and THM FP
analysis. Both methods measure a complex family of compounds
instead of a specific substance as is the case for analysis of
individual HOC. TOC includes some substances that are not THM
precursors at all. It is likely that only a small portion of
the TOC represents precursors for THM. If a single test method
measured the whole family of organics present, the resulting
adsorption breaktheough curve would look quite different from
that of an individual organic breakthrough curve. This effect
is shown in Figure 113 where the total cubic centimeters of the
HOC mixture entering and adsorbed by 0.76 (2.5 feet), 1.52
(5.0 feet), 2.29 (7.5 feet) and 3.05 (10 feet) meters of GAC bed
are plotted. After initial breakthrough the frontal adsorption-
zone has a steep slope, similar to that observed by a specific
HOC. However, the total HOC curve then changes to a gradual
196
-------
slope .that approaches a plateau in some cases. During the 122-
day test period, none of the total HOC curves reached the
influent level curve. As the GAG bed depth increased, the
difference between effluent and influent curves increased. This
was because the more strongly adsorbed compounds were adsorbed
in various degrees up to complete adsorption. The total HOC
curves in Figure 113 representing 19 specific compounds are
similar to those obtained when TOG and THM FP data, which
measure a larger number of substances, are plotted except that
the plateaus of the TOC and THM FP curves are more pronounced.
Examples of such breakthrough are shown in Figures 117 and 118.
It is possible that the plateaus exhibited on TOC and THM FP
breakthrough curves are the result of biological degradation
occurring along with adsorption. Thus the apparent plateaus on
the curvets for the TOC and THM FP may have partially different
explanations than those for the total HOC.
Figure 117 shows the THM FP in yg/L entering the first GAC
column and in the effluent from each of the four 0.76 meter
(2.5 feet) deep columns connected in series. Figure 118 shows
TOC data in mg/L of effluent from the same four GAC columns.
In Figure 118, the TOC breakthrough curves show three distinct
areas. There is some TOC breakthrough from all four columns
right from the beginning of initial flow. This breakthrough
appears to be about equal through all four columns. It is
possible that this represents a nonadsorbable fraction of TOC.
It appears that after this base line breakthrough we observe a
rapid rise in the curves. This rapid rise begins further from
time zero as the bed depth increases. These rapid rise or
steep slope regions then change into apparent plateaus for each
column depth.
The curves in Figure 117 are replotted individually for
each column in Figures 119, 120, 121 and 122. These individ-
ually plotted THM FP curves more clearly show the same patterns
shown by the TOC data in Figure 118. In Figures 119, 120, 121
and 122 we again have three distinct zones of the breakthrough
curves. Starting from a base line of a consistent low level
THM FP passing through all four columns, there is an initial
breakthrough period from start of flow that increases in time
as the bed depth increases. The breakthrough then expands into
a rapid rise or steep slope zone to a plateau. The plateau can
be seen to occur at lower concentrations with increase in column
bed depth. These differences will be discussed later.
The adsorption curve for THM FP adsorbed from raw water by
0.76 (2.5 feet) and 1.52 (5.0 feet) meters of IRA-904 resin is
shown in Figure 123. The level of THM FP passing through the
1.52 meter (5.0 feet) deep bed of IRA-904 resin from start of
initial flow is about equal to the value from the 0.76 meter
(2.5 feet) deep bed. However, in the 1.52 meter (5.0 feet)
deep bed, the steep slope portion of the curve begins four days
197
-------
900-
800-
700-
600-
Finished water
._O— Finished water thru 2.5 feet GAG (0.76 meter )
— A ^ Finished water thru 5 feet GAG (1.52 meters)
...Jk"- Finished water thru 7.5 feet GAG (2.29 meters)
Finished water thru 10 feet GAG (3.05 meters)
Days
28 42 " 56 73 91 105 115
Figure 117. THM FP in finished water and removal by 0.76, -1.52, 2.29 and 3.05 meters
(2.5, 5, 7.5, and 10 feet) of. GAG (ED4) .
-------
vo
V£>
a
Finished water
Finished water thru 2.5 feet GAG (0.76 meter )
Finished water thru 5 feet GAG (1.52 meters)
Finishes water thru 7.5 feet GAG (2.29 meters)
Finished water thru 10 feet GAG (3.05 meters)
Days
u> ~.3
Av. 3.21*
it appears that TOG data on these two sampling
dates are probably in error (too low)
*for plateau segment
7 m 21 29 35 42 H9 59 63 70 77 84 91 98 105
Figure 118. TOG in finished water and removal by 0.76, 1.52, 2.29 and 3.05 meters
(2.5, 5, 7.5, and 10 feet) of GAG (ED4).
-------
0)
-p
-H
to
o
o
p.
04
700
600
500
400
300
200
100
Days
Finished water
Finished water thru 2.5 feet GAC (0.76 meter )
*for plateau segment
II II II 1 II—I IT I—I T—I IT
70 84
Figure 119. THM FP in finished water and removal by 0.76 meter
of GAC (ED4).
98
(2.5 feet)
112
-------
-------
to
o
700'
600-
200 -
100 -
/V___ -M-----^MIZJLipi..._ '...-._ Jr.-" f-
/ V * \
\ /* \
r\ ^ /f \^& A A
/ \ ^-A ^ A
B / X^
3 / / 9 Finished water
1 A-Ar
•^ / >^ —A— Finished water thru 7.5 feet GAC (2.29 meters)
\l
A
250*
* for plateau segment
Days
28 42 ' 56 70
98
112
Figure 121. THM FP in finished water and removal by 2.29 meters (7.5 feet of GAC (ED4) .
-------
Finished water
—— O~— Finished water thru 10 feet GAC (3.05 meters)
to
o
u>
TOO—.
600 __
500 —
400 _
300 —
200_
100 —
Days
Av.
H.T. water
531
Type III
substances
partial saturation plateau
Type II
substances
I Type I
T substances
229
11 2& 1*2 56 73 93 105 115
Figure 122. THM FP in finished water and removal by 3.05 meters (10 feet) of GAC (ED4)
-------
soo
Raw water
Raw water thru 2.5 feet
—D~" Raw water thru 5 feet
Av- 600
Days
Av.377*
*for plateau segment
7 10 Ik IB 21 25 2*8 32 ' \2
Figure 123. THM FP in raw water and removal by 0.76 and
1.52 meters (2.5 and 5 feet) of IRA-904 resin (ED3)
204
-------
after a period of initial breakthrough. Again, partial satu-
ration plateaus are reached and in this case they appear to be
equal for both bed depths, but the plateau for the 1.52 meter
(5.0 feet) deep bed occurs 14 days later than for the 0.76 meter
(2.5 feet) deep bed.
Adsorption data tables were prepared for HOC adsorption
curves. A problem arises when this is attempted with TOC and
THM FP curves. The three zones of the curves and perhaps some
idea of what produces these three zones should be taken into
account. Although the predominant substance that acts as pre-
cursors for THM formation have been cited as being humic acid
or fulvic acid, it is reasonable to assume that such complex
substances may adsorb like a family of different compounds
possessing various adsorption affinities with a given adsorbent.
Likewise, TOC can more easily be seen to comprise a family of
different substances with varying adsorption affinities.
THM FP data will be discussed by assuming that three types
of substances comprise the precursors for THM in the Florida
water tested. The broad definitions for each type of substance
is as follows:
TYPE I - Substances that are not initially adsorbed
under the conditions tested.
TYPE II - Substances that are initially adsorbed and/
or biodegraded within the adsorbent column,
but that eventually breaks through in
increasing concentrations under the
conditions tested.
TYPE III - Substances that are completely adsorbed and/
or biodegraded within the adsorbent column
under the conditions tested.
Figure 122 shows the relationship of the three types of
substances to the three zones of adsorption. In this figure,
' the Type I substances are shown as the fraction that passes
through 3.05 meters (10 feet) of GAC, with a contact time of
24.8 minutes from the start of initial flow of water through
the column. This could be a nonadsorbable fraction or a frac-
tion nonadsorbable at that concentration with the contact time,
type of GAC used and other conditions of the test.
Type II are substances that show a rather classic pattern
of complete adsorption initially, followed by breakthrough and
increasing effluent concentrations up to an apparent plateau
value below equilibrium with the influent concentration. This
could be either a substance with a given adsorption affinity or
a family of substances that have various adsorption affinities
that yield the effluent curve as shown.
205
-------
Type III substances might be those that have a very strong
adsorption affinity such that they are completely adsorbed
within the adsorbent column for the duration that the study was
conducted, a biodegradable fraction of the THM FP, or a combi-
nation of both. Either assumption could be used to provide a
possible explanation of the plateau shown on the curve. We
have limited data to suggest that in our system as studied, that
strong asorption of the Type III substances accounts for more
of the plateau level than could be accounted by biodegredation.
These data are shown in Figure 124. The average values for the
plateau levels in Figures 119, 120, 121 and 122 which extend to
the end of the test period of 115 days, are presented again in
Figure 124. At the end of the test period in ED4 , the columns
were allowed to operate further and two additional samples were
taken on days 176 and 177. These two data points are plotted
in Figure 124 and their average value is shown. For each
carbon bed depth, it is seen that the plateau level rises. For
example, for 0.76 meter (2.5 feet) of GAC, at 115 days, the
plateau level was 84 percent of the influent level and at 177
days it was 94 percent of the influent level. This suggests
that the Type III substances from test day 115 to test day 177
are showing a typical breakthrough pattern. From these limited
data we cannot determine how closely the breakthrough curve at
Type III saturation will approach the influent level. However,
even at test day 177, it is apparent that biodegradation can
only account for a maximum of 6 percent of the influent level if
the Type III breakthrough curve levels off at the 177 day level.
We suspect that the breakthrough will continue further. Since
the test was primarily designed for a four month period, the
frequency of testing beyond that time does not allow more than
speculation that the apparent plateau may be caused more by
adsorption than biodegradation. It is obvious that we do not
know the true reason for the apparent plateau at this time. It
is likely that both factors play a role. Future studies will
collect data relative to factors influencing the apparent
plateau and ways, to enhance it.
In evaluating data in Figures 119, 120, 121 and 122, using
the assumptions regarding the three types of substances, it is
apparent that the difference between the influent THM FP con-
centration and apparent plateau, which defines Type III,
becomes greater as the bed depth and contact time are increased
from 0.76 (2.5 feet) to 3.05 (10 feet) meters (6.2 minutes EBCT
to 24.8 minutes EBCT). Concurrently the Type II substances
defined by the difference between the plateau concentration and
breakthrough concentration decreases with increased bed depth
and contact time. The general definitions state that each
class of substance is defined at a given set of conditions and
each figure present a different condition. However, the
changes in the amounts of Type II and III substances with
different contact times may be explained as the effect of
increased carbon volume and contact times that provide more
206
-------
-H
500-4
10
o
400-
300-
200 _
Av. 365(84% of influent) 2.5 feet GAG
(0.76 meter )
Av. 312(72% of influent) 5 feet GAC
(1.52 meters)
00
Q
Av. 434 influent
Av. 407 (94% of influent)
Av. 380(88% gf influent)
Av. 338 (78% of influent)
Av. 316(73% of influent)
Av. 250(58% of influent)7.5 feet GAC (2.29 meters)
Av. 229(53% of influent} 10 feet GAC p!o5 meters)
100-
DAYS
Figure 124. Test extention data for THM FP removal by 0.76, 1.52, 2.29 and 3.05
meters (2.5, 5, 7.5 and 10 feet) of GAC (ED4).
-------
effective adsorption of Type II substances with adsorption
affinities closer to Type III substances while also more effec-
tively removing substances of weaker adsorption affinities.
One cannot dismiss the possibility that biodegradation is the
cause for the different plateaus. As previously stated, it is
likely a combined effect that causes the plateau and we cannot
as yet discern a primary factor.
Using this devised system of data presentation we have
assumed that Type I substances continuously break through at
constant levels and the Type III substances are completely
adsorbed by a specific bed depth and contact time at the same
level during the test period under the conditions of the study.
This assumption allows analysis of the TOC and THM FP data
and presentation in tabular form such as presented for the halo-
genated organic compounds.
Raw Water Source—
THM FP adsorption data appear in Table 38. Adsorption
curves appear in Figures 125, 126 and 127.
EDI data are plotted in Figure 125. Since collection of
data began on test day 61, complete analysis cannot be made on
this ED. Comparison of the average values from day 61 to the
end of the test period can be made from Figure 125. The
average influent THM FP was 816 yg/L. The average effluent
from the 0.76 meter (2.5 feet) deep GAG column was 757 yg/L,
a reduction of seven percent. The effluent from the 0.76
meters (2.5 feet) deep XE-340 column was 833 yg/L, about two
percent higher than the influent.
Column 0.76 meter (2.5 feet) deep of GAC and XE-340 were
studied in ED1R. Adsorption curves appear in Figure 126. In
Figure 126, average values are shown for the influent and for
the column effluents from a point after Type II saturation on
the plateau portion of the adsorption curves. Considering
these average values, from test day 17 to the end of the test
period, GAC and XE-340 removed 24 percent and 20 percent
respectively of the influent THM FP. Integration of the curves
for the whole test period produced the results in Table 38
(ED1R). During ED1R 0.76 meter (2.5 feet) of GAC adsorbed 29
percent of the THM FP precursors from entering raw water over
the entire test period. Type I substances represented two
percent; Type II, 72 percent; and Type III, 26 percent of the
total precursors entering. Also during ED1R, 0.76 meter
(2.5 feet) of XE-340 adsorbed 24 percent of the precursors from
entering raw water over the entire test period as compared with
29 percent adsorbed by the same bed-depth of GAC. One hundred
grams of GAC adsorbed 1.5 times as much precursor substances as
100 grams of XE-340 during the 119-day test period. When equal
volumes of the two adsorbents are compared, at 119 days, GAC
208
-------
TABLE 38. THM FP ADSORPTION DATA FROM RAW WATER
O
10
O\ *O 13
ED
1
1
1R
1R
3
3
•O ft
33
Feet
2.5
2.5
2.5
2.5
2.5
5
Adsorbent
GAG
XE-
340
GAG
XE-
340
904
904
S
J)
3
H
>H
a
H
% of Total
Enter in
%
ot ca
29
24
(34)
46
55
^
O^g
CO < -rl
Sg«
sIB
SUM
HAH
0 H
H id
(d w CD
•M Oi
P >»>*
6 ffl £
5rams
culate
"
.37
No Sat
.59
1.14
04
fa
0*
o
IIP
*
5
22
43
conti
Type III
Substances
Adsorbed
per Column
Grams
1.85
1.71
(.76)
.98
.99
ued)
3 g
S Of -r
fa >•
M-I a
o § -P
frj j
dp P t
%
26
24
(26)
37
37
2.5 feet=6.76 meter 5 feet=1.52 meters
-------
TABLE 38%(continued)
ED
1
1
1R
1R
3
3
•O ft
Feet
2.5
2.5
2.5
2.5
2.5
5
sorbent
S
GAG
XE-
340
GAG
XE-
340
904
904
_4
&
i_l
t-l O
O cn
*S
%
4
0
45
44
£
t.
rtj
rH 4J
ro G
| w
* 1
%
72
74
(72)
45
45
Type I
Substances
*o
Type II
turation
•P 10
m w
Grams
.21
No Sat
.22
.21
at
!
%
Grams
.56
.33
.44
.26
Per
Col.
w
!
%
Grams
.99
.71
1.21
1.43
2.5 feet=0.76 meter 5 feet=1.52 meters
-------
to
1100-
100CL.
900-
800-
700_
0)
4J
tn
60CH
500_
400-
300-
Raw water thru 2.5 feet GAC (0.76 meter )
— A— Raw water thru 2.5 feet XE-340 (0.76 meters)
"" 9 • Raw water
Days
reduction
_ 61 64 69717678 83 9092 97
Figure 125. THM FP in raw water and removal by 0.76 meter (2.5 feet) of GAC and
0.76 meters (2.5 feet) of XE-340 (EDI).
117
-------
900
20%
reduction
: 527 •
24%
reduction
• Raw water
Raw water thru 2.5 feet GAC (0,76 meter )
Raw water thru 2.5 feet XE-340 (0.76 meters)
Davs 0 14 28 42 56 70 84 98 112
Figure 126. THM FP in raw water and removal by 0,76 meter (2.5 feet) of GAC and
0.76 meter (2.5 feet) of XE-340 (EDlR) .
11 9
-------
900
800
Raw water
Raw water thru 2.5 feet
Raw water thru 5 feet
38% reduction
45% reduction
Davs Figure 127
28 "*2 "»9
THM FP in raw water and removal by 0.76 and 1.52 meters (2.5 and 5 feet) of
IRA-904 resin (ED3).
-------
adsorbed 1.2 times as much precursor substance as XE-340. XE-
340 had about three times the capacity of GAG for HOC adsorption.
If Type II substances have an initial breakthrough and satura-
tion time through XE-340, the values are less than three days
when the first datum point after initial flow was obtained.
In EDS, 0.76 meter (2.5 feet) of IRA-904 resin was evalua-
ted for 49 days on raw water compared with a test period of 119
days for adsorbents used in EDlR. Since the THM FP influent
levels of the EDlR and EDS are quite close, 659 yg/L and 600 yg/L
respectively, the two sets of data can be compared. For pur-
poses of comparison, integration values for 0.76 meters (2.5
feet) of GAG at 49 days in EDlR are shown in parentheses in
Table 38. At 49 days 0.76 meter (2.5 feet) of GAG adsorbed 34
percent and 0.76 meters (2.5 feet) of IRA-904 resin adsorbed 46
percent of the precursors entering each column. A bed of 1.52
meters (5.0 feet) of IRA-904 resin adsorbed 55 percent of pre-
cursors from raw water during EDS. The adsorption curves for
both IRA-904 resin bed depths are shown in Figure 127.
We have no way of knowing the status of the IRA-904 resin
plateau after the end of the 49-day test period. If it remained
at the same level for the same number (119) of test days as in
EDlR, 41 percent would have been adsorbed compared with 29 per-
cent for GAG. In Figure 127 the plateau levels are equal for
both bed depths of IRA-904 resin. We can speculate, but have no
explanation for the two levels being equal.
Initial breakthrough and saturation times for the Type II
substances were longer in the 1.52 meter (5.0 feet) deep IRA-
904 resin bed than in the 0.76 meter (2.5 feet) deep bed. The
level of Type I substances was the same through both bed depths
of IRA-904 resin.
The last four columns of data in Table 38 present the ad-
sorption of total precursors per 100 grams of adsorbent at the
ned of the test period, at Type II saturation, at a common point
in time of 49 days, and adsorption per column at 49 days which
was the shortest test duration. If 0.76 meter (2.5 feet) of
GAG, 0.76 meter (2.5 feet) of IRA-904 resin, and 1.52 meters
(5.0 feet) meters of IRA-904 resin columns were all regenerated
at their respective Type II saturation times of 10, 18, and 34
days, adsorption per 100 grams of adsorbent would be similar;
i.e., 0.21, 0.22, and 0.21 grams respectively per 100 grams of
adsorbent. If all adsorbents tested (0.76 meter [2.5 feet] of
GAG, XE-340, and IRA-904 resin, and 1.52 meters (5.0 feet) of
IRA-904 resin) were generated at the same time, 49 days, adsorp-
tion in grams per 100 grams of adsorbent would be 0.56, 0.33,
0.44, and 0.26 respectively. GAG removed the most total precur-
sor per 100 grams of adsorbent. The last column of values in
Table 38 are amounts of precursors adsorbed by the entire column
of each adsorbent containing equal volumes of the adsorbents.
214
-------
When equal volumes of the three adsorbents (0.76 meter [2.5
feet] of GAC, XE-340 and IRA-904 resin) are considered, the ad-
sorption from raw water at 49 days is 0.99, 0.71, and 1.21 grams
respectively. When compared on a volume basis, IRA-904 resin
adsorbes more precursors than GAC. However, the initial break-
through concentration of precursors was the highest for the IRA-
904 resin at about 100 yg/L at start-up, and the lowest at start-
up was obtained by GAC. At this point, no meaningful conclusions
can be made regarding the total merits of the three adsorbents
applied to raw water without a specific objective in mind and
research on regeneration of each adsorbent and the accompanying
cost.
In 404 days of testing, we found that lime softening of raw
water removed an average of 27 percent of the THM FP. This com-
pares favorably with THM FP removal of 29 percent by GAC and 24
percent by XE-340 for 119 days by 0.76 meter (2.5 feet) deep
beds. A 0.76 meters (2.5 feet) deep bed of IRA-904 resin remov-
ed 46 percent of the precursor from raw water for 49 days when
the test ended. If all three adsorbents were assessed at the
end of 49 days, the expected results would be removal of 34 per-
cent by GAC, 24 percent by XE-340 and 46 percent by IRA-904
resin. As is the case with HOC results, the shape of the adsorp-
tion curves affect data interpretation. Choosing a reference
point for comparison of adsorbent capacity is important.
The flocculated calcium carbonate in the H.T. can be con-
sidered as still another adsorbent and compared with our column
adsorbents. Earlier in the report, it was shown that adsorption
occurring on calcium carbonate floe in the H.T., under normal
operating conditions of the Preston Plant, removed 0.07 grams of
THM FP per 100 grams of floe. The 0.76 meter (2.5 feet) deep
GAC column in EDlR, Table 38, at Type II saturation removed 0.21
grams of THM FP per 100 grams of GAC. It is interesting to note
that the type of adsorption occuring on calcium carbonate floe
removed about 33 percent of the precursors, as measured by THM
FP _as was removed by GAC on a weight basis.
TOC adsorption data from raw water appears in Table 39.
Adsorption curves.appear in Figures 128 and 129.
In Figure 128, from the start of initial flow through the
XE-340 column, essentially no TOC was removed. Over the entire
test period an average of all test data indicates that the XE-
340 column removed only two percent of the influent TOC. The
GAC column removed 8 percent of the influent TOC, calculated
after Type II saturation. In Table 39, the results obtained by
integration of the entire adsorption curves are presented, GAC
and XE-340 removed 12 percent and two percent of the influent
TOC respectively. The same columns removed 29 percent and 24
percent respectively of THM FP. The GAC column removed 10.2
times and 8.4 times as much TOC as the XE-340 column on an equal
215
-------
TABLE 39. TOC ADSORPTION DATA FROM RAW WATER
ro
ED
1R
1R
3
3
,f
W S
Feet
2.5
2.5
2.5
5
iiJ
C!
fi
o
01
a
GAG
XE-
340
904
904
•P
C
0)
3
H
tl_j
£
{•^
g>
2 „
35
C M
Q
CJ
y B 4J
C^ 9 01
S H fl)
H
g ^
Grams
111.1
111.1
40.7
40.7
•d
2
M
O -P
01 <;
»=«; s -p
5 M
(Hi C^
80
O *H
f o
43
o -d
H Rj C
nJ W H
4J
S&
Srams
13.2
2.3
10.5
15.4
8
EH
ty*
3C
*rH
j^l
0 (U
EH -P
c
o
*
12
2
26
38
1
S-P O
(4! -H
K to
3 Ej
H JJ
U O <0
^^ CJ Ul
£H
/3 H
O M
H ID
flj |JF^ (JJ
S &6*
Srams
5.1
No
Sat.
9.2
14.6
0
EH
Cn
•P M
O (U
O
dP
*
5
23
36
(cont
Type III
Substances
S
i^
•d 3
^3
O
CO M
a &
Grams
9.1
2.3
2.4
2.8
.nued)
r-|
id
-P O
WO)
* s i
*
8
2
6
7
2.5 feet=0.76 meter 5 feet=1.52 meters
-------
TABLE 39. (continued)
to
ED
1R
1R
3
3
^••^••••••••••••••••••H
•O Oi
ss
Feet
2.5
2.5
2.5
5
••••••••^^••^•^••••WBBI
Adsorbent
GAC
XE-
340
904
904
4J
1
H
niaav/a.j^«=
Per 100 grams
of Adsorbent
+J
(0
e
m
o
•s
0)
8
Grains
7.5
1.1
3.8
2.8
•***
**
8
Grams
4.3
.42
3.8
2.8
"•^•MW^^^HVOH
Per
Col.
(0
1
o»
<*
8
Grams
7.6
.9
10.5
15.4
•Aq^^^^^^^^^^^
2.5 feet=0.76 meter 5 feet=1.52 meters
-------
Days
2% reduction
8% reduction
O — Raw water thru 2.5 feet GAC (0.76 meter )
Raw water thru 2.5 feet XE-340 (0.76 meter )
to
m
yg/liter
1
/
/
/
O
i
\ 1
V
037
Figure 128.
98 105 112 119 127
TOC in raw water and removal by 0.76 meter (2.5 feet) of GAG and
0.76 meter (2.5 feet) of XE-340 (ED1R) .
-------
M
0)
•M
•H
11-
10-
9 -
8 -
7-
6-
5-
4-
3-
2-
1-
Av. 8.6
Raw water
— D — Raw water thru 2.5 feet IRA-904 resin
(0.76 meter)
Raw water thru 5 feet IRA-904 resin
(1.52 meters)
Days
11
18
25
32
3
-------
volume and equal weight basis respectively.
In EDS, Table 39 and Figure 129, 0.76 (2.5 feet) and 1.52
(5.0 feet) meters of IRA-904 resin removed 26 percent and 38 per-
cent of the influent TOG respectively. Comparing 0.76 meter
(2.5 feet) of IRA-904 resin with 0.76 meter (2.5 feet) of GAG
at an equal time of 49 days, the resin removed 0.9 times and 1.4
times as much TOG as GAG on an equal weight and equal volume
basis.
H.T. Water Source—
THM FP adsorption data appear in Table 40. Adsorption
curves appear in Figures 130, 131, and 132.
XE-340, 0.76 meter (2.5 feet) deep, was studied in EDl and
adsorption data shown in Figure 130. Since collection of data
began on day 61, complete analysis is not possible. From day 61
to the end of the test period, the XE-340 column removed four
percent of the influent THM FP. On raw water, the XE-340 column
removed seven percent during the same time period. However,
while we will continue to compare THM FP and TOG adsorption
across water sources, raw, H.T. and finished, it really is not a
valid way to interpret these data. We will show in the discus-
sion on Finished Water Source, which follows later, that one
cannot compare directly across water sources when the influent
levels of THM FP and TOG change.
In ED1R, Figure 131, 0.76 meter (2.5 feet) of XE-340 was
studied from initial column start-up. The XE-340 column removed
two percent of the influent THM FP compared to 24 percent remov-
al (Table 38) from raw water.
In ED3, 0.76 meter (2.5 feet) of IRA-904 resin was studied,
Figure 132 and Table 40. The IRA-904 resin removed 32 percent
of the influent THM FP compared to 46 percent from raw water.
In Figure 132, the adsorption curve suggest that additional Type
II substances are being removed from H.T. water but no Type III
(at Type II saturation the influent and effluent curves coin-
cide) .
TOG adsorption data appear in Table 41. Adsorption curves
appear in Figures 133 and 134.
In ED1R, 0.76 meter (2.5 feet) of XE-340 removed six per-
cent of the in-fluent TOG compared to 24 percent from raw water.
In ED3, 0.76 meter (2.5 feet) of IRA-904 resin removed 41 per-
cent of the influent TOG compared to 46 percent from raw water.
Finished Water Source—
THM FP adsorption data appear in Table 42. Adsorption
curves appear in Figures 135, 136, 137, 138, 139, 140, 141, and
142.
220
-------
TABLE 40. THM FP ADSORPTION DATA FROM H.T. WATER
ED
1
1R
3
•O ft
SS
Feet
2.5
2.5
2.5
Adsorbent
XE-
340
XE-
340
904
g Average Influent
j> THM FP
580
474
394
Type II
Substances
jj? Column
"^ Breakthrough
<3
0
fi? Column
'to Saturation
Par
<3
39
MT
z
Inch
:ial ru
30+
30
p? Test
"m Duration
i onlj
119
49
$ Total THM FP Entering
| Each Column During
5 Test
- car
5.0
1.72
ft Total THM FP Adsorbed
§ By Each Column At
01 End of Test
not c<
.12
.55
S
H£
$V
o M W
O
%
Iculal
2
32
$ Total THM FP Adsorbed
| By Each Column At
» Type II Saturation
Q
No Sat
.55
&
tp
•S-S
•p M
gs
c
-------
TABLE 40.(continued)
ED
1
1R
3
i!
Feet
2.5
2.5
2.5
Adsorbent
XE-
340
XE-
340
904
jg Average Influent
£> THM FP
580
474
394
Type II
Substances
H Adsorbed
§ per Column
0
.55
i? Passed
I Each Column
4.88
.83
i? Total Entering
p Each Column
4.88
1.38
% of Type II
* Adsorbed
0
40
% of Total
THM FP Entering
98
80
Type I
Substances
tl
-------
N)
to
U)
1000-
900-
800-
0)
.tJ 700-i
rH
\
3.
fc 600.
500,
400
1047
4% reduction
Hydrotreator water
Hydrotreator water thru 2.5 feet XE-340
(0.76 meter .)
6^71 7*678 8*385 ^0^2 9*7
U7
Figure 130. THM FP in raw water, and removal by lime softening and by
0.76 meter (2.5 feet) of XE-340 on H.T. water (EDl).
-------
M
(O
M
(0
4-1
•H
Cn
CM
En
900
800
700
600
500
400
300
200
100
Raw water
H.T. water
H.T. water thru 2.5 feet of XE-340 (0.76 meter )
0 3
Days
7 10 1417 212"+ 28313538 42 »»5, 4952 56 6366 7073 77 84 9'4 98 105 108112115119
Figure 131. THM FP in raw water and removal by lime softening and by
0.76 meter (2.5 feet) of XE-340 on H.T. water (EDI).
-------
to
to
tn
10
Raw water
H.T. water
H.T. water thru 2.5 feet
IRA-904 resin (0.76 meter )
047 1114 1821 2528 3235 3942 4649 53
Figure 132. TKM FP in raw water and removal by lime softening and by 0.76 meter
of IRA-904 resin on H.T. water (EDS).
(2.5 feet)
-------
TABLE .41. TOC ADSORPTION DATA FROM H.T. WATER
ED
1R
3
$
•O D.
3S
Feet
2.5
2.5
Adsorbent
XE-
340
904
|j Average Influent
^ TOC
6.8
6.0
Type II
Substances
jj? Column
*Jj; Breakthrough
?
0
•
p> Column
"iw Saturation
?
46
MT
z
Inch
•)
•
30
tu Test
to Duration
127
53
j^ Total TOC Entering
H Each Column During
to Test
77.1
28.4
•a
.8
n
O 4J
(0 <
•O
•4! S +J
1 co
3 i
EH ffl
Srams
3.1
11.3
% of Total TOC
* Entering
4
40
ft Total TOC Adsorbed
fi By Each Column At
™ Type II Saturation
No
Sat.
10.9
o
% of Total TO
* Entering
38
(conti
2.5 feet=0.76 meter
Type III
Substances
O
g Adsorbed
3 per Column
ul
3.1
3.1
nued)
% of Total
*> TOC
Entering
4
11
-------
TABLE 41. (continued)
to
N)
ED
1R
3
•d ft
0) 0)
n a
Feet
2.5
2.5
Adsorbent
XE-
340
904
•M
C
Q)
5
TOC Entering
65
Type I
Substances
$
§ Total Passed
6.6
% of Total
* TOC Entering
14
j.ww fluawj.M'c;
Per 100 grams
of Adsorbent
•P
to
u
IH
o
•s
0)
^
Grams
1.4
4.1
n At Type II
§ Saturation
2.9
01
1
0\
•&
%
Grams
.6
4
Per
Col.;
(0
1
en
^
$
Grams
1.2
11.1
2.5 feet=0.76 meter
-------
9 -
NJ
N>
00
6 _
0) c
u ->
I
84 -
3 -
1 _
0
4%
reduction
H.T. water
H.T. water through 2.5 feet XE-340 (0.76 meter )
37 11+21 28 35 42 64 71 78 84 91 93 105 112 119
Days Figure 133. TOC in H.T. water and removal by 0.76 meter (2.5 feet) of XE-340 (ED1R) .
127
-------
in
0)
+J
•H
to
to
Av. 6
H.T. water
H.T. water through 2.5 feet
IRA-904 resin (0.76 meter)
Days Figure 134. TOC in H.T. water and removal by 0.76 meter (2.5 feet)
of IRA-904 resin (ED3).
-------
TABLE 42. THM FP ADSORPTION DATA FROM FINISHED WATER
Ul
o
ED
1
1R
3
3
4
4
4
4
•
rC
•^
•o a
S3
Feet
2.5
2.5
2.5
2.5
2.5
5
7.5
10
+J
a
0)
.Q
M
O
CO
S
XE-
340
XE-
340
GAC
904
GAC
GAC
GAC
GAC
t!
2
3
IH
a
H
a)
Cn CM
fd p4
M
<1) §
^^ nQ
ug/L
451
355
274
274
434
434
434
434
Type II
Substances
fi
0^
3
O
S +*
K f^
3 (O
rH
C
MH W
O
%
.ate
3
18
13
22
39
53
60
13
JS
Vt C
O 4J O
M ft -H
*d -P
C 1 1
CM 9 0
i4 «— 1 -M
O rd
2
EH X H
O H
H Id
(d PQ Q)
S&&
Grams
.14
No Sat
.36
.84
1.15
1.41
PM
S
hM
£-4
tn
^-S
•P M
£$
C
M-l FT1
O
*
12
8
19
26
32
(conti
Type III
Substances
S
'0 3
M U
o
CO M
Grams
.138
.11
.15
.71
1.25
1.89
2.11
iued)
H
id
•P tr
o c
EH CM -H
m a)
O 2 4J
<*> EH W
*
3
9
13
16
28
43
48
2.5 feet=0.76 meter 5 feet=1.52 meters 7.5 feet=2.29 meters 10 feet=3.05 meters
-------
TABLE 42. (continued)
to
co
ED
1
1R
3
3
4
4
4
4
fj
•C ft
0) 0)
PQ Q
Feet
2.5
2.5
2.5
2.5
2.5
5
7.5
10
.p
A
fc
o
CO
s
XE-
340
XE-
340
GAG
904
GAG
GAG
GAG
GAG
-P
0)
3
H
M-l
H
9
cr> PJ
jd fa
4) g
rt! EH
ug/t
451
355
274
274
434
434
434
434
Type II
Substances
£4
B
•d 9
as
ou
(0 N
"O flj
ri! ft
Grams
.11
p
.25
.49
.49
.57
rt
c
2
H
O
tJ CJ
0)
W JS
to U
« id
cu w
Grams
.89
?
3.39
2.61
1.97
1.67
C
•H
® I
rt i-«|
w g
H
id ^:
O id
EH W
Grams
1.0
?
3.64
3.1
2.46
2.24
H
H
(U
S a)
IH O
o S
%
11
'
7
16
20
26
H1
•H
I.
iHI Jj
dlj ^
4-* ri
EH P.
M-l
O 21
* 1
%
83
?
82
70
55
50
Type I
Substances
•o
0)
PM
H
id
S
Grams
.09
?
.11
.11
.11
.11
£
-H
d)
3jj
*^
.T2
g o,
h
O 51
* 1
%
8
?
2
2
2
2
Per 100 grams
of Adsorbent
W
rd
ft, CO
Grams
.08
No Sat
.2
.24
.22
.2
CO
^1
(d
-0
5
<5
Grams
.032
.13
.07
.3
.29
.25
.21
Per
Col.
CO
^i
id
•O
5
<
Grams
.057
.23
.19
.53
1.02
1.32
1.48
2.5 feet=0.76 meter 5 feet=1.52 meters 7.5 feet=2.29 meters 10 feet=3.05 meters
-------
to
00
to
800 -I
700 -
600 -
U
0)
4J
•rl
500 -
p.
PH
400 -
300 -
200 -
Finished water
Finished water thru 2.5 feet XE-340 (0.76 meter 0
Av. 521
100 -
Days w ' " 61 6\ 6971 7^788385 90 ^2 97 Tl
Figure 135. THM FP in finished water and removal by 0.76 meter (2.5 feet)
of XE-340 (EDI).
117
-------
800
700
to
CJ
Av.
H.T. water
471
23% reduction
362
349
3% reduction
Finished water
Finished water thru 2.5 feet XE-340 (0.76 meter )
03 7 10 l«t 17 2124 28 31 3538 4245 4952 56 63 66 7073 77 84 94 98 112 119
Figure 136. THM FP in finished water and removal by0.76 meter (2.5 feet) of XE-340 (ED1R)
-------
800
700
600
500
Finished water
Finished water thru 2.5 feet GAC (0.76 meter: )
Finished water thru 2.5 feet IRA-904 resin (0.76 meter )
tn
to
400
300
200
100
Days
Av. 274
Av. 230
16% reduction
T
11 14
-i—i—n—r~i—n—r—r
18 2125 28 3235 3942 4649
T
53
Figure 137. THM FP in finished water and removal by 0.76 meter
and 0.76 meter (2.5 feet) of IRA-904 resin (&D3).
(2.5 feet) of GAC
-------
Finished water
Finished water thru 2.5 feet GAC (0.76 meter )
_ Finished wate* thru 5 feet GAC (1.52 meters)
».A-'< Finished water thru 7.5 feet GAC (2.29 meters)
_.Q._ Finished water thru 10 feet GAC (3.05 meters)
700
Aye.
Days o 1* 28 «t2 56 73 91 105 115
Figure 138. THM FP in finished water and removal by 0.76, 1.52, 2.29 and 3.05 meters
(2.5, 5, 7.5, and 10 feet) of GAC (ED4).
-------
Finished water
to
U)
en
thru 2.5 feet GAC (0.76 meter' )
Days 0 I't ~ 28 " ~ ~
-------
Finished water
Finished water thru 5 feet GAC (1.52 meters)
to
10
600-
M
-------
Finished water
r... Finished water thru 7.5 feet GAG (2.29 meters)
7001
to
oj
00
Av.
H.T. water
531
0
ll 28 42 56 73 ' 93 1051 'll5
Figure 141. THM FP in finished water and removal by 2.29 meters (7.5 feet) of GAG (ED4).
-------
NJ
U)
700-.
9 Finished water
—O— Finished water thru 10 feet GAC (3.05 meters)
Type III Substances
Saturation
Type I Substances
*for plateau
setmen
Days
0 14 28 42 56 70 at 93 112
Figure 142. THM FP in finished water and removal by 3.05 meters (10 feet) of GAC (ED4) .
-------
In EDI, Figure 135, from day 61 to the end of the test per-
iod, the average THM FP effluent from the 0.76 meter (2.5 feet)
deep XE-340 column was greater than the influent. In ED1R,
Figure 136, the average THM FP effluent from the 0.76 meters
(2.5 feet) deep XE-340 column was three percent below the influ-
ent, indicating essentially nil removal from finished water. Ad-
sorption curves from ED3 for 0.76 meter (2.5 feet) of GAG and
IRA-904 resin appear in Figure 137. It is interesting to note
that the GAC column is removing additional Type II substances,
as evidenced by the 0 and 7 day test point portion of the adsorp-
tion curve. From time 0, this portion of the adsorption curve
is absent from the IRA-904 resin column. After Type II satura-
tion, the average THM FP reduction was 10 percent for GAC and 16
percent for IRA-904 resin (calculated from time 0). Because of
the Type II portion of the adsorption curve, the GAC column at
the end of the test period removed 18 percent of the influent
THM FP compared to 13 percent for IRA-904 resin (Table 42). At
a common point in time of 49 days, GAC removed 1.9 times and 1.2
times as much THM FP as IRA-904 resin on an equal weight and
equal volume basis respectively.
Unlike the 0.76 meter (2.5 feet) deep bed of IRA-904 resin
evaluated in ED3 on finished water, the carbon bed showed a
typical Type II substance removal zone. The time of Type II
saturation was only 11 days. However, of the three adsorbents
tested on finished water, GAC was the only one to exhibit low
Type I bleed, some Type II adsorption and greater total precur-
sor removal. Because of this, GAC was selected for study in ED4.
Four 0.76 meter (2.5 feet) deep columns were connected in
series providing carbon bed depths of 0.76 (2.5 feet), 1.52 (5.0
feet), 2.29 (7.5 feet) and 3.05 (10 feet) meters. THM FP ad-
sorption curves for all four bed depths for ED4 appear in Figure
138. The curves for individual bed depths appear in Figures 139,
140, 141 and 142. The data from the 0.76 meter (2.5 feet) deep
bed in ED4 can be compared with the data from the same bed depth
in ED3. The average influent level of THM FP was 274 yg/L in
ED3 and 434 yg/L in ED4. Total precursor adsorption at Type II
saturation was 0.08 grams per 100 grams of carbon in ED3 and 0.2
grams per 100 grams of carbon in ED4, This indicates that ad-
sorptive capacity for precursor substances increases as adsor-
bate concentration increases, as was found with HOC adsorption.
GAC beds, deeper than 0.76 meter (2.5 feet), ED4 showed in-
creased Type II breakthrough and saturation time; i.e., 14 and
46 days respectively in the 3.05 meters (10 feet) deep bed.
Total THM FP adsorption data for the four columns are summarized
in Figure 143. Column 1 received 4.46 grams of THM FP substances
and adsorbed 0.97 grams. The other columns received and adsorb-
ed less. However, about 20 percent of the influent THM FP to
each column was uniformly removed. The THM FP adsorbed per 100
grams of carbon is also shown in Figure 143 for each 0.76 meter
(2.5 feet) column on Curve II and the 0.76 (2.5 feet), 1.52
meters (5 feet), 2.29 meters (7.5 feet) and 3.05 meters (10 feet)
240
-------
to
4 -
2 -
CM
B
1 -
4.46
1.2
.68
.55*
*Per 100 grams of GAC
.3
-fc
17*
GAC Column 1 2 3-4
Figure 143. THM FP substances in grams entering and adsorbed by each GAC column in 115 days (ED4) .
-------
dolumns on Curve III. Extension of the bottom curve to the "Y"
axis may provide a rough estimate of the.maximum adsorptive
capacity of GAC at 1.2 grams per column and 0.68 grams per 100
grams. Further work is needed to verify the usefulness of such
an approach.
It was mentioned earlier that comparing THM FP adsorption
by adsorbents across water sources was not really a valid method
of interpreting these data because of the change of THM FP in-
fluent across water sources. Some of the THM FP adsorption data
from Tables 38, 40 and 42 appear plotted in Figures 144 and 145
for further consideration. Figure 144 presents THM FP adsorp-
tion per column (0.76 meter [2.5 feet] deep) by GAC, XE-340 and
IRA-904 resin at a common point in time of 49 days. The three
data points from raw, H.T. and finished water for GAC and IRA-
904 resin fall on a straight line. The three data points for
XE-340 do not fall on a straight line. The XE-340, H.T. and
finished water data points are small numbers, 0.053 grams and.
0.057 grams respectively. These values were obtained by inte-
grating the adsorption curves, and it is possible that the minor
adsorption over the test period of only two percent and three
percent respectively is itself not a very accurate base. In
Figure 144, it is probably-not very important which XE-340 line
is accepted, A-B through the H.T. water point, A-D through the
finished water point, or an average line A-C. For our first
discussion, we can eliminate the XE-340 curve. The GAC and IRA-
904 resin curves, both containing three data points, appear to
be a straight line. This might indicate that the change in THM
FP influent across water sources, decreasing from raw to H.T. to
finished water, is the predominant cause of decreased adsorption
per column. To compare the effectiveness of different adsorb-
ents across water sources, the curves in Figure 144 may lead to
a more accurate interpretation of data. Raw water adsorption
data for the three adsorbents is compared in Table 43, with
varying THM FP Influent levels and at a constant THM FP level.
In Figure 144, the horizontal line of 600 yg/L may indicate
what the adsorption would be at that uniform THM FP level. Line
A-C was chosen for the XE-340 curve for this discussion. The
data in Table 43 show that the effectiveness of IRA-904 resin
compared to GAC and XE-340 varies when influent concentration is
considered. The slopes of the GAC and IRA-904 resin curves are
different, and as the influent THM FP level decreases the curves
cross. At low influent levels, GAC becomes more effective than
IRA-904 resin. Comparison of adsorbents would be different if a
different common time point was chosen. The results are also
different at 49 days when compared at an equal adsorbent weight
basis, Figure 145. In Figure 145, GAC is more effective than
IRA-904 resin in all three water sources. In both Figures 144
and 145, XE-340 has considerably less adsorptive capacity than
GAC or IRA-904 resin.
242
-------
to
>t»
U)
700-1
600
500,
400 -
300 .
200 .
100 -
GAC data points
•O— IRA-904 resin data points
- XE-340 data points
0
.9
1.0
1.1
1.2
.5 .6 .7 .8
Grams of THM FP adsorbed per column
Figure 144. THM FP adsorption by GAC, XE-340 and IRA-904 resin per column 0.76 meter deep (2.5 feet)
at 49 days.
-------
TABLE 4 3.
EFFECT OF THM FP INFLUENT
CONCENTRATION IN RAW WATER ON
ADSORPTION DATA INTERPRETATION
tO
*»
*»
Adsorption
per column
at 49 days
at varying
THM FP
influent
levels
(from
Table 38 )
Adsorptive
capacity of
904 resin
campaiced to
GAC and
XE-340
Adsorption
per column
at 49 days
at 600 yg/
of THM FP
influent
level (from
Figure 144 )
Adsorptive
capacity
of 904
resin
compared
to GAC
and XE-340
adsorbent grams of THM FP grams of THM FP
2.5 feet 904 resin
2.5 feet GAC
2.5 feet XE-340
1.21 1.21
.99 1.22 times .88 1.38 times
.71 1.70 times .57 2.12 times
2.5 feet=0.76 meter
-------
800 _
GAC data points
" IRA-904 resin data
points
— XE-340 data points
0
71
.i
.3 .k 5 . .7 .6
Grains of THM FP adsorbed per 100 grams of adsorbent
Figure 145. THM FP adsorption by GAC, XE-340 and IRA-904 resin per 100 grams of adsorbent at 49 days.
-------
TOG adsorption data from finished water appear in Table 44.
Adsorption curves appear in Figures 146, 147 and 118.
XE-340, 0.76 meter (2.5 feet) deep, was studied in EDlR,
Table 44 and Figure 146. TOC removal averaged three percent of
the influent. Removal was also very low in raw and H.T. water,
five percent and four percent respectively.
GAC and IRA-904 resin, 0.76 meter (2.5 feet), were studied
in EDS. In Figure 147, both GAC and IRA-904 resin appear to re-
move some Type II TOC substances. In Figure 137, the IRA-904
resin did not appear to remove Type II THM FP substances. In
EDS, the IRA-904 resin removed 1.3 times as much TOC from finish-
ed water as GAC on an equal volume basis. On an equal weight
basis, GAC removed 1.2 times as much as IRA-904 resin. GAC, 0.76
meter (2.5 feet) deep, was studied in both EDS and ED4. As the
influent TOC level decreased, the adsorptive capacity decreased.
In Figure 118, it is clear that increased GAC bed depth resulted
in larger periods of time to breakthrough and saturation. I«
Figure 118, it appears that the TOC data collected on test days
29 and 35 are too low and should probably be discarded. When
adsorptive data for the three adsorbents are compared across
water .sources some problems arise. TOC adsorption data from
Tables 44, 46 (see page 257), and 49 (see page 300) are plotted
in Figures 148 and 149. Straight lines were not obtained with the
three data points for each adsorbent. In both Figures 148 and 149,
there appears to be a trend to increased adsorption for each ad-
sorbent as the influent TOC level decreases, which does not appear
reasonable. Perhaps additional work with TOC data in future re-
search is necessary before conclusions can be reached. In Figure
148, the adsorptive capacity (equal volume basis) of IRA-904 resin
was always better than GAC and GAC always better than XE-340.
In Figure 149, on an equal weight basis, GAC was generally bet-
ter vthan IRA-904 resin and IRA-904 resin was always better than
XE-340.
Other Parameters
Chlorine—
The effect of free and combined chlorine in finished water
as it passes through adsorbent columns is included in this sec-
tion with the chlorine data on finished water. XE-340 removes
essentially all free chlorine for 17 days. Free chlorine in the
effluent than steadily rose and by the end of the test (122 days)
was about one-third the influent level. Throughout the test
period the level of combined chlorine in XE-340 effluent remain-
ed about one-half of the influent. IRA-904 resin removed all
free chlorine and 93 percent of the combined chlorine throughout
the 53-day test. In the same test period, GAC removed all free
chlorine and initially 90 percent of the combined chlorine. Com-
bined chlorine gradually increased in the effluent, reaching 72
percent removal at the end of the test. Increased amounts of
246
-------
TABLE 4 4 . TOC ADSORPTION DATA FROM FINISHED WATER
ED
1R
^^^^MM^^^H^^Hm
3
3
4
4
4
4
s
*w Qi
SS
Feet
2.5
^^H^^^^^^^^M
2.5
2.5
2.5
5
7.5
10
•P
C
d)
8
a
XE-
J340
GAC
904
GAC
GAC
GAC
GAC
£
Q)
3
rH
H-l
C
H
0)
(^
2
Q) CJ
£ 8
mg/L
6.5
^^^^^^^HBBHHBIIIH
5.9
5.9
5.4
5.4
5.4
5.4
Type II
Substances
A
a
fl 5
I*
rH 0)
O rl
O ffl
Days
ca
sa
0
0
0
0
7
14
0
C rt
I 3
rH +J
0 Hi
u w
Days
i't
L£U
32
39
24
32
43
60
MT
Inch
BBVVriHWriBBHIHBBBBBBHBIBBVBBBHB
30
30
30
30
25
23
C
O
•H
-P
(0 rl
Days
127
53
53
127
127
127
127
-S
a! t?
c n
w 3
p*
o||
O EH
O
H
£ W
Grams
73.7
IHHIHHIHVVI^HIBB
rt -H
O 0)
EH -P
i K
EH ffl EH
Grams
8.6
11.3
5.9
9.7
14
21.2
0
Q
F~i
tn
i f+
(0 -rl
-P rl
S5
•H S
O
%
HBIBBBflBBBBHHBflBBH^v^pvv
31
41
10
16
23
35
iconti
Type III
Substances
'0 1
fl) rH
X) O
W rl
•O 0)
<; ft
Grams
2.2
2.8
4.7
4
10.9
18.9
25.1
ued)
rH
j fr 1
EH -'
M
4-1 {
dp EH &
%
3
10
17
7
18
31
41
2.5 feet=0.76 meter 5 feet=1.52 meters 7.5 feet=2.29 meters 10 feet=3.05 meters
-------
TABLE 44* (continued)
to
it*.
CO
1
ED
1R
3
3
4
4
4
4
•d ft
Feet
2.5
2.5
2.5
2.5
5
7.5
10
^
(3
2
Adsor
XE-
340
GAC
904
GAC
GAC
GAC
GAC
4J
3
r-l
MH
C
H
(1)
nt
aj u
mg/L
6.5
5.9
5.9
5.4
5.4
5.4
5.4
Type II
Substances
§
•d 3
M U
0
01 M
•d a)
(=< ft
Grains
6.9
7.8
5.1
6.6
7.6
9.4
fl
1
rH
•d o
Passe
Each
Grains
15.8
13.4
48.4
40
31
23
*J*
•S
M a
<
%
44
37
10
14
20
29
C*
•H
m
H "
id g
g
O U
%
81
76
87
76
63
53
Type I
Substances
•d
0)
0)
01
id
ft
Total
Grams
2.4
2
4
4
4
4
CT>
g
•H
(U
•a J
4J r *i
s
%
9
7
6
6
6
6
Per 100 grams
of Adsorbent
^J
0)
(U
E-i
-------
to
*>.
6 -
M
0>
•P
•H
4 -
2 -
1 _
Av. 6.5
6.3
3%
reduction
Finished water
Finished water thru 2.5 feet XE-340 (0.76 meter )
Days
37 1"+21 28 35 42 64 71 78 84 91 98 105 112 119 122
Figure 146. TOG in finished water and removal by 0.76 meter (2.5 feet) of XE-340 (ED1R) .
-------
0)
-p
•H
rH
u
(Jl
O
8 -
7 _
6 -
5 -
4 -
3 -
Av. 5.9
Finished water
Finished water thru 2.5'IRA-905
resin (0.76 meter )
Finished water thru 2.5' GAG
(0.76 meters)
Days
11
25
18
Figure 147.
32
39
53
TOC in finished water and removal by 0.76 meter (2.5 feet) of
IRA-904 resin and by 0.76 meter (2.5 feet) of GAG (ED3).
-------
GAC data points
Q "1RA-904 resin data points
— Q— XE-340 data points
U1
10 —
•P 9 -
I
H
IM
•S 8-
Ir
7 „
^
6 .
5 .
Raw
D ED1R
l\
l\
1\
\
\
H.T. « \
EDlRg \ Fin.
Q ED1R
Raw
ED3
O
o
H.T.
ED3
•o
Fin.
ED3
•^i
1 -
%
Illllllillll'l
1234 56 78 9 10 11 12 13
Grains of TOC adsorbed per column
Figure 148. TOC adsorption by GAC, XE-340 and IRA-904 resin per column-0.76 meter
at 49 days.
(2.5 feet) deep
-------
10 -i
•M
§
-------
combined chlorine were removed with increased GAC bed depth.
Turbidity—
The effect on turbidity when raw, H.T. and finished water
pass through adsorbents and turbidity in the distribution system
are included in Appendix A.
Adsorbents removed turbidity from H.T. effluent but turbid-
ity of the raw water increased considerably after passing through
XE-340. Turbidity also increased as raw water passed through
IRA-904 resin. This same resin removed turbidity from finished
water and H.T. effluent.
Color—
No color was removed from raw water by XE-340 or GAC.
pH—
The effect on pH of raw, H.T. and finished water as they
pass through adsorbent columns is included in this section with
other pH data. The average pH of raw water through GAC, XE-340
and IRA-904 resin increased by approximately 0.1. The average
pH through XE-340 was one-tenth lower than H.T. water; through
IRA-904 resin it was three-tenths lower. There was no change
through GAC. The average pH of finished water decreased 0.1
when passed through XE-340, and 0.2 through GAC and IRA-904
resin.
Comparison of Laboratory and Distribution
System Aging
It is important to know if bottle aging in the laboratory
to determine total THM, terminal THM or THM FP correlates with
actual distribution system formation of these parameters. To
avoid confusion, we will refer to the THM growth of two-day aged
samples as total THM growth. The parameter, terminal THM will
be reserved for THM growth occurring in six-day aged samples
which is the time factor used throughout this study in deter-
mining THM FP (THM FP = term. THM - inst THM) . Based on mea-
surements of the length of time it took for abrupt changes in
fluoride concentrations made at the plant to reach certain points
in the distribution system, we estimated that it takes two days
for finished water leaving the Preston Plant to reach the Red
Road sampling point. A sampling procedure to determine THM
growth was established as part of ED3 to compare laboratory
bottle aged samples with samples taken directly from the distri-
bution system sampling point at Red Road. The sampling proce-
dure appears in Table 45.
The data obtained are presented in Figure 150. During the
53-day test period of ED3, the Inst. THM levels of finished
water at the Preston Plant are shown by the lower curve in Fig-
ure 150. Finished water leaving the Preston Plant had an aver-
age Inst. THM level of 128 yg/L. Total THM levels of finished
253
-------
TABLE 45. SAMPLING PROCEDURE TO COMPARE LABORATORY AND DISTRIBUTION
SYSTEM AGING
WATER SAMPLE
TREATMENT
Finished water at Preston Plant
aged 2 days in bottle with no additional
chlorine
KJ
Ul
Finished water at Preston Plant
pH 9 buffered and excess free chlorine
added then aged 6 days in bottle
Red Road distribution system sample
has been aged approximately 2 days in
distribution system
Red Road distribution system sample
sample taken and treated
after approximately 2 days in distri-
bution system with pH 9 buffer and excess
free chlorine and then stored in bottle
in laboratory for 4 days before analysis
(represents a total of 6 days aging)
-------
500
A.
o-
Pinished water at Preston Plant
bottle aged 6 days
Red Road water (buffer + chlorine)
'aged 4 days in bottle.
Av.436
Av.410
Av.261
——Q Av.255
——O Av.128
0 Red Road water
_-O-« Finished water at Preston Plant
— -D*— Finished water at Preston Plant
aged 2 days in bottle .
39 4*2
4 } f i 14 il h h is ^2 a'h 39 42 w 49
Figure 150 . Comparison of laboratory bottle aged and
distribution system THM growth (EDS).
255
-------
water at the Preston Plant aged two days in a bottle also are
plotted in Figure 150, indicating an average total THM level of
255 yg/L, a growth of 1.99 times. The Red Road distribution sys-
tem sample had an average total THM level of 261 yg/L, a growth
of 2.04 times, which is very close to the average value of the
bottle aged growth. Although the specific two-day bottle and
actual distribution sample values vary more widely than .the aver-
age values, the general magnitude of the distribution water
values can be roughly estimated by laboratory aged samples.
Curves II and III in Figure 155, show additional data on the com-
parison of bottle aged and distribution system samples.
Finished water from the Preston Plant aged six days in a
bottle with a pH 9 buffer and excess free chlorine (data plotted
in Figure 150) had an average terminal THM level of 410 yg/L, a
growth of 3.2 times. The Red Road sample, aged in a bottle for
four days with a pH 9 buffer and excess free chlorine (a total
of six days from leaving the Preston Plant when including the
two days travel in the distribution system plus four days bottle
storage) had an average terminal THM level of 436 yg/L (data
plotted in Figure 150), a growth of 3.4 times, which again is
very close to the six-day bottle aged sample. Therefore, labora-
tory bottle aging compares fairly closely with actual distribu-
tion system aging. It should be noted that the finished water
sample from the Preston Plant that was aged two days and the
Red Road sample that had been in the actual distribution system
for two days, had no free chlorine left at the end of two days,
and would have had higher total THM values if free chlorine had
been maintained.
Total THM Growth in Adsorbent
Column Effluents
As finished water passes through an adsorbent column, it
loses all of its free chlorine and various amounts of its com-
bined chlorine. The column effluent would have to be rechlori-
nated to a free chlorine level of approximately 2.5 ppm before
entering the distribution system. To study the effect of total
THM growth with two days of chlorine contact on such effluent
samples from 0.76 meter (2.5 feet) deep columns of GAC and IRA-
904 resin, a sampling procedure was established as part of ED3.
The sampling procedure appears in Table 46.
256
-------
TABLE 46. TOTAL THM GROWTH IN ADSORBENT COLUMN EFFLUENTS
WATER SAMPLE
Finished water through
0.76 (2.5 feet) meter of
IRA-904 resin
TREATMENT
•*
buffered at pH 9 and free
chlorine added to 2.5 ppm,
bottle aged 2 days
Finished water through
0.76 (2.5 feet) meter
of GAC
buffered at pH 9 and free
chlorine added to 2.5 ppm,
bottle aged 2 days
The data obtained are presented in Figures 151, 152, 153
and 154. In Figure 151, as a reference base, curves are presen-
ted showing the inst. THM in finished water at the Preston Plant
and the total THM growth occurring after two days of bottle ag-
ing. The average inst. THm level in finished water was 128 yg/L
and the total THM after two days of bottle aging averaged 255
yg/L. The inst. THM levels in the 0.76 meters (2.5 feet) deep
IRA-904 resin column effluent is also presented in Figure 151.
The average level was 176 yg/L. This level is higher than in
the finished water entering the IRA-904 resin column due to the
catalytic generation of THM in the column as previously discuss-
ed. The average THM growth in the column was 48 yg/L (176 yg/L-
128 yg/L). Total THM levels in the IRA-904 resin column efflu-
ent, buffered to pH 9, rechlorinated to 2.5 ppm of free chlorine
and bottle aged two days are presented in Figure 151. The aver-
age value of THM growth resulting from catalytic generation, 48
yg/L, is subtracted from the 309 yg/L average above, an average
value of 261 yg/L is obtained. This average value is very close
to the 255 yg/L average value (Figure 151) for the Total THM
growth obtained by aging finished water for two days. These
data indicate that the IRA-904 resin column did not remove much
THM FP from finished water in ED3. This separate information
confirms the information reported in Table 42, that the IRA-904
resin column in ED3 0.76 meter (2.5 feet) deep removed only 13
percent of the influent THM FP.
The curve shown in Figure 152 was obtained by subtracting
Curve III from Curve IV in Figure 151. It represents the total
THM growth in the IRA-904 resin column effluent due only to THM
FP conversion. It clearly shows that throughout the test period
even from initial startup, the IRA-904 resin never removed
enough THM FP to keep THM regrowth below 100 yg/L.
In Figure 153, the total THM growth in a 0.76 meter (2.5
feet) deep GAC column effluent which had been buffered, rechlor-
257
-------
to
en
CO
Curve I - Inst, THM in finished water at Preston Plant
Curve II - Total THM in finished water at Preston Plant bottle aged 2 days
Curve III - Inst. THM in finished water thru 2.5 feet of IRA-904 resin (0.76 meter)
Curve IV - Total THM in finished water thru 2.5 feet IRA-904 resin, pH 9 buffer
2.5 ppm free chlorine, bottle aged 2 days (0.76 meter )
7 li lV 18 21 25 28 32 35 39 42 46 49 53
Figure 151. Total THM growth in rechlorinated - 2 day aged IRA-904 resin column effluent(EDS).
-------
M
0)
•P
O>
I
Total THM growth due to THM PP conversion in 0.76 meter
(2.5 feet) deep IRA-904 resin" column effluent
buffered at pH 9, rechlorinated to 2.5 ppm
free chlorine and aged 2 days
300-i
to
en
ID
200 -
100 -
Days
Av. 167
4649 53
Figure 152. Total THM growth in IRA-904 resin column effluent due to THM FP conversion (ED3).
-------
400 —,
300 —
-------
400 .
300
0)
-P
•H
200
rti
•M
O
100
Total THM growth due to THM FP conversion in
2.5 feet deep GAC column effluent buffered at
pH 9, rechlorinated to 2.5 ppm free chlorine
and aged 2 days (0.76 meter )
Days
0 4711
18 21 25 28 32 35 39 42 46 49 53
Figure 154. Total THM growth in GAC column effluent due to
THM FP conversion (EDS).
261
-------
inated and aged two days is shown. Curve IV represents the
total THM present after two days aging. Unlike the similarly
treated IRA-904 resin column effluent, Curve IV in Figure 151,
the GAC curve indicates that for a period of time after initial
column flow, the GAC removed sufficient THM FP precursors to
keep the total THM below 100 yg/L. The curve shown in Figure
154 was obtained by subtracting Curve III from Curve IV in
Figure 153. It represents the total THM growth in the GAC
column effluent due only to THM FP conversion. It indicates
that up to some point between 7 and 14 days, the GAC column
removes sufficient THM FP precursor to keep the total THM growth
below 100 yg/L. Of the three adsorbents tested, GAC, XE-340 and
IRA-904 resin, GAC was the only adsorbent to show this charac-
teristic. This was the basic reason for studying deeper GAC
columns in ED4.
Bed Life Criteria in Deep GAC Columns
In ED4, the effluents from the 2.29 meters (7.5 feet) and
3.05 meters (10 feet) deep GAC columns were buffered to pH 9,
rechlorinated to 2.5 ppm of free chlorine and aged two days to
study total THM growth. The results should give some idea of
the bed life one could expect from such columns. GAC exhaustion
criteria at this point can be expressed in various ways. The
EPA tentatively has proposed basing exhaustion time of GAC to
assure maximum protection for the consumer (8).
/
In this report we are determining GAC exhaustion based only
on the proposed MCL regulations that the total THM should not
exceed 100 yg/L. As mentioned previously, finished water pass-
ing through a GAC column will have nil free chlorine. It would
therefore have to be rechlorinated to approximately 2.5 ppm of
free chlorine before it could enter the distribution system.
Since precursors are still present, the addition of free
chlorine would cause THM regrowth. We therefore define GAC
exhaustion or bed life at the point where THM regrowth in a
sample of GAC column effluent after rechlorination to 2.5 ppm
of free chlorine and aging for two days, reaches the THM MCL
level of 0.1 mg/L (100 ppb). The results obtained in ED4 on the
2.29 meters (7.5 feet) and 3.05 meters (10 feet) deep GAC
columns are shown in Figure 155. Instantaneous THM level varia-
tions in Preston Plant finished water (carbon column influent)
are shown in Curve I. The average value was 147 yg/L. When
Preston Plant finished water.was bottle laged two days, addi-
tional THM growth occurred and the Total THM present is shown
by Curve II. The average was 243 yg/L, an increase of 1.7
times. Bottle aging and distribution system aging was compared
in ED4. Samples were again taken at the Red Road sampling point
in the distribution system. Total THM levels are represented by
Curve III. The average value was 218 yg/L, an increase of 1.5
times. This demonstrates again that the laboratory bottle aging
approximates Total THM growth in the distribution system.
262
-------
400
M
U)
20
Curve I - Finished water at the Preston Plant
— Curve II -Preston Plant, finished water aged 2 days
— Curve III - Red Road sample point
• A--- Curve IV - 7.5 feet deep GAG column effluent - rechlorinated - aged 2 days (2.29
•O — Curve V - 10 feet deep GAC column effluent - rechlorinated - aged 2 days (3.05
meters)
100
100
Days
1 /^. ,A' ^
/V ' V-*' ,
) k ............. * ... . i
/
N>--°''
65
Days
1
81
Days
14 28 42 56 70 80 94 112
Figure 155. Total THM growth in rechlorinated - 2 day aged 2,29 and 3.05 meters
(7.5 and 10 feet) GAG-
122
-------
In Figure 155, data points prior to day 42 were not deter-
mined for the 2.29 meters (7,5 feet) deep GAG column. In Figure
155, GAG bed life (based on when the rechlorinated GAG effluent
reaches 100 yg/L) is shown to be 65 days for 2.29 meters (7.5
feet) and 81 days for 3.05 meters (10 feet). The total THM in
the two-day aged 2.29 meters (7.5 feet) and 3.05 meters (10
feet) deep GAG column effluents are the result of two condi-
tions, 1) increasing amounts of inst. THM in the effluent due
to column breakthrough and 2) increasing amounts of THM growth
due to the THM FP breakthrough and conversion. Curve I in
Figure 156, shown the total THM in the rechlorinated two-day
aged 2.29 meters (7.5 feet) deep GAG column effluent and Curve
II, is obtained by subtracting the inst. THM levels in the
column effluent, breakthrough and indicates the contribution of
total THM resulting from THM FP breakthrough and conversion.
On test day 56, inst. THM began to breakthrough the column. It
is seen that at day 65 when Curve I reached the 100 yg/L level,
approximately 63 percent of the THM was due to THM FP and 37
percent due to inst. THM breakthrough. From test day 65 to the
end of the test period the average of the data points for Curve
II is 93 yg/L, which is below the 100 yg/L limit. It would
appear that the 2.29 meters (7.5 feet) deep GAG column removes
sufficient THM FP precursor to keep the average THM regrowth
below 100 yg/L at least up to the 115 day test period. Thus,
one might attribute column failure at 65 days to inst. THM
breakthrough, since if the THM did not break through, the
column might last over 115 days. In Figure 158, it is apparent
that THM breakthrough alone would cause bed life failure (reach-
ing 100 yg/L) in 94 days.
Curve I in Figure 157, shows the total THM in the rechlori-
nated two-day aged 3.05 meters (10 feet) deep GAG column efflu-
ent. Curve II, obtained by subtracting the inst. THM levels in
the column effluent, indicates the contribution of total THM
resulting from THM FP breakthrough and conversion. On test day
70, inst. THM began to break through the column. It is seen
that at day 81, when Curve I reached the 100 yg/L level, only a
very small amount of the total THM was due to THM breakthrough,
approximately 8 percent. Thus one might attribute bed failure
at 81 days to THM FP breakthrough. However, from test day 81 to
the end of the test period the average of the data points for
Curve II is 95 yg/L, which is below the 100 yg/L limit. As with
the 2.29 meters (7.5 feet) deep GAC column, it appears that the
3.05 meters (10 feet) deep column removes sufficient THM FP
precursor to keep the average THM growth below 100 yg/L at least
up to the 115 day test period. Again, one might attribute
column failure at 81 days and beyond for the 3.05 meters (10
feet) deep GAC bed to inst. THM breakthrough. In Figure 158,
it is apparent that THM breakthrough alone would cause bed life
failure in 119 days. GAC bed failure is thus caused by a com--
bination of THM and THM FP breakthrough and the ratio of the
two components at bed failure probably varies with bed depth and
264
-------
--D-" Curve
I - total THM in rechlorinated and
two day aged 7.5 feet deep GAC column effluent (2.29 meters)
Curve II - Curve I minus inst-THM in the 7.5 feet
deep GAC column effluent (2.29 meters)
250 -,
to
cr\
01
Days
7 10 14 17 21 24 28 31 35 38 42 45 49 52 56 59 63 66 70 73 77 80 84 87 91 94 98
105
112
122
Figure 156. THM breakthrough and THM FP conversion components of Total-THM in
two day aged effluent from 2.29 meters (7.5 feet) deep GAC column (ED4) .
-------
0)
-p
ON
200 -
150 -
Curve I - total THM in rechlorinated and
two day aged 10 feet deep GAC column effluent
Curve II - Curve I minus inst-THM in the 10 feet
deep GAC column effluent (3.05 meters)
'p.-
100
50 -
Days
7 10 14 17 21 2H 2831 35 38 1*245 4952 5659 63 66 7073 7780 84 87 9194 98101
112 122
Figure 157. THM breakthrough and THM FP conversion components of total-THM
in 2 day aged effluent from 3.05 meters (10 feet) deep GAC column (ED4)
-------
200
to
Inst. THM xn finished water
Inst. THM in finished water thru 7.5 feet of GAG
(2.29 meters)
Inst. THM in finished water thru 10 feet of GAC
(3.05 meters)
r T T i
52 56 59 63 66
77 80 8k 87
0
Days
q q? qp qqgqi qg gp-^
3 7 10 14 17 21 24 28 31 35 38 42 45 49 5
9194 98101105108112115119122
Figure 158. Inst. THM in finished water and finished water through2.29 and 3.05 meters
(7.5 and 10 feet) of GAC (ED4).
-------
influent concentrations of both.
The curves in Figure 143, show that the first column,
which received the highest concentration of adsorbate, adsorbed
the greatest amount over the test period. Therefore, for great-
est GAG efficiency, as much of the carbon in a bed as possible
should at some time be exposed to the highest possible concen-
tration of adsorbate. If 3.05 meters (10 feet) of bed is
necessary, by using four beds 0.76 meter (2.5 feet) deep in
series, and removing the first bed at saturation while placing
a new bed at the end of the series, we expect 31 percent
increased adsorptive capacity. This would increase the bed life
from 81 days to 106 days. A continuous in-out bed would increase
bed life by 43 percent to 116 days.
In Figure 155, it is apparent that if a MCL below 100 yg/L
were chosen, bed life would seriously be reduced. At 50 yg/L
and 25 yg/L, bed life for a 3.05 meters (10 feet) deep GAG bed
would be approximately 18 days and 4 days respectively.
Relationship of TOC and THM FP Data
Determination of the THM FP of water is a new analytical
method. When a new method is introduced the question of its
relationship or correlation with an existing method or methods
usually arises. In this case, the relationship to the deter-
mination of TOC arises. From our research work, we can divide
a discussion on this subject into two parts, 1) relationship
in the treatment plant and 2) relationship in GAG column
effluents. TOC and THM FP data were simultaneously collected
in ED1R, ED3 and ED4. TOC and THM FP data from Tables 10 and
11 are plotted in Figure 159. ED1R data points for raw, H.T.
and finished water are connected by line segments A-B and B-C.
As one might expect, the slope of these two segments are differ-
ent. A-B is the result of adsorption on precipitated calcium
carbonate, and B-C is the result of conversion by chlorination,
oxidation and sand filtration. The two segments in EDS, A1- B1
and B1- C', are quite similar to those in ED1R. The length of
the segments differ in proportion to the percent removal data in
Tables 14 and.15. If only these two sets of data were available,
ED1R and ED3, one might conclude that correlation between TOC
and THM FP results are quite close. However, the ED4 data
points show a considerable shift to the right. The two segments,
in ED4, A"- B" and B"- C", are relatively parallel to the seg-
ments in ED1R and EDS. If an unknown point "X" were chosen
within the triangle formed by the three raw water data points
and parallels were drawn to all segments in the three ED, the
shaded area represents the predicted zone of results from data
point "X". While there may be some degree of confidence in this
prediction, the value of it is unknown. If point "X" were
chosen outside the triangle, the degree of confidence might be
less. In our treatment plant data, one could say that there is
268
-------
to
CT»
VO
10-
9-1
8.
7-
6-
5-
A,A1,A" = Raw water data points from EDlR, 3 and 4
respectively *
B,B',B" = H.T. water,data points from EDlR, 3 and 4
respectively *
C,C',C" = Finished water data points from EDlR, 3 and 4
respectively *
—0~
EDlR data points
ED3 data points
ED4 data points
Shaded area - predicted result
from data point X
B1
200 300 400 500 600 700
THM FP ygAiter
Figure 159. Relationship of TOC and THM FP data in raw, H.T. and finished water.
-------
a relationship between TOC and THM FP data, but one cannot be
converted into the other by a single or simple conversion
factor.
TOC and THM FP data from the 3.05 meters (10 feet) GAG
effluent in ED4 are plotted in Figure 160. The average level of
TOC for the first 14 days was 0.37 mg/L. The average THM FP
for this same period was 17 yg/L. From day 14 to day 49, both
curves show a steady rise. The THM FP curve has a steeper
slope. From day 49 to the end of the test period, when the
plateau portion of both curves was attained, the TOC level
averaged 3.0 mg/L and the THM FP level averaged 240 ug/L. The
TOC concentration increased 8.1 fold, from 0.37 to 3.0 mg/L.
The THM FP concentration increased 14.1 fold from 17 yg/L to
240 yg/L. There is a relationship in the segments of the two
curves, but, again, no single or simple conversion factor for
converting one to the other. As GAG bed depth changes, a seg-
ment relationship exists, but is different for each bed depth.
We can probably conclude that while there is some degree of
predictable relationship between TOC and THM FP data in both our
treatment plant and a separate degree of predictable relation-
ship in our GAG effluent waters, both tests yield separate and
valuable information which are not convertible one into the
other by a single or simple conversion factor. It is unlikely
that this situation will change as more data are collected,
especially when different geographical relationships are con-
sidered.
Leaching Study on XE-340 Resin Column
The experimental design of the leaching study on a 0.76
meter (2.5 feet) deep XE-340 column is discussed under ED2.
In ED2, a fresh XE-340 column was installed, preceding the
partially exhausted XE-340 column on the finished water line of
ED1R, to supply halogenated organic free water for the leaching
study on the partially exhausted column. In this discussion we
will first present data on chlorodibromomethane. The level of
chlorodibromomethane entering and leaving the leaching study
column in the 63-day test period is plotted in Figure 161. A
weakness in the experimental design is apparent. The lower
curve indicates the chlorodibromomethane in the fresh XE-340
column effluent and entering the leaching study column. Ideally,
to obtain a mass balance of.leached substances, we should have
replaced the fresh XE-340 column with a new fresh XE-340 column
before breakthrough of HOC occurred. Actually, it also would
have been better to use a deeper XE-340 column, because as seen
on the chloroform data to follow, the MTZ for chloroform is
greater than 0.76 meter (2.5 feet). Nevertheless, despite the
breakthrough of chlorodibromomethane we can still draw some
conclusion on leaching from Figure 161. If we consider the data
up to test day 53, the level of chlorodibromomethane entering
270
-------
400-
300-
M
-------
Curve I - Chlorodibromomethane leaving the
exhausted column
— — O— — Curve II - Chlorodibromomethane entering the
exhausted column
:
:
Days ° * 7
\r
__^~~-i
11 14 25 32 53 56 63
Figure 161. Level of Chlorodibromomethane entering and leaving
the partially exhausted 0.76 meter (2.5 feet)
deep XE-340 column (ED2).
272
-------
the leaching study column for most of the period was below 1
Ug/L. Of this 53 day period, the data point at day 53 on Curve
I is in greatest error. It would not be scientifically correct
merely to subtract Curve II from Curve I to obtain the true
leaching curve without the interference of entering chloro-
dibromomethane. We can at this point only be aware that the
test point on Curve I at test day 53 is considerably lower than
plotted. Curve I is replotted in Figure 162, with the full
chlorodibromomethane adsorption curve indicating the level of
breakthrough leaving the XE-340 column at the end of the ED1R
study. In Figure 162, it is seen that the breakthorugh level of
the column on test day 122 of ED1R and the leaching from the
column on test day 0 in ED2 are approximately equal at about
8 ug/L- As the leaching study continued (Curve II) it is
apparent that the leaching curve is just the reverse of the
adsorption curve (Curve I).
The level of bromodichloromethane entering and leaving the
leaching study column is plotted in Figure 163. Leaching data
points to test day 32 are plotted in Figure 164. Again, it is
apparent that the breakthrough level of the column on test day
122 of ED1R is approximately equal to the leaching level on test
day 0 of ED2 (20 yg/L) . Again, the leaching curve appears to
be the reverse of the adsorption curve. Similar results are
shown for chloroform in Figures 165 and 166 and for cis 1,2-
dichloroethane in Figures 167 and 168. It is not surprising to
find that on XE-340, desorption of the four HOC discussed above
appears to be the reverse curve of adsorption. It is well known
that adsorption and desorption on GAG follows this pattern with
some substances while other substances may exhibit some hyster-
esis on desorption.
273
-------
Curve I
0 adsorption from finished
water (ED1R)
-O—Curve II
leaching (ED2)
-P
-H
Days
7 10 14 17 2i 2k 28 31 35 3 4245 49 2 6 6§ 66 7b 3 7 8 8 87
8 105 112 122
I I
Figure 162. Adsorption and leaching of chlorodibromomethand on a 0.76 meter
(2.5 feet) deep XE-340 column (ED1R and ED2).
-------
25-*
20-
K)
>»J
Ul
N
3.
15-
10 _
5 -
0
Days
Curve I- Dichlorobromo-
fr methane leaving the
exhausted column
_ Q fit rye II-Dichlorobromo-
w methane leaving the
exhausted column
o-o
--o
X
X
X
X
M-q-rr""
i+ 7 11 14 25 32 53 56 63
Figure 163. Level of Bromodichloromethane entering and leaving the partially
exhausted 0.76 meter (2.5 feet) deep XE-340 column (ED2).
-------
to
_. Curve I - adsorption from finished water
(ED1R)
... Curve II - leaching (ED2)
Days 03 7 10 14 17 21 24 28 31 35 38 4245 49 52 56 59 63 66 7073 77 80 84 87 91 94 98 101105108112115119122
Curve I
6^ 56 5*332 ?5
Figure 164. Adsorption and leaching of bromodichloromethane on a 0.76 meter (2.5 feet) deep
XE-340 column (ED1R and ED2).
Tf-ri—r—i—i Days
1411 74 0 Curve
II
-------
60 -i
to
10 -
• Curve I - Chloroform
leaving the column
—O — Curve II - Chloroform
entering the column
0
Days
63
% 7 11 14 25 32 53 56
Figure 165. Level of chloroform entering and leaving the partially
exhausted 0.76 meter (2.5 feet) deep XE-340 column (ED2)
-------
00
60 -
50 _
40-
30_
M
<0
•P
-H
Curve I - adsorption from finished water (ED1R)
—O—Curve II - leaching (ED2)
Days Curve I
32 25 1411 740
Figure 166. Adsorption and leaching of chloroform on a 0.76 meter (2.5 feet) deep XE-340
column (ED1R and ED2) .
, Days
-------
15—i
to
-j
vo
3
Curve I - cis l,2-0ichloroethene
0 leaving the column
—o—Curve II- cis 1,2-oichloroethene
entering the column
7 11 1^ 25 30 53 56 63
Figure 167. Level of cis 1,2-Dichloroethene entering and leaving the partially exhausted
0.76 meter (2.5 feet) deep XE-340 column (ED2).
-------
Curve I - adsorption from finished water
(ED1R)
_O-Curve XI " leaching (ED2)
to
00
o
15
0 3
Days Curve I
Figure 168. Adsorption and leaching of cis 1,2-dichloroethene on a 0.76 meter (2.5 feet)
deep XE-340 column (ED1R and ED2).
-------
Biological Activated Carbon (BAG)
An aquatic microbiology laboratory has been established in
the Drinking Water Quality Research Center at Florida Inter-
national University. At the beginning of ED4, Dr. Frances
Parsons began a bacteria profile study of raw and finished water
at the Preston Plant and the effluent from each of the four GAC
columns. Samples of water from the distribution system were
.also analyzed. This work was contributed to this project by
Florida International University to demonstrate their capability
and interest in this area. Two reports by Dr. Parsons are
available in Appendix A. It is obvious from this work that we
had a BAG system. Since no additional oxygen was added to the
water, we call it a partial BAG system to differentiate it from
the oxygenated European system. We do not know how our adsorp-
tion results would differ if the bacterial growth had not
developed. Despite massive bacterial growth that hindered and
finally prevented backwashing of the columns, the adsorptive
capacity of the GAC for HOC did not seem to be affected. The
initial breakthrough and saturation times for each HOC through
each column were too consistent to suggest blocking of active
sites or the pore openings in the carbon by the bacteria.
The two reports in Appendix A indicate the inadequacy of
the standard plate count method in bacterial analysis of drink-
ing water. It stresses the need for longer incubation times at
various temperatures and with different media for specific
species as first indicated by Van Der Kooij (9).
Conclusion and tentative recommendations from the two
reports by Dr. Frances Parsons are as follows:
Report I:
Bacteria that occur in small numbers in raw water survive
treatment and colonize granular activated carbon (GAC) columns
used to remove organic solutes from treated water. The bacteria
multiply, form slime that interferes with column maintenance by
preventing backwashing, and slough off in large numbers into
the water passing through the columns.
Based on influent and effluent sampling, the size and com-
position of the microbial populations in GAC columns appear to
change with time. The composition of the microbial population
of the raw water apparently influenced the population in the
columns. Each column had a somewhat different population com-
position and size on each sample date.
Some of the organisms that multiply in the GAC columns may
pose a health hazard because of the vast numbers Pfesent if the
column effluent is ingested or comes in contact with suscept-
ible body surfaces such as the otic canal or the naso-pharyngeal
281
-------
mucosa. The possibility of a consumer incurring enteritis,
intoxication, and/or an opportunistic infection should be
studied. Because of the large numbers of Gram-negative organ-
isms that colonize GAC columns, endotoxin should be assayed
using the LAL method. Staphylococci sp. sometimes present in
finished water should be tested for coagulase.
The large numbers of noncoliform bacteria found in column
effluents will suppress coliform growth and interfere with
interpretation of the standard coliform detection test.
Effects of rechlorination of column effluents on the sub-
sequent bacterial population of this water over a period of
time is being studied. Preliminary results indicate that enter-
ing organisms that survive treatment plant processes become a
major component of the microbial population in GAC columns. A
count of 300/100 mL of sample of Enterobacter agglomerans was
obtained in a GAC column effluent .sample two days after
rechlorination to 3 ppm. Two colonies of Enterobacter
agglomerans per 100 mL of sample were isolated from a sample of
effluent that was chlorinated to 10 ppm and held at 25°C for six
days. These results suggest that small numbers of bacteria can
survive chlorination probably inside of cell aggregates.
Report II:
Chlorination of the effluent from granular activated carbon
(GAC) columns apparently kills bacteria that grow on the carbon
granules and slough off into the effluent, but the initial dose
of chlorine must be adequate to combine with the bacteria and
leave sufficient free chlorine to prevent regrowth. The con-
centration of chlorine necessary would vary with the bacterial
biomass and chlorine demand due to all constituents of the
water.
This study was of a cursory nature and was only intended to
suggest a more complete study. Shorter sampling intervals
(daily), for a period of time longer than six days (end point
determination), with more than these two concentrations
(especially less than 3 ppm free chlorine) of several disin-
fectants (chlorine, chloramines, chlorine dioxide, ozone,
ferrates) should be examined. Certainly, the minimum level of
chlorine needed and the time that it is effective for several
bacterial population sizes should be determined. All of these
factors; i.e., dose size, contact time, regrowth rate and size
and composition of the bacterial population should be studied
and compared with parallel determination of the bacteriology of
the distribution system.
282
-------
Polanyi-Manes Adsorption Theory
Theory Development—
Theories of adsorption from solution were developed by
Polanyi. (10) They were modified by Hansen and Fackler. (11)
Manes and Hofer (12) made modifications for predicting relative
adsorption potentials from the refractive index of a substance.
With these modifications one can estimate the adsorptive capa-
ity at saturation of a variety of miscible organic liquids by
activated carbon over a wide concentration range.
Deviations from the Polanyi Theory and its modifications
are ascribable to specific chemical interactions or steric
effects. Work on adsorption of miscible liquids from water has
been done by Wohleber and Manes. (13,14) Their work has been
expanded by Chiou and Manes (15) to include solids from solu-
tion. Adsorption of Binary Organic Liquids was studied by
Schenz and Manes. (16) The most recent work on competitive
adsorption was reported by Rosene and Manes. (17,18,19,20)
Application of the theory usually starts with adsorption
in the gas phase. The mechanisms involved are illustrated in
the following drawing of a micro pore in a carbon particle.
0 molecules in the gas phase
O
more concentrated gas phase
liquified
283
-------
Two molecules in the gas phase are attracted by
Van der Waals forces. The distance between two molecules is
"r" and the attracting force decreases by -T-. As the mole-
r
cules enter the pore they become more concentrated, and when
"r" becomes small enough they will condense into a liquid.
Molecules of both gas and liquid are held to the carbon surface
molecules by Van der Waals forces.
There are two parts to the driving force responsible for a
substance going from the gas to the liquid phase. One is the
energy part of the attractive force. It is related to the
polarizability. The other part of the driving force is entropy,
which is related to solubility. The less soluble a substance
is, the easier it is to bring to saturation (condensation). The
driving force is expressed as e.
P_
e = RTln s
R = ideal gas constant 1.987 cal/deg/mole
T = absolute temperature
P = saturated pressure
S
P = equilibrium pressure
The polarizability of a substance is ps~
s ni
P = T
where n. = refractive index.
The refractive index for all hydrocarbons is about the
same, therefore pb, where b stands for butane , equals 0.236.
To determine the adsorption isotherm curve for heptane adsorbed
on a particular GAC (gas phase) , the volume of heptane adsorbed
(cc per 100 gm of GAC) is measured at various concentrations of
heptane in the gas phase. This is usually plotted with concen-
tration expressed as e on the horizontal axis.
4.6V
The natural log is converted to base 10 log by:
In X = 2.3 log X
P P
e = 1.987 T 2.3 log -- =4.6 T log
molar volume = V molecular weight
density
284
-------
divide both sides of equation by V
divide both sides of equation by 4.6
The adsorption isotherm (gas phase) for butane on GAC
Filtrasorb 400 is shown in Figure 169. On the horizontal axis,
"0" is adsorption on GAC from pure butane in the gas phase. The
higher numbers on the axis represent lower concentrations of
butane (usually in nitrogen gas) . The seven data points making
up the Butane Gas |>hase curve in Figure 169 (and the four data
points for the Butane Gas Phase on XE-340 in Figure 170) were
supplied by Rosene at the Calgon Corporation. His data extended
to the 0.1 cc range on the "Y" axis. Later he indicated that
subsequent work on GAC resulted in data points falling on a
tangential straight line extension of the curve below 0.1 cc.
This extension is drawn in Figure 169 as a dash line segment.
Later in this discussion we will extend this straight line still
further for both GAC and XE-340 to get some idea of the behavior
of very low concentrations .
To determine how another gas substance, other than hydro-
carbon, will adsorb on the same GAC (if it is a substance that
follows the modified Polanyi Theory), one can calculate a theo-
retical adsorption isotherm using a scale factor YS/ based on the
ratio of the polarizability of a given substance p to the polariz-
ability of butane
The theoretical curve is obtained by multiplying any point
on the butane curve by ye •
O
Using the Polanyi Theory modifications of Hansen and
Fackler, as well as the modifications of Manes and Hofer, one
can calculate a theoretical adsorption isotherm curve for a
substance from a solution, in our case water, by the following:
where C = solubility of the substance in water (gm/lOOcc)
water) .s
C = concentration of the substance in the influent water.
The scale factor for a liquid substance (s) adsorbed from
another liquid (in our case from water) is Ysl-
285
-------
100
4.6V
Figure 169. Chloroform adsorption by 0.76 meter (2.5 feet)of GAC.
286
-------
100.
SemMxifiarUhiSc
4 Cycles z 10 to the Inch
Figure 170, Butane adsorption curves for K-400 GAG and XE-340 resin.
287
-------
Ysl = Ys -
= scale factor for substance to be adsorbed
= scale factor for water
Y »
e
(1.33281) +2
= 0.206
Y , = EL - 0.206
S P
For chloroform the calculated y , is:
ni2-!
ni2 + 2
(1.44643)2+2
= .2669
(refractive index
from Table 4 7 )
^2669.
. ^ Jo
Using this scale factor (.93), take as many points on the butane
curve (Figure 169 ) as needed to produce the predicted curve for
adsorption of chloroform from pure water on Filtrasorb 400. For
example, take the lowest data point "A" on the Butane Gas Phase
(GAG) curve having coordinates 0.1 cc and 26.2 for . gv. The
horizontal coordinate corresponding to 0.1 cc on the 'predicted
chloroform curve is .93 (26.2) = 24.4. This is point "B".
Continuing in this manner generates the predicted curve for
chloroform shown in Figure 169. This curve indicates the cc of
chloroform that will be adsorbed from purified water per 100
grams of Filtrasorb 400. Physical data for the halogenated
organic compounds required for application of the Polanyi-Manes
adsorption theory are provided in Table 47. As seen in Table 47,
the solubility of chloroform is 0.82 grams/100 cc of water.
e _ T Cs
4.6V - v 10g C~
T = 295
V = 80 (from Table 47 )
288
-------
TABLE 47. PHYSICAL DATA FOR POLANYI-MANES CALCULATIONS
N3
00
VO
Cpd.
No. Compound
1
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Methylene chloride
Trans 1 , 2-dichleroethene
1 , 1-dichloroethane
Cis-1 , 2-dichloroethene
Chloroform
1,1, 1-trichloroe thane.
1,2-dichloroethane sum
Carbon tetrachloride
Trichloroethylene
Bromodichloromethane
Tetrachloroethylene
Chlorodibromome thane
Chlorobenzene
Bromoform
p-chlorotoluene
m-dichlorobenzene
p-dich lorobenzene
o-dichlorobenzene
Vinyl chloride
Total
ED4
Cone,
in fin.
water
ug/L
SD
.77
.4
24.1
67.3
7.7
.68
47
.003
33.6
.36
2.5
.1
nil
.21
.14
6.2
191.6
ED4
Cone, of
cpds.5,6, Ysl
11 S 13 Scale
pg/L Factor
.875
.931
24.1 .937
67.3 .93
.906
.928
.961
.993
47 1.033
1.052
33.6 1.14
1.092
1.24
1.081
1.135
1.084
1.148
.752
172
90* of Total
Solubility
e in water
4.6V gra/100ce
22°
22.68 .63
20.17 .35
18.75 .8222*
.8215
.86920
.082°
20.86 .1
18.45 .60622*
29*
18.80 .519
16.68 -MS%1°
21.00 *,g30
.012325
13.88 .007925
15.75 .014525
24.06 .2825
17.7 009
Molecular
wt.
84.
96.
98.
96.
119.
133.
98.
153.
131.
163.
165.
208.
112.
252.
126.
147.
147.
147.
62.
93
94
97
94
38
42
96
82
39
83
85
29
56
75
58
01
01
01
6 lig.
gas
Density
gm/cc
1.
1.
1.
1.
1.
1.
1.
1.
1.
2.
1.
2.
1.
325
26
177
284
492
338
256
594
464
006
623
451
106
2.89
1.
1.
1.
1.
.
m
07
288
241
305
9013
00279
Molar
Volume Dipole
cc/mole Moment
64.1 1.54
76.
84.
75.
80.
99.
78.
96.
89.
81.
102.
85.
101.
87.
118.
114.
118.
112.
69.
9
1
5 1.9
0 1.02
7 1.79
8 1.19
5 0
8 1.22
7
2
0
8 1.7(1.55)
5 1.8
3
1 1.72
5 0
7 2.52
34
Refractive
Index
1.424
1.4490
1.4519
1.44643
1.4377
1.4448
1.46305
1.4777
1.4964
1.5055
1.5482
1.52479
1.5980
1.5193
1.54570
1.52104
1.5518
1.370
ND = not determined
*0ur analysis in tap water at 22°C
-------
V -82
C = concentration of chloroform in water
= 67.3 ug/liter from ED4
=6.73 ug/100 cc
= .00000673 gm/100 cc
£ 295 i _ _. .82 -I Q -jj-
T76V ~ ~W i0g .00000673 ~ 18-75
This means that at 18.75 on the horizontal axis (Figure 169)
which corresponds to a concentration of 67.3 ug/L of chloro-
form in purified water, we expect Filtrasorb 400 to adsorb about
0.68 cc (1.015 grams) per 100 grams of GAC. Manes and his asso-
ciates have studied the adsorption of several halogenated organic
compounds, including chloroform, from pure water on GAC. Actual
data points on chloroform coincided very well with predicted
values.
Theory Application—
Application of the Polanyi-Manes Theory to interpretation of
data in this report first considers chloroform adsorption from
finished water by 0.76 (2.5 feet) meter deep columns of GAC.
Two runs were made, EDS and ED4, Table 48 . In these two ED,
average influent levels were 57 yg/L and 67.3 yg/L respectively.
Adsorption per 100 grams of adsorbent at saturation were
0.0280 cc and 0.0358 cc respectively. The adsorptive capacity
increased as the concentration increased, as predicted by the
Polanyi-Manes Theory. These two finished water data points are
plotted in Figure 169. The two vertical lines from the "X" axis
were drawn from the appropriate k >v values shown in Table 48.
These vertical lines were then projected from their intercept
points on the Chloroform Predicted (GAC) curve to the "Y" axis.
For ED3 and ED4, respective predicted adsorption values were
0.64cc and 0.68 cc from pure water compared to actual adsorption
values of 0.028 cc and 0.0358 cc. Respective actual values are
4.4 percent and 5.3 percent of the predicted values. These data
appear in Table 48. This reduction is the result of competitive
adsorption by DOM, including other HOC. To construct an actual
adsorption curve through the two finished water data points it
would be better if the two points were further apart. However,
working with what is available, we can calculate the actual YS!
value for these two points from our actual water. Projections
horizontally from the two data points to the Butane Gas Phase
(GAC) curve indicate z values of 28.8 and 28.4 respectively.
4. bv
290
-------
TABLE 48. CHLOROFORM ADSORPTION DATA
FROM FINISHED WATER
ED
1
1R
'2
3
3
4
4
4
4
£
t3 a
<]> (U
ffl T3
Feet
2.5
2.5
2.5
2.5
2.5
2.5
5
7.5
10
Adsorbent
XE-
340
XE-
340
XE-
340
GAG
904
GAG
GAG
GAG
GAG
r Average
*x influent
f
80.2
69.3
64
57
57
67.3
67.3
67.3
67.3
«
1 Adsorption per
w 100 grams adsorbent
at saturation
.265
.22
.2
.042
.0534
.067
.071
.073
e
CC
.177
.148
.134
.028
.0358
.049
.048
.049
4.6V
18
18
18
19
18
18
18
18
.45
.71
.83
.02
.75
.75
.75
.75
Predicted Polanyi-
o Manes adsorption
for GAG
.64
.68
.68
.68
.68
Percent adsorption
* of predicted value
4.
5.
7.
7.
7.
§
M
0
-------
Respective actual YS! values are 0.660 and 0.660. Using this
Ysl value, the actual adsorption curve, shown in Figure 169, was
generated, and it passed through the two data points. We predict
that for our water, this generated actual chloroform adsorption
curve can be used to predict chloroform results on our water over
the entire chloroform concentration range we will experience. To
generate this curve, only one actual data point is required to
calculate the actual Ysl value. Since there was nil chloroform
in our raw water, we cannot test this prediction from our data on
chloroform. In the discussion on XE-340 which follows we will
show that it appears to work down to a very low chloroform level.
The curves in Figure 169 are a log-log plot since the equation
for _!•— contains a log function. In our water, the concentra-
tions'of HOC are within the tangential straight line portion of
the adsorption curve, therefore the log-log plot of data points
mentioned earlier in this report, fall on a straight line and the
line can be used to predict adsorption values at different con-
centrations .
Chloroform adsorption from finished water by 0.76 (2.5 feet)
meter deep columns of XE-340 were studied in three runs, EDI,
ED1R, and ED2, Table 48 . The Butane Gas Phase (XE-340) curve
and the generated Chloroform Predicted (XE-340) curve for pure
water appear in Figure 170- The predicted curve was again
generated from the gas phase curve by using the 0.93 Ysl scale
factor (Table 47) for chloroform. The average influent level
varied from 80.2, 69.3 to 64 yg/L. For these three runs, compare
the grams of chloroform adsorbed per 100 grams of adsorbent at
saturation with their respective average influent level. The
adsorptive capacity of XE-340 for chloroform decreases (0.177 cc,
0.148 cc and 0.134 cc) as the influent concentration decreases,
as predicted by the Polanyi-Manes Theory. Using £ values
4.6V
(from Table 48) corresponding to the three chloroform concentra-
tions of 80.2, 69.3 and 64 yg/L, the three levels of cc's
adsorbed per 100 grams of adsorbent at saturation (also shown in
Table 48) are plotted in Figure 171. The three vertical lines
from "X" axis correspond to the three e values. The three
finished water data points are replotted'in Figure 172 on an
expanded "X" and "Y" axis scale for greater accuracy to show that
they fall approximately on a straight line. To minimize drawing
error the slope of this was transferred to the scale in Figure
171 and the actual adsorption curve drawn as shown through the
three finished water data points. It is apparent that the actual
adsorption curve is not parallel to the Butane Gas Phase (XE-340)
curve nor to the Chloroform Predicted (XE-340) curve. The
predicted curve predicts adsorption from pure water of 0.105 cc,
0.094 cc and0.09cc instead of the 0.177 cc, 0.148 cc and
0.134 cc actually adsorbed from our finished water. Obviously,
the Polanyi-Manes Theory does not apply to XE-340, an adsorbent
292
-------
9_
8..
7_
.... " " ! ' "~~" i "1! I——|-
_water_dat;a_Eoints;i'... _l
"
.01
0.76 me
(2.5 feet)of XE-340 (EDI, ED1R and ED2).
293
-------
I
CO
=3
O
01
O
O
i-H
a
V
•a
O
(0
•d
(8
u
O
19.0
4.6V
Figure 172. Chloroform adsorption by 0.76 meter (2.5 feet)
of XE-340.
294
-------
that not only allows micropore surface adsorption but adsorption
into the polymer matrix. However, if at least two actual ad-
sorption data points are obtained on a water system, the actual
adsroption curve could be drawn to predict adsorption at some
other concentration on the straight line portion of the curve.
We do not have finished water experimental data to test this
possibility but we do have a data point on H.T. water in EDlR,
Table 19. Competitive adsorption by HOC and TOC will'be differ-
ent in H.T. and finished water so one cannot expect too much
from this comparison. The H.T. data point is plotted in Figure
173 at the appropriate e value of 25.2. From the prediction
4.6V
curve we predict an adsorptive capacity of 0.0023 cc and we reported
an observed capacity of 0.0027 cc. Granted, error possibilities
are great, but at least even at this very low concentration
(1.2 yg/L of chloroform) adsorption does occur, can be measured
and the adsorptive capacity predicted with some degree of
success.
Adsorption of cis 1,2-dichloroethene on 0.76 (2.5 feet)
meter of XE-340 was studied in EDI, EDlR and ED2. The Butane
Gas Phase (XE-340) curve appears in Figure 174. From this curve,
the cis 1,2-dichloroethene predicted (XE-340) curve was genera-
ted by using the scale factor ys^ of 0.937 appearing in Table
47. As with chloroform adsorption on XE-340, Figure 171, plot-
ted cis 1,2-dichloroethene adsorption points in Figure 174 lie
above the predicted curve and, in the case of raw and H.T. data
points, lie above the Butane Gas Phase (XE-340) curve. Lines
drawn through the data points are not parallel to the predicted
curve. As with chloroform XE-340 data, the Polanyi-Manes Theory
does not apply to adsorption of cis 1,2-dichloroethene by XE-
340. Again however, two actual data points on raw, H.T. and
finished water should be sufficient to generate an actual ad-
sorption curve for our samples. In Figure 174, notice how ad-
sorption data points from raw and H.T. water fall nearly on the
same straight line. Notice also how adsorption points for fin-
ished water fall on a displaced straight line, indicating con-
siderably less adsorptive capacity per 100 grams of adsorbent.
This reduction in adsorption of about 30 percent is due, we
believe, to increased competitive HOC adsorption, as was dis-
cussed for chloroform on page 112 and for cis 1,2-dichloroethene
on pages 113 and 1S3. The adsorptive capacity of XE-340 is
about the same from raw and H.T. water, indicated by the actual
data points falling on almost the same curve. The capacity of
XE-340 to adsorb from finished water was less. The competition
of TOC adsorption decreased from raw to H.T. to finished water,
corresponding to TOC values of 8.3, 5.8, and 5.4 mg/L. This
would indicate possibly greater adsorptive capacity of cis 1,2-
dichloroethene from finished, water. The level of HOC in finish-
ed water is approximately 150 yg/L compared to cis 1,2-dichloro-
ethene levels 25 yg/L in raw and H.T. water. The competitive
adsorption of other HOC would indicate less adsorption capacity
295
-------
.00*-
A
18.83 25.2
Figure 173. Chloroform adsorption by 0.76 meter (2
XE-340 (EDI, ED1R and ED2).
5 feet) of
296
-------
1..0..
points:: (EDl .and :EDlR)_;r ^-p:tr
.".Finished v/ater_data_^Epints_(EDl
' |
17 18 19 20 21
Figure 174. cis 1,2-Dichloroethene adsorption by 0.76 meter
(2.5 feet) of XE-340.(ED1, ED1R and ED2).
297
-------
in finished water. The data showed less adsorption; therefore,
this may suggest that competitive adsorption of HOC has a great-
er influence than TOC competitive adsorption from finished water.
Adsorption of cis 1,2-dichloroethene on 0.76 (2.5 feet)
meter of GAG was studied in ED3 and ED4. Plotted curves and
actual data points appear in Figure 175. Raw and finished water
data points, as with XE-340 in Figure 174, fall on displaced
lines, indicating again the reduction in adsorptive capacity due
to increased competitive HOC adsorption. The lines through the
actual data points in Figure 175 were drawn through the points
by sight and by calculating the ysl value for each point and
generating the line from Butane Gas Phase (GAG) curve. Since
the two methods produced the same lines, on GAG columns, one
actual data point is enough to generate an actual adsorption
curve. The predicted Polanyi-Manes adsorptive capacities for
GAG in EDI, EDlR, EDS and ED4 appear in Table 49. The percent
of actual adsorption in our water is also shown. In two separ-
ate runs, columns 0.76 (2.5 feet) meter deep on raw water both
adsorbed 8.8 percent of the predicted adsorption value from pure
water. On finished water two runs at the same bed depth adsorb-
ed 6.4 and 6.5 percent of the predicted adsorption value from
pure water. In ED4, as bed depth increased, the adsorptive
capacity increased as expected due to less competitive HOC ad-
sorption with increasing bed depth.
Adsorption data from Table 50 for bromodichloromethane by
0.76 (2.5 feet) meter of XE-340 from H.T. and finished water
are plotted in Figure 176. Since the H.T. water data point is
probably not on the same straight line as the finished water
data points, the dashed line drawn in Figure 176 represents only
an approximate actual adsorption curve for this HOC in our sys-
tem. Adsorption data from Table 50 for bromodichloromethane by
0.76 (2.5 feet) meter of GAG from finished water are plotted in
Figure 177. In this case the bromodichloromethane predicted
(GAC) curve lies above the Butane Gas Phase (GAG) curve because
the YS! value for this HOC in Table 47 is 1.033. The calculated
Ysi value for both data points, obtained from the curves, was
0.652. This indicated that the two points fall on a straight
line parallel to the Butane Gas Phase (GAC) curve. Thus, one
data point would have been sufficient to generate the actual
adsorption curve as shown. The actual adsorption of this HOC
in our water is only about 3.6 percent of the predicted value
from pure water.
Adsorption data from Table 51 for chlorodibromomethane by
0.76 (2.5 feet) meter of XE-340 from H.T. and finished water
are plotted in Figure 178. The approximate actual adsorption
curve is shown. Adsorption data from Table 51 for adsorption
by 0.76 (2.5 feet) meter of GAC from finished water are plotted
in Figure 179. The calculated ysl value for both data points,
obtained from the curves, were 0.602 and 0.598, averaging 0.6.
298
-------
_O _Baw._water. data, points. ;_P l.U.a. I|\-!_M.. -l-)-|-(
; • I ! ; • \ X :.....:
D Finished water data points - •. \ ^
Figure 175. cis 1,2-Dichloroethene adsorption by 0.76 meter
(2.5 feet) of GAX: (ED3 'and ED4) .
299
-------
TABLE 49. CIS 1,2-DICHLOROETHENE ADSORPTION DATA FROM
RAW, H.T., AND FINISHED WATER
ED
Bed
depth
feet
Adsorption per
100 grams Predicted Percent
Average adsorbent at Polanyi adsorption
influent saturation adsorption of
E predicted
Adsorl ant
yg/L Grams
CC 4.6V CC
value
HAW WATER
1
1R
1R
1R
2.5
2.5
2.5
2.5
GAC
XE-340
GAC
XE-340
21
21
29
29
.048
.15
.065
.181
.037 20.4 .42
.117 20.4
.0509 19.36 .58
.141 19.86
8.g
8.8
H.T. WATER
1
1R
2.5
2.5
XE-340
XE-340
?0
25.4
.134
.157
.104 20.49
.122 20.08
FINISHED WATER
1
1R
2
3
4
4
4
4
2.5
2.5
2.5
2.5
2.5
5
7.5
10
XE-340
XE-340
XE-340
GAC
GAC
GAC
GAC
GAC
10.9
19.4
18.4
18.3
19.9
19.9
19.9
19.9
.043
.093
.087
.033
.039
.045
.049
.003
.0724
.068
.0257 20.57 .4
.03 20.17 .46
.035 20.17 .46
.0357 20.17 .46
20.17
6.4
6.5
7.6
7.8
2.5 feet=0.76 meter
5 feet=1.52 meters
7.5 feet=2.29 meters
10 feet=3.05 meters
300
-------
TABLE 50. BROMODICHLOROMETHANE ADSORPTION DATA FROM
H.T. AND FINISHED WATER
ED
1
1R
Bed
depth
feet
2.5
2.5
Adsorption per
100 grams
Average adsorbent at
influent saturation
Adsorbent
XE-340
XE-340
Vg/L Grams
variable
see Pig.
erratic
see Fig.
H.T.
39
40
CC
WATER
.0028
Predicted Percent
Polanyi adsorption
adsorption of
e
4.6V CC
23.1
26.9
predicted
value
FINISHED WATER
1
1R
2
3
4
4
4
4
2.5
2.5
2.5
2.5
2.5
5
7.5
10
XE-340
XE-340
XE-340
GAC
GAC
GAC
GAC
GAC
37.1
42.7
42.4
39
47
47
47
47
.182
.204
.069
.084
.083
.084
.0907
.1017
.0344
.0419
.0414
.0419
18.82
18.60
18.74 1.13
18.45 1.21
18.75 1.14
18.75 1.14
18.75
3.04
3.46
3.63
3.68
2.5 feet=0.76 meter
5 feet=1.52 meters
7.5 feet=2.29 meters
10 feet=3.05 meters
301
-------
SpiffliiSi
liijjjffr -^-jjg|^^f£y£L:l3y^5i3 ^T^rtt^
^J^S^^I^S^¥S^':^^:^^^^^
j^^^i^-ff^iBK^^^^^^f^^
302
-------
• 1-!
; • ' I-TJ j si" f =• 11 = ;
4^-^-r~4i44^4^u44^ ^r •': ;:;1
-•—i--H' . —i 1—^ *..j .j-
?}^f^E!^
'•—'-tH- :4-t-(-tJ:F
iv - ----- --•-
-4-V-H-Hh
-pj-ye^-t—r-i-
.01
18
Figure 177. Bromodichloromethane adsorption by 0.76 meter
(2.5 feet) of GAC (ED3 and.ED4).
303
-------
TABLE 51. CHLORODIBROMOMETHANE ADSORPTION DATA
FROM H.T. AND FINISHED WATER
Bed
depth
ED feet
Adsorbent
Adsorption per
100 grams
Average adsorbent at
influent saturation
Grams CC
Predicted Percent
Polanyi adsorption
adsorption of
_£ predicted
4.6V CC value
1 2.5 XE-340
1R 2.5 XE-340
H.T. WATER
1.0 .0009 23.31
.25 .0005 25.39
FINISHED WATER
1
1R
2
3
4
4
4
4
2.5
2.5
2.5
2.5
2.5
5
7.5
10
XE-340
XE-340
XE-340
GAC
GAC
GAC
GAC
GAC
12 .077 .031 19.56
24.5 .155 .063 18.48
26.7
27 .067 .027 18.34 2.2
33.6 .104 .042 18.01 2.3
33.6 .1 .041 18.01
33.6
33.6
1.2
1.8
2.5 feet=0.76 meter
5 feet=1.52 meters
7.5 feet=2.29 meters
10 feet=3.05 meters
304
-------
ts
a)
•8
o
CQ
O
O
•a
I
.001
.0001
Figure 178. Chlorodibromomethane adsorption by 0.76 meter
(2.5 feet) of XB-340 (EDI and ED1R).
305
-------
en data points
_IjEi)3.!and'jED4 T.TI
18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
4.&V
Figure 17.9. Chlorodibromomethane adsorption from 0.76 meter
(2.5 feet) of GAC (ED3 and ED4).
306
-------
Once again, this indicated that both points lie on the same
straight line parallel to the predicted curve. The actual ad-
sorption of this HOC in our water is only about two percent of
the predicted value from pure water.
Adsorption data in this part of the report for the HOC dis-
cussed above and for additional HOC appear in Table 52. The
chemical identification number given in Table 8, for each HOC
studied correspond to their elution time on a GC through a
Tenax column (except for vinyl chloride, Chem. No. 20, which
elutes first). Chemical No. 3 is eluted first (after vinyl
chloride), and 17, 18, and 19 are eluted last. In Table 52, it
is seen that the Ysl values in most cases predict this order of
elution from a Tenax column. This table supports and adds to
the data in Table 37, which is discussed beginning on page 193.
We have found the Polanyi-Manes Adsorption Theory useful in
interpreting and explaining the data we have obtained. It is
obvious that more fundamental research is needed in the labora-
tory to determine the actual adsorption of all our HOC from
pure water. This will indicate whether or not our observed
reduction in adsorption as the molecular weight of the HOC in-
creased, is due to steric exclusion or largely to competitive
HOC adsorption. We found the Theory useful in predicting re-
sults in a given system where the concentration of dissolved
substances change in magnitude but not in ratio.
307
-------
TABLE 52. SUMMATION OF ADSORPTION PARAMETERS BY GAC
u>
Cpd.
No. Chemical
3 Trans 1, 2-dichloroethene
5 Cis-1 , 2-dichloroethene
6 Chloroform
7>v 1 f 1 / 1- trichloroethane
\ \
8 ^>l,2-dichloroe thane sum
y Carbon tetrachloride
10 Trichloroethylene
11 Bromodichloromethane
13 Chlorodibromomethane
15 Bromoform
17v m-dichlorobenzene
18 ^ p-dichlorobenzene sum
1» o-dichlorobenzene
Column
Bleed
Time
Days
52
18
7
29
38
15
21
42
none
Column
Saturation
Time z
Days Inches
? 16+?
65 22
23 21
76 18
56 9.6
54 21
89 22
94 16.6
>122 <10
Ysl
.931
.937
.93
.906
.928
.961
.993
1.003
1.14
1.24
1.135
1.084
1.148
Polanyi-Manes
Predicted
Capacity
cc
.03
.46
.68
.65
1.14
2.2
1.7
4.9
Observed
Capacity Percent of
Predicted
eg
.002 6.7
.029 6.5
032 5.0
.0012 .19
.04 3.6
.04 1.8
.002 .12
Total .1462
-------
GC/MS HOC Confirmation Data
Periodically during the two-year study, GC/MS determinations
were made on both raw and finished water to confirm GC peaks of
HOC. All nineteen HOC have been confirmed. Sample dates and
results are summarized in Table 53. Analyses on sample dates
August and November 4, 1976; and May 3 and May 20, 1977 were
determined on a Hewlett-Packard Mass Spectrograph, Model 5981,
with a Model 5933 Dual Disc Inter-Active Data System. The data
on October 11, 1977 were obtained by EPA Laboratories in
Cincinnati. Analyses on sample dates March 11 and April 17, 1978
were made on a Varian MAT Model 112S Magnetic Sector, double-
focusing, high resolution mass spectrograph coupled to a MAT 166
data system.
309
-------
TABLE 53. GC/MS HOC CONFUTATION DATA
U)
M
O
I 4/17/78
Cham Raw
No. vg/
20 14.2
. 1 .45
3 j 1.6
4 | .6
5 J24.3
6 j nil
7 £
8 jj.13
9
10
11
12
13
14
15
16
17-W
ben-
^^ns
tol-
uene
I
.37
nil
.003
nil
1.2
nil
.03
nil
.57.19
NR
MR ,
Wtr.
KS*
Y
Y
Y
Y
Y
ND
ND
Y
Y
Y
Y
Y
ND
Y
ND
Y
ND
J2L
Y
Y
Pin.
pc/
11.3
int
_st<3
.87
.5
22.1
69
a.i
.63
46
.003
34.3
.81
2.7
.11
sy*
NR
NR
Wtr.
MS
Y
NR
Y
Y
Y
' Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y/Y
Y
Y
3/11/78 . .
Raw
\ig/
12.8
.52
1.3
.7
27.4
nil
j.17
.41
nil
.003
nil
1.3
nil
.04
JB&
NR
NR
Wtr.
MS
Y
Y
Y
Y
Y
ND
ND
Y
Y
Y
ND
Y
ND
ND
Y
Y^Y
Y
Y
Fin.
pg/
11.7
int.
stcU
.78
.61
25.2
73
7.8
.72
43.4
.002
37.2
2.9
.13
JV-19
NR
NR
Wtr.
MS
Y
NR
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
!ND
1X/Y
Y
Y
10/11/77.**
Raw
V<3/
10.8
.13
2.1
X.O
19.8
nil
.16
.58
nil
nil
nil
nil
.87
1.2
ND
Wtr.
MS
Y
Y
Y
Y
Y
ND
ND
Y?
ND
Y
ND
N0
ND
ND
A
A
Y
•Fin.
v,
-------
REFERENCES
1. Symons, J.M., et al. National Organics Reconnaissance
Survey for Halogenated Organics in Drinking Water. JAWWA,
67(11) :634-647, 1975. Update, 67 (12):708-709, 1975.
2. Bellar, T.A. and J.J. Lichtenberg. Determining Volatile
Organics at Microgram per Liter Levels in Water by Gas
Chromatograph. JAWWA 66 (12):739-744, 1974.
3. Dressman, R.C. and E.F. McFarren. Sample bottle purging
method for determination of vinyl chloride in water at sub-
microgram per liter levels. J. of Chrom. Science, Vol. '15,
pp. 69-72, 1977.
4. Stevens, A.A. and J.M. Symons. "Trihalomethanes and Precur-
sor Concentration Changes Occurring During Water Treatment
and Distribution." JAWWA 69(10) :546 (Oct. 1977).
5. Aldrich Chemical Co., Inc. Phase-Transfer Catalysis in
Organic Synthesis. Aldrichemica Acta, Vol. 9, No. 3 (Final
Issue), 904 W. St. Paul Ave., Milwaukee, WI 53233, 1976.
6. Neely, J.W. A model for the removal of trihalomethanes from
water at Ambersorb XE-340. Rohm and Haas Company, Research
Laboratories, Spring House, PA 19477. IN: Proceedings of
the ACS Annual Meeting, Environmental Section: Activated
Carbon Adsorption of Organics from the Aqueous Phase, Miami
Beach, FL, 1978.
7. DeMarco, J. and P. Wood. Design Data for Organics Removal
by Carbon Beds. IN: Proceedings of National Conference on
Environmental Engineering, Research Development and Design,
Kansas City, MO, 1978, pp. 149-156.
8. Symons, James M. Interim treatment guide for controlling
organic contaminants in drinking water using granular acti-
vated carbon. Water Supply Research Division, Municipal
Environmental Research Laboratory, Office of Research and
Development, Cincinnati, Ohio, January 1978.
9. Van Der Kooij , D. Some investigations into the presence
and behavior of bacteria in activated carbon filters. IN:
Translation of reports on special problems of water tech-
nology, H. Sontheimer, ed. Vol. 9, Adsorption, op. cit.,
MERL, USEPA, Cincinnati, OH, 1976, pp. 348-354.
311
-------
10. Polanyi, M., Verb. dent. Physik. Ges., 18, 55, 1976 and
M. Polanyi, Physik, 2, 111, 1920.
11. Hansen, R. and W. Fackler, Jr. A Generalization of the
Polanyi Theory of Adsorption from Solution. The Journal of
Physical Chemistry, Vol. 57:634-637, 1953.
12. Manes, Milton and L. Hofer. Application of the Polanyi
Adsorption Potential Theory to Adsorption from Solution on
Activated Carbon. J. Phys. Chem., 73(2):584-590, 1969.
13. Wohleber, D. and M. Manes. Application of the Polanyi Ad-
sorption Potential Theory to Adsorption from Solution on
Activated Carbon. II. Adsorption of Partially Miscible
Organic Liquids from Water Solution. J. Phys. Chem., 75(1):
61-64, 1971.
14. Wohleber, D. and M. Manes. Application of the Polanyi Ad-
sorption Potential Theory to Adsorption from Solution on
Activated Carbon III. Adsorption of Miscible Organic
Liquids from Water Solution. J. Phys. Chem., 75(24) :3720-
3723, 1971.
15. Chiou, C. and M. Manes. Application of the Polanyi Adsorp-
tion Potential Theory to Adsorption from Solution on Acti-
vated Carbon. V. Adsorption from Water of Some Solids and
Their Melts, and a Comparison of Bulk and Adsorbate Melting
Points. J. Phys. Chem., 78(6):622-626, 1975.
16. Schenz, T. and M. Manes. Application of the Polanyi Adsorp-
tion Theory to Adsorption from Solution on Activated Carbon.
VI. Adsorption of Some Binary Organic Liquid Mixtures. J.
Phys. Chem., 79(6):604-609, 1975.
17. Rosene, M. and M. Manes. Application of the Polanyi Adsorp-
tion Potential Theory to Adsorption from Solution of Activa-
ted Carbon. VII. Competitive Adsorption of Solids from
Water Solution. J. Phys. Chem., 8 (9) :953-959, 1976.
18. Rosene, M. and M. Manes. Application of the Polanyi Adsorp-
tion Potential Theory to Adsorption from Solution on Activa-
ted Carbon. VIII. Ideal, Non-ideal, and Competitive Ad-
sorption of Some Solids from Water Solution. J. Phys. Chem,
80(23):2586, 1976.
19. Rosene, M. and M. Manes. Application of the Polanyi Adsorp-
tion Potential Theory to Adsorption from Solution on Activa-
ted Carbon. IX. Competitive Adsorption of Ternary Solid
Solutes from Water Solution. J. Phys. Chem., 81:1646, 1977.
312
-------
20. Rosene, M. and M. Manes. Application of the Polanyi Adsorp-
tion Potential Theory to Adsorption from Solution on Activa-
ted Carbon. X. pH Effects and Hydrolytic Adsorption in
Aqueous Mixtures of Organic Acids and Their Salts. J. Phys.
Chem., 81:1651, 1977.
313
-------
MICROBIAL FLORA OF GRANULATED ACTIVATED CARBON
COLUMNS USED IN WATER TREATMENT
(Part I)
by
Frances Parsons
Drinking Water Quality Research Center
Florida International University
Tamiami Campus
Miami, Florida
ABSTRACT
Differential bacteria counts were made on samples of efflu-
ents of granulated activated carbon (GAC) columns used to remove
dissolved organic material from drinking water. The membrane
filter procedure, four primary media, and incubation at 25°C for
six days were used to isolate colonies. Identification was done
using Roche Diagnostics systems and additional diagnostic tests.
Most of the growth occurred on tryptone glucose extract agar and
Czapek Dox agar after four days incubation at 25°C. Most bacte-
rial growth was not detected when standard methods were used.
Raw water organisms, which apparently can survive existing
treatment plant processes, colonized the initially bacteria-free
GAC columns, and released vast numbers of bacteria into the
water flowing through the columns. Some of the organisms,
though innocuous in small numbers, may pose a threat to human
health when they are present in drinking water in large numbers.
These organisms include Pseudomonas-like bacteria, Acinetobac-
ter, Alcaligenes faecalis, Moraxella, Enterobacter agglomerans,
and Flavobacterium (probably aquatile).
The development of bacterial growth in the GAC columns
interfered with backflushing the columns. Preliminary results
indicate that the GAC system provides an ecological advantage
for entering organisms that survive treatment plant processes.
Enterobacter agglomerans in GAC column effluent survived expo-
sure to 3ppm chlorine. This suggests that at least in our sub-
tropical environment a careful study is required to insure prop-
er bacterial control before installing GAC adsorbers in treat-
ment plants.
314
-------
CONCLUSION AND RECOMMENDATIONS (TENTATIVE)
Bacteria that occur in small numbers in raw water survive
treatment and colonize granulated activated carbon (GAG) columns
used to remove organic solutes from treated water. The bacteria
multiply, form slime that interferes with column maintenance by
preventing backflushing, and slough off in large numbers into
the water passing through the columns.
The size and composition of the microbial populations in
GAG columns changed with time. The composition of the microbial
population of the raw water apparently influenced the population
in the columns. Each column had a somewhat different population
composition and size on each sample date.
Some of the organisms that multiply in the GAG columns may
pose a health hazard because of the vast numbers present if the
column effluent is ingested or comes in contact with susceptible
bbdy surfaces such as the otic canal or the naso-pharyngeal mu-
cosa. The possibility of a consumer incurring enteritis, in-
toxication, and/or an opportunistic infection should be studied.
Because of the large numbers of Gram-negative organisms that
colonize GAC columns, endotoxin should be assayed using the LAL
method. Staphylococci sp. sometimes present in finished water
should be tested for coagulase.
The large numbers of noncoliform bacteria found in column
effluents will suppress coliform growth and interfere with in-
terpretation of the standard coliform detection test.
Effects of rechlorination of column effluents on the sub-
sequent bacterial population of this water over a period of time
is being studied. Preliminary results indicate that entering
organisms that survive treatment plant processes become a major
component of the microbial population in GAC columns. A count
of 300/100 ml of sample of Enterobacter agglomerans was obtained
in a GAC column effluent sample two days after rechlorination to
3 ppm. When the concentration of chlorine was increased to 10
ppm, two colonies of Enterobacter agglomerans were recovered
from 100 ml of sample held for six days at 25°C. This experiment
was repeated and supplementary survival tests using organisms
isolated from chlorinated column effluent were done to verify
these results. Cursory examination of results of these experi-
ments indicate that small numbers of bacteria can survive in
315
-------
water containing 3 ppm chlorine for as long as six days, probably
inside of cell aggregates.
The Standard Plate Count method (APHA, AWWA, WPCF 1975) is
inadequate for enumerating these aquatic bacteria. Longer incu-
bation time and lower temperatures than specified by Standard
Methods are needed. New media that would support more kinds of
heterotrophic organisms should be developed and tested. Better
methods for identification of these organisms need to be devised.
316
-------
INTRODUCTION
Granulated activated carbon (GAG) columns that are capable
of retaining organic material and removing chlorine from water can be
expected to diminish the bacteriocidal property of treated water
passed through them and to provide metabolic substrate for
microorganisms that survive chlorination. Controlling bacteri-
al populations in treated water is important because bacteria
that may be harmless in small numbers may be capable of causing
disease under certain conditions (Geldreich 1973, Peterson and
Favero 1975). Wallis et al. (1974) pointed out that charcoal
filters used in domestic water supplies released large numbers
of bacteria to water flowing through them. Allen et al. (1977)
demonstrated that excessive bacterial populations mask coliform
growth in the standard method for determining potability of
water. Fiore and Babineau (1977) stated that activated carbon
filters in household use had no effect on bacteria counts in
water passed through them. They used the pour plate method and
incubated the cultures at 30. *€ for a 48-hour period. Klotz et
al. (1976) demonstrated that 48 hours was inadequate for the
slow-growing micro flora that developed on activated carbon fil-
ters, and incubated their cultures at 27°C for seven days. Our
work supports that of Klotz.
Although American workers in the field of water quality
are concerned about the increase in numbers of bacteria in
water filtered through GAC columns, European workers encourage
bacterial growth in carbon filters used at various points in the
treatment process (Eberhart 1976). By doing so, adsorbed carbon
material is mineralized and soluble inorganic ions are immobi-
lized in bacterial biomass. This can be an attractive feature
to American designers of water systems. Van der Kooij (1976),
Klotz et al. (1976), and Eberhardt (1976) described the develop-
ment of bacterial populations on carbon filters in Europe and
their activity in mineralizing organic substances from the water
flow.
This study is being done to determine the changes in micro-
bial population composition and size in treated water subjected
to filtration through GAC beds of various depths. It is neces-
sary to determine successional changes in bacterial populations
to determine if a potential health hazard can result 1) from
development of massive populations of ordinarily harmless bacte-
ria, and 2) from a change in kind of bacteria multiplying in
the carbon filters from innocuous species to chlorine resistant
317
-------
pathogenic species that ordinarily may be present in small num-
bers in raw water.
It will also become increasingly important to understand
how bacterial growth on carbon filter material can be controlled
to facilitate maintenance of the filters. Bacterial growth
tends to develop slime within the carbon granules, which makes
back-flushing difficult, and after a time impossible.
318
-------
METHODS AND MATERIALS
The bench model column adsorption system used in Experimen-
tal Design No. 4 of EPA Grant Project R804521-01 (Wood et al.
1979) is shown diagrammatically in Figure 1. Sample dates and
bed depths are shown also. The four 1" ID columns were connect-
ed in series and packed with 2.5 feet of GAG Filtrasorb 400 to
give bed depths of 2.5, 5, 7.5, and 10 feet. Sample ports were
located at the effluent end of each column. Two-liter samples
were collected on the dates shown and analyzed within two hours.
The membrane filter technique was used to isolate bacteria from
water samples (APHA, AWWA, WPCF 1975). Figure 2 is a flow dia-
gram of the procedure used. Table 1 is a list of diagnostic
tests and media used in identification. Membranes used to fil-
ter water samples were placed on several primary media, incuba-
ted at 25°C, and read at 1-, 2-, 3-, and 6-day intervals as this
temperature and these incubation times were shown previously to
yield higher counts and greater diversity of bacteria than those
specified by Standard Methods (APHA, AWWA, WPCF 1975).
When the primary cultures were examined, every recogniz-
ably different colony type on each primary medium was described,
assigned a number, and counted. At least two colonies of each
type were picked and each was streaked on a new plate of medium
to insure isolation. Dissimilar colonies were expected to be
sometimes identical as bacteria express different morphologies
on different media. Gram stains were made of each colony type
and examined for purity of cell morphology. When the purity of
the isolates was assured, they were inoculated into differential
media (Figure 2), incubated at 25°C, and examined daily for six
days. The pattern of biochemical reactions for each colony type
was compared with those listed in Sergey's Manual (1974), and
names were assigned.
To determine the bacterial flora of the carbon granules
used to fill the filter columns, a 250 ml volume of new carbon
granules in 500 ml sterile, buffered, deionized water was shaken
for one hour and then allowed to settle. Two hundred ml of the
rinse water was passed through each of two membranes, which were
then placed on TGE and Endo's media.
When it became apparent that large numbers of bacteria were
being generated in the granulated activated carbon columns, it
was necessary to determine the effect of rechlorination of the
effluent to control this bacterial growth. One-liter samples of
319
-------
effluent from the end of the series of columns (10 feet of gran-
ulated activated carbon) were collected and treated as follows:
1) effluent plated on primary media following membrane
filtration;
2) effluent plus 3 ppm chlorine plated after one hour con-
tact time;
3) effluent plus 3 ppm chlorine aged for two days before
plating to simulate residence time of water in a dis-
tribution system with the possibility of depletion of
residual chlorine;
4) effluent plus 10 ppm chlorine aged for six days to simu-
late a condition of overabundance of chlorine, assured
residual chlorine, and a lengthy residence time in a
distribution system.
320
-------
RESULTS
Table 2 lists the numbers of different kinds of microorga-
nisms isolated from raw and finished water from the treatment
plant prior to sampling the GAG columns. Table 3 gives total
colony counts on the different primary media obtained from raw,
finished, and filtered water samples on 11/21/77 after the GAC
columns had been in use for 19 days. Counts are reported for
100 ml of filtered sample because the low numbers of bacteria
isolated would result in fractional values if they were ex-
pressed per ml as Standard Methods (APHA, AWWA, WPCF 1975) spec-
ifies. No bacteria were isolated from water in which new carbon
granules had been shaken. Table 4 lists all the different colo-
nies picked for identification, shows the primary medium and
system location where each colony was found, and the colony num-
ber assigned for reference during the identification process.
Table 5 is a list of groups of colonies with similar biochemical
reactions that resulted from growth in differential media.
Table 6 lists the different identified organisms and their popu-
lation size in raw and finished water and column effluents.
Values given in Table 6 are the counts obtained from the most
favorable medium. Each medium favors a different population of
microorganisms and no single medium can yield an accurate esti-
mate of population size.
Table 7 gives total counts on different primary media ob-
tained from raw, finished, and effluent samples taken on 12/5/77.
Table 8 is the initial description of colonies isolated from
samples taken on 12/5/77. The original 11 colonies were mixed
cultures; they were streaked on agar plates to separate the co-
habitants, and resulted in 29 colonies that were subjected to
identification procedures. Table 9 shows the identification
process, including colony number assigned during the identifica-
tion process, colony description, identity, and population size
in each part of the GAC system. Table 10 shows the numbers of
identified organisms isolated from samples taken on 12/5/77 in
each part of the GAC system. Table 11 is the initial descrip-
tion of colonies isolated from samples taken on 1/6/78, their
numbers and location in the GAC system. Table 12 shows the num-
ber of identified organisms isolated on 1/6/78 from each part of
the GAC system.
Group (genus) names only were assigned in the tables show-
ing organism identification (Tables 6, 10, 12) as many of the
organisms isolated fit no single taxonomic category with the
321
-------
diagnostic tests used. Where species epithets are used in the
tables, identification is reasonably certain. Several organisms
were identified to genus only and some to group; e.g., "Pseudo-
monas-like," "pseudomonad," and Alcaligenes-like."
Orange colonies that constituted a majority of the popula-
tion on 12/5/77 (Table 8) and a large proportion of the popula-
tion on 1/6/78 (Table 12), were composed of Flavobacterium sp.
(red colonies) and Enterobacter agglomerans (yellow colonies).
The orange colonies were streaked repeatedly to separate the co-
habitants in order to identify them. The values given for popu-
lation size of each of these two organisms is the same as the
number of orange colonies that they originally formed. Entero-
bacter agglomerans also appeared by itself.
Results of one test for regeneration of chlorine resistant
bacteria are given in Tables 17, 18, and 19. Table 17 shows
residual total and free chlorine concentrations after a one-hour
contact time, two days aging, and six days aging of rechlori-
nated column effluent. Table 18 is a descriptive list of colo-
nies isolated from unchlorinated and chlorinated column effluent
samples. Table 19 shows the distribution of identified orga-
nisms .
322
-------
DISCUSSION
m 0*;9anisms found in the effluent from GAG columns (Tables 6,
10, and 12) are those normal to raw water sources and finished
water degraded by standing in distribution systems as described
by Geldreich (1973). The bacteria survived the treatment
plant process and colonized the
initially bacteria-free activated carbon granules. Their numbers
in raw water and in treated drinking water are often too few to
detect when relatively small volumes (100-200 ml) of water are
examined by membrane filter techniques (Tables 3, 6, 7, and 12),
but they multiply on the carbon granules in the columns and many
then slough off into the water stream. Some of the organisms
isolated may have health significance because of kind and/or
number (Geldreich 1973, Wallis 1974). Pseudomonas aeruginosa,
a possible health threat (Hoadley 1977), was present in numbers
too few to be detected in raw and finished water, but multiplied
to give counts of 25 per 100 ml of column effluent on 11/21/77
(Table 6). Whether this concentration constitutes an infectious
dose to people would depend on the circumstances of exposure.
Other Pseudomonas species that may have clinical significance
(von Graevenitz and Grehn 1977) are present in greater numbers
(120/100 ml sample on 11/21/77). Acinetobacter, Alcaligenes,
Moraxella, Flavobacterium and Enterobacter, which constituted a
great part of the population in column effluent on 11/21/77,
with more than one million colonies per 100 ml of sample, were
present in raw and finished water in small numbers when 100-200
ml samples were filtered (Table 14).
Population size in each column and in raw water is not
static. Although the total colony counts obtained from column
effluents increased within 19 days after the columns were put
into use (11/1/77 to 11/21/77), counts in those columns de-
creased in the following 14 days (Table 13). Klotz et al.(1976)
reported a rapid increase in bacterial populations followed by a
decline to a lower and stable level in carbon beds they studied.
They suggested that the decrease in bacterial numbers may have
been due to an increase in numbers of bacteria that do not con-
tribute to the plate count. This may be more a reflection of
culture technique that favor some portion of the population than
an actual shift in proportion of the population held by any one
species in the population.
The decrease in bacterial numbers (Table 13) from 11/21/77
to 12/5/77 occurred when sampling was done four days after back-
323
-------
flushing of the columns. Backflushing during this period was
routinely done twice a week. As time passed, backflushing be-
came impossible because of formation of bacterial slime in the
carbon columns. The columns were last backflushed on 12/8/77.
Increased numbers of bacteria were isolated from samples col-
lected on 1/6/78 when backflushing had not been done for 29 days
prior to sampling. It has been assumed that backflushing only
removed surficial deposits of calcium carbonate; its effect,
if any, on the resident bacterial culture is unknown.
The populations sampled on different dates were not com-
posed of the same organisms (Table 14). There was an apparent
increase in population size with increasing length of carbon bed,
but the organisms isolated from samples at different points along
the length of the bed were not always the same (Table 15}.
Klotz et al. (1976) stated that changes in the bacterial popula-
tion composition of the raw water influences the character of
the bacterial populations in the carbon beds they studied. The
bacterial composition of the raw water was not always the same
as that of the GAC columns in this study (Table 15). In most
cases the bacterial population composition in the GAC columns
did change with time and with the bacterial composition of the
raw water. Flavobacterium, Staphylococcus, Moraxella, Alcali-
genes, Citrobacter, Klebsiella, Pseudomonas sp. were found in
the effluents from the columns when they appeared in the raw
water. Pseudomonas aeruginosa was found in raw water once on
11/2/77 and was always found in column effluent. Erwinia was
found in column effluents, but not in raw water. Aeromonas was
found in raw water, but not in column effluents. Enterobacter
agglomerans was found in column effluents when it was not found
in raw water on 12/5/77, but on 1/6/78 it was found in raw water
as well as in column effluent. Enterobacter cloacae was found
in raw water on 12/5/77 and on 1/6/78; it was found in column
effluent on 12/5/77, but noton 1/6/78. Acinetobacter was found
in raw water and in column effluents on 11/21/77.It was not
found at all on 12/5/77. On 1/6/78 Acinetobacter was not found
in raw water, but was present in column effluents. Raw water
bacteria apparently do affect the composition of populations in
the columns, and these resident bacteria apparently determine
which of the incoming organisms can colonize and coexist with
them. In most cases the bacteria count in raw water was smaller
than in column effluent. Finished water had detectable bacteria
only on 11/2/77 and 12/5/77 (Table 15).
The population size of Acinetobacter, Moraxella, Pseudo-
monas, Alcaligenes, Enterobacter agglomerans, and Flavobacterium
increased with increasing bed length (from Column 1 to Column 4).
This suggests that increasing distance from the point of chlori-
nation may affect the numbers and kinds of bacteria able to
colonize the carbon bed. Table 16 gives residual chlorine values
in parts-per-million of water passed through the columns for the
period of this study. Dates on which samples were taken for
324
-------
bacterial analysis are shown inserted in Table 16. The finished
water entering the first column had from 0.2 to 3.3 ppm free
chlorine, which was taken up by the first column. Combined
chlorine, however, passed through the first column in consistent
quantities. The amount of combined chlorine entering the first
column ranged from 0.5 to 3.2 ppm. Effluent from the first col-
umn, after the initial 10 days of negligible quantities (0.05
ppm), had from 0.15 to 0.20 ppm combined chlorine. The column
retained from 0.3 to 3.0 ppm total chlorine, which may account
for the lower bacteria counts in the effluent from Column 1.
This suggests that combined chlorine may be used to control bac-
terial growth in carbon columns. Column 1 on occasion did have
greater numbers of several bacteria species than subsequent col-
umns; these were Staphylococcus on 12/5/77, Flavobacterium on
1/6/78, aud Erwinia on 1/6/78 (Table 15).
The microorganisms exhibit succession of species as do most
dynamic plant communities. The conditions supporting this suc-
cession have not yet been studied. It probably depends on over-
growth and death of a pioneer species, which supplies necessary
metabolites for the succeeding species.
Only Enterobacter agglomerans was found in chlorinated col-
umn effluent (Table 19). Column effluent chlorinated to 3 ppm
had its population of Enterobacter agglomerans greatly reduced
within one hour. These regenerated within two days (300/100 ml
sample), as residual chlorine was depleted within 24 hours
(Table 17). Column effluent with 10 ppm chlorine added had 4.5
ppm free and 6.3 ppm total chlorine remaining at the end of six
days (Table 17). Enterobacter agglomerans was found in this
highly chlorinated effluent at the end of six days; however, only
three were isolated from 100 ml of sample. The rechlorination
experiment is being repeated to determine if the isolated colo-
nies were indeed survivors or if they were merely contaminants.
Additionally, a survival experiment is being done to determine
if selection and adaptation is taking place in this species. In
this experiment chlorinated finished water is seeded with orga-
nisms to give a concentration of about 10,000/ml. The suspension
is held at room temperature (25°C) for various periods of time
and filtered to isolate and enumerate the surviving bacteria.
Enterobacter aqglomerans was not found in finished water
durini this study even though 200 ml volumes (twice the volume
suggested by Standard Methods) were passed through membrane
filters.
325
-------
REFERENCES
Allen, M. J., R. H. Taylor, and E. E. Geldreich. 1977. The im-
pact of excessive bacterial populations on coliform method-
ology. Microbiological Treatment Branch, Water Supply Re-
search Division, MERL, USEPA, Cincinnati, Ohio. In Press.
APHA, AWWA, WPCF. 1975. Standard Methods for the examination of
water and wastewater, 14th edition. Am. Publ. Hlth.Assoc,,
Washington, DC. 1193 pp.
Sergey's Manual of Determinative Bacteriology, 8th edition.
1974. Williams and Wilkins, Baltimore. 1246 pp.
••\
Eberhardt, M. 1976. Experience with the use of biologically
effective activated carbon. Pages 331-347. IN H. Sontheijner,
ed. Translation of reports on special problems of water
technology, MERL, USEPA, Cincinnati, Ohio. 453 pp.
Fiore, J. V. and R. A. Babineau. 1977. Effect of an activated
carbon filter on the microbial quality of water. Appl. and
Env. Microbiol. 34 (5):541-546.
Geldreich, E. E. 1973. Is the total count necessary? Presented
at the AWWA First Water Quality Technology Conference.
December 1973, Cincinnati, Ohio. 11 pp.
Hoadley, A. W. 1977. Pseudomonas aeruginosa in surface waters.
Pages 31-57. IN Viola Mae Young, ed. Pseudomonas aeruginosa;
Ecological aspects and patient colonization. Raven Press,
NY. 137 pp.
Klotz, M., P. Werner, and R. Schweisfurth. 1976. Investigations
concerning the microbiology of activated carbon filters.
Pages 312-330. IN H. Sontheimer, ed. Translation of reports
on special problems of water technology, MERL, USEPA,
Cincinnati, Ohio. 453 pp.
Peterson, N. and M. Favero. 1975. Significance of Gram-negative
bacteria in water supplies. Presented at the Third AWWA
Water Quality Conference. December 1975. Cincinnati, Ohio
10 pp.
van der Kooij, D. 1976. Some investigations into the presence
and behavior of bacteria in activated carbon filters. Pages
326
-------
348-354. IN H. Sontheiraer, ed. Translation of reports on
special problems of water technology,- MERL, USEPA, Cincin-
nati, Ohio. 453 pp.
von Graeveniz, A. and M. Grehn. 1977. Clinical microbiology
of unusual Pseudomonas species. Pages 50-134. IN ASM
Committee on Continuing Education. Unusual organisms of
clinical significance. ASM, Washington, DC. 184 pp.
Wallis, D., C. H. Stagg, and J. L. Melnick. 1974. The hazards
of incorporating charcoal filters into domestic water
systems. Water Res. 8:111-113.
Wood, P., L. Kaplan, J. A. Gervers, D. Waddell, andD.F. Jackson.
IN preparation. Removing potential organic carcinogens and
precursors from drinking water. Report to EPA, Grant Pro-
ject R804521-01-2, Florida International University, Miami,
Florida.
327
-------
PRESTON PLANT
to
00
WELLS "
D CLEAR
1 WELL
WATER FINISHED
WATER
T
30"
1 1
_L
c
o
L.
1
G
A
C
~D
(
L
•»
^
0
L.
2
(
i
•*
j
v
c
t
1 23
Sample Point 1 I 1 _
•••••MM
MBMBM
1
c
o
L.
3
G
A
C
1
C
*
Raw Finished 1st. Col. 2nd. 3z
Effluent
Sampled:
11/2/77 Raw and Finished Water (Columns installed and flow begun)
11/21/77 All Points
12/5/77
1/6/78
~1
c
0
L.
4
G
A
C
1
6
•d. 4th.
Figure 1 . Bench scale column adsorption unit (Experimental Design No. 4) .
Sampling points and date of sampling at each point are indicated,
-------
WATER SAMPLE (RAW, FINISHED, AND COLUMN EFFLUENTS)
MEMBRANE FILTER (1-, 10-, 100-, 200-ml volumes)
t
PRIMARY MEDIA (Table 1)
—•f —•t v » , r
TGE BHI ENDO'S DBS Ps Cz
Colony #1 Colony 1, etc. Colony 1, etc.
Colony #2
1) Differential colony counts made on each
primary culture.
2) Each colony type (numbered) picked for
isolation and identification.
Colony #23
Diagnostic Tests and Media
i
Colony Identification: Name applied to each colony type isolated
from each medium; duplicates combined.
Figure 2 . Flow diagram of procedure
used to isolate and identify
organisms from raw, finished,
and filtered water.
329
-------
TABLE 1 . PRIMARY MEDIA AND DIAGNOSTIC TESTS USED TO ISOLATE AND
IDENTIFY BACTERIA IN RAW, FINISHED, AND FILTERED WATER
Primary media*
Endo's broth (Endos) Pseudosal agar (Ps)
Desoxycholate lactose Brain heart infusion
agar (DES) broth (BHI)
Tryptone glucose yeast Czapeck agar (Cz)
extract agar (TGE)
Diagnostic tests
Roche Oxiferm system Hanging drop/motility
Gram stain Motility agar
Triple Sugar Iron agar Flagella stain
Oxidase Casein digestion
Nitrate reduction Catalase
Roche Enterotube system Starch digestion
*TGE and Cz gave highest counts; other media used with poor results were
Simmons citrate agar, potato dextrose agar, nutrient agar, PA agar, and
milk agar.
TABLE 2 BACTERIAL POPULATION COUNTS (COLONIES/100ml SAMPLE)
OBTAINED FROM RAW AND FINISHED WATER SAMPLES COLLECTED
ON 11/2/77
Name
Ac inetobac ter
Moraxella
Pseudomonas aeruginosa
Pseudomonas sp«
(other than P. aeruginosa)
Penicillium
Raw
120
4
2
22
0
Finished
12
2
3
0
85
330
-------
TABLE 3 . BACTERIAL POPULATION COUNTS (COLONIES/lOOml SAMPLE) OBTAINED FROM SAMPLES
OF RAW AND FINISHED WATER AND COLUMN EFFLUENTS COLLECTED ON 11/21/77
U)
to
Medium
TGE
Endo's coliform
non-coliform
Ps
Raw
96
0
41
27
Finished
0
0
0
0
Col. 1 Col. 2
6*
972 10
0 0
27 100
0 520
Col. 3
6*
10
0
120
25
Col. 4
6*
10
0
131
171
Cz
None was evident in 24, 48, 72 hours, but when growth developed in four to
five days, the plates were overgrown with minute yellow colonies. Finished
water had none.
DES
58
92
208
420
240
BHI
62
120
480
640
483
*estimated by counting 10cm
-------
TABLE 4 . BACTERIAL POPULATION (COLONIES/lOOml SAMPLE) OF RAW
AND FINISHED WATER* AND COLUMN EFFLUENTS COLLECTED
ON 11/21/77,
Colony type
Endo ' s
2mm, dark red
0.7mm, dark red
<0.3mm, It. red
2mm, It. red
DES
3mm, gray
2mm, tan
2mm, black
10 >10 >10
1 2
12 2
5 25
10 11
7 520 19 160
5 6
*No growth was obtained from 200ml volumes of finished water on any medium.
Blanks opposite colony description indicate no growth.
2
**Estimated by counting 10cm
332
-------
TABLE 5 . GROUPING OF COLONIES WITH SIMILAR BIOCHEMICAL
REACTIONS, ISOLATED FROM SAMPLES TAKEN 11/21/77
Group number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Identity
Acinetobacter
Aeromonas
Alcaligenes
Citrobacter
Enterobacter
Klebsiella
Moraxella-like
Pleisomonas
Proteus
Pseudomonas
aeruginosa
Pseudomonas sp-
Pseudomonas
maltophiTa
Pseudomonas
stutzeri
Pseudomonas-like
arouo 5E-1
Includes colony number
2, 16,
If 2
13
11
29, 32
10
14, 17
22, 24
23
0
4, 5,
3, 6,
28
7, 8
15
27
, 19, 20, 21,
, 24A, 26, 31
12, 18, 25, 30
9
333
-------
TABLE 6. DISTRIBUTION OF IDENTIFIED BACTERIA (COLONIES/10Oml
SAMPLE) ISOLATED FROM WATER SAMPLES TAKEN ON 11/21/77
Group name
Acinetobacter
Aeromonas
Alcaligenes
Citrobacter
Enterobacter
agglomerans
Klebsiella
Moraxella
Pseudomonas
aeruginosa
Pseudomonas spp.
(other than
Ps. aeruginosa)
Raw
19
37
54
54
25
4
67
0
2
Finished
0
0
0
0
0
0
0
0
0
Sample
Col. 1 Col. 2
2 1000
0 0
0 80
0 80
0 0
0 2
970 >106
0 0
27 100
Col. 3 Col. 4
670 630
0 0
0 2
0 2
6 0
0 0
>io6 >io6
1 25
120 110
TABLE 7 BACTERIAL
OBTAINED
EFFLUENT
POPULATION COUNTS (COLONIES/lOOml SAMPLE)
FROM RAW AND FINISHED WATER AND COLUMN
SAMPLES COLLECTED ON 12/5/77
Medium
TGE
Endo's coliforms
Endo's noncoliform
Ps
Cz
Raw
179
0
10
3
40
Finished
5
0
0
0
0
Sample
Col. 1 Col. 2
15,180 20,000
0 0
0 0
0 0
2,100 8,160
Col. 3 Col. 4
25,000 25,000
0 0
0 8
4 17
6,950 8,880
334
-------
TABLE 8
INITIAL DESCRIPTION OF COLONIES COUNTED AND
SELECTED FOR ISOLATION AND IDENTIFICATION
FROM SAMPLES TAKEN 12/5/77
Colony no.
1
2
3
4
5
6
7
7a
8
8a
9
10
11
12
a,b
a,b,c
5b
a,b,c,d,e
a,b,c
(a) (b)
a,b
(a) (b)
a,c
a,b,c
a,b,
,13,14
Source Primary
sample medium
1
1
5
5
1
1
1
1
5
6
1
6
6
TGE
TGE
TGE
TGE
Endos
Endos
Ps
Ps
Ps
Ps
Cz
TGE
TGE
Colony
description
2mm, yellow
1mm, white
<. 3mm, yellow
5mm, light yellow
3mm red, mucoid
5mm, light red,
wrinkled
3mm, cream-yellow
3mm , creamy ,
mucoid, stinks
5mm, green
3mm, green son. ,
fruity odor
3-5mm, white
3-5mm, (greenish,
yellow)
<.3mm yellow
-— "•- '•• '• • -—
Gram stains
G-
mixed sizes
G+,staph
G- mixed?
G-
G-
G-
G-
G-
G-
G-
G-
G-
mixed
mixed
mixed
mixed
mixed
mixed
mixed
mixed
sizes
sizes
sizes
sizes
sizes
sizes
sizes
sizes
mixed sizes &
&
&
&
&
&
&
&
&
shapes
shapes
shapes
shapes
shapes
shapes
shapes
shapes
shapes
Tiny Orange=
Yellow +
Red
Cz <.3mm orange,
developed 6 days
after plating
mixed sizes & shapes
*Colony numbers followed by letters were suspected of being composed of
more than one species when the Gram stain was examined. They were
streaked on TGE and SMA to effect separation and subcultures were made
of several isolated colonies.
335
-------
TABLE 9. POPULATION DISTRIBUTION (COLONIES/lOOml SAMPLE)* AND COLONIES SELECTED
FOR ISOLATION AND IDENTIFICATION FROM SAMPLES TAKEN 12/5/77
u>
U>
CT>
Colony
old no.
TGE 1
v,
2
3
4
Endo's
5
6
Ps 7
8
C£ 9
10
11
Tiny Orange ~")
(6 days later!
yellow f
red J
Includes
new no.
1
2
4
(5al
bbj
H
L6bJ
f7(SMA)l
{8 (7TSM)\
(9<7A) j
10(asbf|
il J
12
13
14
Sample
source
Raw
Raw
Col. 3
Col. 3
Raw
Col. 4
Col. 3
Col. 3
Raw
Col. 4
Col. 4
Col. 4
Colony
description
2mm, yellow
3mm, white
3mm, yellow-
green
3mm, yellow-
green
•3mm, It. red
2mm,med.red
2mm, white
3mm, green
2mm , wh ite
3-5mm, green
<0.3mm,
yellow
< 1mm, orange
Identity
Group, 2K-1
Pseu.-like
Staph . (sapro-
phyticus)
Enterobacter
agglomerans
Ps . aeruginosa
Ps .putida [
Ent. cloacae!
Serratia -'
marcescens
pilcaligenes~^
\ faecalis /
/Enterobacter 1
\ cloacae ?
Enterobacter 1
cloacae J
"PseudomonasJ
, aeruginosa \
Ps.cepacia J
Enterbacter
cloacae
Ps . aeruginosa
Enterobacter
agglomerans
Enterobacter ~"J
agglomerans 1
Flavobacteriure/
(aquatile) _J
Raw
14
144
4
6
40
Fin.
2
3
1
Col. 1
180
15,000
3
8
2,140
Col. 2
30
20,000
57
8,100
Col. 3
40
25,000
1
/>
3
1
250
6,000
Col. 4
60
25,000
3
8
17
3
380
8,500
*Blank spaces indicate none was isolated
-------
TABLE 10. BACTERIAL POPULATION DISTRIBUTION (COLONIES/lOOml)
SAMPLE OF ALL ORGANISMS ISOLATED FROM RAW AND
FINISHED WATER AND COLUMN EFFLUENTS COLLECTED ON
JLfL/ J/ / /
' " ' " • 'I „„...,.
Group name/specie* Raw Finished Col. 1 Col. 2 Col. 3 Col. 4
Alcaligenes faecalis 3
Enterobacter
agglomerans 15,000 20,000 25,000 25,000
Enterobacter
cloacae 40 133
Pseudomonas
aeruginosa 1 17
Pseudomonas
cepacia 1 17
P_. putida 4
Pseudomonas-like,
Group 2K-1 14 2
Serratia marcescens 6 8
Staphylococcus
(saprophyticus) 144 3 180 30 30 60
Flavobacterium
(aguatile) ~ 20 2,140 8,100 6,700 8,500
*Species names given are accurate within the limits of the Roche
system and additional media and tests listed in Table 1.
Blank spaces in columns opposite organism name indicates none was isolated.
337
-------
U!
CO
00
TABLE 11. INITIAL DESCRIPTION OF COLONIES COUNTED AND SELECTED FOR
ISOLATION AND IDENTIFICATION ON 1/6/78
Colony Colony Sample Primary
number description source med.(vol.)
1 3mm, light red Raw water Endos(lO)
2 3mm, rose beige " TGE (100)
3 1mm, yellow " TGE (10)
4 1mm, white " TGE (10)
5 4mm, white " Cz (100)
6 1mm, yellow Column 1 TGE (10)
7
-------
TABLE 12. POPULATION DISTRIBUTION (COLONIES/100ml SAMPLE)* OP
rmrmif ™J™ ISOLATED FROM RAW AND FINISHED WATER
COLUMN EFFLUENTS TAKEN ON 1/6/78
Organism isolated
Raw Finished Col. 1 Col. 2 Col. 3 Col. 4
Enterobacter agglomerans 780
Enterobacter cloacae 100
Ac inetobacter
Erwina (Stewartii)
Pseudomonas aeruginosa
Group, 5A-2,
Pseudomonas-like
Pseudomonas (syringae) 20
Flavobacterium
(aquatile) 90
1,730 910 610 8,000
2,800 80,000 850,000
60,000 2,040
i
2
3,000 70,800
1,730 910 610 8,000
*Blank spaces opposite organism name indicates none was isolated.
TABLE 13 BACTERIAL POPULATION (COLONIES/lOOml SAMPLE) GROWN
ON MOST FAVORABLE MEDIUM IN EACH COLUMN EFFLUENT,
RAW, AND FINISHED WATER (TOTALS OF TABLES 2, 6, 12)
Sample date
11/2/77
11/21/77
12/5/77
l/6/78b
Raw
148
96
179
990
Finished Col. 1
17 (+85
moulds) a
0 999
5 15,180
0 66,460
Col. 2 Col. 3 Col. 4
a a a
6 6 6
>10 >10 >10
20,000 25,000 25,000
75,820 83,260 866,000
aNo growth was obtained from 200ml of water taken from 500ml water shaken
for one hour with a 250ml volume of dry unused carbon granules.
bBackflushed 30 days prior to sampling. Other samples were taken four days
after backflushing.
339
-------
TABLE 14 . BACTERIAL POPULATION (COLONIES/lOOral SAMPLE) OF THE
SAMPLE POINT GIVING THE HIGHEST VALUE FOR THE LISTED
SPECIES ON THE DIFFERENT SAMPLE DATES COMPARED WITH
POPULATION OF INCOMING RAW WATER*
Date sampled
Organism 11/2/77
Acinetobacter (120) 12
Moraxella (4) 2
Ps. aeruginosa (2) 3
Pseudomonas sp. (22) 0
Penicillium (0) 85
Aeromonas
Alcaligenes
Citrobacter
Klebsiella
Enterobacter
agglomerans
Enterobacter
cloacae
Pseudomonas-like
Serratia xnarcescens
Staphylococcus
Flavobacterium
Erwinia
11/21/77 12/5/77 1/6/78
(90) 1,000 (0) 850,000
(67) >106
(0) 25 (0) 17 (0) 2
(2) 120 (4) 0 (20) 0
(37) 0
(54) 80 (0) 3
(54) 80
(4) 2
(25) 6 (0) 25,000 (780) 8,000
(40) 3 (100) 0
(14) 3 (0) 70,000
(6) 8
(144) 180
(20) 8,500 (90) 8,000
(0) 60,000
*Raw water values given in parentheses
340
-------
TABLE 15. DISTRIBUTION OF ORGANISMS ISOLATED FROM
COLUMN EFFLUENTS
Organism
Acinetobacter
Moraxella
Pseudomonas
aeruginosa
Pseudomonas
species
Aeromonas
Alcaligenes
Citrobacter
S
m
P
1
e
R
1
2
3
4
R
1
2
3
4
R
1
3
4
R
1
2
•j
j
4
p
*x
1
2
4
R
1
3
4
R
1
2
3
4
Sample date
11/2/77 11/21/77 12/5/77
120{12)* 19
2
1000
670
630
4(2)* 67
1 970
>io6
>io6
>io6
2(3)* 0
0
o
™* ^/
1
25
22(0)* 2
27
100
120
110
0 37
fl
w \J
0
0
0 54
0
80
0
-)
^ £t
0 54
0
80
^^ ^f\f
o
2
^^ &»
0
0
0
0
0
0
0
0
0
0
0
0
0
1
17
4
0
0
1
17
0
0
0
0
0
0
0
0
3
0
0
0
0
0
0
1/6/78
0
0
2800
80,000
850,000
0
0
0
0
0
0
0
2
20
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
(continued)
341
-------
TABLE 15. (continued)
Organism
Klebsiella
Enterobacter
agglomerans
Enterobacter
cloacae
Pseudomonas-like
Serratia
marcescens
Staphylococcus
Plavobacterium
S
a
m
P
1
e
R
1
2
3
4
R
1
2
3
4
R
1
2
3
4
R
1
2
3
4
R
1
2
3
4
R
1
2
3
4
R
1
2
3
4
Sample date
11/2/77
0
-
-
-
-
0
-
-
-
-
0
-
-
-
-
0
-
-
-
-
0
-
-
-
-
0
-
-
-
-
0
-
-
-
-
11/21/77
4
0
2
0
0
25
0
0
6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
12/5/77
0
0
0
0
0
0
15,000
20,000
25,000
25,000
40
1
3
3
0
14(2)*
0
0
0
0
6
0
0
0
8
144(3)*
180 t
30
40
60
20
2140
8100
6700
8500
1/6/78
0
0
0
0
0
780
1730t
910
610
8000
100
0
0
0
0
0
3000
70,800
0
0
0
0
0
0
0
0
0
0
0
0
90
1730t
910
610
8000
(continued)
342
-------
TABLE 15. (continued)
s
m
Organism
p
1
e
j^
Erwinia stewarti ,
1 J-
2
3
4
11/2/77
0
0
-
—
Sample
11/21/77
0
0
0
0
0
date
12/5/77
0
0
0
0
0
1/6/78
0
60,000t
0
2040
0
*Values in parentheses are for finished water, all other values for
finished water were less than 1. Zero values indicate less than 1,
- indicates not done, R = raw water, 1 through 4 are column numbers.
tColumn 1 effluent had higher counts than following columns.
343
-------
TABLE 16. CONCENTRATIONS OF RESIDUAL FREE AND TOTAL CHLORINE IN FINISHED
WATER AND COLUMN EFFLUENT FOR THE STUDY PERIOD
11
Finish
Date
ll/l/77a
ll/2/77b
11/3
11/4
11/73
11/8
n/ioa
11/11
11/14
11/15
ll/17a
11/18
11/25
11/29
Free
2.25
1.60
1.80
1.60
1.20
2.00
1.35
1.75
Total
3.25
2.30
2.60
2.40
2.10
2.65
1.85
2.40
19
Col. 1
Free
0
0
0
0
0
0
0
0
Total
0.05
0.05
0.05
0.10
0.10
0.10
0.15
0.15
21 23 25
Col. 2 Col. 3 Col. 4
Free
0
0
0
0
0
0
0
0
Total Free
0.05 0
0.05 0
0.05 0
0.05 0
0.05 0
0.05 0
0.05 0
0.05 0
Total Free
0.05 0
0
0.05 0
0.05 0
0.05 0
0.05 0
0.05 0
0.05 0
Total
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
17
M. Lakes
Free Total
0.10 0.55
0.05 0.50
0.05 0.40
(continued)
-------
TABLE 16. (continued)
11
Finish
Date Free Total
12/1 a
12/2 2.00 2.65
12/5b 1.90 2.50
12/8a
12/9 1.40 2.10
12/13 1.95 2.40
12/15
12/16 1.70 2.10
12/20 3.30 6.55
12/22
12/23 2.40 2.90
12/27/77 0.20 2.70
J^
I/O
19 21 23 25 17
Col. 1 Col. 2 Col. 3 Col. 4 M. Lakes
Free Total Free Total Free Total Free Total Free Total
0.05 0.45
0 0.15 0 0.05 0 0.05 0 0.05
0 0.20 0 0.05 0 0.05 0 0.05
0.10 0.50
0 0.20 0 0.05 0 0.05 0 0.05
0 0.15 0 0.05 0 0.05 0 0.05
.05 .40
0 0.15 0 0.05 0 .05 0 .05
0 0.15 0 .05 0 .05 0 .05
.05 .50
0 .20 0 .05 0 .05 0 .05
0 0.15 0 0.05 0 0.05 0 .05
sackflushed .columns
Samples taken for bacterial analysis
-------
TABLE 17 . RESIDUAL FREE AND TOTAL CHLORINE CONCENTRATIONS
IN GAG COLUMN EFFLUENT AFTER AGING
Sample
Rechlorination
concentration
Age of sample
(Contact time)
Chlorine, ppm
Free Total
Col. 4
effluent
Col. 4
effluent
Col. 4
effluent
3ppm
3ppm
lOppm
24 hours
48 hours
6 days
0.05
0.05
4.5
0.2
0.3
6.3
346
-------
TABLE 18. INITIAL COLONY DESCRIPTION AND IDENTITY OF BACTERIAL
ISOLATES FROM RECHLORINATED, AGED GAG COLUMN EFFLUENT
Colony
Medium number
Endo ' s
Endo ' s
Endo ' s
Endo ' s
Endo ' s
TGE
TGE
TGE
TGE
TGE
TGE
TGE
Cz
\f1U
Cz
Cz
El
E2
E3
E4
4
1
2
3
5
6
10
11
7
/
8
9
Sample*
4+0 (0)
4+0 (0)
4+0 (0)
4+0 (0)
4+2 (3)
4+0 (0)
4+3 (3)
4+6 (10)
duplicate
4+2 (3)
4+6 (10)
4+0 (3)
4+0 (0)
4+0 (0)
4+0 (0)
4+0 (0)
-» • ^^ \ ** /
4+0 (3)
4+2 (3)
4+6 (10)
4+0 (3)
4+0 (0)
— — — —
Colony Colonies/
description 100ml sample
3mm, white,
moist
2mm, red, domed
2mm, white/halo
<0. 3mm, red
2mm, red
-------
TABLE 19 . BACTERIAL POPULATION (COLONIES/100 ml SAMPLE) OF GAC
COLUMN EFFLUENT CHLORINATED AND AGED TO SIMULATE
DISTRIBUTION SYSTEM CONDITIONS
Sample
Organism
4+Oa
(0)
4+Ob
(3ppm)
4+2C
(3ppm)
4+6d
(lOppm)
Acinetobacter
Clwoffi) 600
Enterobacter
agglomerans
Moraxella
Pseudomonas-like
59,000 4 300
48,000 0 0
500
3
0
0
Effluent from Column
collection .
4, no chlorine added, analyzed one hour after
collection.
Effluent from Column 4, 3ppm chlorine added, aged 2 days before analysis.
Effluent from Column 4, lOppm chlorine added, aged 6 days before analysis.
eMore than one-half (55,800) of the colonies identified as Enterobacter
agglomerans were orange. Previously orange colonies yielded Flavobacterium
sp. as well, but Flavobacterium was not isolated from orange colonies that
developed from these samples.
348
-------
CHLORINATION OF GRANULATED ACTIVATED CARBON (GAG)
COLUMN EFFLUENT TO CONTROL BACTERIA
(Part II)
by
Frances Parsons
Drinking Water Quality Research Center
Florida International University
Tamiami Campus
Miami, Florida
ABSTRACT
Granulated activated carbon (GAG) columns used in water
treatment produce effluents with bacteria counts up to 10,000/ml.
A cursory examination was made of the effect of chlorination of
effluents on these bacteria. Samples of GAG column effluent
chlorinated to 3 ppm and 10 ppm were held at 25°C and plated at
two-day intervals for six days to test for survival and regrowth
of bacteria. At 3 ppm, chlorine killed most of the bacteria
(none recovered initially) and controlled regrowth to less than
500/ml for up to five days. When the 3 ppm chlorine initially
added to GAC column effluent was depleted during aging for six
days, bacterial regrowth reached 36,000/ml or greater than half
that of six-day-old unchlorinated GAC column effluent, which was
62,000/ml. When GAC column effluent had 10 ppm chlorine added,
residual free chlorine was 3 ppm throughout the aging period
(1.0 ppm in one sample on the sixth day), and no bacteria were
recovered.
Finished water to which 8,000 bacteria per ml and 10 ppm
chlorine was added, which retained 3 ppm residual free chlorine,
had three colonies per ml after six days. Finished water to
which 8300 bacteria per ml and 10 ppm chlorine was added, which
had less than 0.4 ppm residual free chlorine, had 83,000 colonies
per ml after five days aging.
349
-------
CONCLUSION
Chlorination of the effluent from granulated activated
carbon (GAC) columns apparently kills bacteria that grow on the
carbon granules and slough off into the effluent, but the initial
dose of chlorine must be adequate to combine with the bacteria
and leave sufficient free chlorine to prevent regrowth. The
concentration of chlorine necessary would vary with the bacterial
biomass and chlorine demand due to all constituents of the water.
This study was of a cursory nature and was only intended to
suggest a more complete study. Shorter sampling intervals
(daily), for a period of time longer than six days (end point
determination), with more than these two concentrations (espe-
cially less than 3 ppm free chlorine) of several disinfectants
(chlorine, chloramines, chlorine dioxide, ozone, ferrates)
should be examined. Certainly the minimum level of chlorine
needed and the time that it is effective for several bacterial
population sizes should be determined. All of these factors;
i.e., dose size, contact time, regrowth rate and size and compo-
sition of the bacterial population should be studied and compared
with parallel determinations of the bacteriology of the dis-
tribution system.
350
-------
INTRODUCTION
Granulated activated carbon (GAG) columns used to remove
organic substances from water are colonized by bacteria that
survive water treatment plant processes. Large populations of
these bacteria develop on the carbon granules and slough off into
the water flowing through them. Some of these bacterial popula-
tions have been characterized (Parsons 1978) . In an attempt to
find a way to control bacterial numbers, GAG column effluent was
chlorinated and analyzed for bacterial survivors and for regrowth
Finding Enterobacter agglomerans in highly chlorinated (10
ppm) GAG column effluent prompted this study to determine if the
few colonies isolated represented selection of a resistant strain
or survived because of protection offered by cell aggregation
or slime.
351
-------
METHODS AND MATERIALS
The following samples were taken from the water treatment
plant and from the bench scale model GAG system described earlier
(Wood et al. 1979) :
Finished water referred to as: Finished
(containing 3 mg/1 chlorine)
Column 4 effluent referred to as: Column 4 (0)
(with no added chlorine)
Column 4 effluent referred to as: Column 4 (3)
(with 3 mg/1 chlorine added)
Column 4 effluent referred to as: Column 4 (10)
(with 10 mg/1 chlorine added)
Finished water referred to as: Finished Water
(with 10 mg/1 chlorine added) + Bacteria (10)
(bacteria were added to this
sample in the laboratory)
The samples were taken to the laboratory where bacteria were
added to the sample of finished water containing 10 mg/1 chlorine
to give a concentration of approximately 10,000 cells per ml.
A culture of Enterobacter agglomerans isolated from chlorinated
Column 4 effluent was used to seed the chlorinated (10 ppm) fin-
ished water on 2/2/78. An Acinetobacter isolated from the 2/2/78
samples was used on 2/17/78 when the experiment was repeated.
Chlorine contact time for this sample was approximately ten
minutes as plating was begun shortly.after addition of the bac-
teria to the water. The time interval (contact time) for the
other samples on 2/2/78 was approximately three hours after col-
lection. The remainders of the samples, after initial plating,
were stored at room temperature (25°C) to simulate conditions in
a distribution system, and plated for bacteria at intervals dur-
ing the following six days. The samples were aged to determine
bacterial regrowth potential of chlorinated GAG column effluent.
Enumeration and isolation of bacteria from the water samples
were done by the membrane filter technique (APHA 1976). Volumes
from 0.01 to 200 ml were passed through Gelman GN6 (0.45 ym pore
size) filter membranes, which were then placed on tryptone glu-
cose extract (TGE) agar in petri dishes. Volumes less than 100
352
-------
ml were diluted with 100 ml sterile, buffered, glass-distilled
water to assure uniform distribution of the organisms on the
surface of the membrane. The cultures were incubated at 25°C
and examined daily. Cumulative colony counts were made daily for
six days of all different types of colonies observed. Different
colony types were described, assigned numbers, and picked for
identification. The API system (Analytab Products, Division of
Ayerst Laboratories, Inc., Plainview, NY), and supplementary
media and tests (Parsons 1978) were used in identification. Gram
positive cocci were planted in glucose OF medium. Staphylococci
were tested for coagulase.
Residual chlorine was determined at the time of plating
using a Taylor Basic 2000 test kit (DPD).
353
-------
RESULTS
Table 1 is a summary table of total bacteria counts obtain-
ed from samples taken on 2/2/78. Concentrations of residual free
chlorine are also given. The distribution of the different kinds
of bacteria isolated in the samples is shown in Table 2. Table
3 lists the original colony description, the sample from which
it was isolated, the estimated population size, and the identity
of the bacterial isolates from samples taken 2/2/78.
Table 4 is a summary table of bacteria counts and free
chlorine concentrations for samples taken 2/17/78. Table 5
shows" the different kinds of bacteria isolated and the sample
from which they came. Table 6 is a list of colonies, with their
description, chosen for isolation. Their final identity is also
given.
Comparison of large numbers and small numbers precludes
uniform rounding. Small numbers (less than 100) are not
rounded; large ones are.
354
-------
DISCUSSION
Three mg/1 added chlorine was sufficient to reduce the bac-
teria count in Column 4(3) effluent to the level found in treat-
ment plant Finished Water when both were plated shortly after
chlonnation. Table 1 shows that both samples had 0 colonies per
ml isolated on 2/2/78. Table 4 shows that Finished Water had one
colony per ml and Column 4(3) effluent had none on 2/17/78.
After four days aging of the samples taken on 2/2/78, Column 4(3)
had 330 colonies per ml and Finished Water had 22 colonies per ml.
After five days aging of samples taken on 2/17/78, Column 4(3)-
effluent had 250 colonies per ml and Finished Water had four
colonies per ml. These numbers have little health significance
as there is no standard for noncoliform bacteria, but do indicate
regrowth potential when chlorine is depleted by initially high
bacteria counts. The initial bacteria counts for Column 4 efflu-
ent with no added chlorine, 4(0), were 1100 colonies per ml on
2/2/78 (Table 1) and 8300 on 2/17/78 (Table 4) .
On 2/2/78 Column 4 effluent with 3 ppm chlorine added ini-
tially had less than 0.4 mg/1 free chlorine three hours later
when initial plating was done. Finished water, which had 3 ppm
chlorine added by the treatment plant process, had 0.5 mg/1
(Table 1) . At four days age, Column 4(3) had ten times more
bacteria than finished water, but the count was only 330/ml
(Table 1) . On 2/17/78 when finished water had less than 0.4 mg/1
free chlorine at the time of initial plating, it had, a higher
bacterial count (though still a small number: 42/ml) at the end
of three days age than did Column 4(3) effluent (Table 4).
Column 4 effluent without added chlorine had high counts (83,000/ml
initially that became larger (120,000/ml) with aging. When the
concentration of chlorine added to Column 4 effluent was in-
creased to 10 ppm, it effectively killed bacteria; one per ml was
recovered after a contact time of 3 hours, none after six days
aging. There was sufficient chlorine at this-level (10 ppm) to
sustain a residual of 1 ppm at the end of six days after sample
preparation on 2/2/78 (Table 1) . Only one colony was recovered
from this sample initially. Chlorine residual was 3 ppm at the
end of five days after sample preparation on 2/16/78 (Table 4).
No bacteria were recovered from this sample. No attempt was made
to determine the products formed by the action of chlorine on
bacteria.
Bacteria added to finished water (Finished Water + Bacteria
355
-------
(10) to determine if chlorine-resistant bacteria were developing
in GAC columns bathed in chlorinated finished water were effec-
tively controlled when 3 ppm residual free chlorine was present.
Three colonies/ml were recovered after six days (Table 1). When
free chlorine was initially depleted in column effluent rechlor-
inated to 10 ppm on 2/17/78, regrowth occurred and reached 83,000
colonies per ml in five days (Table 4).
Staphylococcus sp. and Acinetobacter sp. grew in Finished
Water + Bacteria (10) although only Enterobacter agglomerans was
added to the sample (Tables 2, 3), which suggests that these
organisms, survivors of the treatment plant processes, survived
the subsequent rechlorination to 10 ppm because the added Entero-
bacter agglomerans cells (approximately 10,000 cells/ml were
added to the Finished Water) combined with and depleted the avail-
able free chlorine in the sample. On 2/17/78 only Acinetobacter
sp. was added to the Finished Water + Bacteria (10) sample and
only Acinetobacter sp. was recovered (Tables 5, 6).
Bacteria in GAC column effluent could deplete chlorine,
which in turn could allow pathogenic survivors of the treatment
process to grow in the distribution system.
The results do not follow uniform patterns. Cell aggregates
examined briefly by microscope, make it exceedingly difficult to
obtain ten-fold counts from Log.Q dilutions.
There seems to be a tendency among water bacteria to form
mixed-culture colonies that appear well isolated on solid medium
and for them to have similar cellular morphologies. For example,
colonies that first appear as small, white, and entire become
yellow after two or three days. Staining only discloses Gram-
negative rods of sizes and shapes within the range of individual
variability. When diagnostic media are inoculated, the results
often indicate a mixed culture. Upon re-isolation by streaking
from a diagnostic medium and from the original colony, a mixed
culture often results. Succession of species, as often occurs
in higher plants, is suspected.
356
-------
REFERENCES
APHA, AWWA, WPCF. 1975. Standard Methods for the examination of
water and wastewater, 14th edition. Am. Publ. Hlth. Assoc.,
Washington, B.C. 1193 pp.
Wood, P., L. Kaplan, J. A. Gervers, D. Waddell, and D. F. Jack-
son. 1978. Removing potential organic carcinogens and
precursors from drinking water. Report to EPA, Grant
Project R804521-01. Florida International University,
Miami, Florida.
Parsons, F. 1978. Microbial flora of granulated activated car-
bon columns used in water treatment. IN P. Wood et al.
Report to EPA, Grant Project R804521-01. Florida Interna-
tional University, Miami, Florida.
357
-------
U)
en
oo
TABLE 1. BACTERIAL POPULATIONS (COLONIES ISOLATED/ml SAMPLE) OBTAINED IN SAMPLES
OF FINISHED WATER AND GRANULATED ACTIVATED CARBON (GAG) COLUMN EFFLUENT
TREATED IN SEVERAL WAYS. SAMPLES TAKEN 2/2/78
Finished 4 (3)
water (3)
Initial
Residual Cl2d 0.5 <0.4
Colonies/ml 0 0
2 days
Residual Cl_ ND6 ND
2
Colonies/ml 0 29
4 days
Residual Cl <0.4 <0.4
Colonies/ml 22 330
6 days
Residual Cl ND ND
Colonies/ml 330 36000
4 (10) Bact. +
fin. (10)°
3 3
1 8000
ND ND
0 130
3.0 <0.4
0 1
1.0 ND
0 3
4 (0)
0
1100
1600
19000
62000
lumber in parenthesis is ppm (mg/1) Cl added initially.
Contact time about 3 hours. Plating was done two hours after sample collection.
f"
Contact time, 10 minutes. Bacteria were added in the laboratory just prior to plating.
Initial residual chlorine values, mg/1, determined at plating time.
@
ND = not done
-------
TABLE 2. DISTRIBUTION OF BACTERIA BY TYPES IN SAMPLES TAKEN
2/2/78.* ONLY COUNTS GREATER THAN I/ml ARE GIVEN
Finished (3) Col.4 (3) Col.4 (10) Bact-+ Col.4(10)
Pin.
Day 1
Acinetobacter
Enterobacter
agglomerans
Aged 2 days
Ac inetobacter
Enterobacter
agglomerans
Pseudomonas-1ike
Staphylococcus
Aged 4 days
Acinetobacter
Enterobacter
agglomerans
Aged 6 days
Acinetobacter
Moraxella
28
22**
330
36000
8000
10
120
940
140
1300
300
260
19000
62000
330
*Blank spaces in table indicate no growth.
**^ese colonies were initially white and developed a yellow&
two days. It is suspected that the Knterobacter aqqlomerans overgrew and
replaced another organism.
359
-------
TABLE 3. COLONIES CHOSEN FROM ISOLATION MEDIUM FOR
IDENTIFICATION. SAMPLES. COLLECTED 2/2/78
Colony
number
la
2a
3a
4a
5a
6a
7a
8a
1
2
3
4
5
6
7
8
9
Description
1mm, yellow
1mm, golden
1mm, lemon
1mm, white
< 1mm, white
malodorous
5mm, white
slimey,
malodorous
1mm, yellow
< 1mm, white turns
yellow, malodorous
5mm , white , mucoid
1mm, yellow
-------
TABLE 3. (continued)
woj.ony
number
•• '••' 1 11
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Description
_
1mm, golden
1mm, white
5mm, white,
mucoid
1mm, yellow
-------
TABLE 3. (continued)
Colony Sample
number Description source Colonies/ml Identity
25 1mm,white,
mucoid,malodorous Fin.+Bact. 3 Acinetobacter
26 2mm,white,
mucoid,malodorous Fin.+Bact. 10000 Acinetobacter
27 2mm,white,
mucoid,malodorous Fin.+Bact. 62000 Acinetobacter
362
-------
U)
TABLE 4. BACTERIAL POPULATIONS (COLONIES ISOLATED/ml SAMPLE) OBTAINED IN SAMPLES
OF FINISHED WATER AND GRANULATED ACTIVATED CARBON (GAG) COLUMN EFFLUENT
TREATED IN SEVERAL WAYS. SAMPLES TAKEN 2/17/78
Finished 4 (3)
water (3)a
Initial13
Residual Cl d <0.4 <0.4
Colonies/ml 1 0
Aged 3 days
Residual Cl NDe ND
Colonies/ml 42 0
Aged 5 days
Residual Cl,, ND ND
2
Colonies/ml 4 250
4 (10)
3
0
3
0
3
0
Bact. + 4 (0)
fin. (10)
<0.4
8300 8300
ND
2 25000
ND
83000 12000
Number in parenthesis is ppm (mg/1) C12 added initially.
Contact time about 3 hours. Plating was done two hours after sample collection.
CContact time, 10 minutes. Bacteria were added in the laboratory just prior to plating.
Initial residual chlorine values, mg/1, determined at plating time.
ND
not done
-------
TABLE 5. DISTRIBUTION OF BACTERIA (COLONIES/ml) BY TYPES IN
SAMPLES TAKEN 2/7/78.* ONLY COUNTS GREATER THAN
I/ml ARE GIVEN
Bact.+
Finished (3) Col.4 (3) Col.4 (10) Fin. (10) Col.4 (0)
Day 1
Ac inetobacter
Aged 3 days
Acinetobacter
Aged 5 days
Acinetobacter
Enterobacter
agglomerans
Staphylococci
II K 2
42
2
2
250
8300
83000
8300
25000
7500
4150
*Blank spaces in table indicate no growth.
364
-------
TABLE 6. QOLONIES CHOSEN FROM ISOLATION MEDIUM FOR
IDENTIFICATION. SAMPLES COLLECTED 2/17/78
Colony
number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Description
2mm , white , mucoid
1mm, white , entire
2mm, golden , entire
1mm, lemon , entire
< 1mm, orange
-------
TABLE 6. (Continued)
Colony
number Description
Sample
source
Colonies/ml
sample
Identity
15
-------
REPORT NO.
EPA-600/2-80-130a
TECHNICAL REPORT DATA
(flease read Instructions on the reverse before completing)
2.
3. RECIPIENT'S ACCESSION-NO.
riTLE AND SUBTITLE
REMOVING POTENTIAL ORGANIC CARCINOGENS AND PRECURSORS
FROM DRINKING WATER
Volume I and Appendix A
S. REPORT DATE
August 1980 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
Paul R. Wood, Daniel F. Jackson, James A. Gervers,
Doris H. Waddell, and Louis Kaplan
8. PERFORMING ORGANIZATION REPORT NO
'ERFORMING ORGANIZATION NAME AND ADDRESS
Drinking Water Quality Research Center
Florida International University
Miami, Florida 33199
10. PROGRAM ELEMENT NO.
61C1C, SOS #1, TASK 42
11. CONTRACT/GRANT NO.
R-804521
. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory—Gin.,OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final 6/22/76-6/30/80
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
See also Volume II, EPA-600/2-80-130b
Project Officer: Jack DeMarco (513) 684-7282
16. ABSTRACT
Feasible and economical methodologies were needed to remove existing organic contaminants—specifically,
four trihalomethanes (chloroform, bromodichloromethane, chlorodibromomethane, and bromoform)—from and
prevent development of potential carcinogens in the public water supplies in Dade County, Florida. A
four-phase study was designed to evaluate the efficiency of three adsorbents in removing 19 individual
halogenated organics and trihalomethane precursors. These adsorbents were XE-340—a carbonized polymeric
macroreticular resin; IRS-904—a strong base catonic resin designed to remove large molecular weight sub-
stances such as precursors from water; and granular activated carbon (GAG). Adsorbent columns were
placed at various stages in the water processing system: the raw-water, the lime-softened at the up-flow
Hydrotreator effluent, and the finished water stage. Four 0.76-meter-deep GAC Filtrasorb 400 columns,
arranged in series on the finished water line, were most effective in reducing the level of the trihalo-
methanes present in the finished water. The Polanyi-Manes theory of adsorption was applied and found
helpful in interpreting results.
Appendix A contains the preliminary studies made of the bacterial profile of raw and finished water and
effluent from four GAC columns from the Preston Water Treatment Plant. Raw water organisms, which appar-
ently, can survive existing treatment plant processes, colonized the initially bacteria-free GAC columns
and released vast numbers of bacteria into the water flowing through the columns. The development of
bacterial growth in the GAC columns interfered with backflushing the columns.
Appendix B (Volume II of this report) contains the supporting data for the study.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERM
c. COS AT I Field/Group
Adsorbents, Granular Activated Carbon
Treatment, Synthetic Resin Treatment,
Potable Water, Organics Control
Adsorption, Specific
Organic compounds,
General Organic Para-
13B
meters
13. DISTRIBUTION STATEMENT
Release to Public
•MWMM^HMMMMMMWMMMM
EPA Form 2220-1 (9-73)
19. SECl
389
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
367
«, U S. GOWWMEm PRINTING OFFICE: 1990-657-165/0114
------- |