P383-121731
EPA-600/2-82-087a
October 1982
FEASIBILITY STUDY OF GRANULAR ACTIVATED
CARBON ADSORPTION AND ON-SITE REGENERATION
Volume 1. Detailed Report.
by
Richard Miller
David J. Hartman
Cincinnati Water Works
Cincinnati, Ohio 45232
Cooperative Agreement No. CR805443
Project Officers
Jack DeMarco and Ben W. Lykins, Jr.
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'
REPRODUCED BY
NATIONAL TECHNICAL
INFORMATION SERVICE
U.S. DtPARIWHT OF COMMERCt
S>8WGf (£10. YA 22161
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-82-087a
2.
3. RECIPIENT'S ACCESSION«NO.
j "•* - •"* *
.'••'_ 3 i£ J. /
4. TITLE AND SUBTITLE
Feasibility Study of Granular Activated Carbon and
On-Site Regeneration
Volume 1. Detailed Report
&. REPORT DATE
October 1982
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Richard Miller, David J. Hartman
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
Cincinnati Water Works
4747 Spring Grove
Cincinnati, Ohio 45232
11. CONTRACT/GRANT NO.
CR805443
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory- Cin., OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
Project Officer: Ben W. Lykins, Jr. (513-684-7460)
16. ABSTRACT
Most research pertaining to water quality and treatment methods conducted in the
United States in the last decade utilized pilot-scale components. This project em-
ployed full-sized filters, post-filtration contactors and carbon regeneration furnace
at one site to study carbon's ability to remove organics from Ohio River Water.
Various GAC bed depths and types were studied in order to compare organic removal
efficiencies, bed lives, general water quality characteristics, the need of a sand
underlayer and operational problems. Pilot-scale GAC components were also used to
determine the reliability of pilot columns as indicators of the performance of full-
scale components. The relative performance of lignite and bituminous-based GAC was
also studied in pilot columns.
In the most important phase of this project, the relative performance of GAC filters
to post-filtration GAC contactors was studied along with the most advantageous empty
bed contact time for the GAC contactors and the effectiveness of on-site GAC regener-
ation. Finally, a significant aspect of this project was the development of prelim-
inary cost estimates for full-plant conversion to GAC.
Volume I is the detailed report. Volume II includes data <
Volume I and microfilm images of all raw data developed un<
iraphs not included
ler the project.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COS AT I Field/Group
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS {This Report}
Unclassified
21. NO. OF PAGES
305
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
-------
NOTICE
THIS DOCUMENT HAS BEEN REPRODUCED
FROM THE BEST COPY FURNISHED US BY
THE SPONSORING AGENCY. ALTHOUGH IT
IS RECOGNIZED THAT CERTAIN PORTIONS
ARE ILLEGIBLE, IT IS BEING RELEASED
IN THE INTEREST OF MAKING AVAILABLE
AS MUCH INFORMATION AS POSSIBLE.
-------
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publication. Mention of trade names
or commercial products does not constitute endorse-
ment or recommendation for use.
11
-------
FOREWORD
The U.S. Environmental Protection Agency was created because of increas-
ing 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 Environmental Research Laboratory
develops new and improved technology 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 communication link between the researcher
and the user community.
This report presents the results of a field-scale research effort to
evaluate granular activated carbon (GAC) adsorption and on-site fluidized
bed reactivation. GAC units consisting of converted sand filters and
contactors were studied to determine their performance relative to virgin
and reactivated granular carbon. Extensive organic analyses were performed
along with general water treatment plant parameters. Cost of GAC treatment
and reactivation were collected and preliminary cost estimates of full-scale
plant conversion to GAC were developed.
Francis T. Mayo
Director
Municipal Environmental Research Laboratory
ill
-------
ABSTRACT
This project determines whether the use of granular activated carbon (GAC)
is feasible for removing certain trace organics from Ohio River water while
treating it for human consumption. The study used either deep-bed contact-
ors or conventional-depth gravity filters and on-site GAC regeneration. To
be considered feasible, the facility had to be able to remove selected
organics to a predesignated level at a cost acceptable to consumers without
adversely affecting the level of treatment provided by the existing plant.
A secondary objective was to develop plant design and operating parameters
for full-scale plant conversion to GAC treatment. The study was unusual in
that it employed full-sized filters, contactors, and carbon regeneration
furnace instead of the pilot-scale components used by most water quality
researchers.
In the first phase of the project, three existing rapid sand filters
were converted to GAC filter adsorbers. Various GAC bed depths and types
were studied to compare organic removal efficiencies, bed lives, general
water quality characteristics, the need of a sand underlayer and operational
problems.
The second phase involved the use of pilot-scale GAC components to investi-
gate the effects of regeneration on the carbon's adsorptive capability and
to determine the reliability of polot column as performance indicators for
full-scale components. The relative performances of lignite and bituminous-
based GAC were also studied.
The last phase of this project studied the relative performance of GAC
filters to post-filtration GAC contactors, the most advantageous empty bed
contact time for the contactors and the effectiveness of on-site GAC regener-
ation. Pilot columns were also operated in parallel with the full-sized
units to assess the usefulness of pilot columns as predictors of full-scale
operation. During this phase, an attempt was made to maximize the use of
currently available organic analysis techniques. Additional organic analyti-
cal techniques such as acid extract GC/FID profiles, Grob closed loop strip-
ping analyses and carbon-adsorbable organohalides provided a broad data
base. Finally, a significant aspect of this project was the development of
preliminary cost estimates for full-plant conversion to GAC.
This report was submitted in fulfillment of Cooperative Agreement
No. CR805443 by the Cincinnati Water Works under the sponsorship of the U.S.
Environmental Protection Agency. This report covers the period from
August, 1977 to May, 1982, and work was completed as of April, 1982.
IV
-------
CONTENTS
Disclaimer ii
Foreword iii
Abstract iv
Figures vi
Tables xiv
Abbreviations and Symbols xviii
Acknowledgments xxi
1. Introduction 1
2. Conclusions 4
3. Recommendations 6
4. Desc'ription of the Treatment Complex 7
Filters 9
Pilot columns 15
Contactors 21
Regenerator 30
Contactor/regenerator building 35
Laboratory 35
Data management 37
5. Methods and Operating Procedures 40
Physical plant 40
Laboratory 51
6. Results and Discussion 80
Phase 1. Full-scale GAC Filters 80
Objective 1: Relative performance of GAC filters. . . 80
Objective 2: Sand vs. GAC for filter media 90
Phase 2. Pilot GAC Filters and Contactors. .• 94
Objective 3: Virgin vs. regenerated GAC 94
Objective 4: Bituminous vs. lignite-based GAC .... 94
Objective 5: Prediction of full-scale performance . . 107
Phase 3. Full-scale GAC Filters and Contactors
With On-site Regeneration 121
Objective 6: Full-scale GAC filters vs. contactors. . 121
Objective 7: Effect of regeneration 158
Objective 8: Pilot vs. full-scale systems 173
Objective 9: GAC exhaustion criteria 193
Objective 10: Regenerated GAC characteristics 215
Objective 11: GAC regeneration and transport losses. . 215
Objective 12: Design parameters 220
Objective 13: Costs 239
Other Observations 254
References 265
Appendix
A. Computerized Data Dictionary 267
-------
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
FIGURES
Treatment process
Filter and pipe gallery cross-section
Filter gallery layout
Filter carbon transport piping
Pilot GAC filters, processing flow
Pilot GAC filter, backwashing flow
Lignite pilot contactor, processing flow
Lignite pilot contactor, backwashing flow
Bituminous pilot contactor, processing flow
Bituminous pilot contactor, backwashing flow
Contactor Building layout
Contactor cross-section
Regenerator process
Treatment process sample point locations
Full-scale GAC system sample point locations
Maximum THM (MTTT) breakthrough curves for GAC Filters 21A
level 2 and 19A effluent, Phase 1-0
Total organic carbon (TOC) breakthrough curves for GAC
Filter 19A and 21A effluents, Phase 1-0
Maximum THM (MTTT) breakthrough curves for GAC Filter 19A
and 21A effluents, Phase 1-0
Instantaneous THM (ITTT) concentration curves for GAC
Filter 19A and 21A effluents, Phase 1-0
Page
8
11
12
13
16
17
19
20
22
23
24
25
31
52
53
81
83
84
85
VI
-------
20 Total organic carbon (TOC) breakthrough curves for GAC
Filter 21A and 23A effluents, Phase 1-0 88
21 Maximum THM (MTTT) breakthrough curves for GAC Filter 21A
and 23A effluents, Phase 1-0 89
22 Instantaneous THM (ITTT) concentration curves for GAC
Filter 21A and 23A effluents, Phase 1-0 91
23 Total organic carbon (TOC) percent removal curves for virgin
and once regenerated pilot GAC filter effluents, Phase 2-1 . 95
24 Instantaneous THM (ITTT) percent removal curves for virgin and
once regenerated pilot GAC filter effluents. Phase 2-1 ... 96
25 Seven-day simulated distribution syst. THM (STT7) percent
removal curves for virgin and once regenerated pilot GAC
filter effluents, Phase 2-1 97
26 THM formation potential (FTTT) percent removal curves for
virgin and once regenerated pilot GAC filter effluents,
Phase 2-1 98
27 Total organic carbon (TOC) percent removal curves for virgin
and twice regenerated pilot GAC filter effluents, Phase 2-2. 99
28 Instantaneous THM (ITTT) percent removal curves for virgin and
twice regenerated pilot GAC filter effluents, Phase 2-2. . . 100
29 Seven-day simulated distribution syst. THM (STT7) percent
removal curves for virgin and twice regenerated pilot GAC
filter effluents, Phase 2-2 101
30 THM formation potential (FTTT) percent removal curves for
virgin and twice regenerated pilot GAC filter effluents,
Phase 2-2 102
31 Total organic carbon (TOC) breakthrough curves for bituminous
and lignite pilot contactor effluents, Phase 2-0 105
32 Seven-day simulated distribution system THM (STT7)
breakthrough curves for bituminous and lignite pilot
contactor effluents, Phase 2-0 106
33 Total organic carbon (TOC) breakthrough curve for virgin pilot
GAC filter, Phase 2-1 108
34 Instantaneous THM (ITTT) breakthrough curve for virgin pilot
GAC filter, Phase 2-1 109
vn
-------
35 Seven-day simulated distribution system THM (STT7)
breakthrough curve for virgin pilot GAC filter, Phase 2-1. . 110
36 THM formation potential (FTTT) breakthrough curve for
virgin pilot GAC filter, Phase 2-1 Ill
37 Total organic carbon (TOC) breakthrough curve for virgin
pilot GAC filter, Phase 2-2 112
38 Instantaneous THM (ITTT) breakthrough curve for virgin
pilot GAC filter, Phase 2-2 113
39 Seven-day simulated distribution system THM (STT7)
breakthrough curve for virgin pilot GAC filter, Phase 2-2. . 114
40 THM formation potential (FTTT) breakthrough curve for virgin
pilot GAC filter, Phase 2-2 115
41 Total organic carbon (TOC) breakthrough curve for
bituminous pilot contactor, Phase 2-0 117
42 Instantaneous THM (ITTT) breakthrough curve for
bituminous pilot contactor, Phase 2-0 118
43 Seven-day simulated distribution system THM (STT7) breakthrough
curve for bituminous pilot contactor, Phase 2-0 119
44 THM formation potential (FTTT) breakthrough curve for
bituminous pilot contactor, Phase 2-0 120
45 Total organic carbon (TOC) breakthrough curves for GAC
Filter ISA effluent and Contactor D level 7, Phase 3-0 . . . 122
46 Three-day simulated distribution system THM (STT3) breakthrough
curves for GAC Filter ISA effluent and Contactor D level 7,
Phase 3-0 123
47 Instantaneous THM (ITTT) breakthrough curves for GAC Filter 15A
effluent and Contactor D level 7, Phase 3-0 124
48 THM formation potential (FTTT) breakthrough curves for
GAC Filter 15A effluent and Contactor D level 7, Phase 3-0 . 125
49 Instantaneous THM (ITTT) breakthrough curves for GAC Filter ISA
effluent and Contactor C level 7, Phase 3-1 126
50 Instantaneous THM (ITTT) percent removal curves for GAC
Filter 15A effluent and Contactor C level 7, Phase 3-1 . . . 128
51 THM formation potential (FTTT) percent removal curves for GAC
Filter ISA effluent and Contactor D level 7, Phase 3-0 ... 129
Vlll
-------
52 Carbon adsorbable organohalides (CAOX) breakthrough curves for
GAC Filter ISA effluent and Contactor D level 7, Phase 3-0 . 130
53 Acid extract profiles for filter influent, GAC Filter ISA
effluent and Contactor D level 7, Phase 3-0, runday eight. . 136
54 Acid extract profiles for filter influent, GAC Filter ISA
effluent and Contactor D level 7, Phase 3-0, runday 85 ... 137
55 Acid extract profiles for filter influent, GAC Filter ISA
effluent and Contactor D level 7, Phase 3-0, runday 113. . . 138
56 Acid extract profiles for filter influent, GAC Filter ISA
effluent and Contactor D level 7, Phase 3-0, runday 141. . . 139
57 Instantaneous chloroform (ICLR) breakthrough curves for
Contactor D, Phase 3-0 145
58 Instantaneous bromodichloromethane (ICL2) breakthrough curves
for Contactor D, Phase 3-0 146
59 Total organic carbon (TOC) breakthrough curves for Contactor D,
Phase 3-0 147
60 THM formation potential (FTTT) breakthrough curves for
Contactor D, Phase 3-0 148
61 Three-day simulated distribution system THM (STT3)
breakthrough curves for Contactor D, Phase 3-0 149
62 Carbon adsorbable organohalides (CAOX) breakthrough curves
for Contactor D, Phase 3-0 150
63 Instantaneous chloroform (ICLR) carbon use rates for
multiple runs of contactors, Phase 3 151
64 Instantaneous bromodichloromethane (ICL2) carbon use rates
for multiple runs of contactors, Phase 3 153
65 Total organic carbon (TOC) carbon use rates for multiple
runs of contactors, Phase 3 154
66 Acid extract profiles for Contactor D, Phase 3-0, runday 29 . . 155
67 Acid extract profiles for Contactor D, Phase 3-0, runday 113. . 156
68 Total organic carbon (TOC) percent removal curves for
multiple runs of Contactor D effluent, Phase 3 162
IX
-------
69 Three-day simulated distribution system THM (STT3) percent
removal-curves for multiple runs of Contactor D effluent,
Phase 3 163
70 Total organic carbon (TOC) percent removal curves for
multiple runs of GAC Filter ISA effluent, Phase 3 164
71 Instantaneous chloroform (ICLR) percent removal curves for
multiple runs of GAC Filter ISA effluent, Phase 3 165
72 Instantaneous chloroform (ICLR) adsorbed per GAC weight for
multiple runs of Contactor D effluent, Phase 3 167
73 Instantaneous bromodichloromethane (ICL2) adsorbed per GAC
weight for multiple runs of Contactor D effluent, Phase 3. . 168
74 Total organic carbon (TOC) adsorbed per GAC weight for
multiple runs of Contactor D effluent, Phase 3 169
75 Instantaneous chloroform (ICLR) adsorbed per GAC weight for
multiple runs of GAC Filter 15A effluent, Phase 3 170
76 Total organic carbon (TOC) adsorbed per GAC weight for
multiple runs of GAC Filter ISA effluent, Phase 3 171
77 THM formation potential (FTTT) adsorbed per GAC weight for
multiple runs of GAC Filter ISA effluent, Phase 3 172
78 Three-day simulated distribution system THM (STT3) percent
removal curves for pilot GAC filter and GAC Filter ISA
effluents, Phase 3-0 174
79 Total organic carbon (TOC) percent removal curves for pilot
GAC filter and GAC Filter ISA effluents, Phase-3-0 175
80 Three-day simulated distribution system THM (STT3) breakthrough
curves for pilot GAC filter and GAC Filter ISA effluents.
Phase 3-0 177
81 Total organic carbon (TOC) breakthrough curves for pilot
GAC filter and GAC Filter ISA effluents, Phase 3-0 178
82 Three-day simulated distribution system THM (STT3) percent
removal curves for pilot GAC filter and GAC Filter ISA
effluents, Phase 3-1 179
83 Total organic carbon (TOC) percent removal curves for pilot
GAC filter and GAC Filter ISA effluents, Phase 3-1 180
-------
84 Three-day simulated distribution system THM (STT3) breakthrough
curves for pilot GAC filter and GAC Filter ISA effluents,
Phase 3-1 181
85 Total organic carbon (TOC) breakthrough curves for pilot GAC
filter and GAC Filter ISA effluents, Phase 3-1 182
86 Three-day simulated distribution system THM (STT3) percent
removal curves for pilot GAC filter and GAC Filter ISA
effluents, Phase 3-2 183
87 Total organic carbon (TOC) percent removal curves for pilot
GAC filter and GAC Filter ISA effluents, Phase 3-2 184
88 Three-day simulated distribution system THM (STT3) breakthrough
curves for pilot GAC filter and GAC Filter ISA effluents,
Phase 3-2 185
89 Total organic carbon TOC breakthrough curves for pilot GAC
filter and GAC Filter ISA effluents, Phase 3-2 186
90 Three-day simulated distribution system THM (STT3) percent
removal curves for pilot GAC filter and GAC Filter ISA
effluents, Phase 3-3 187
91 Total organic carbon (TOC) percent removal curves for pilot
GAC filter and GAC Filter ISA effluents, Phase 3-3 188
92 Three-day simulated distribution system THM (STT3) breakthrough
curves for pilot GAC filter and GAC Filter ISA effluents,
Phase 3-3 189
93 Total organic carbon (TOC) breakthrough curves for pilot
GAC filter and GAC Filter ISA effluents, Phase-3-3 190
94 Acid extract profiles for GAC Filter ISA and pilot GAC filter,
Phase 3-0, runday 57 191
95 Carbon adsorbable organohalides (CAOX) breakthrough curves for
pilot GAC filter and GAC Filter 15A effluents, Phase 3-0 . . 192
96 Three-day simulated distribution syst. THM (STT3) percent
removal curves for bituminous pilot contactor and Contactor D
effluents, Phase 3-0 194
97 Total organic carbon (TOC) percent removal curves for
bituminous pilot contactor and Contactor D effluents,
Phase 3-0 195
XI
-------
98 Three-day simulated distribution system THM (STT3) breakthough
curves for bituminous pilot contactor and Contactor D
effluents, Phase 3-0 196
99 Total organic carbon (TOC) breakthrough curves for bituminous
pilot contactor and Contactor D effluents, Phase 3-0 .... 197
100 Three-day simulated distribution syst. THM (STT3) percent
removal curves for bituminous pilot contactor and
Contactor D effluents, Phase 3-1 198
101 Total organic carbon (TOC) percent removal curves for
bituminous pilot contactor and Contactor D effluents,
Phase 3-1 199
102 Three-day simulated distribution system THM (STT3) breakthrough
curves for bituminous pilot contactor and Contactor D
effluents, Phase 3-1 200
103 Total organic carbon (TOC) breakthrough curves for bituminous
pilot contactor and Contactor D effluents, Phase 3-1 .... 201
104 Three-day simulated distribution syst. THM (STT3) percent
removal curves for bituminous pilot contactor & Contactor D
effluents, Phase 3-2 202
105 Total organic carbon (TOC) percent removal curves for
bituminous pilot contactor and Contactor D effluents,
Phase 3-2 203
106 Three-day simulated distribution system THM (STT3) breakthrough
curves for bituminous pilot contactor and Contactor D
effluents, Phase 3-2 204
107 Total organic carbon (TOC) breakthrough curves for bituminous
pilot contactor and Contactor D effluents, Phase 3-2 .... 205
108 Acid extract profiles for Contactor D and pilot contactor,
Phase 3-0, runday 113 206
109 Carbon adsorbable organohalides (CAOX) breakthrough curves for
pilot contactor and Contactor D effluents, Phase 3-0 .... 207
110 Total organic carbon (TOC) in raw, filter influent and sand
filter effluent locations for the year 1980 221
111 Three-day simulated distribution system THM (STT3) in raw,
filter influent and sand filter effluent locations for the
year 1980 222
Xll
-------
112 Total organic carbon (TOC) and three-day simulated
distribution system THM (STT3) seasonal service time curves
for GAC filters 224
113 GAC filter carbon use rates, 1980 227
114 GAC filter service lives, 1980 228
115 Total organic carbon (TOC) carbon use rates for multiple runs
of contactors, Phase 3 (TOC=1000 ug/1) 230
116 Total organic carbon (TOC) seasonal service time curves for
contactors, Phase 3 232
117 Three-day simulated distribution system THM (STT3) seasonal
service time curves for contactors, Phase 3 233
118 Contactor carbon use rate, 1980 234
119 Contactor service life, 1980 235
120 Acid extract profiles for quench and scrubber samples,
May 13, 1981 261
121 Acid extract profiles for quench and scrubber samples,
May 15, 1981 262
122 THM comparison of actual distribution system (ITTT) vs
three-day simulated distribution system (STT3), 1980 .... 264
Xlll
-------
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
TABLES
Contactor design specifications
Washwater pump specifications
Process pump specifications
Instrument air system specifications. .
Pre-grant costs of establishing organics laboratory
Laboratory equipment inventory
Graphics equipment
Actual operating conditions for full-scale GAC systems. .
Actual operating conditions for GAC pilot columns ....
Nominal operational parameters for GAC filters. .
Nominal operational parameters for pilot GAC filters. . .
Nominal operational parameters for pilot contactors . . .
Nominal operational parameters for contactors
Nominal GAC regenerator parameters
Phase, run, runday schedule
Phase 1 sample plan overview
Phase 2 sample plan overview
Phase 3 sample plan overview
GC/MS lower detection limits
Analytical methods reference list
Average percent relative deviation of THM analyses. . . .
Page
27
28
. . 28
29
... 36
38
. . 39
... 41
42
42
42
... 43
44
46
54
55
57
59
64
67
... 68
XIV
-------
22
23
24
25
26
27
?a
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Average percent relative deviation for THM control samples. . .
USEPA Performance Evaluation Standards WP005, March, 1979 . . .
USEPA Performance Evaluation Standards WS006, February, 1980. .
USEPA Performance Evaluation Standards WS008, May, 1981 ....
USEPA Performance Evaluation Standards March 1979
USEPA Performance Evaluation Standards April 1979
USEPA Performance Evaluation Standards November 1979 . .
USEPA Performance Evaluation Standards June, 1980
USEPA Performance Evaluation Standards February 1981
Average percent relative deviation of purgeable
non-halogenateds . . ..
TOC results from USEPA Performance Evaluation Standards, WS005.
Purgeable halogenated organics for GAC Filters 19A and 21A,
Phase 1-0 . .
Purgeable halogenated organics for GAC Filters 21A and 23A,
Phase 1-0
Comparison of filter run between GAC filters and plant average.
Comparison of filter run times during typical period when
sand filter runs were less than 60 hours . . .
Bituminous and lignite GAC characteristics
Bituminous and lignite GAC prices
Comparison of bituminous and lignite pilot contactors for
selected exhaustion criteria
Purgeable halogenated organics for Contactor D level 7 and
GAC Filter 15A effluent, Phase 3-0
Purgeable halogenated organics for Contactor C level 7 and
GAC Filter 15A effluent, Phase 3-1
Purgeable non-halogenated organics for Contactor D level 7 and
GAC Filter 15A effluent. Phase 3-0
69
70
71
71
73
74
75
76
77
78
79
87
92
93
93
104
104
104
131
132
133
XV
-------
43 Purgeable non-halogenated organics for Contactor C level 7 and
GAC Filter ISA effluent, Phase 3-1 134
44 Acid extract compounds tentatively identified by GC/MS from
GAC Filter ISA 141
45 Exhaustion criteria for Contactor D level 7 and GAC Filter ISA
effluent, Phase 3-0 143
46 Exhaustion criteria for Contactor C level 7 and GAC Filter ISA
effluent, Phase 3-1 144
47 Acid extract compounds tentatively identified by GC/MS
from Contactor D 157
48 Established MCLs for organic compounds 158
49 Percent removal data summary of Grob CLSA results
Contactor A, Phase 3-0, rundays 1 through 134 159
50 Percent removal data summary of GROB CLSA results
Contactor A, Phase 3-0, rundays 162 through 302 160
51 Average influent concentration for Contactor D and
GAC Filter ISA 166
52 Comparison of pilot and full-scale GAC filter systems 173
53 Comparison of pilot and full-scale contactor system 193
54 Application of various exhaustion criteria to
GAC Filter ISA and Contactor D, Phase 3-0 209
55 Application of various exhaustion criteria to
GAC Filter 15A and Contactor D, Phase 3-1 210
56 Application of various exhaustion criteria to
GAC Filter ISA and Contactor D, Phase 3-2 211
57 GAC analysis data 216
58 Comparison of CWW and GIT analyses of GAC characteristics . . . 217
59 Total GAC losses 218
60 GAC losses across furnace 218
61 Contactor GAC transport losses 219
62 Sand separator, percent sand in GAC 219
xvi
-------
63 GAG filter service lives for selected TOG treatment goals,
225
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
GAG filter service lives for selected THMSIMDIST treatment
goals , in days
Contactor service life, in days
Useful life estimates for capital cost data
Cost factors
Detailed capital costs
Capital cost summary
Preliminary annual operating and maintenance costs
Comparison of annual amortized capital cost
Total annual costs recap
Unit costs recap
Operating and maintenance cost factors
Conceptual design summary for GAC filters
Cost estimates for full-scale conversion of plant to
GAC filters
Conceptual design summary for contactors
Cost estimates for full-scale conversion of plant to contactors
Cost of alternative treatment goals
Standard plate counts per ml for Phase 3-0
Regenerator off-gas analyses results
Physical, chemical and organic data from May 13, 1981 quench
and scrubber samples
Physical, chemical and organic data from May 15, 1981 quench
and scrubber samples
THMMAX data from May 13 and 15, 1981 quench and scrubber
samples
225
231
240
241
242
243
244
245
245
246
247
249
250
252
253
254
255
257
258
259
260
XVll
-------
ABBREVIATIONS AND SYMBOLS
BET Brunauer-Emmett-Teller total
surface area
BIT bituminous
BVL bed volume losses
BTU British thermal unit
CAOX carbon adsorbable organohalide
CBE Chemical Building East
CBW Chemical Building West
ccf hundred cubic foot
CHBr bromoform
CHBrCl_ bromodichloromethane
CHBr^CI dibromochloromethane
CHCl chloroform
CLSA closed loop stripping analysis
CWW Cincinnati Water Works
cu ft cubic foot
cu m cubic meter
cu m/d cubic meter per day
cu m/s cubic meter per second
cu yd cubic yard
DFTPP decafluorotriphenylphosphine
EBCT empty bed contact time
EBV empty bed volume
EF exhausted freeboard
ELECT Electrical
EMSL Environmental Monitoring and
Support Laboratory
EWODS Early Warning Organic Detection System
EXH exhaustion
ft foot
GAC granular activated carbon
gal gallon
GC gas chromatograph
GC/FID gas chromatograph/flame ionization
detector
GC/MS gas chromatograph/mass spectrometer
GIT Georgia Institute of Technology
gpd gallon per day
gpd/sq ft gallon per day per square foot
gpm gallon per minute
XVlll
-------
hr hour
ICI ICI Americas, Inc.
in inch
INSTTHM trihalomethanes present at time of sample collection
k cu m 1000 cubic meter
kg kilogram
km kilometer
kPa kilo Pascals
kWh kilowatt-hour
Ib pound
LIG lignite carbon
1 liter
Ipd liter per day
Ipd/cu m liter per day per cubic meter
Ipm/sqm liter per minute per square meter
Ips liter per second
m meter
MCL maximum contaminant level
MERL Municipal Environmental Research
Laboratory
mg milligram
mgd million gallon per day
mg/1 milligram per liter
mgy million gallon per year
mi mile
mil gal million gallon
min minute
mph mile per hour
mt metric ton
NI new interface
ng nanogram
NPSH net positive suction head
NTU nephelometric turbidity unit
OEPA Ohio Environmental Protection Agency
01 original interface
ORP Ohio River Plant
ORSANCO Ohio River Valley Water Sanitation Commission
PAC powdered activated carbon
PCS polychlorinatedbiphenyl
PFA post filter adsorber, pilot columns
pH index of acidity or alkalinity
psi pound per square inch
psig pounds per square inch gauge
PUMP STA pumping station
RBV regenerated bed volume
RCC Regional Computer Center
RESV reservoir
RF regenerated freeboard
xix
-------
RV regenerated/virgin, pilot columns
SCFM standard cubic foot per minute
sq ft square foot
sq km square kilometer
sq m square meter
sq mi square mile
ss stainless steel
ST1 Storage Tank 1
ST2 Storage Tank 2
SWORCC Southwest Ohio Regional Computer
Center
tdh total dynamic head
THM trihalomethane
THMFP trihalomethane formation potential
THMSIMDIST simulated distribution system trihalomethane
TOC total organic carbon
ton short ton
USEPA United States Environmental
Protection Agency
TBE tetrabromoethane
TFE polytetrachloroethylene
VMR volume of material removed
yr year
-------
ACKNOWLEDGEMENTS
This research project was accomplished through the efforts of many
individuals working toward a common goal. Many employees of the Cincinnati
Water Works (CWW), although not specifically trained for scientific research,
played key roles and contributed greatly to the success of this project.
This was a total team effort, with all CWW Divisions participating: the
Administration Division for financial control and secretarial support in
preparation of the various reports,- the Commercial Division for typing and
preparation of data for entry into the computer; the Supply Division for
construction, maintenance and operation of associated facilities; the Distri-
bution Division for occasional pipe work; the Water Quality and Research
Division for conducting the analytical work; and the Engineering Division for
design, GAC measurement and graphics activities.
The support and assistance of several people at the U.S. Environmental
Protection Agency, Municipal Environmental Research Laboratory (MERL) was
greatly appreciated. In particular the advice and guidance of Jack DeMarco,
Project Officer, was instrumental in keeping the project on track and goal
oriented. Alan A. Stevens, Ronald C. Dressman and Dennis R. Seeger performed
some of the acid extracts and organohalide analysis that gave added dimension
to the total findings. Frederich C. Kopfler, Robert G. Melton and
W. Emile Coleman performed the Grob closed loop stripping analyses, which
again broadened the total findings. Analytical training and assistance was
provided by the Environmental Monitoring and Support Laboratories (EMSL).
Also much appreciated was the assistance of the Office of Public Water
Supply Division of the Ohio Environmental Protection Agency (OEPA) in conduct-
ing pesticide , herbicide and PCB analyses.
Additional analytical work was performed under contract by the University
of Cincinnati and PEDCo Environmental.
The loan of pilot colums by Westvaco and I.C.I. Americas Inc., which
expedited the pilot scale work, was sincerely appreciated.
Design and construction supervision for the contactors and regeneration
furnace was performed under contract by Black & Veatch, Consulting Engineers.
Finally, we wish to gratefully acknowledge the significant contributions
of the late Aaron A. Rosen, Principal Research Investigator for the project
until his untimely death in December, 1979. A noted expert on water pollution
and a pioneer in the study of organic contaminants, Dr. Rosen contributed
immeasurably to the success and professionalism of this project.
xxi
-------
SECTION 1
INTRODUCTION
The Cincinnati Water Works (CWW) serves all of corporate Cincinnati,
most of Hamilton County, and a portion of Butler and Warren Counties in
Ohio. The water service area covers approximately 1,165 sq km (450 sq mi)
and serves more than 860,000 people by means of 195,300 water accounts.
Average daily pumping rates are currently 6.1 cu m/s (139 mgd). As such,
this system is the largest community water system located on the Ohio
River.
Water intakes for the system are located 744.8 km (462.8 mi) downstream
from the headwaters of the Ohio River. These headwaters begin at the con-
fluence of the Allegheny and Monongahela Rivers. The uppermost 560 km
(350 mi) of the Ohio River and some of its tributaries, particularly the
Kanawha River, accommodate many of the nation's chemical industries. Each
year, more than 72 billion kg (80 million tons) of coal, petroleum products,
grain and chemical products traverse that portion of the river upstream of
Cincinnati . The Ohio River is truly a commercial giant subject to a
variety of pollutants.
In their concern for water quality, the CWW had previously unilaterally
funded research for reducing trihalomethane formation and had also made
initial inquiries to expand their research capability and to seek financial
assistance at the Federal level. In early 1977, approximately 63.5 metric
tons (70 short tons) of carbon tetrachloride were discharged into the
Kanawha River and found its way into many water intakes along the Ohio
River. This incident gave impetus to concluding these efforts.
On August 3, 1977, a research grant was awarded to CWW by the United
States Environmental Protection Agency (USEPA) entitled "Feasibility Study
of Granular Activated Carbon Adsorption and On-site Regeneration."
The primary objective of the project was to determine if the use of
granular activated carbon (GAC), utilizing either deep bed contactors or
conventional depth gravity filters with on-site GAC regeneration, is
feasible for removing specific trace organics from Ohio River water while
treating it for human consumption. To be considered feasible, the facility
had to remove selected organics to a predesignated level at a cost acceptable
to the consumers without adversely affecting the level of treatment provided
by the existing plant. A secondary objective was the development of plant
design and operating parameters for full-scale plant conversion to GAC
treatment.
-------
As a result, a three-phase study was designed to accomplish the following
comparisons and determinations:
Phase 1, full-scale GAC filters:
- Compare the relative performance of conventional depth gravity GAC
filters (sometimes referred to as filter-adsorbers) for organics removal
using three different configurations: 45.7-cm (18-in) 12 x 40 GAC with
30.5-cm (12-in) filter sand underlayer, 76.2-cm (30-in) 12 x 40 GAC,
and 76.2-cm (30-in) 20 x 50 GAC.
- Determine the need for the 30.5-cm (12-in) layer of filter sand
currently required by the OEPA for rapid sand filters that are con-
verted to GAC filters.
- Determine the need for using a GAC mesh size similar to that of sand
for the removal of particulate matter in a GAC filter.
Phase 2, pilot-scale GAC filters and pressure contactors:
- Compare the relative performance of virgin GAC and regenerated GAC in
pilot columns simulating gravity GAC filters.
- Compare the relative performance of bituminous-based GAC to lignite-
based GAC in a pilot column simulating post-filtration contactors
(hereinafter referred to as contactors).
- Predict expected performance of full-scale contactors in Phase 3 through
operating data and experience gained with pilot columns.
Phase 3, full-scale contactors and GAC filters with on-site GAC regeneration:
- Determine the extent of trace organics removal by full-scale contactors
and GAC filters operating concurrently.
- Determine whether a positive correlation exists between results obtained
from pilot columns simulating GAC filters and those obtained from
actual full-scale GAC filters.
- Determine whether a positive correlation exists between results obtained
from pilot columns simulating contactors and those obtained from actual
full-scale contactors.
- Compare the performance of GAC filters and contactors after successive
regenerations.
- Determine GAC regeneration criteria for full-scale contactors and GAC
filters.
- Determine GAC losses under on-site regeneration conditions.
-------
- Determine the costs associated with organics removal by contactors and
GAC filters.
- Determine the costs associated with on-site GAC regeneration.
- Determine various facility design parameters for full-scale applica-
tions.
Numerous acronyms and abbreviations have been used throughout this
report. Their meaning can be found in either the Abbreviations List in the
preliminary pages or in the Computerized Data Dictionary in Appendix A.
Volume 1 of this report contains representative figures and tables of
data discussed herein; Volume 2 presents figures and tables of other relevant
data and contains on microfilm a comprehensive listing of data developed
under the cooperative agreement.
-------
SECTION 2
CONCLUSIONS
The conclusions stated below are based solely on the governing condi-
tions and findings of this particular project. Although certain findings
may very well apply at other locations, particularly on the Ohio River, the
reader is cautioned not to make conclusions that are all-encompassing and
may be inappropriate under differing conditions.
1. No turbidity reduction benefit was derived from the requirement of
Section 4.2.1.6 in the Recommended Standards for Water Works , commonly
referred to as the "Ten State Standards" for a 30.5-cm (12-in) sand under-
layer to GAC or to the requirement that replacement media be the same effec-
tive size as filter sand.
2. Bacterial growth within the GAC filters and contactors was experienced.
Harmful bacteria were eliminated by post-chlorination.
3. Post-chlorination would be an absolute necessity if the entire plant
were converted to GAC.
4. Bituminous-based GAC outperformed lignite-based GAC with respect to
service life, weight of contaminants adsorbed, and cost per weight of contam-
inants adsorbed.
5. Pilot columns were reasonably predictive of full-scale GAC systems for
organics removal.
6. Floe removal by GAC filters had little effect on the carbon's adsorp-
tive ability.
7. The optimum GAC empty bed contact time (EBCT) would be between 7.0 and
15 minutes during annual average conditions and greater than 15 minutes
during critical summer conditions.
8. Regeneration restored the GAC to its virgin adsorptive capacity.
9. GAC regeneration losses averaged 15% by volume for ten contactor regen-
erations and 18.5% by volume for six GAC filter regenerations.
10. No adverse impact from regeneration was experienced relative to air
pollution, wastewater discharge or worker environment.
-------
COSTS
Cost conclusions made in this report are based on preliminary estimates
using cost curves developed for general application. These cost curves were
applied to site-specific design criteria. Actual costs could be considerably
different when determined for this or other sites based on a detailed engin-
eering design.
1. The preliminary cost estimates for full-scale design, construction and
operation of a GAC system at the CWW treatment plant indicate that a capital
investment of approximately $40 million (based on 1981 dollars) may be
required to reduce total organic carbon (TOC) concentrations to a specified
criterion of 1,000 ug/1 using either GAC filters or contactors.
2. A GAC filter system compared to a contactor system will annually cost
(in 1981 dollars) about twice as much for operating and maintenance costs
($8.0 vs. $4.0 million) and about 1.5 times as much for total costs, includ-
ing capital amortization ($12.4 vs. $8.5 million).
3. The estimated increase in the unit production costs of water to reduce
finished water from an average of about 2,100^ig/l to a treatment goal of
1,000 >ig/l TOC will be $0.06 per cu m ($0.24 per 1,000 gal) for GAC filters
(7.5 min EBCT) and $0.04 per cu m ($0.165 per 1,000 gal) for contactors
(15.0 min EBCT), all in 1981 dollars.
4. The cost to regenerate GAC on-site over the life of the project averaged
about $0.46 per kg ($0.21 per Ib).
-------
SECTION 3
RECOMMENDATIONS
1. The contactors used for this project were carbon steel lined with a
20 mil coat of Cook Phenicon 980 Epoxy Phenolic paint. The lining was pur-
ported to be resistant to organic leaching and carbon abrasion. This config-
uration was used in lieu of stainless steel for economic reasons. After two
years, the lining failed and parted from the steel. Additional work should
be done to identify suitable lining materials or application methods that
will produce a longer service life and yet not leach organic contaminants.
2. The sand separator supplied with the regeneration furnace was very
inefficient. A more efficient means of separating filter sand from GAC
should be developed for sites that use filter conversion to a GAC configura-
tion.
3. Health effects were not a part of this project, therefore, toxicity
tests were not conducted. The effectiveness of GAC for removing toxic sub-
stances should be determined by performing toxicity tests on influent water
and GAC effluent water of various contact times and various stages of carbon
exhaustion.
4. Lignite-based GAC was only used in pilot-scale studies during this
project. Full-scale use of lignite-based GAC should be conducted to deter-
mine the regenerability characteristics and the losses due to transport and
regeneration.
5. The hydraulic loading to the contactors was established to obtain a
maximum 15 minute EBCT. Carbon use rates for 20 minute EBCT were based
solely on extrapolation which indicated no economic advantage. Additional
work should be done to establish the optimum EBCT for the CWW which appears
to fall somewhere between 7.5 and 20 minutes.
-------
SECTION 4
DESCRIPTION OF THE WATER TREATMENT COMPLEX
The CWW has the Ohio River as its primary water source. The original
steam powered pumping station, constructed at the turn of the century, is
served by an intake pier situated approximately 100 m (300 ft) from the
Kentucky shore (Figure 1). A gravity tunnel beneath the river supplies four
0.7 cu m/s (15 mgd) electric pumps which are the only active units at the
River Pumping Station (RS) since the deactivation of the steam plant in
1963. A second raw water pumping station, the electrically powered Ohio
River Plant (ORP), was put into service at that time. It too is supplied by
a gravity tunnel beneath the river which has a subsurface intake crib located
directly upstream from the intake pier. ORP houses four 1,492 kW (2000 hp)
electrically driven pumps having a capacity of 3 cu m/s (70 mgd) each. The
raw water is conducted from the two low pressure pumping stations through a
pair of 1.5 m (60 in) cast iron mains installed at the time of construction
of the RS and a single 1.8 m (72 in) concrete main put in service with ORP.
Adjacent to the cast iron mains is Chemical Building West (CBW), a
treatment facility, which includes a powdered activated carbon (PAC) storage
and slurry feed system, liquid alum storage and feed equipment and four
active and one standby 50 metric ton (55 short ton) liquid chlorine storage
tanks with a battery of chlorine evaporators, feeders and injection pumps.
From this installation, all feed systems may inject directly into two
1.5 m (60 in) and one 1.8 m (72 in) mains. Chlorine solution and PAC slurry
may also be fed at the hydraulic jumps of Chemical Building East (CBE) or at
the Filter Plant influent flume from CBW.
The raw water from the pumping stations discharges into two large
presettling basins having a combined capacity of 1.4 mil cu m (372 mil gal).
This rather unique feature among water plants provides 2 to 3 days retention
time. In past years, chlorine was injected into the raw water mains ahead
of these basins. A study in 1975 revealed that improved water quality was
achieved by moving the point of chlorination downstream from the basins.
This resulted in a 75% reduction of THM formation and a 65% reduction of
chlorine consumption.
From the two large reservoirs, the water flows through CBE passing
through two parallel water turbine powered generators which serve as a
velocity breaker while, at the same time, recovering some of the pumping
power costs. The turbine discharge flow is split into two hydraulic jumps,
thereby providing a rapid mixing of chemicals being fed at that point. CBE
-------
LEGEND
03
INTAKE PIER
TUNNEL
RIVER PUMPWO STATION
SUBMOtOEO INTAKE PIER
TUNNEL
OHIO MVEft PLANT
CHEMICAL BULDINa WEST
UNDEROROUHD WASH WATER CHAMBERS
SETTLIN9 BASINS
CHEMICAL BULDINa EAST
HYDRAULIC JUMPS (RAPID MIX 1
FLOCOULATORS FOUR PER BASIN
OLARIFtRS TWO PER *ASIN
CHLORINE OPTIONAL
FILTRATION PLANT (4TBAPTD BAND
FILTER*)
POST OHLORINATION KROOIO
UNOEROROUND FINISHED WATER RE8V
• • • •
enAvmr TUNNELTO MAIN puuctN STA.
SHAV1TY TUNNELTO TENNYSON PUMOi
STA.
Figure \. Treatment process.
-------
incorporates a PAC facility equal to that of CBW, lime storage, slaking and
feed equipment, ferric sulfate storage and feed, and soda ash storage and
feed. From here PAC may be fed directly to the hydraulic jumps or to the
raw water mains prior to the presettling basins. Fluoride solution is also
fed at the hydraulic jumps.
The hydraulic jumps direct the water to two parallel basins, each con-
taining a flocculation section having a series of four horizontal shaft,
paddle wheel flocculators followed by clarification with settled coagulant
being collected by two radial sweep boom clarifiers. Each basin has a
capacity of 49,215 cu m (13 mil gal).
From the clarifiers, the water enters the Filter Plant through two
flumes. The Filter Plant consists of 47 rapid sand gravity filters which
are described in detail in a subsequent section of this report. The building
also houses a fluoride storage and feed facility and a post chlorination
installation supplied by 0.91 metric ton (1 short ton) capacity cylinders.
The plant effluent flume transmits the finished water to either of two
underground clearwells. The effluent flume serves as a contact chamber when
post chlorination is required.
Clearwell No. 1 was constructed as an open reservoir in the original
plant construction (circa 1900) and was covered in the late 1930s. It has a
capacity of 86,315 cu m (22.8 mil gal) and primarily serves the Main Pumping
Station (MS) through a gravity tunnel. Clearwell No. 2, put into service in
the early 1950s, has a capacity of 20,821 cu m (5.5 mil gal) and serves
Tennyson Pumping Station (TS)through a separate gravity tunnel. The MS and
TS increase the water pressure to provide distribution throughout the system.
Filter backwash water is supplied by two underground reservoirs located
adjacent to the presettling basins. They have a combined capacity of
3,500 cu m (0.92 mil gal).
FILTERS
The CWW utilizes the rapid sand filtration method in its treatment
process. The Filtration Plant houses 47 filter units, each having an effec-
tive area of 130 sq m (1400 sq ft) with a normal operating rate of
1.7 Ips/sq m (2.5 gpm/sq ft or 5 mgd).
The filters are of reinforced concrete construction, each comprised of
two boxes 4.3 m (14 ft) wide by 15 m (50 ft) long, separated by a 0.76 m
(2.5 ft) wide gullet running the long dimension. The gullet serves as the
filter influent and backwash discharge flume. A series of four troughs per
boK conducts the backwash to the gullet and serves to attain uniform effect-
iveness of the backwash operation. Seven filters, of most recent construc-
tion, have the Leopold perforated tile bottoms and surface wash. The other
40 filters have a network of perforated cast iron pipe laterals which conduct
the filtered water to collection headers beneath the filter structure.
-------
The filter bed is built up of gradations of gravel and sand in diminish-
ing sizes from bottom to top (Figure 2) as follows: 15.2 cm (6 in), of 3.8
to 6.4 cm (1.5 to 2.5 in) gravel covering the laterals, 12.7 cm (5 in) of
1.9 to 3.8 cm (0.75 to 1.5 in) gravel, 10.2 cm (4 in) of 0.25 to 0.6 cm (0.1
to 0.25 in) pea gravel, 7.6 cm (3 in) of 1.0 mm (0.04 in) torpedo sand, and
76.2 cm (30 in) of 0.45 mm (0.02 in) effective grain size filter sand.
The top surfaces are graded off with the top lip of the troughs as a
reference with the clearance termed as freeboard. A finished filter bed
initially has a 68.6 cm (27 in) freeboard.
Each filter has an adjacent control panel with switch controls and
position indicators for operating the influent, effluent, wash water, and
sewer valves. The control panel also has meters showing the rate of flow,
loss of head, and elapsed time. The actual rate of flow is accomplished by
modulating a discharge valve controlled by pressure differential through a
venturi in the effluent pipe.
Large meters indicating the rate of rise for controlling the backwash
operation are situated near each end of both filter galleries clearly visible
from the filter control panel. Normal plant operation requires backwashing
when a head loss of 1.8 m (6 ft) is attained or after 60 hr of service,
whichever comes first. The procedure calls for a rate of 12.7 cm or
122 Ipm/sq m (5 in or 3 gpm/sq ft of rise) for one min followed by 61 cm or
611 Ipm/sq m (24 in of rise or 15 gpm/sq ft) for 3.5 min.
Construction and Modification
For the purpose of this study, five of the Water Treatment Plant's 47
filters were selected in an area (Figure 3) where the most recent filter
rebuilding had taken place in order to minimize the possibility of a break-
through of turbidity during the program. A single concentration of filters
was considered desirable to simplify the media transport system and as a
convenience in collecting samples (Figure 4).
One empty filter box was set aside for use as a storage facility for
regenerated GAC in one half and virgin GAC in the other. Three GAC filters
(19A, 21A, 23A) were put into service on February 14, 1978 in Phase 1 of the
study using various configurations and GAC types. Another GAC filter (15A)
was used during Phase 3 for direct comparison with contactor operation.
No actual structural changes were made to the filters involved, however,
some features were added to facilitate the various roles in the study.
Foremost of these was the installation of a sample pump which could be
selectively connected to stainless steel pipe probes which draw water from
the GAC bed. The depth of the inlet ports were adjustable to permit testing
of the water quality at any desired level in the GAC bed. Initially the
probes were set at the depth of 30.5 cm (1 ft) and at the interface of the
GAC and support sand. A third probe in GAC Filters 21A and 23A was set at a
depth of 45.7 cm (18 in) in order to get a direct comparison between them
and GAC Filter 19A's interface.
10
-------
*rj •.''''••'•*••/'•'-.' ' 'UNDER DRAJ
•.-:• •.-.•.••:•••:;•:•• .w«*,-
FILTER WATER
/"COLLECTION
( PIPING
EFFLUENt;.'.'. I :
fVALVE /.. ' .'
Figure 2. Filter and pipe gallery cross-section.
-------
1
CD
i
CO
J* GALLERY PROCESS -#«
SUPPLY CONNECTIONS
ro
i
CD
CD
I
Ol
i
01
i
CD
i
T
CD
9
CD
ro
CD
2
E
>
A" GALLERY
ro
>
>
CO
O
N
1
| CROSS GALLERY ' j
o»
i
00
i
CD
CO
.2
ro
CO
ro
CM
m
&(&*-. — WASH WATER SUPPLY
. DRAIN CONNECTIONS
00
CT>
CD
CD
CD
ro
O
CO
ro
ro
i
CD
ro
A
CO
ro
Ul
CO
ro
01
i
cc
13-A
STORAGE
FILTER
U>
TO
>
^A,,,.,, „. r* nt i*r
>
i
DISCV
I
^*~*t M3
Ci(JI -njJ*
*
0 _
ACTOR
URGE
O
j) ^-
B
•n
<-- aro1
0 _
ro
EFFLUENT
POINT
fe
-------
FILTERS
23A
2IA
19 A
!7A
ISA
i3A
is
CJX
£8
.££!
Jj
IO
D
-------
The sample pump, probes, and priming tank are situated on a structural
steel frame and grating platform which extends from the walkway out over the
filter opening. Safety rails with a kick plate were installed along the
heretofore unprotected walkways to permit safe access to the sampling loca-
tion.
In addition, a connection was tapped into the filter effluent pipe to
permit sampling at that point as well as provide a water source for a Hach
turbidimeter which continuously monitored and recorded the output quality on
a remote recorder located in the laboratory.
Modification of Sand Filter 19A
In view of the fact that the State of Ohio is signatory to the Recomm-
ended Standards for Water Works commonly referred to as the "Ten State
Standards" for potable water treatment, it was necessary to study one
filter structured in compliance with those standards which require a minimum
of 30.5 cm (12 in) of filter sand supporting the GAC bed. Accordingly,
Filter 19A had only 45.7 cm (18 in) of sand removed and replaced with
45.7 cm (18 in) of Westvaco 12 x 40 WV-G GAC. Filter 19A was taken out of
service on March 21, 1980, and the GAC was regenerated during shakedown
operation of the furnace. The regenerated GAC was stored for use as make-up
GAC. A new 76.2 cm (30 in) bed of filter sand was installed and the filter
returned to normal plant service.
Modification of Sand Filter 23A
The Ohio Environmental Protection Agency (OEPA) approved the project
contingent upon inclusion of at least one filter containing GAC of the same
effective grain size (0.45 to 0.55 mm) as the sand which was removed. While
not specifically recommended by the supplier for this application, the
20 x 50 size met the State's requirement. Modifying Filter 23A to comply
with this request entailed removal of the 76.2 cm (30 in) deep bed of filter
sand, regrading the torpedo sand to establish a 1.5 m (57 in) freeboard and
installing 76.2 cm (30 in) of Westvaco 20 x 50 WV-W GAC.
Modification of Sand Filters 21A and ISA
Filter 21A was utilized as a filter adsorber throughout the program
while Filter 15A was in use only during Phase 3 for direct comparison to the
contactors. In order to obtain the greatest possible contact time for the
water passing through the GAC bed, the full 76.2 cm (30 in) bed of filter
sand was removed and the torpedo sand screeded off to a uniform 1.5 m
(57 in) below the lip of the filter troughs. This was replaced with an
equal volume of Westvaco 12 x 40 WV-G GAC. This grade of GAC was recommended
by the supplier for a broad spectrum of organic removal.
Modification of Filter 13A
Filter 13A was selected as the GAC storage facility. The top 76.2 cm
(30 in) of fine filter sand was removed. A panel of polypropylene mesh of
14
-------
0.018 cm (0.007 in) filament, having a maximum opening of 0.45 mm, was
placed on top of the torpedo sand to serve as a barrier separating the sand
from the GAC, thereby preventing mining or inclusion of the sand when remov-
ing the GAC.
Modifications to Filter 11B
Initial discussions provided for the dedication of a single 0.22 cu m/s
(5 mgd) filter to supply the water for the contactor process. Filter 11B
was selected for this purpose. After considering the effects of a possible
turbidity breakthrough of the filter bed and the interruption of supply
caused during normal filter backwash operations, this idea was discarded in
favor of drawing from Filter Plant "B"-Gallery effluent. In order to accomp-
lish this without structural changes to the effluent flume, it was necessary
to remove the existing rate-of-flow controller with its companion
square-to-round transition elbow and replace it with one of a current design
with a shorter length and standard pipe connections. The elbow was replaced
with a tee which provided the additional connecting point as a supply source
for the project.
Modifications to Filter 13B
A supply of water for backwashing the contactors was obtained by remov-
ing a 76.2 cm (30 in) elbow from the washwater supply to Filter 13B and
replacing it with a 76.2 cm (30 in) tee. The additional branch was reduced
down to the desired 25.4 cm (10 in) pipe which was then run into the pump
room to the suction end of the 82 1/s (1,300 gpm) wash water pump. The
discharge piping from the pump was routed through the access tunnel to the
contactor building where it joined the multiple manifold network which
interconnected contactor influent and effluent lateral systems. The backwash
discharge from a contactor was directed to a 25.4 cm (10 in) backwash sewer
line header, back through the access tunnel, pump room and into the filter
pipe gallery where it was discharged through an existing connection on the
plant washwater sewer system at Filter 13B.
PILOT COLUMNS
Description of 7.6 cm (3 in) Diameter Pilot Columns
Two units of 7.6 cm (3 in) diameter pilot columns were utilized during
the course of the project (Figures 5 and 6). The two units were operated
with a 76.2 cm (30 in) column of Westvaco 12 x 40 WV-G GAC.
Three glass tube sections of 45.7 cm (18 in) length were coupled
together with companion flanges, resulting in a single tube of 1.4 m
(54 in) . A conical cap was fabricated to close the top to assure even flow
over the cross section and eliminate dead space where filtrate may be
trapped while being backwashed.
Between the flanges connecting the bottom and second sections a stain-
less steel ring was inserted. This was drilled through radially, the inside
15
-------
O
w
(A
<3>
O
O
(1)
t+-
o
CD
LO
u_
16
-------
-n
(D
O)
CD
O
~n
-------
opening was fitted with a fine-screened sample probe, while the outer opening
provided a sample line connection referred to as Regenerated/Virgin (RV) for
identification of samples. By regulating the depth of the supporting gravel,
this port could match the contact time of a 45.7 cm (18 in) deep GAC filter
bed.
The bottom was capped with a flat plate drilled for a pipe connection
at the center. A fine mesh screen covered the bottom to prevent break-
through of GAC. An influent valve and backwash waste valve at the top
connection in conjunction with an effluent valve and backwash supply valve
at the bottom connection made possible complete simulation of filter opera-
tion.
The columns were mounted on a plywood panel with the effluent directed
to a container which was emptied through a solenoid valve when sensing
electrodes determined that the container was full. The flow for the entire
operation was calibrated and the number of cycles recorded by an automatic
counter.
Feed was intended to be regulated through a rotameter mounted on the
board with the column. Due to the limited available head, the added head-
loss caused by the rotameter was too great to yield the desired throughput
so it was bypassed.
Description of the Lignite 10.2 cm (4 in) Pilot Column
The pilot column, loaned by ICI Americas, Inc. for testing their Hydro-
darco 1030 GAC product, consisted of five tubes of 10.2 cm (4 in) diameter
and 1.8 m (6 ft) length (Figures 7 and 8). They were mounted side-by-side
on a free-standing framework of Unistrut-type channel. Both ends of each
tube were capped. The top cap had a quick-connect gauge fitting and a
two-valved fitting for connections for process and backwash flow. The first
tube in the process sequence was fitted with a dead weight safety valve.
The lower end had only the connector with the two valves, one for process
flow and the other for backwash flow. Inside the column, the port opening
was fitted with a Johnson well screen to prevent loss of GAC. Pea gravel,
3 x 6 mm (1/8 x 1/4 in) was placed in the tube to a depth of 2.5 cm (1 in)
above the screen.
Four of the tubes were connected in series, with the interconnecting
tubing running from the bottom of one to the top of the next, resulting in
downward flow through each. The first tube in the sequence had 0.9 m (3 ft)
of GAC, each of the three successive tubes had 1.2 m (4 ft) of GAC, for a
cumulative column depth of 4.6 m (15 ft), which equals that of the contactors.
A sample valve was spliced into each length of the series piping providing
intermediate points for collecting samples identified as post filter
adsorbers (PFAs).
The fifth 10.2 cm (4 in) tube was completely filled with GAC to give
the maximum contact time possible. This tube was utilized to treat the
backwash water for the four process tubes so as not to load the tubes with
18
-------
LEGEND
OA/ -
SAMPLE. VALVE.
t^] 3AFC.TY V~(.V£
><1 3-WAX VALVE.
0 PRC3SUKE. 6AI/6C.
QO3 QUICK CtlSC.ONHC.C7 SOC.
C'G/tC
Figure 7. Lignite pilot contactor, processing flow.
-------
BACKWASH St/PPLV
L/fJ£ -
&A.CK.WA.3H WATCH
fiLTEK rtsoe
ON-Off VALVE.
3AFC.TY VALVE.
3-WAY VALVE.
PBESSUKE &AU6C.
QUICK OtSCQNNZCTiOC.
— PLLKiGCO FITTING
C • GAC
-------
organics from the bottom. The effluent was connected individually and
selectively to the valved connection provided at the bottom of each of the
four tubes. Influent flows were regulated by a rotameter and total through-
put indicated on a standard water meter in the supply line.
Description of the Bituminous GAC 10.2 cm (4 in) Pilot Column
The GAC column was comprised of four 1.5 m (5 ft) sections of 10.2 cm
(4 in) diameter glass tubes (Figures 9 and 10). These were mounted side-by-
side on a plywood panel board affixed to a free-standing frame.
Each tube was capped on both ends with a blind flange which was drilled
and tapped to receive the required fittings. Each end was piped and valved
to permit downflow for process use and upflow for backwash mode. Process
piping connected the four tubes in series from the outlet at the bottom of
the first one to the inlet at the top of the next, continuing through the
sequence of the four tubes. Each series pipe had two branch connections,
one for a PFA sample tap and the other for a pressure gauge, all of which
were mounted on the board adjacent to their respective tubes. There was a
pressure gauge on the influent line to the first tube and a dead weight
safety valve on the cap of the first tube. Flow rates were controlled by a
rotameter mounted on the face of the board. Total throughput was indicated
on a wall-mounted water meter.
The tubes were charged with Westvaco WV-G 12 x 40 GAC. The first tube
contained 0.9 m (3 ft) of GAC while each of the three succeeding tubes
contained 1.2 m (4 ft), giving a cumulative column depth of 4.6 m (15 ft).
Rather than make all four tubes equal at 1.1 m (45 in), the existing config-
uration was established to provide ample freeboard for expansion during
backwash of the first tube which, in effect, is the top portion of the GAC
column and would entrain all of the suspended matter. A fine mesh screen
and shallow gravel bed supported the GAC in each tube.
CONTACTORS
Description
Four contactor tanks were placed in the south half of the contactor
building (Figure 11). Each contactor unit was basically a cylindrical shell
construction of 9.55 mm (3/8 in) thick carbon steel plate (Figure 12). The
ends were closed with standard ASME flanged and dished heads. At the lower
end of the cylindrical section, a pipe was welded in place. The exposed
flanged end served as the connecting point for the effluent system (downflow
mode). There was a short flanged length projecting into the tank. A
20.3 cm (8 in) stainless steel pipe was bolted to that flange and extended
diametrically across the tank with its free end being capped and supported
by an angle iron cradle welded tc the interior wall. This was the header to
which eight flanged, stainless steel laterals, arranged into four horizontal-
ly opposed pairs, were connected. The lateral pipes had varying lengths
dictated by the curvature of the contactor tank shell. Each lateral had a
pattern of 2.5 cm (1 in) diameter holes drilled through its wall equally
21
-------
VENT
NJ
K)
f^O <3^C
.9m
S. 7 em (I. 3,»)GtAvet.
[X] ON-OFF VALVE.
Xh SAMPLE VALVE.
t^J SAFETY VALVE.
Q
Figure 9. Bituminous pilot contactor, processing flow
-------
Figure 10. Bituminous pilot contactor, backwashing flow.
-------
SCM-t
MAKC _-,
rnrmfs, M-Sr /H
Figure it. Contactor Building layout.
-------
10.2cm (03ft)D!A CARBON I
CARBON DISTRIBUTOR PLATE
.20.3cm (0.7ft) OIA.
OBSERVATION PORT
ro
ui
DISTRIBUTOR
20.3cm(07fl) INFLUENT.
HEADER
(03ft)LONG W/50 MESH
SCREEN TYPICAL 6 PLACES • - -
3.1cm 10. 3ft ) OIA SAMPLE
TUBE W/50 MESH SCREEN
TVPir.il 1 PI AOFS
HEADER -— -^____^
IO.2cwD.3fl) OIA. LATERAL ASS£MBLI__-,
W/SO MESH SCREEN ~~~
-^
-
3.4M(I1.0H) 00
CARBON BED SURFACE
if* '\ r vO* if""* s-s ^ .v>>v'""'**Vi i -*,~"')?"'Jv>-j^
i*j7<
^^lr«l^
j
11 437 cm
j (1.5 ft)
"" 45.7 cm
. (15 ft)
-« 406 cm
j (1.3 ft)
•» 406 cm
j (13 ft)
"* 55.9cm
. (1 8 ft)
i (To ft)
61.0 cm
(2.0 ft)
^ 45.7 cm
-« (1.5 ft)
"* 45.7 cm
• (1.5 ft) .
11 45 7 cm
PI 1^ 3fl)
0.2cm—
(0.3ft)
§£
*«o
goj
2C
ION
— <»•
zg
IMt-
zS
»«
I-O
OJ
Figure 12. Contactor cross-section.
-------
spaced in seven annular rings with three holes radially 120 degrees apart in
each ring. About each lateral was a 50 mesh stainless steel screen sleeve
secured by stainless steel hose clamps. Directly above this arrangement was
an identical header/lateral system. The centerlines of the two systems have
a vertical separation of 6.5 m (21.2 ft). This provided the external connec-
tion for the influent piping. The separation was calculated to permit a 50
percent expansion of the 4.6m (15 ft) deep GAC bed when backwashed.
A series of 10.2 cm (4 in) flanged pipe stubs which projected from the
outer face of the tank were along a vertical line 90 degrees from the
influent/effluent connections. The purpose and position of these sample
taps were dictated by the analytical requirements of the program, basically
on 45 cm (1.5 ft) increments through the GAC bed with some variations to
match the contact time experienced at the sample points in the converted
sand filters. Not all taps were utilized on a continuous basis. Five prime
points were equipped with probes which extended across the full diameter of
the bed to assure a representative sample of the water. The sample probe
was a 5.1 cm (2 in) pipe with a pattern of twenty holes 6 mm (1/4 in) in
diameter and having a 50 mesh screen sleeve in the same manner as a lateral.
The remaining sample ports were fitted with short stubs with only six holes
and a screen sleeve in the same fashion. These projected only 12 cm
(4.75 in) into the bed. Each contactor had its own sample sink with individ-
ual faucets connected to the five primary sample taps.
Each tank had a "GAC carbon-in" connection at the center of the top
head, and just below it a cone which distributed incoming GAC and protected
the influent lateral system. At the top of each tank was an air/vacuum
release which vented air as the tank was being filled and admitted air when
it was being drained. A steel ladder, with safety cage extending the full
height of the tank, provided access to a 50.8 cm (20 in) manway which was
located on the side of the tank above the fully charged GAC bed level. Each
tank was supported by four wide flanged (WF) structural steel legs. Specifi-
cations for the contactor tanks are shown in Table 1.
Although the "A", "B", "C" , and "D" labeling on the tanks and instrumen-
tation appears to be random, it was not. This resulted from the identifica-
tion being assigned in the order that each unit became available from the
installing contractor and put into service.
Each tank had a 7.6 cm (3 in) ball valve at the top and bottom GAC
transport connections. From the bottom or carbon-out connection, the 7.6 cm
(3 in) line joined an overhead common line or header. No increase in size
was necessary since only one unit at a time was discharged into it and a
constant velocity was desirable. In the same manner, the carbon-in port at
the top was connected to the 7.6 cm (3 in) carbon-in header.
Contactors could be charged directly from the GAC tank truck, the
regenerated GAC storage tank or the make-up GAC filter (13A). Each
contactor carbon-in valve control was a pneumatic cylinder operated valve.
26
-------
TABLE 1. CONTACTOR DESIGN SPECIFICATIONS
Diameter
Vertical Sidewall Height
GAC Depth
Design Capacity
Hydraulic Loading
Contact Time
Design Pressure
Test Pressure
Backwash Hydraulic Loading
3.4 m (11 ft)
6.7 m (22 ft)
4.6 m (15 ft)
0.04 cu rr,/s (1.0 mgd)
285 1pm/sq m (7.4 gpm/sq ft)
15.3 min
517.1 kPa (75 psig)
620.5 kPa (90 psig)
407 1pm/sq m (10.0 gpm/sq ft)
Spent GAC was removed from the contactors and transferred to the spent
GAC storage tank. The GAC slurry consistency was maintained at 10-15% to
transfer in or out of the contactors. Water injection nozzles were welded
into the GAC transport pipes to accomplish this. The carbon-out valves were
pneumatic cylinder operated valves. Control for the valves were the same as
the carbon-in valves.
The lower lateral collector flange of each tank was connected to three
elevated headers, one 25.4 cm (10 in) washwater supply line and two 40.6 cm
(16 in) diameter lines which functioned alternately as influent or effluent
lines dependent upon the operational mode selected. The top lateral collec-
tor flange was similarly piped to the two larger headers. By proper opera-
tion of the valves at each juncture, water could selectively be directed to
either system for upflow or downflow mode operation. A flow meter, which
could measure flows in either direction, was included in the riser pipe of
each contactor.
Backwash System--
The backwash water and drain system, previously described in associa-
tion with the modifications to Filter 11-B, was also interconnected with
this overhead network. Valving arrangements permitted backwash water to
enter the contactor through the effluent laterals and exit through the
influent laterals.
A pnuematically operated backwash rate valve was common to the four
contactors. It was located on the 25.4 cm (10 in) backwash drain line. The
backwash drain valve was used to maintain a backwash rate through the
contactor of 407 Ipm/sq m (10 gpm/sq ft).
The backwash pump was used to pump the backwash water through the
contactor. The pump took its suction from the 91.4 cm (36 in) washwater
line that supplied the backwash water for the B-gallery sand filters. The
pump was driven by an induction type motor operating on 460 volts, 3 phase.
Pump specifications are listed in Table 2.
27
-------
TABLE 2. WASHWATER PUMP SPECIFICATIONS
Total Rated Head 12.2 m (40 ft)
Capacity at Rated Head 82 1/s (1300 gpm)
Minimum Operating Head 10.7 m (35 ft)
Maximum Power at Minimum Operating Head 16.3 kW (21.8 bhp)
Net Positive Suction Head (NPSH) 3.05 m (10 ft)
Influent Piping--
The connection for supplying sand filter effluent water was previously
described under modifications to Filter 11B.
From this point of connection, a 40.6 cm (16 in) pipeline was run to
the suction connection of the two 0.1 cu m/s (3 mgd) process pumps. The
elevation of much of this line was above the hydraulic gradient of the
flume, therefore, it became necessary to add a vacuum priming system.
One process pump serves the contactors with the second acting as a
standby. The process pumps were driven by a horizontal, totally enclosed,
460 volt, 3 phase induction motor. The process pump specifications are
shown in Table 3. The pump discharges were manifolded into a single 40.6 cm
(16 in) line going through the access tunnel to the contactor building.
Effluent Piping--
The water from the effluent of the contactors flowed through a 40.6 cm
(16 in) effluent line. This line was routed through the access tunnel, pump
room and across the filter building where it discharged into the A-gallery
effluent flume through a connection welded to the cover plate of an access
manhole near Filter 13A. The long run of pipe was planned in order to
introduce the contactor effluent as far downstream as practicable in order
to eliminate the possibility of recirculating water, through the contactor,
which had previously been through the process.
The control valves on the influent/effluent and washwater systems were
pneumatically operated. Compressed air for this purpose was provided by a
TABLE 3. PROCESS PUMP SPECIFICATIONS
Total Rated Head 38.4 m (126 ft)
Capacity at Rated Head 132.5 1/s (2100 gpm)
Approximate Capacity Range 63.1-151.4 1/s (1000-2400 gpm)
Operating Head Range 29.9- 39.6 m (98-130 ft)
Maximum Power 64.9 kW (87 bhp)
Pump Rotation Counterclockwise
NPSH 4.6 M (15 ft)
28
-------
dual compressor package. The compressed air system also served the inst
rumentation and control requirements of the regenerator. The contactor
control cabinet housed the flow and pressure indicators with recorders.
Flowmeter--Each contactor was equipped with a 15.2 cm (6 in) magnetic
flowmeter capable of measuring the flow of filtered water in both directions
without any moving parts in the meter coming in contact with the water. The
flowmeters were designed for a maximum pressure of 1,172 KPa (170 psig).
Each contactor had a two-pen indicating recorder with a chart and scale
which read 0 - 0.06 cu m/s (0 - 1.5 mgd). This meter also measured the rate
of backwash.
Flow indicating controller—Each contactor had a flow indicating con-
troller to set the rate through the contactor. The flow could be controlled
manually or automatically.
Pressure recorders—Each, contactor had two-pen pressure recorders and
indicators with a chart and scale which reads 0 - 700-KPa (0 - 100 psig).
One pen recorded the pressure at the top of the contactor and the other the
bottom pressure.
Instrument Air System--
The instrument air system was used for the metering and control of the
contactors. The compressed air equipment consisted of two heavy duty,
single-acting, two-stage, air-cooled, V-belt driven compressors, a vertical
receiver and two refrigeration-type, self-contained, air dryers. The speci-
fications for the instrument air system are shown in Table 4. The control
cabinet for the system which contained indicating lights, switches, relays,
solenoid valves and gauges was located immediately adjacent to the compres-
sor.
TABLE 4. INSTRUMENT AIR SYSTEM SPECIFICATIONS
Compressor Discharge Pressure 861.9 kPa (125 psig)
Motor Speed 1,800 rpm
Power Requirement 18.6 kW (25 hp)
Receiver Test Pressure 1,034.2 kPa (150 psig)
Receiver Volume 1.1 cu m (40 cu ft)
Dryer Capacity 0.84 mm (30 scfm)
Dryer Power 0.38 kW (0.5 hp)
Receiver Dewpoint 2 °C (35°F)
29
-------
REGENERATOR
Acquisition of the Regeneration Furnace
Once the eKtent of the Research Project was established, the required
capacity ,-for regenerating spent GAC was determined by extrapolating
published carbon use rates. Every known manufacturer of such equipment was
contacted for information which would be helpful in writing a performance
specification and to stimulate the interest of possible bidders of such
equipment.
A broad-based specification was drawn calling for a 227 kg (500 Ib) per
hour dry regeneration capacity furnace of either the infrared or fluidized
bed technology. Evaluation considerations included such items as initital
capital outlay, projected operating and maintenance costs, and expected GAC
loss in the process. Five bids were received, one electric infrared and
four fluidized bed units. The lowest three bids were determined and a
sample of GAC, previously exhausted in a USEPA pilot column study on
Cincinnati water, was submitted to each of the bidders for regeneration to
verify their proposed performance data. In every case the fuel requirement
was increased due to the fact that the GAC loading was not as great as
advance publicity lead them to believe.
Although the aggregate cost of other units over the claimed equipment
life (of up to 20 years) would have been lower, Westvaco's bid of $604,000
coupled with guaranteed performance was the lowest bid for the term of the
project.
Description of the Fluidized Bed Regenerator and Support Equipment
The fluidized bed regeneration process, sometimes referred to as "reac-
tivation", is a vertical process where the GAC progresses downward through
the regenerator counterflow to rising hot gases which carry off volatiles as
it dries the spent GAC and pyrolyzes the adsorbate.
The regenerator vessel was 6.9 m (22.5 ft) tall and 2.1 m (7 ft) at its
widest diameter (Figure 13). Typically, the outer steel shell was lined
with 10 cm (4 in) of mineral wool blocks covered with a cast-in-place refrac-
tory of up to 0.36 m (14 in) thickness. From bottom to top, it was divided
into three compartments which house four functional areas (Figure 13). The
bottom section was the combustion chamber into which a stoichiometrically
balanced stream of fuel and oxygen (air) flowed. These expanding gases of
combustion provided the heat and fluidizing medium. Steam was also injected
into this chamber, while it was part of the regeneration process the volumes
were predicated upon the need of additional fluidizing gases. Temperatures
reached approximately 1,040°C (1,900°F) in the combustion chamber which had
a water-sealed pressure relief vent. The burners were of a dual fuel type
which used either fuel oil or natural gas.
30
-------
w
w
o
o
o
o
-+j
o
L.
d>
C
d)
a>
CO
3
u_
31
-------
The regeneration zone was separated from the combustion chamber by a
stainless steel diaphragm fitted with a number of nozzles which distributed
the gases uniformly over the cross section of the furnace. It also supported
the GAC when at rest. A series of weir plates rested upon the diaphragm
plate to assure sufficient residence time for the GAC in process. The upper
part of this chamber was the incineration zone. Additional air and fuel
were injected by the secondary burner to incinerate organics from both the
regenerator and dryer zones. Temperatures across this chamber averaged
816°C (1,500°F).
The furnace off-gas was conducted from the regeneration/incineration
chamber through a breaching, a venturi scrubber, a tray scrubber and dis-
charged to the atmosphere by a stack through the roof. The breaching was
equipped with a safety blow-out panel which would release excessive pressures
through a relief stack.
The third chamber, smallest in volume, was the dryer which was separated
from the regenerator/incinerator by 316L stainless steel plate perforated
with a uniform pattern of holes to distribute the gas flow to attain uniform
fluidization of the GAC bed above it. The gas flow through this chamber was
induced by the suction of a fan, creating a negative pressure within the
chamber. The gas flow went through a cyclone separator which dropped out
much of the suspended particulates (GAC fines). The blower then discharged
into the incinerator zone where the entrained volatiles, which were driven
off the GAC in the drying process, and the remaining particulate matter were
incinerated. The temperature within the dryer was approximately 150°C
(300°F) and was controlled with cooling water sprays located above the GAC
bed. The dryer chamber also had emergency water sprays and a blow-out panel
type pressure relief vent connected to the relief stack.
The furnace contract also included two cylindrically-walled, conically-
bottomed tanks, each had a capacity of 23.9 cu m (855 cu ft), fabricated of
316L stainless steel, one for receiving and holding spent GAC prior to
regeneration (ST-1) and the other to receive and hold regenerated GAC (ST-2)
until it could be returned to the filter or contactor from which it came
(Figure 11). Spent GAC was discharged from ST-1 through a variable speed
rotary valve (RV-1), the rotational speed of which was controlled by the
operator. This determined the rate of feed to the furnace. The GAC was
moved from RV-1 by an eductor to either the sand separator or to the dewater-
ing screw.
A sand separator, which was an application of a standard piece of
placer mining equipment, was utilized only when GAC from a filter was being
regenerated. It was a vibrating table designed to separate material through
their difference in specific gravity. The sand was discharged to a portable
dump hopper for disposal and the GAC was discharged into a stainless steel
vessel which was fitted with an eductor which transported the GAC to the
dewatering screw.
The dewatering screw, which was an inclined screw conveyor, discharged
wet GAC directly into the dryer section of the regenerator. The GAC
32
-------
traversed across the dryer plate and through a side pipe. It passed through
a rotary valve (RV-2) and reentered the furnace at the regeneration zone.
The rotary valve served to isolate the two pressure zones within the process.
The GAC flowed across the regenerator bed to an outlet port and discharge
chute to the quench tank. The discharge chute was submerged to prevent
entry of excess oxygen into the process. An eductor at the bottom of the
quench tank transported the GAC to ST-2. An A-S-H rubber-lined slurry pump
was used to transfer the regenerated GAC to a filter or a contactor.
Most of the equipment to support direct operations of the regenerator
were a part of Westvaco's contract. Foremost of these was a Bristol micro-
processor/controller and a recording unit. This was housed in a controlled
environment room within the main building. The instrumentation and control
systems were designed to require only the minimum of attention by the opera-
tor. However, it was seldom that all of the automated systems were function-
al and the successful operation relied heavily upon the skills of the CWW
operating and maintenance personnel.
A three-stage turbo blower supplied the combustion air for both the
regenerator and incinerator burners. The recycled gas blower which handled
the dryer off-gas and reinjected it into the incinerator section, also
provided the gas movement required to fluidize the dryer bed. Particulate
matter in the recycled gases were removed by a cyclone separator. Two
blowers in series handled the furnace off-gases which also passed through a
venturi and a tray-type scrubber. All parts of the various gas handling
apparatus that were in contact with the gas stream were fabricated of 316L
stainless steel.
There was also a steam generator to provide the steam required in the
process. It, like the furnace, may be fired by either gas or fuel oil.
Appurtenant to the regenerator were the fuel oil day tank and two
0.2 1/s (3.2 gpm) centrifugal pumps which supplied contactor effluent water
for steam generation, quenching and transporting regenerated GAC.
GAC Treated Process Water Piping and Pumps
Two 12.6 1/s (200 gpm) pumps (EP-1 & EP-2) provided contactor effluent
for the steam generator, tempering water, quenching water, and transporting
regenerated GAC, thereby reducing exposure to organic laden water.
The suction was initially tapped into the 40 cm (16 in) contactor
effluent line just before it left the contactor building. With all four
contactors in service the pumps became air-bound and lost their prime. The
suction connection was moved to the drop leg as the 40 cm (16 in) line
entered the tunnel area making a trap configuration, assuring a full line.
Apparently, turbulance entrained sufficient air to cause the pumps to become
air bound.
Reliable service was achieved only by relocating the pumps to the floor
of the stairwell which connects the building to the access tunnel (Figure 11).
33
-------
The 40 cm (16 in) contactor effluent pipe was tapped adjacent to the pumps,
providing a positive suction head with only two contactors in service. A
7.6 cm (3 in) discharge header was routed to the north end of the building
along the east wall.
GAG Transport Piping
A 7.6 cm (3 in) carbon-in pipe line was connected to the top of each
contactor tank from a central header which terminated near the center of the
building. Similarly, a 7.6 cm (3 in) carbon-out line ran from the bottom of
each contactor to a header which terminated at a point approximately 30 cm
(1 ft) below the carbon-in line. A third line which began outside the
building provided a connection for unloading or loading GAC trucks and
terminated on a 90 degree alignment to the carbon-in and carbon-out lines
and at the same elevation [approximately 3.7 m (12 ft)] above the floor. A
quick-disconnect hose permitted various modes of interconnection between
these and the GAC transport piping provided by others (Figures 4 and 12).
The furnace installation included GAC transport piping that originated
at a point on the extended centerline and 1.8 m (6 ft) from the terminous of
the carbon-out line described above. This line discharged vertically down-
ward into the top of the spent GAC tank (ST-1).
From the bottom of ST-1, the GAC, motivated by water pressure through
an eductor, was selectively conducted to the dewatering screw which fed the
furnace or to a sand separator which discharged GAC into a receiving tank
equipped with a second eductor which was then connected to the dewatering
screw feed line.
GAC coming out of the furnace dropped into a quench tank (QT-1). An
eductor at the bottom of QT-1 propelled the GAC through a pipe which dis-
charged at the top of the regenerated GAC storage tank (ST-2) . A slurry
pump at the bottom of ST-2 moved the regenerated GAC through a transport
pipe which terminated facing the carbon-in line with a 1.8 m (6 ft) gap
between them where the quick-disconnect hose, described above, could direct
the product to a selected contactor or filter.
Two 10.2 cm (4 in) GAC transport lines traversed the Filter Building
from the area of the filters in A-gallery, to the GAC building. One line
terminated at the top of ST-1, bringing GAC from the filters, the other
terminated in the area of the other open ends of GAC piping in the center of
the building. By use of a flexible hose connection, a number of combinations
could be selected including: truck to contactor, truck to filter, filter to
truck, contactor to filter, filter to contactor, regenerator (ST-2) to
contactor, regenerator to filter, contactor to regenerator (ST-1), truck to
regenerator (ST-1), or regenerator (ST-2) to truck.
These lines were assembled with Victaulic connections to permit ease of
rearrangement within the filter area and removal of sections for clean out
in the event of plugging. All bends in the GAC transport system were of a
long radius design in order to reduce frictional loss of GAC.
34
-------
CONTACTOR/REGENERATOR BUILDING
Design work covering all aspects of the project began in early
September, 1977. In the interest of establishing the earliest possible
progress, and in hopes of having the concrete work done before severe winter
weather set in, a contract was let for the construction of the building
foundation, one half of the floor slab and a utility tunnel connecting the
proposed building to the existing Filter Plant. A standard modular building
was selected which could house the contactor tanks, whose size and number
were established, and could accomodate either an infrared electrical furnace
(horizontal process) or a fluidized bed gas fired furnace (vertical process)
which would ultimately be situated on the other half of the floor slab. The
tunnel was actually an adaptation of a large prefabricated culvert pipe. A
severe winter and late delivery of the tunnel line delayed completion of
this phase until August, 1978.
LABORATORY
Establishment of an Organics Laboratory
One condition imposed by the USEPA before they would approve the grant
was that CWW must establish, at their own cost, a laboratory capable of
conducting organic analyses of the type to be involved during the research
project. The existing laboratory was overtaxed with routine analytical work
to the point where a capital expansion project was underway, but it would
not be available in time to meet the schedule of the proposed program. A
lecture room (5.8 m x 10.1 m, 19 ft x 33 ft) was converted to an organics
laboratory.
Work on the conversion began in the spring of 1977. Progress of the
work was dependent upon delivery of the larger items of equipment which
included a fume hood, an ultra-high temperature oven (muffle furnace), a
Varian 3700 gas chromatograph (GC) with a Hall detector, a Linear recorder
and a Tekmar LSC-1 concentrator. At the same time, A. Dohrman DC-50 TOC
analyzer, existing in the general laboratory, was being retrofitted to the
DC-54 configuration with an ultra-low level analyzer.
It immediately became apparent that the air conditioning system designed
for the lecture room would not be adequate for the laboratory conversion due
to the heat-producing equipment that would be in use. As a temporary
measure, the CWW welding shop designed, fabricated, and installed additional
fresh air ducts and vents to the existing system to maximize air circulation
to the room. In addition, one of the laboratory windows was modified and a
10.1 MJ/hr (9,600 BTU/hr) window unit was installed.
In order to simulate the same conditions that the finished water in the
distribution system would encounter, three ambient temperature units were
designed and constructed. These consisted of three 18.9 1 (20 quart) coolers
fitted with sample bottle racks, an air-gapped water supply and a drain. A
steady stream of water flowing through these units would maintain the samples
at approximately the same temperature as that of water in the distribution
35
-------
system. An incubator that was in storage was reconditioned and placed in
service.
Storage for liquid nitrogen, necessary for subambient operation of one
of the Varian 3700 GC's was provided in the front cross gallery of the
filter building. An access hole was cored through an adjoining wall into
the laboratory and an insulated pipe was installed from the storage tank
through the wall to the instrument.
Table 5 lists the major items of work and equipment associated with the
pre-grant commitment.
New Laboratory Addition
7
After passage of the Safe Drinking Water Act the CWW administration
determined it would be necessary to modernize and enpand the existing facil-
ities to accommodate increased staff and new instrumentation necessitated by
the new regulations. Consulting engineers were commissioned to design the
laboratory addition. The new laboratory, not funded by the project grant,
became available as the grant study progressed and was utilized to relieve
TABLE 5. PRE-GRANT COSTS OF ESTABLISHING ORGANICS LABORATORY
Item Material Labor Total
Electrical Services $ 1,500 $ 9,400 $10,900
Incubator - 200 200
Cold Water Piping 20 200 220
Hot Water Piping 20 200 220
Natural Gas Piping 30 250 280
Drains and Cup Sinks 80 300 380
Distilled Water Piping 50 150 200
Ambient Temperature Units 120 . 900 1,020
Fume Hood 4,500 4,000 8,500
Ultra-high Temperature Oven 2,230 80 2,310
Vacuum Pumps and Piping 400 240 640
Air Compressors and Piping 800 480 1,280
Gas Tank Rack and Piping 300 550 850
Temporary Air Conditioning Provisions 100 330 430
Permanent Air Conditioning 2,000 5,780 7,780
Liquid Nitrogen Facilities 30 100 130
Initial Cleaning and Painting 300 1,000 1,300
Miscellaneous Items of Work 500 1,500 2,000
Table Tops and Consoles 900 3,100 4,000
Gas Chromatograph 16,600 600 17,200
Ultra-low Level TOC Module 7,590 80 7,670
$38,070 $29,440 $67,510
36
-------
overcrowded conditions in the existing laboratory. A gas chromatograph/mass
spectrometer (GC/MS), which was funded by and utilized in the final phase of
the study, was housed in the new laboratory. Use of this equipment was
delayed until the laboratory was completed and the environmental control
system balanced and regulated.
The laboratory addition provided an additional 427.3 sq m (4,600 sq ft)
of laboratory and office space. A new heating, air conditioning and ventil-
ating system for both the new and existing laboratory space was provided in
the construction contracts.
The new laboratory provided space for radiological equipment, GC/MS,
atomic absorption; expanded physical, chemical, and microbiological testing,
GAC analyses, extractions, and other related support equipment. The remod-
eled organics laboratory provides the CWW with an excellent facility for
water quality and research analyses. This facility was dedicated on
May 5, 1980.
Laboratory Equipment Acquisition and Installation
In the grant application, CWW committed itself to procure and/or install
certain facilities and equipment in advance of the grant period and outside
of the grant funding. Accordingly, the CWW requisitioned selected pieces of
analytical equipment and established an organics laboratory as previously
discussed.
Table 6 identifies the various vital statistics associated with major
pieces of lab equipment acquired or installed. There are always consider-
able unanticipated delays associated with the acquisition and installation
of laboratory equipment. CWW's experience was no exception. Such being the
case, only those equipment acquisitions, installations or shakedown problems
that significantly affected our ability to meet grant objectives are high-
lighted in footnotes to Table 6.
DATA MANAGEMENT
During grant planning, costs and problems associated with the recording,
sorting, reporting and evaluation of data were under-estimated. In sheer
numbers alone, close to 150,000 pieces of data have been computerized.
Originally grant data were recorded in the same fashion as was done with
routine laboratory data; that is, in log books, diaries and on a few
specially prepared forms. Reporting of these data in required quarterly
reports was initially done by photocopying handwritten forms, then through
typewritten versions of the handwritten forms.
Although CWW uses computer services provided by the Regional Computer
Center (a City and County sponsored computer center), the availability of
needed services was extremely difficult to obtain and development time was
usually long. The inadequacy of this system quickly became apparent. In
order to evaluate grant operations, graphs would be needed and doing these
37
-------
10
CO
TABLE 6. LABORATORY EQUIPMENT INVENTORY
Item Description
GC/ Electron
Capture
EC/ Electrolytic
Cond. Detector
GC Liquid Sample
Concentrator
GC Dual Pen Recorder
Fume Hood
Ultra-Low-Level TOC
Analyzer Module
Muffle Furnace
Analytical Balance
GC/FID
GC Microprocessor
GC/MS Cool Water
Recirculator
Water Bath (Ambient)
Safety Refrigerator
GC/FID
GC/Electronic
Integrator
GC/FID Capillary
Column
GC/MS
GC/Misc. Accessories
GC Integrator
Mixer Mill
, Not available.
Manufacturer
Varian
Tracor
Hall
Tekmar
Linear
Hamilton Ind.
Dohrman
Blue M
Sartorius
Perkin-Elmer
Perkin-Elmer
Neslab
Instrumts.
CWW Fabric.
Labline
Varian
Spectra
Physics
Varian
Finnigan
Varian
Spectra
Physics.
Spex
' lT»_J«n CT* nKA>
Model //
3700
T700
LSC-1
385
540M947
PR-1
BFD-20F-3
2434
Sigma 1
Sigma 10
CFT-25
-
355-10
3740
SP4100
3700
OWA-20
NA
SP4100
8000
Req.
Date
03/31/77
03/31/77
03/31/77
03/31/77
05/31/77
03/31/77
09/14/77
10/25/77
a
a
03/19/79
a
01/05/78
01/16/79
01/16/79
c
11/17/78
07/12/79
c
a
P.O. Date
04/21/77
04/21/77
04/22/77
04/22/77
07/15/77
05/19/77
09/30/77
12/05/77
12/01/77
12/01/77
04/30/79
a
02/13/78
08/20/79
08/01/79
c
02/15/79
10/11/79
c
a
Equip.
Rec'd.
Date
07/05/77
05/10/77
06/16/77
06/16/77
10/06/77
a
11/14/77
12/20/77
06/05/78
03/13/78
05/29/79
a
03/07/78
01/30/80
09/27/79
02/01/80
07/20/79
01/02/80
02/01/80
08/29/80
Purchase
Price
$ 9,465
3,514
3,619
b
4,455
7,591
2,232
1,440
16,937
d
815
128
593
8,158
6,550
c
85,755
4,230
c
1,141
Install.
Cost
$ 403
100
100
b
3,957
82
80
-
a
d
ft
l,402e
-
a
a
687
a
a
-
~*
Total
Acq.
Cost
$9,868
3,614
3,719
b
8,412
7,673
2,312
1,440
a
d
815
1,530
593
a
a
c
a
a
c
1,141
C Due to numerous problems with the Perkin-Elmer instrument and detector, this unit was eventually traded
evenly to ORSANCO for the Varian 3700 GC and integrator.
Purchased by ORSANCO for CWW use under the Early Warning Organics Detection System (EWODS), used also
for grant work.
e In order to store water samples at ambient water temperatures, CWW fabricated water bath equipment
from picnic coolers and tubing. Considerable plumbing was necessary to make unit operational.
-------
graphs manually was cumbersome, costly and did not produce the graphs in
sufficient quantity on a timely basis. Even output reports showing -related
data on the same listing were not possible.
Considerable difficulty would be encountered later in the project. For that
reason other possibilities were considered. A local computer center,
Southwestern Ohio Regional Computer Center (SWORCC), in conjunction with the
USEPA, was exploring the feasibility of using graphic computer software.
Later, under a contract with another local firm, computing capabilities were
developed at the University of Michigan. By acquiring a graphics terminal,
the CWW was able to access its data (stored at the University of Michigan)
and produce the outputs required. Table 7 lists the equipment purchased as
part of the cooperative agreement.
TABLE 7. GRAPHICS EQUIPMENT
Tektronix
Equipment Model tt Cost
Graphics Computer 4051 $ 9,892
Printer 4642 2,609
Plotter 4662 4,621
Accessories — 703
$17,825
39
-------
SECTION 5
METHODS AND OPERATING PROCEDURES
PHYSICAL PLANT
Operational parameters were standardized to establish, as near as
possible, constant or reproducible conditions to generate valid data. These
parameters were determined by experimentation and/or equipment manufactur-
ers' recommendations. Tables 8 and 9 contain actual operating conditions
for the full and pilot scale systems. Nominal operating parameters are
described in the following subsections.
GAC Filters
The most significant deviation from standard plant operations was the
reduction of the backwash rate from 5.3 1/min/sq m (15 gpm/sq ft) to
2.8 1/min/sq m (8 gpm/sq ft) to reduce GAC losses. During experimentation,
backwash rates that approached standard plant operation resulted in observ-
able GAC losses. These and other GAC filter operating parameters appear in
Table 10.
Pilot Columns
Pilot GAC Filters —
During Phase 2, two 7.6 cm (3 in) pilot GAC filters were used, one
contained virgin GAC, the other contained regenerated GAC. During Phase 3,
only one column was in service. Initially the column contained virgin GAC
and on subsequent runs regenerated GAC. This column always matched GAC
Filter ISA. Operating parameters for both phases are shown in Table 11.
Pilot GAC Contactors--
During Phase 2, two sets of 10.2 cm (4 in) diameter pilot contactors
were on line, one contained lignite GAC, the other contained bituminous GAC.
During Phase 3, one bituminous GAC pilot contactor was on line. The operat-
ing parameters for the two pilot-scale units are contained in Table 12.
On the bituminous pilot GAC contactor, the second, third and forth
tubes in the series had limited freeboard which required reduction of back-
wash rates in order to reduce bed expansion. Except for the fact that the
lignite pilot contactors had longer tubes with sufficient freeboard to
40
-------
TABLE 8. ACTUAL OPERATING CONDITIONS FOR FULL-SCALE GAC SYSTEMS
GAC Phase-
System3 Run3
15AE
15AE
15AE
15AE
15AE
19AE
21AE
21AE
21AE
21AE
23AE
AE
AE
BE
BBE
BBE
CE
CE
DE
DE
DE
3-0
3-1
3-2
3-3
3-4
1-0
1-0
3-1
3-2
3-3
1-0
3-0
3-1
3-0
3-0
3-1
3-0
3-1
3-0
3-1
3-2
GAC ,
Depth
m
0.76
0.75
0.75
0.77
0.74
0.46
0.76
0.77
0.77
0.74
0.76
4.61
4.55
4.58
4.55
4.60
4.59
4.60
4.59
4.57
4.61
GAC
Weight
kg
43,185
44,500
47,719
47,525
47,689
26,655
44,418
47,715
48,939
49,133
50,173
19,936
19,659
19,922
19,659
19,880
19,811
19,880
19,825
19,756
19,936
Water
Thruput
mil 1
3,407
852
814
1,221
3,709
13,854
14,725
814
1,779
3,558
16,390
1,210
393
870
427
504
855
355
778
401
485
Length
of Run
days
180
45
43
64
196
732
778
43
94
188
866
317
103
228
112
132
224
93
204
105
127
Avg
EBCT
min
7.56
7.43
7.40
7.60
7.30
4.53
7.56
7.63
7.66
7.31
7.56
15.28
15.20
15.37
15.32
15.37
15.32
15.26
15.42
101
101
101
101
101
101
15.41 300
15.20 300
300
300
300
300
300
300
300
300
3 For description of GAC System and Phase-Run refer to Appendix A.
GAC depth in contactors was measured from centerline of bottom laterals.
Based on flow controller settings of 3.8 mil Ipd (1 mgd)for contactors and
,18.9 mil Ipd (5 mgd) for GAC filters.
Actual run time, not including the time the units were out of service for
backwashing, etc.
41
-------
TABLE 9. ACTUAL OPERATING CONDITIONS FOR GAC PILOT COLUMNS
GAC Phase-
System Run
RV3
RV3
RV3
RV3
RV3
RV5
RV5
RV5
2-1
2-2
3-1
3-2
3-3
2-1
2-2
3-0
GAC
Depth
m
0.76
0.78
0.76
0.76
0.76
0.76
0.78
0.76
Water, Length
45
46
55
78
73
65
51
1.41
Thruput
kl
66.5
59.8
30.2
27.4
41.5
67.3
62.8
116.3
of Run
days
135
128
46
43
64
135
128
180
Avg
EBCT
min
10.18
10.87
55
83
7.69
10.22
10.37
7.59
Avg
Hyd.Ldg.
1pm/sq m
73.9
71.5
101.0
97.6
99.3
74.8
75.0
100.6
PFA5
PFA5
PFA5
PFA5
1-0
3-0
3-1
3-2
4.57
4.57
4.57
4.57
18.09
18.24
18.09
16.92
864.0
758.3
356.8
380.3
237
210
106
128
14.70
14.80
15.80
16.40
311.9
309.1
289.5
279.0
PFA9
1-0
4.57
14.43
889.3
237
14.30
321.2
, For description of GAC System and Phase-Run refer to Appendix A.
Based on meter or counter readings.
Actual run time, not including the time units were out of service for
backwashing, etc.
TABLE 10. NOMINAL OPERATIONAL PARAMETERS FOR GAC FILTERS
Hydraulic Loading Rate
Contact Time (Empty Bed)
Backwash Rate (Max.)
Frequency of Backwash
Backwash Time
Backwash Water Used
102 1/min/sq m (2.5 gpm/sq ft)
7.48 min
327 1/min/sq m (8 gpm/sq ft)
1.8 m (6 ft) head loss
12 min/wash
378,500 I/wash (100,000 gal/wash)
TABLE 11. NOMINAL OPERATIONAL PARAMETERS FOR PILOT GAC FILTERS
Hydraulic Loading Rate
Contact Time (Empty Bed)
Backwash Rates
Frequency of Backwash
Backwash Time
Backwash Water Used
Static Head Above Carbon Bed
Head loss not measured.
102 1/min.sq m (2.5 gpm/sq ft)
7.6 min
745 1/min/sq m (18.3 gpm/sq ft)
3 times/wk
20 min/wash
75.7 I/wash (20 gal/wash)
2.5 m (9 ft)
42
-------
TABLE 12. NOMINAL OPERATIONAL PARAMETERS FOR PILOT CONTACTORS
Hydraulic Loading Rate
Contact Time, Empty Bed
Backwash Rates, Bituminous
Tube 1
Tube 2
Tube 3
Tube 4
Backwash Rates, Lignite
Tube 1
Tube 2
Tube 3
Tube 4
Backwash Water, 10 min
Backwash Water, 20 min
Backwash Criterion
302 1/min/sq m (7.4 mgd/sq ft)
15 min
409 1/min/sq m (10 gpm/sq ft)
192 1/min/sq m (4.7 gpm/sq ft)
192 1/min/sq m (4.7 gpm/sq ft)
192 1/min/sq m (4.7 gpm/sq ft)
409 1/min/sq m (10 gpm/sq ft)
409 1/min/sq m (10 gpm/sq/ft)
409 1/min/sq m (10 gpm/sq ft)
409 1/min/sq m (10 gpm/sq ft)
189 1 (50 gal)
378 1 (100 gal)
138 kPa (20 psig)
3 Not required during some runs of Phase 3, see text for modifications.
accomodate full backwash rates on all four tubes, all operating parameters
for the two pilot contactor systems were the same.
Initially in Phase 2, an arbitrary 172 kPa (25 psig) head loss across a
pilot contactor was selected as the criterion for backwashing. On the first
attempt to backwash a pilot contactor, it became apparent that the selected
head loss was too great. The GAC had been compressed into a plug which was
difficult to break up. The criterion was revised to 138 kPa (20 psig) back
pressure, which proved to be acceptable. Eventually the gauge ports became
clogged with GAC and subsequently the tubes were backwashed when throughput
diminished.
The contactors had an effective depth of 4.6 cm (15 ft) from the center-
line of the bottom laterals. Below the bottom laterals was an additional
3.8 cu m (136 cu ft) volume of GAC (Figure 12).
Contactors
The backwash parameter of 138 kPa (20 psig) pressure drop across the
bed was never attained on the contactors during Phase 3 through normal
operation. The one occasion which required backwashing was due to air
binding following a process pump failure. All contactors were backwashed
prior to being put into service in order to purge fines which might clog the
lateral screens or cause turbidity problems. Each contactor was backwashed
again prior to removing the GAC for regeneration. The purpose of the final
wash was to loosen the bed to facilitate hydraulic movement of the GAC and
to relax and level the bed for accurate measurements for determing GAC
43
-------
losses. A turbidimeter was installed on the combined contactor effluent
line.
A standardized backwash procedure, designed to obtain a reproducible
bed volume for measurement purposes, was recommended by the GAC supplier.
This method, described below, was utilized for all backwashing operations.
Backwash was started at a rate, of 122 1/min/sq m (3 gpm/sq ft) then
gradually increased to 407 1/min/sq m (10 gpm/sq ft) and held there for
thirty minutes. The rate was then gradually reduced to zero and the bed
permitted to settle for five minutes. Backwashing was again started and
gradually reduced to zero and the bed permitted to settle for five minutes.
Backwashing was again started and gradually increased to 407 1/min/sq m
(10/gpm/sq ft), held for fifteen minutes and then gradually reduced to zero.
Backwashing and other nominal operating parameters are shown on Table 13.
Regenerator
The regenerator was controlled by a Bristol UCS 3000 microprocessor.
Nine pneumatic control loops were provided for the process:
1. regenerator bed temperature,
2. air fuel ratio,
3. steam flow (total gas flow to combustion chamber),
4. incinerator gas temperature,
5. incinerator off-gas oxygen content,
6. gas flow rate through the dryer,
7. dryer bed temperature,
8. dryer pressure,
9. scrubber's venturi pressure drop.
Start-up--
The microprocessor was loaded with one of two diskette programs, natural
gas or fuel oil. The dryer was charged with 63.5 kg (140 Ibs) of virgin
GAC. Prior to igniting the primary burner, the furnace was purged with air
for two minutes. On completion of the purge cycle, the pilot light was
ignited. The fuel blocking valves were opened to light the primary burner
(Figure 13). The temperature in the regenerator bed section was raised to
TABLE 13. NOMINAL OPERATIONAL PARAMETERS FOR CONTACTORS
Hydraulic Loading Rate 302 1/min/sq m (7.4 gpm/sq ft or 1 mgd)
Contact Time (Empty Bed) 15.3 min
Backwash rates (Max) 409 1/min/sq m (10 gpm/sq ft)
Frequency of Backwash 138 kPa (20 psig) head loss
Backwash Time 45 min
Backwash water per wash 143,100 1 (37,800 gal)
44
-------
260°C (500°F) at which time steam was admitted to the combustion chamber.
When the temperature in the incinerator section reached 370°C (700°F), the
incinerator pilot was lit, the fuel blocking valves were opened and the
incinerator burner was lit.
After the regenerator bed section reached 815°C (1,500°F) and enough
heat was transferred from the incinerator section to the dryer section to
require 2.6 to 3.4 1/min (0.7 to 0.9 gpm) water flow through the dryer spray
cooling water meter to maintain 150°C (300°F) in the dryer bed, GAC feed was
started at approximately 63.5 kg (140 lbs)/hr rate. The GAC feed rate was
increased, in increments of 77 kg (170 lbs)/hr at intervals of thirty minutes.
As the GAC feed rate was increased the temperature was increased in the
incinerator section to maintain 0.76 to 1.5 1/min (0.2 to 0.4 gpm) cooling
water flow to the dryer to maintain proper dryer bed temperature. Approxi-
mately four hours elapsed from start-up to full-feed rate.
The degree of regeneration was determined by apparent density and
iodine number analyses which were performed on three samples daily. Apparent
density was the primary controlling factor and dictated regenerator bed
temperature adjustments, in 5.5°C (10°F) increments, as needed. Furnace
operations were fine tuned using iodine number results. The system tempera-
tures were allowed to stablize after each incremental temperature increase
with the maximum temperature of 1,090°C (2,000°F) in the combustion chamber.
Another factor limiting the amount of spent GAC that could be fed to the
regenerator was the dryer temperature. If the spent GAC contained excessive
moisture, the feed rate was reduced to maintain 150°C (300°F) in the dryer.
See Table 14 for operational parameters.
All control loops of the regenerator were designed to operate in the
automatic mode. Manual control of all control loops, except primary air and
steam, was possible when the need arose.
At times, gas flow to the dryer was restricted due to the holes in the
dryer plate becoming clogged. The holes were cleared by drilling them out
with a 3.2 mm (1/8 in) drill bit. At times the seal between the incinerator
and dryer section leaked, the hot gases bypassing the dryer plate, thereby
reducing bed fluidization. When this condition progressed to the point
where the control valve in the dryer off-gas line reached 100% open, the
gasket between the incinerator and dryer section had to be replaced. When
the seal between the combustion chamber and regenerator bed section deterior-
ated, the hot gas flow bypassed the regenerator plate necessitating replace-
ment of this gasket.
When a GAC filter was being regenerated, sand and gravel entered the
dryer and regenerator sections due to the inefficiency of the sand separator.
The sand and gravel migrated to the bottom of the fluidized bed thus insulat-
ing the GAC in the upper part of the bed from the fluidizing gas flow.
After approximately 120 hours, the regenerator had to be shut down and the
sand and gravel removed from the dryer and regenerator plates.
45
-------
TABLE 14. NOMINAL GAC REGENERATOR PARAMETERS
10.1
30.5
6.4
7.6
5.1
14.2
7.6
-2.5
3.3
18.4
5.1
10.1
to
to
to
to
to
to
to
to
to
to
to
to
20
48
8
17
15
16
15
-7
4
20
30
22
.3
.3
.9
.8
.2
.1
.2
.6
.6
.4
.5
.9
Feed Rate (Spent GAC) 265
Regenerated GAC Output 231
Regenerator Distributor
Plate DPa
Combustion Chamber Pressure
Regenerator Fluid Bed DP
Incinerator Pressure
Dryer Plate DP
Gas Flow Rate
Dryer Bed DP
Dryer Pressure
Dryer Off-Gas Annubar
DP (Off-Gas Flow Rate)
Venturi Scrubber DP
Tray Scrubber DP
Combustion Chamber
Temperature 982 to 1,093
Regenerator Bed Temperature
(at Discharge Port) 788 to 843
Incinerator Temperature 677 to 871
Dryer Bed Temperature 149 to 163
Combustion Chamber Gas
Analysis 0.4% to
Incinerator Gas Analysis 4.0% to
Regenerator Fuel Oil
Flow Rate
Exhaust Gas Annubar DP 0.4 to
(Exhaust Gas Flow Rate) 12.7 to
Regenerator Natural Gas
Flow Rate 0.4 to 0.5
Regenerator Combustion Air 0.4 to 0.5
Steam 136 to 181
Regenerator Total
Fluidizing Gas Flow Rate0 7.1 to
Incinerator Natural Gas
Flow Rate 0 to 0.5
Incinerator Combustion Air 6.8 to 9.9
Incinerator "Flame"
Temperator 816 to 982
Dryer Off-Gas Temperature 121 to 177
Off-Gas Temperature
Venturi Scrubber Inlet 316 to 538
Venturi Scrubber Outlet 93 to 204
Tray Scrubber Outlet 43 to 82
Regenerator Combustion
Air/Fuel Ratio
Natural Gas 8.3 to 9.1
Fuel Oil 8.6 to 9.1
Incinerator Fuel Oil 13.2 to 26.5
kg (585 lbs)/hr
kg (510 lbs)/hr
cm (4 to 8 in) H_0
cm (12 to 19 in) HO
cm (2.5 to 3.5 in)TI 0
cm (3 to 7 in) HO
cm (2 to 6 in) KO
cu m/min (500 to 570 SCFK)
cm (3.0 to 6.0 in) HO
cm (-1.0 to -3.0 in)TI 0
cm (1.3 to 1.8 in) HO
cu m/min (650 to 720 SCFM)
cm (2 to 12 in) HO
cm (4 to 9) HO
°C (1800 to 2000°F)
°C (1450 to 1500°F)
°C (1250 to 1600°F)
°C (300 to 325°F)
0.0% excess 0
2.0% excess 0^
25.7 1/hr (6.8 GPH) + 10%
0.8 cm (0.15 to 0.3 in) HO
17.0 cu m/min (450 to 600 SCFM)
cu m/min (14 to 18 SCFM)
cu m/min (140 to 180 SCFM)
kg (300 to 400 lb)/hr
8.5 cu m/min (250 to 300 SCFM)
cu m/min (0 to 16 SCFM)
cu m/min (240 to 350 SCFM)
°C (1500 to 1800°F)
°C (250 to 350°F)
°C (600 to 100°F)
°C (200 to 400°F)
°C (110 to 180°F)
1/hr (3.5 to 7.0 gpm)
, Differential pressure.
Set point is determined by fuel-air ratio and oxygen trim controllers
in software.
Set point in determined by total fluidizing gas flow controller in
software.
-------
Shut Down--
Prior to shutting the regenerator down, the GAC bed remaining in the
regenerator section was gasified. This was accomplished by extinguishing
the incinerator burner, raising the temperature in the regenerator bed
section to 870°C (1,600°F) and increasing the oxygen to the combustion
chamber. Under these conditions the GAC bed was gasified (burned away).
After the bed was gasified, the temperature in the regenerator section
was lowered gradually to 580°C (1,000°F) at which time the primary burner
was extinguished. When the temperature in the regenerator bed section
reached 260°C (500°F), the GAC remaining in the dryer was removed by vacuum-
ing.
Purging--
If the regenerator was shut down and restarted at temperatures above
260°C (500°F) (hot restart), the regenerator was purged with steam. At
temperatures below 260°C (500°F) air was used for purging.
GAC Loss Measurements—
Losses were generally measured "bed-to-bed," that is, a measurement was
taken before GAC was removed for regeneration and after the regenerated GAC
was placed back in the bed. In addition, attempts were made to measure
losses due to transporting the GAC and those that occurred only within the
furnace battery limits. The furnace battery limits included everything
between the spent carbon storage tank and the regenerated carbon storage
tank (Figure 13). All loss measurements were based on volume. The proced-
ures for the various measurement technigues appear below.
GAC filter measurement--
1. Close the filter drain.
2. Open the backwash supply valve slowly in order to level the GAC evenly
over the filter area without discharging GAC into the washwater troughs.
3. Backwash the filter, described previously.
4. After the backwashing, drain the filter until the water level is below
the GAC surface.
5. Place a straight edge across the top of the washwater trough and measure
the distance from the straight edge to the GAC surface. Take three
measurements, one on each side and one in the middle of each of the
eight filter openings for a total of 24 measurements.
47
-------
Contactor measurement--
1. Backwash contactor.
2. Stop the backwash pump and allow the water in the contactor to drain
below the top laterals. Release vacuum.
3. After vacuum is released from the contactor, open the observation port.
4. Lower the tape measure into the port until the attached plate rests on
the GAC surface. Measure the distance the plate dropped, using the top
of the port flange for a reference point.
5. Measure at 4 points around the circumference of the port and take one
measurement at the center of the port. Average the measurements.
6. Subtract this measurement taken from 7.1 m (23.2 ft). This represents
the number of feet of GAC above the bottom laterals.
GAC loss measurement for furnace battery limits—Two methods were
employed depending upon the operational status of the furnace.
Furnace not operational--
1. Transfer GAC from a contactor or a filter to spent storage tank (ST1) .
2. Prior to beginning the regeneration process, measure the contour of GAC
(and possibly sand) in STl with a leveling rod at 30.5 cm (1 ft) incre-
ments along a square grid placed on top of STl. The volume of GAC was
determined using the following method referred to as the "borrow-pit"
method. The horizontal and vertical lines form a grid of square and
triangular sections. The squares have an area of 930 sq cm (1 sq ft)
and the varying triangular sections around the circular tank wall have
areas which were computed individually. The depth to the GAC surface at
each section's vertex is averaged and multiplied by the section area to
compute the volume of carbon (volume of tank - calculated volume -
carbon volume). All individual sections are then summed to obtain the
total volume of GAC in the tank. Also prior to regeneration, ensure
that an empty sand hopper is in place.
3. Regenerate the GAC in STl as a batch until STl is empty and the regener-
ated storage tank (ST2) is full.
4. Place the furnace on idle heat and measure the contours of ST2 in the
same fashion that STl was measured.
5. Determine the volume of sand in the sand hopper.
6. Determine the volume of GAC in STl and ST2 according to the "borrow-pit"
method described earlier.
48
-------
Furnace operational--
1. Transfer GAC from a contactor or filter to STl.
2. Regenerate the GAC in STl until the edge of the GAC in ST2 is above the
conical section of ST2 (in the cylindrical portion).
3. With the furnace operating, measure the contours of STl and ST2 using
the grids and leveling rod identified above. At the start of these
measurements, an empty sand hopper should be placed in the regeneration
system.
4. Regenerate the GAC in STl until the level of STl is just above the cone
at the sidewall.
5. With the furnace operating, again measure STl and ST2 as before.
6. Determine the sand volume in the sand hopper.
7. Determine the volume of GAC in STl and ST2 according to the "borrow-pit"
method described earlier.
Transport loss measurement--
1. Remove gate from the rotary valve at the bottom of STl (Figure 12).
2. Bypass the dewatering screw, furnace and quench tank by running the
discharge from the eductor on STl to the GAC inlet line on ST2.
3. Backwash and measure the GAC level in the contactor.
4. Transfer as much GAC as possible from the contactor to STl.
5. Transfer the GAC from STl to ST2.
6. Transfer the remainder of the GAC from the contactor to STl.
7. After the contactor has been flushed, add water to cushion the returning
GAC.
8. Transfer GAC from ST2 to the contactor.
9. When transfer is complete, backwash and measure the GAC depth.
GAC loss calculation—Formulas used to calculate GAC losses from meas-
urements taken according to procedures described above are listed below
along with definitions of terms.
49
-------
Exhausted Bed Volume (EBV)
Regenerated Bed Volume (RBV) -
Volume of Materials
Removed (VMR)
Sand Volume Removed (SVR)
Volume in cu ft of GAC in
system just before regeneration,
based on the measured GAC depth
after backwashing.
Volume in cu ft of GAC in
system after regeneration, based
on the measured GAC depth after
backwashing.
In a contactor, the same as the
Regenerated Bed Volume. In a
filter, the difference in cu ft
between the freeboard measurements
before and after removing GAC (and
some sand) for regeneration.
Volume in cu ft of
sand discharged from the sand
separators.
01 = original interface ( in), the distance between the top of the washwater
trough and the bed surface before GAC added.
EF = exhausted freeboard ( in), the distance between the top of the washwater
trough and the bed surface before materials removed for regeneration.
NI = new interface ( in), the distance between the top of the washwater
trough and the bed surface after materials removed for regeneration.
RF = regenerated GAC freeboard ( in), the distance between the top of the wash-
water trough and the bed surface after regenerated GAC returned to
filter.
GAC filter
Exhausted
Bed Volume
GAC filter
Volume of
Materials
Removed
GAC filter
Regenerated
Bed Volume
GAC Filter Bed
Volume Loss (%)
Contactor
Volume
of Materials
Removed
EBV = (01 - EF) (1400/12) = cu ft
VMR = (NI - EF) (1400/12) = cu ft
RBV = (NI - RF) (1400/12) = cu ft
BVL = (VMR - RBV - SVR) 100
VMR-SVR
VMR = EBV = (EF) (95) + 136 = cu ft
50
-------
Contactor RBV = (RF) (95) + 136 = cu ft
Regenerated
Bed Volume
Contactor BVL = (VMR - RBV) 100
Bed Volume VMR
Loss (%)
Note: Multiplier to convert cubic feet to cubic meters is 0.028.
LABORATORY
Sampling Plan Overview
Throughout this and later sections, reference will be made to phase and
run. Phases have been previously explained in the Introduction. The term
"run" is used to describe the period of time that a GAC system was on line
between start-up and regeneration. A numbering system was used to describe
any particular run, e.g. 3-0 or 3-1. The "3" identifies the phase. The "0"
or "1" identifies the run, with "0" meaning virgin GAC and "1" meaning once
regenerated GAC. The dictionary contained in Appendix A describes other
combinations.
Once the organics laboratory was established and three sand filters
were converted to GAC as detailed in Section 4, the project was ready to
begin. A sampling and analytical plan was establishedand the GAC filters
were put into operation. The locations of the sample points utilized for
the study are indicated in Figures 5, 7, 9, 14 and 15. The first two phases
of the project, in addition to accomplishing specific objectives also
provided experience which helped make the most important third phase even
more successful.
Phase 1: Full-Scale GAC Filters --
This initial phase of the work was dedicated to evaluating various GAC
filter configurations with respect to both GAC depths and GAC types.
The runday schedule for this phase can be found in Table 15. Sample
point locations and sample plan overview can be found in Figures 14 and 15
and in Table 16. Samples were often collected on a more frequent basis than
listed in Table 16. For a more precise sampling schedule, refer to the data
listing located in Volume 3.
Phase 2: Pilot Scale GAC Systems--
Pilot GAC Filter—In this portion of Phase 2, pilot GAC filters were
utilized to compare virgin bituminous-based GAC to regenerated bituminous-
based GAC. The comparison extended through two regeneration cycles. Glass
columns 7.6 cm (3 in) in diameter were used. Each contained 76 cm (30 in)
51
-------
gAVPLF, LOCATION
®RAW WATER
SCHEM. EAST (SETTLED)
MO-BASIN CLARIF1ERS
©FILTER INFLUENT
©GAC RLTER
(§) CONTACTOR INFLUENT
©COMTACTOR
(DCLEARWELL (Tor G)
©PORT
© DISTRIBUTION SYSTEM
NTAKE
RER
SLBMHRGED
N.ET
Figure 14. Treatment process sample point locations
52
-------
FILTER INFLUSNT
1
*
M: FILTER-' 2
3CTCAC-W2
I
1
1
1 I
j— j_
1-
3A
OiSQ
•ILTE«-«2IA
3 COARSE
IFF.-I
SAMPLE
i— n
i
2
*o
t'^H
o '
FILTER-' ISA
SANO
3 SANO 3
EFK J EFF J
1— -
LZ
SAMPLE PORTS
^
i
z
FILTER-USA
SO'CAC-CB»4O
3
JEFF.-)
POINT DEPTH
_OCATION DEPTH
1 .30M
2 .46M
3 .76 M
EFF .76M
(12")
(18")
(30")
(30")
«*, -. '"-".i;^,,
CONTACT
Tl
2.99
MEjfaM
4.49
7.48
7.48
* Saiftt Polnti
CARBON FILTERS SAMPLE COLLECTION
j
4 -
5 .
® 7-
9 -
II -
>^ ^
4.
i
(D
'^^ ^-
Uud
RLTER
FLOORT
^y
POINTS
CONTACTCH INFLUENT
(SANO FILTER EFFLUENT)
'•-
• -
3-
1-
^ 5-0
^ 9- CC
II -0
i-
Lul
1
4
/^\ 5
^ 9
II
{-EFF [-EFF -££F____^
To Cleorwell
(
JS.
Sample Points
SAMPLE POINT DEPTH
LOCATION
4 '
5
7 *
9 *
II
EFF '
Loading Rat*
DEPTH CONTACT
.87M (34.2")
ISM (51.5")
2.2M (848")
32M (1272")
4.IM (1628")
4.6M (I80O")
2
4.
TIMECmin}
83
36
7.17
10.77
13.77
15.23
*°°^$^>, ' «"-»»
Poinlt IH.d
3AC CONTACTORS SAMPLE COLLECTION POINTS
Figure 15. Full-scale GAC system sample point locations
53
-------
TABLE 15. PHASE, RUN, RUNDAY SCHEDULE
GAC
a
System
ISA
ISA
ISA
ISA
ISA
19A
21A
2 LA
21A
21A
23A
A
A
B
BB
BB
C
C
D
D
D
PFA5
PFA5
PFA5
PFA5
PFA9
RV3
RV3
RV3
RV3
RV3
RV5
RV5
RV5
For description
Phase-
a
Run
3-0
3-1
3-2
3-3
3-4
1-0
1-0
3-1
3-2
3-3
1-0
3-0
3-1
3-0
3-0
3-1
3-0
3-1
3-0
3-1
3-2
2-0
3-0
3-1
3-2
2-0
2-1
2-2
3-1
3-2
3-3
2-1
2-2
3-0
of GAC system
Run
Start
Date
01/14/80
08/11/80
10/20/80
12/22/80
03/23/81
02/14/78
02/14/78
07/28/80
10/06/80
02/09/81
02/14/78
10/01/79
09/15/80
10/29/79
06/23/80
11/03/80
12/17/79
08/11/80
01-14/80
09/01/80
12/29/80
10/24/78
01/14/80
09/01/80
12/29/80
10/24/78
10/24/78
05/14/79
08/11/80
10/20/80
12/22/80
10/24/78
05/14/79
01/14/80
and Phase -Run
Run
Stop
Date
07/14/80
09/26/80
12/02/80
02/24/81
10/07/81
03/21/80
04/22/80
09/09/80
01/12/81
09/04/81
10/07/80
08/19/80
12/30/80
06/17/80
10/14/80
03/16/81
07/29/80
11/12/80
08/07/80
12/16/80
05/06/81
06/18/79
08/07/80
12/16/80
05/06/81
06/18/79 •
03/09/79
09/14/79
09/26/80
12/02/80
02/24/81
03/09/79
09/14/79
07/14/80
Length
of Run
days
181
46
44
64
198
766
799
43
98
207
967
323
106
231
113
133
225
93
206
106
128
237
206
106
128
237
137
123
46
44
64
137
123
182
refer to Appendix A.
54
-------
TABLE 16. PHASE 1 SAMPLE PLAN OVERVIEW
a
TOC
1
1
for*
2
2
2
2
2bc
2bC
2
_bc
2
1
1
1
ITTT
1
1
2 ,
d
1
1
2
ld
1
ld
2 T
ld
^
1^
2
1
1
1
FTTT
1
1
2g
e
1
1
2g
-Ld
I
leg
le
1
I6
2g
1
1
1
STT7
1
1
f
1
1
j
•^f
-^
jf
1.
lf
j
1^
1
1
1
-
Purg
Non
Halo
lh
_
lh
-
1
lh
lh
lh
~h
1
—
lh
-
-
-
B/N
Extr TEMP
i 2
2
i 2
- -
-
i
— "1
1
1
i 1
- -
_ —
i
1
2
-
DO
1/mo
-
2
1
2
2
2
2
2
2
2
2
2
2
-
-
-
PH
2
2
2
.. c
1
1
2
1C
1
1C
2
1C
1
lc
2
2
2
-
TURB
2
2
2
—
-
2
-
_
-
2
-
—
-
2
2
2
-
C12
Free
Comb
Tot
„
-
2
., c
1
1
2
1C
1
1C
2
1C
1
1C
2
2
2
-
ODOR
2
-
2
""
2
-
-
-
2
-
-
—
2
-
-
2
-
Bact
SPC
and
TCOL
1
-
2
""
2
2
-
-
2
2
-
—
2
2
-
daily
-
TSOL
and
TSF
1
-
1
~
-
-
-
-
-
-
-
*"
-
-
-
1
-
Location
RAW
SETT
FLIN
19A1
19A2
19AE
21A1
21A2
21A3
21AE
23A1
23A2
23A3
23AE
SFEF
CT
DIST
Samples of frequencies in table are "per
week" except where otherwise noted. For
explanation of abbreviations refer to
Appendix A.
Samples collected I/week after 3/28/78.
Sampling discontinued after 5/9/78.
Sampling discontinued after 3/21/78.
Sampling discontinued after 3/7/78.
Sampling discontinued after 3/14/78.
g Samples reduced to I/week after 4/25/78.
Samples collected and stored until a contract was
secured with the University of Cincinnati.
Unfortunately, few samples were analyzed once the
contract was secured due to problems encountered
using the USEPA recommended method.
1 Samples collected 3/3/78, 5/23/78, and 6/20/78.
Only the 3/3/78 sample was analyzed.
-------
of Westvaco Nuchar WVG 12 x 40 GAC and had an empty bed contact time (EBCT)
of 7.8 min at an hydraulic loading of 1.7 Ips/sq m (2.5 gpm/sq ft). The
sample ports were referred to as Regenerated/Virgin (RV) for identification
of samples.
Pilot Contactors—In this portion of Phase 2, twin pilot contactors
were operated in parallel to compare bituminous-based GAC to lignite-based
GAC. One column contained 4.57 m (15.0 ft) of Westvaco Nuchar WVG 12 x 40
bituminous-based GAC. The other contained 4.57 m (15.0 ft) of ICI
Hydrodarco 1030, a lignite-based GAC. Both columns were 10.2 cm (4 in)
diameter and had an EBCT of 16 min at a hydraulic loading of 0.78 Ips/sq m
(7 gpm/sq ft). Both columns had ports located at various depths for sampling
purposes, referred to as Post Filter Adsorbers (PFA) for identification of
samples.
The runday schedule for Phase 2 can be found in Table 15. Sample point
locations can be found in Figures 5, 7, and 9 and the sample plan overview
in Table 17. Samples were often collected on a more frequent basis than
detailed in Table 17. For a more precise sampling schedule, refer to the
data listing located in Volume 3.
Phase 3: Pilot and Full-Scale Contactors and GAC Filters With On-Site
Regeneration--In this phase, contactors operated simultaneously with GAC
Filters 21A and ISA. Also during this phase, a pilot GAC filter was run in
parallel with GAC Filter ISA and a pilot contactor in parallel with
Contactor D.
Four contactors, A, B, C and D, were constructed. Each had a rated
capacity of 0.04 cu m/s (1 mgd) at a hydraulic loading of 5.03 Ips/sq m
(7.41 gpin/sq ft). The contactors received effluent water from the normal
sand filters (SFEF). Each contactor contained 4.6 m (15 ft) of Westvaco
Nuchar WVG 12 x 40 GAC and had an EBCT of 15.2 min. Filters 21A and ISA
contained the same type and grade of GAC as the pilot units with each having
an EBCT of 7.5 min. After Phase 3-0, the spent GAC from Contactor B was
removed and replaced with virgin GAC and subsequently referred to as
Contactor BB.
During Phase 3 an attempt was made to maximize the use of currently
available organic techniques rather than restricting the monitoring to the
small number of currently regulated contaminants included in the Interim
Primary Drinking Water Regulation . This approach provided a broader assess-
ment of the usefulness of GAC for water treatment for organic
contaminants. Specifically, the purpose in performing the broader scope of
analyses was three fold:
1. A data base of the occurrence of numerous specific organic compounds
was provided that was useful in defining whether a problem substance
occurred with a sufficient degree of frequency and concentration to
merit concern.
56
-------
TABLE 17. PHASE 2 SAMPLE PLAN OVERVIEW
Location
RAW
SFEF
PFA2
PFA3
PFA4
PFA5
PFA6
PFA7
PFA8
PFA9
FLIN
RV3
RV5
DIST
Purg
Non B/N
TOC ITTT FTTT STT7 Halo Extr TEMP DO PH
1/mo I/day 1/mo 1/mo 1/mo
2, 2, 1, 1, 1/mo
2b 2b lb lb l/moC
2 2 1, 1. l/moC
2 2 1 I3 l/moc
2 2 1, 1, l/moC
2, 2, 1, 1, 1/mo'
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
l/moc
1/mo
c
1/mo
d
d
d
d
1/mo 1/mo
C12
Free
Comb
TURB Tot
1/mo 1/mo 1/mo 1/mo
1111
Bact
SPC TSOL
and and
ODOR TCOL TSF
£k
1/mo 1/mo 1/mo
11-
Sampling frequencies in table are "per week" except where otherwise noted. For explanation of
abbreviations refer to Appendix A.
Sampling discontinued after 5/8/79.
C Samples collected and stored until a contract was secured with the University of Cincinnati.
Unfortunately, very few samples were analyzed once the contract was secured due to problems
, encountered using the EPA recommended method.
Samples collected only once on 5/13/79.
Sampling discontinued after 1/16/79.
-------
2. The degree of removal of many specific compounds achieved by a given
unit process in an ambient environment was determined.
3. The potential for using some easily measured general organic parameters
as operating surrogates in lieu of the more specific, and often more
costly, test could be evaluated.
The runday schedule for Phase 3 can be found in Table 15. The sample
locations can be found in Figures 5, 7, 9, 14 and 15 and the sample plan
overview in Table 18. Samples were often collected on a more frequent basis
than detailed in Table 18. For a more precise sampling schedule, refer to
the data listings located in Volume 3.
Analytical Procedures
Total Organic Carbon--
The TOC was analyzed using a Dohrmann Envirotech DC-54 Ultra Low Level
TOC analyzer. The method used was as described in the DC-54 manual with
the following variations:
1. Double-distilled water was prepared using an AG-3 Corning all glass
still.
2. The instrument blank was determined by recycling a double-distilled
water run.
3. The POC-2 sparger, P/N 511-220, was used because of the presence of
particulate matter in some of the samples.
4. Samples with a TOC value greater than 3.0 mg/1 were rerun using 300 ul
of persulfate reagent per 8.0 ml sample. This provides sufficient
reagent for TOC levels of up to 10.0 mg/1.
The following gases were used: hydrogen, grade zero; helium, grade
five; and air, grade zero.
Purgeable Halogenated Organics--
Three water samples from various locations were collected that were
ultimately analyzed for halogenated organics. Four data elements resulted
from these three samples:
1. The instantaneous trihalomethane (INSTTHM) sample was analyzed to
determine the concentrations of the four THMs, chloroform (CHC1 ),
bromodichloromethane (CHBrCl ), dibromochloromethane (CHBr?Cl) and
bromoform (CHBr_) present at the time of sampling. The summation of
the concentrations of these four individual THMs represented the total
THM concentration. This sample was analyzed for eight other purgeable
-------
TABLE 18. PHASE 3 SAMPLE PLAN OVERVIEW8
Location TOC
RAW 1
SFEF 1
M 1
A7 1.
A9 1J
AE 1
B4 1
B7 1
B9 1
BE 1
BB4 1
BB7 1
BBE 1
C4 1
C7 1
CE 1
D4 1
D7 1
1)E 1
PFA2 1
PFA3 1
PFA5 1
FI.1N \b
21AE T
15AE 1
RV3 1
RV5 1
CT
D1ST 1
Purg.
Non B/N
ITTT FTTT STT3 Halo" Exlr
1 1/mo 1/mo 1/mo
1. 1. 1. l/moC
l' I1 I1 l -
!i 'i J 'i J i J "
,k ,k ,k k , . c
1 1 1 1/mo
1 1 1
1 1 1 - -
1 1 1 - -
1111 l/moC
1111-
1111-
1 1 1 I
. les
iles
>l es
iles
lies
>les
Liles
> les
lU'S
)1 es
It appeared that
CO
CO
no
CO
CO
no
CO
CO
CO
Cl)
no
( 0
lected
lectcd
( ol 1 e<
e< led
ei
I O
t't
«lr
t'<
(•(.
<•-d
CO 1 1 f(
If.-li-il
Irom I'll
only dn
led dur
2/monl h
2/n>onlli
li-.l all
din nif>
I/.1 Wl'(
Mi vi-e
I/ 15/80
ti'.l din
most of the i
ase
ring
1»8
-------
halogenated compounds-. dichlorome thane, carbon tetrachloride,
1,2-dichloroethane, trichloroethylene, tetrachloroethylene, 1,1,1-tri-
chloroethane, chlorobenzene, and o-dichlorobenzene.
2. The simulated distribution THM (THMSIMDIST) sample resulted in an
estimate of the THM concentration which would be found in the distribu-
tion system (with or without GAC treatment).
The THMSIMDIST sample was dosed with excess free chlorine and stored
under controlled conditions. Chlorine dose, pH, temperature and storage
or reaction time were controlled to match those present in the actual
distributed water at the time of sampling. After the specified storage
time, the THMSIMDIST concentration was measured. This measurement
assisted in evaluating the effect of GAC treatment on the actual quality
of distributed water.
3. The maximum THM (THMMAX) sample resulted in a measure of the THM concen-
tration found at the end of a set time period having been held under
controlled conditions. The THMMAX sample was dosed with excess free
chlorine residual and treated similarily to the THMSIMDIST sample
except that the storage conditions were constant and represented the
maximum conditions experienced in the Cincinnati distribution system.
4. THM formation potential (THMFP) was a practical measure of the precursor
concentration, generally regarded as aquatic humic materials, present
at the time of sampling. For this data element determination, the
INSTTHM sample concentration was subtracted from a THMMAX sample concen-
tration.
Sample preparation--
INSTTHM--This sample was collected in a muffled 40 ml screw cap vial in
the presence of sodium sulfite to chemically reduce chlorine residual thus
arresting further formation of THMs.
THMSIMDIST--This procedure involved placing the sample in a 325 ml
glass stoppered amber bottle containing pH 8.2 boric acid - borax buffer and
sufficient calcium hypochlorite solution to result in an additional free
chlorine residual of 2.5 mg/1. The bottles were then stored in an insulated
waterbath for three days. Finished water from the water treatment plant
continuously passed through the waterbath to maintain the temperature of the
samples at plant finished water temperature. After the proper storage
period, the THM reaction was stopped by collecting a sample in a muffled
40 ml screw cap vial containing sodium sulfite. The sample was then analyzed
to determine the THMSIMDIST concentration.
During Phases 1 and 2 the storage or reaction time was seven days.
This seven-day storage time was thought to represent the travel time neces-
sary to reach the remote points of the distribution system. However, a
subsequent test utilizing fluoride as an indicator showed a travel time of
60
-------
three days to the most remote points of the Cincinnati distribution system.
Therefore, during Phase 3, a three-day storage time was used.
THMFP—This procedure involved placing a THMMAX sample in a 325 ml
glass stoppered amber bottle containing pH 9.5 borax-NaOH buffer and suffic-
ient calcium hypochlorite solution to result in an additional free chlorine
residual of 15 mg/1. The bottles were then stored at 29.4°C (85°F) for
seven days after which time the THM reaction was stopped by collecting a
sample in a muffled 40 ml screw cap vial containing sodium sulfite. The
THMFP was determined by subtracting the INSTTHM concentration, determined
earlier, from the THMMAX concentration.
Sample analysis—Purgeable halogenatad organics were measured by the
USEPA purge/trap THM analytic procedure . THM and similar volatile com-
pounds were stripped from a 5 ml water sample with a 40 ml/min stream of
helium for 11 minutes. The organics were collected on a trap of Tenax-GC.
The adsorbed organics were backflushed for six minutes at 200°C (329°F) with
helium at 50 ml/min and collected on a cool, 40°C (104°F) gas chromatographic
packed column. This process was facilitated by the use of a Tekmar LSC-1
liquid sample concentrator and a Varian 3700 GC. The chromatographic column
was a 1.8 m (6 ft) x 2 mm ID glass column packed with 0.4% Carbowax 1500 on
Carbopack-A (80/100 mesh). The chromatogram was then developed by tempera-
ture programming to 175°C (347°F) at a rate of 6°C/min (ll°F/min) and held
for eight minutes. A Hall Model 700 electrolytic conductivity detector in
the halogen specific mode was used for detection.
Purgeable Non-Halogenated Organic Compounds--
Analyses for purgeable non-halogenated compounds were initially per-
formed by the University of Cincinnati (UC) with CWW eventually taking over
the analyses. The purgeable non-halogenated compounds of interest included:
hexane, ethylbenzene, benzene, o-xylene, toluene and 1,2,3,4 -tetrahydro-
napthalene (tetralin).
UC employed a modification of the USEPA purge/trap THM method and
experienced many problems. Unfortunately, the results from the samples
analyzed using this method were not reliable.
The following method was used by UC. The volatile compounds were
stripped from a 10 to 20 ml volume, contained in a 50 ml purge device, with
a 40 ml/min stream of ultra-pure nitrogen for 11 minutes. The organics were
collected on a trap of Tenax-GC. The trap was then placed in the carrier
gas system of a Varian 1700 GC by means of quick-connectors. A heating tape
was wrapped around the trap. The adsorbed organics were backflushed for six
minutes at 230°C (446°F) with ultra-pure nitrogen onto a 60°C (140°F) Chromo-
sorb 101 column. The chromatogram was then developed by rapidly heating the
column oven to 125°C (257°F), holding for six minutes and then temperature
programming to 220°C (428°F) at a rate of 6°C/min (ll°F/min).
61
-------
The following method was used at CWW. The volatile compounds were
stripped from a 5 ml water sample with a 40 ml/min stream of helium for
11 minutes. The organics were collected on a trap of Tenax-GC. The adsorbed
organics were backflushed for six minutes at 180°C (356°F) with helium at 50
ml/min and collected on a 60°C (140°F) gas chromatographic packed column.
This process was facilitated by the use of a Chemical Data System Inc.,
Model 310, concentrator and a Varian 3700 GC. The chromatographic column
was a 1.83 m (6 ft) x 2 mm ID glass column packed with 1% SP 1000 on
Carbopack B (60/80 mesh). The chromatogram was then developed by temperature
programming to 200°C (392°F) at a rate of 10°C/min (18°F/'min) and held for
33 minutes. An FID was used for detection.
Acid Extract Capillary Column GC/FID Profiles--
Organic profiles were generated by a liquid/liquid extraction of 3.8 1
(1 gal) samples adjusted to pH 2 with hydrochloric acid. Fresh volumes of
250, 100, and 100 ml of dichloromethane were shaken consecutively with the
sample in a six liter separatory funnel, combined after separation, dried
with anhydrous sodium sulfate, and reduced in volume by boiling to 0.5 ml by
use of the Kuderna-Danish apparatus. One ug of anthracene (corresponding to
0.25 ;ug/l in the original sample) was added to the final volume after concen-
tration as an internal standard. Blanks corresponding to each batch of
redistilled solvent with internal standard added were also analyzed.
Analysis was initially performed by the USEPA, Cincinnati, Ohio using a
microprocessor controlled GC (H.P. Model 5840) equipped with a splitless
capillary injection system, a 30 m, SP-2100, wall-coated-open-tubular glass
capillary column, and a flame ionization detector. Analysis was eventually
performed at the CWW using a Varian 3700 GC equipped as stated above except
with a fused-silfca capillary column and a Spectra Physics SP4100 computing
integrator.
Gas Chromatograph/Mass Spectrometer (GC/MS) Identification of Acid Extract
Compounds--
The primary purpose of the GC/MS was to attempt to identify and estab-
lish the frequency of occurrence of high molecular weight compounds before
and after full-scale GAC treatment. A secondary purpose was to determine if
typical water utility laboratory personnel are capable of operating and
maintaining a GC/MS.
A Finnigan Organics in Water (OWA) GC/MS was purchased by CWW. This
was new generation equipment and one of the first few sold in the country.
This instrument was selected because of its reasonable capital and operating
costs and because it was advertized as being capable of operation without
the degree of attention required by more sophisticated, high-priced equip-
ment being used in basic research laboratories.
A number of problems were encountered that deferred obtaining useful
results from the instrument:
62
-------
1. Air conditioning in the new laboratory, in which the GC/MS was
housed, was not initially operational. Later it was found neces-
sary to add a supportive unit.
2. Replacement of damaged door frames in the new laboratory required
shutting down and dust proofing of the instrument for several
weeks.
3. Normal software and hardware problems typically experienced with
any new generation equipment required, in the aggregate, many
months of delays awaiting field visits and parts.
4. Operator training was necessary.
5. GC/MS operator turnover occurred at the peak of the project
requiring replacement and subsequent training of a new operator.
6. Upgrading the capillary injector to a Grob-type split/splitless
injector was required to permit a more controlled injection.
In spite of the problems mentioned, CVW experiences with the Finnigan, OWA,
GC/MS would indicate that water utility staff, who are competent in running
and maintaining chromatographic equipment, should be capable of operating a
non-research, less sophisticated GC/MS.
The GC/MS intended purpose was to tentatively identify the peaks observed in
the acid extract, GC/FID profiles. After the GC/FID capillary column profile
was obtained, the acid extract sample was further concentrated to 0.2 ml by
passing helium above the sample vial. Analysis was then performed using the
GC/MS equipped with a splitless capillary injection system and a 30 m SP-2100
wall-coated-open-tubular, fused-silicon capillary column. Four ul of sample
were slowly injected with the splitter off. After 45 seconds, the splitter
was opened. The initial oven temperature of 20°C (68°F) was maintained for
five minutes. The oven was then programmed to 240°C (464°F) at a rate of
2°C/min (3.6°F/min) and held for 30 minutes.
The GC/MS was used to report only tentative identification of comnmmds and
not concentrations. The USEPA evaluated a Finnigan, OWA, GC/MS using a
packed column and reported a lower detection limit of between 10 and 20 ng
decafluorotriphenylphosphine (DFTPP). CWW used a capillary column which
should be superior to a packed column. The lower detection limits for
several phenolic compounds were calculated and found to have a concentration
range of 0.002 - 0.13 ng (Table 19). This lower detection limit, along with
the fact that the GC/MS was not able to identify many of the compounds
observed in the acid extract profiles, should give an indication of the
concentrations present.
Carbon Adsorbable Organohalides (CAOX)--
The CAOX method used for rarbon adsorbable organohalides was described
in detail by Dressman, et al. The CAOX analyses were performed by the
63
-------
TABLE 19. FINNIGAN GC/MS LOWER DETECTION LIMITS
Compound Detection Limit (ng)
2-Chlorophenol 0.002
Phenol 0.004
2-Nitrophenol 0.006
2,4-Dimethylphenol 0.004
2,4-Dichlorophenol 0.002
p-Chloro-m-cresol 0.002
2,4,6-Trichlorophenol 0.002
2,4-Dinitrophenol 0.13
4-Nitrophenol 0.026
4,6-Dinitro-o-cresol 0.025
Pentachlorophenol 0.004
Calculated using capillary column.
USEPA, Cincinnati, Ohio. In general, the organic material was adsorbed on
40 mg of activated carbon from a 100 ml sample to which nitric acid had been
added to adjust to pH 2 thus improving adsorption of organics. Also, sodium
sulfite had been added to chemically reduce the chlorine residual. Adsorp-
tion was facilitated by use of a mini-column assembly and a commercially
available Dohrmann AD2 adsorption module. Following the adsorption step,
the activated GAC was washed with a nitrate solution to remove interference
of chloride. Halide ions were formed by combustion of the sample in a
controlled atmosphere and measured by microcoulometric titration. This
process was accomplished with either a Dohrmann Envirotech 11CTS-20 or MC-1
system. Results were expressed in >ig/L as chloride.
Grob Closed Loop Stripping--
Sample preparation by a variation of the closed loop stripping analysis
(CLSA), first described by Grob, °' followed by glass capillary GC/MS, was
considered to be a sensitive and convenient approach to the measurement of
individual organic compounds in the intermediate volatility range. Thg
exact method used in this study was described in detail by Coleman, et al.
Samples of 3.8 1 (1 gal) were purged at 30°C (86°F) for two hours with their
own headspace gas continuously recirculating. By use of an inline activated
carbon filter, the purged organics were adsorbed from the gas phase.
Adsorbed organics were later desorbed from the carbon with microliter
quantities of carbon disulfide. Aliquots were then analyzed by GC/MS with
results interpreted relative to the appropriate added internal standards.
Base/Neutral Extractables--
Analyses for base/neutral extractable compounds were performed by PEDCo
Environmental, Inc. as a contract service. The samples were analyzed for
64
-------
the following compounds: isophorone; napthalene; bis(2-ethylhexyl)phthalate;
di-n-butyl phthalate; dimethyl phthalate; diethyl phthalate; nitrobenzene;
2,4-dinitrotoluene; 2,6-dinitrotoluene,- aniline; alpha-napthylamine;
butylbenzylphthalate; pyridine; alpha-picoline and 2,4-dimethylpyridine.
PEDCo employed the USEPA method for extraction and concentration of
the samples with the following exceptions.-
1. 1.5 1 of the sample were extracted.
2. Samples were first evaporated to less than 5.0 ml using the Kuderna-
Danish apparatus then further evaporated to 0.5 ml using a stream of
nitrogen.
Subsequent qualitative and quantitative analyses were performed using
either a GC/FID or a GC/MS.
Pesticides/Herbicides/Polychlorinated Biphenyls (PCB)--
The Ohio Environmental Protection Agency (OEPA) was asked to participate
in the project by analyzing 14 samples for pesticides, herbicides and PCBs.
All of these submitted samples showed less than detectable levels for seven
parameters: endrin, lindane, methoxychlor, toxaphene, ?2,4-D, silvex, and
PCBs. The OEPA Industrial Chemical Laboratory used USEPA methods.
Sampling Glassware Preparation for All Organic Analyses--
All glass sample bottles were washed, rinsed, covered with aluminum
foil, and heated to 400°C (752°F) for at least 1/2 hour in a muffle furnace.
Vials and bottles with Teflon-faced, silicone-rubber septa and plastic screw
caps were used for the THM, CAOK, TOC and non-halogenated samples. Bottles
of 3.8 1 (1 gal), employing plastic screw tops with Teflon liners, were used
for the acid extractable organic GC/FID profile, base/neutral extractable
organic, pesticide, herbicide and Grob CLSA samples. Samples were returned
to the laboratory and stored at 4°C (39°F) until analyzed or otherwise
treated prior to analysis. At the time of sampling, sufficient sodium
sulfite was added to chemically reduce the chlorine residual thereby pre-
venting further formation of disinfectant by-products (except THMMAX,
THMSIMDIST, non-halogenated, base/neutral and pesticide/herbicide/PCB
samples). Mercuric chloride was added at approximately 10 mg/1 in dry
powder form to the 3.8 1 (1 gal) Grob CLSA and GC/FID profile samples to
retard microbiological activity. Hydrochloric acid was added to the 40 ml
non-halogenated sample vial to retard microbiological activity. The chlorine
residual in the base/neutral extractable samples was chemically reduced by
the addition of sodium sulfite before being extracted.
Reagents--
Chlorine dosing solution—Calcium hypochlorite (0.5 g) was dissolved in
500 ml of helium-stripped, double-distilled water. This resulted in approxi-
mately 500 mg/1 of free available chlorine.
65
-------
Chlorine neutralizing solution--0.2 N Sodium sulfite (2.5 g) was dis-
solved in 100 ml of helium-stripped, double-distilled water with 0.5 ml
added to a 40 ml sample vial.
Boric acid/borax, pH 8.2 buffer—A mixture of 3.1 g boric acid and 0.7
g borax was dissolved in one 1 of helium-stripped, double-distilled water
with 15 ml of this buffer added to each 325 ml sample bottle.
Borax, pH 9.5 buffer-- A mixture of 4.8 g borax and 0.7 g sodium hydrox-
ide was dissolved in one 1 of helium-stripped, double-distilled water with
15 ml of this buffer added to each 325 ml sample bottle.
Bacterial, Physical and Chemical--
The methods used in performing the bacterial, physical and chemical
analyses were from "Standard Methods."
Regenerator Off-Gas--
The regenerator specifications required that emissions during the
operation would comply with all regulations to the satisfaction of the local
air pollution authority. In order to ascertain that this requirement was
met, a local laboratory (PEDCo) was engaged to conduct stack gas analysis.
All sampling and analytical procedures were conducted .in accordance with the
methodology protocol set forth in the Federal Register.
Analytical methods used in the evaluation of GAC qualities can be found in
Table 20.
Sand in Carbon--
The following procedure J was used to determine the ratio of sand to
GAC contained in sand table discharges.
1. Prepare 10 g of a dried representative GAC/sand sample by riffling
entire sample.
2. Pour 100 ml of a 54% tetrabromoethane and 46% carbon tetrachloride
solution into a 500 ml separatory funnel. The funnel should be the
open-end variety with a 0.64 cm (1/4 in) opening for use with rubber
tubing and a tubing clamp.
3. Pour the sample into the separatory funnel and swirl the sample until
particles no longer settle out (approximately one min).
4. Drain off 50 ml of the solution which contains the sand into a 100 ml
beaker.
5. Dilute the solution containing the sand with approximately 20 ml of
acetone and decant the solution being very careful not to carry over
any sand. Repeat this step several times.
66
-------
TABLE 20. ANALYTICAL METHODS REFERENCE LIST
Test Method Reference
Iodine Number 5
Molasses Number 5
Decolorizing Index 5
Abrasion Number (Ro-Tap) 5
Apparent Density 5
Sieve Analysis (dry) 5
Effective Size and Uniformity Coefficient 5
Moisture 5
Moisture (Alternate Method) 5
Total Ash 5
BET Total Surface Area 20
Phenol Value - AWWA Modified 21
Phenol Value - Westvaco Modified 22
6. Dry this sample in 100°C (212°F)) oven for at least one hour, cool and
weigh in a preweighed aluminum dish.
7. Perform calculation: % sand = dry sand vt. (g) K 100.
original sample wt. (g)
Analytical Quality Control—
Bacterial, physical and chemical—The CWW Laboratory and its personnel
are certified by the OEPA for physical and chemical analyses under Approval
No. 882, and for bacterial analyses under OEPA Approval No. 130.
Once a year, all certified physical and chemical analyses are evaluated
under the USEPA quality control standards program. As recommended by OEPA,
quality control analyses are continuously performed.
Purgeable halogenated--Purgeable halogenate.d organics were measured by
the USEPA purge/trap THM analytical procedure. The analytical equipment
was standardized by the external standard method. If the average response
factor for the standard was not within 10% of the previous standard run, the
calibration was not accepted. The standard was then rerun until the average
response factor was within 5% of the previous run.
The integrity of the analytical procedure throughout the day was
checked by performing a duplicate analysis of every tenth sample. A duplicate
analysis involved two successive analyses of the same sample. The average
percent relative deviations on duplicate THM analyses can be found in
Table 21. Percent relative deviation (RD) was calculated by determining the
percent_ difference between _the higher of the two data points (A) and their
mean (X) , RD = [(A - X)/X]100. The average percent relative deviation
67
-------
TABLE 21. AVERAGE PERCENT RELATIVE DEVIATION OF THM ANALYSES
Duplicate Samples
Concentration (jug/1)
0.4 - 10
10 - 100
100
Overall Average
Concentration (jug/1)
0.4 - 10
10 - 100
100
Overall Average
CHC1 % CHBrCl , %
O £
1.87
1.74
1.84
1.79
Replicate
1.55
1.56
1.56
Samoles
CHC13, % CHBrCl2, %
6.23
4.04
2.98
4.24
6.55
4.15
4.84
CHBr0Cl, %
1.90
1.45
1.73
CHBr2Cl, %
6.03
4.58
5.55
3.26
3.26
CHBr3, %
8.94
8.94
Based on one value.
represents the average of the relative deviations of all the data sets in
the stated concentration range. Thus, an average percent relative deviation
of 1% would mean that the data was only 1% from the mean. An average percent
relative deviation of 10% was accepted. The average percent relative devia-
tions from the mean values over the entire range of values were as follows:
CHC13 (1.79), CHBrCl (1.56), CHBr Cl (1.73) and CHBr3 (3.26). These values
show little variability from analysis to analysis and overall excellent
precision in the method.
The integrity of the analytical procedure from day to day and, there-
fore, from standardization to standardization was performed using replicate
analyses. A replicate analysis involved the collection of two separate
vials at the time of sampling. The two vials were then analyzed on different
days and the results obtained based on two different standardizations. One
replicate analysis was performed for every 20 samples of each of the follow-
ing: INSTTHM, THMSIMDIST and THHMAX. The average percent relative deviation
of replicate trihalomethane analyses can be found in Table 21. The average
percent relative deviations from the mean values over the entire range of
values were as follows: CHC1 (4.24), CHBrCl2 (4.84), CHBr Cl (5.55) and
CHBr (8.94). As expected, these values are slightly higher than the dupli-
cate results but continue to show good precision in the method.
The accuracy of the formation potential procedure was determined using
the control sample analysis. Control sample analyses were performed both on
the THMMAX samples and on the THMSIMDIST samples. The procedure involved
the preparation of two 325 ml bottles, rather than the normal one, according
68
-------
to the corresponding method using the proper chlorine dose, buffer, tempera-
ture and time for the two types of formation potentials performed. One vial
was collected from each of the sample bottles. The vials were then analyzed
in succession on the same day. One control sample analysis was performed on
every 20th sample. The average percent relative deviation from control
sample THM analyses can be found in Table 22. The average percent deviations
from the mean values over the entire range of concentrations for the THMMAX
control samples were as follows: CHC13 <2-17)' CHBrCl2 (2.58),
CHBr Cl (2.67) and CHBr (4.32). The average percent deviations from the
mean values over the entire range of concentrations for the THMSIMDIST con-
trol samples were as follows: CHC1 (3.17), CHBrCl2 (2.41), CHBr Cl (2.20)
and CHBr, (3.84). As expected, these values were between the duplicate and
the replicate results and show excellent accuracy in the methods used in the
preparation of the control samples.
Evaluation of CWW1s ability to correctly identify and quantify organic
compounds was accomplished by analyzing USEPA performance standards. Eight
USEPA performance evaluation standards were analyzed during the length of
the project. The standards consist of two sealed ampules containing various
concentrations of unknown volatile halogenated organic compounds. The vola-
tile organic compound concentrates were spiked into organic free water and
analyzed.
Three sets of samples were received directly from the USEPA. The
results from these samples are contained in Tables 23, 24 and 25 and show
that all 31 compounds present in the vials were correctly identified and
that their concentrations were within the acceptable limits established.
The acceptance limits used were as reported by the USEPA. Table 23 limits
TABLE 22. AVERAGE PERCENT RELATIVE DEVIATION FOR THM CONTROL SAMPLES
Maximum THM
Concentration (ug/1) CHC1_. % CHBrCl_, % CHBr^Cl, % CHBr
0.4 - 10
10 - 100
100
Overall Average
CHC13, %
2.47
2.05
2.32
2.17
CHBrCl ,
3.12
2.42
-
2.58
% CHBr Cl,
3.06
2.20
-
2.67
0
O
Simulated Distribution System THM
3'
Concentration (jag/1)
0.4 -
10
Overall 1
10
100
100
We rage
CHC1, , %
3
3.12
3.57
2.94
3.17
CHBrCl2/
2.88
2.37
2.41
% CHBr2Cl, %
2.0
2.34
2.20
4.32
4.32
CHBr
3'
3.84
3.84
69
-------
TABLE 23. USEPA PERFORMANCE
Parameter
Vial 1
Chloroform
Bromodichlorome thane
Dibromochlorome thane
Bromoform
Total Trihalomethane
1, 2-Dichloroe thane
1,1, 1-Trichloroethane
Carbon Tetrachloride
Trichloroethylene
Tetrachloroethylene
Vial 2
Chloroform
Bromodichlorome thane
Dibromochlorome thane
Bromoform
Total Trihalomethane
1 , 2-Dichloroe thane
1 , 1, 1-Trichloroethane
Carbon Tetrachloride
Trichloroethylene
Tetrachloroethylene
, Received from USEPA.
Based on 99% confidence
True
Value
Qug/l)
22.8
2.4
2.1
2.8
30.1
2.1
1.7
4.2
1.9
1.8
91.3
23.8
10.3
11.4
136.8
17.3
16.8
EVALUATION
Reported
Value
(jug/1)
21.1
2.7
1.4
2.5
27.7
1.8
1.0
3.2
1.7
1.6
83.0
24.8
8.8
10.1
126.7
13.4
12.8
STANDARDS3 WP005,
Acceptance
Limits
0
0
0
0
0
0
0
0
0
29
0
0
0
4
0
(ug/1)
- 45.1
8.2
.1 - 3.4
6.0
5.9
9.8
- 13.2
6.4
4.5
.8 - 137.0
- 55.7
- 17.1
.7 - 21.3
.2 - 32.6
- 43.3
MARCH, 1979
Performance
Evaluation
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
interval.
70
-------
TABLE 24. USEPA PERFORMANCE EVALUATION STANDARDS3 WS006, FEBRUARY, 1980
Parameter
Vial 1
Chloroform
Bromodichlorome thane
Dibromochlorome thane
Bromoform
Total Trihalomethane
Vial 2
Chloroform
Bromodichlorome thane
Dibromochlorome thane
Bromoform
Total Trihalomethane
3 Received from USEPA.
Based on 20% of true
TABLE 25. USEPA
Parameter
Vial 1
Chloroform
B r omodichlo r ome thane
Dibromochlorome thane
Bromoform
Total Trihalomethane
Vial 2
Chloroform
Bromodichlorome thane
Dibromochlorome thane
Bromoform
Total Trihalomethane
3 Received from USEPA.
Based on 20% of true
True
Value
(ug/D
7.5
42.5
17.0
81.8
148.8
92.3
4.7
50.9
12.3
161.1
value .
PERFORMANCE
True
Value
4am
76.6
91.2
71.1
98.7
337.6
10.2
22.8
11.8
32.9
77.7
values.
Reported
Value
(ug/D
8.2
46.3
18.5
82.9
156.0
99.4
5.1
49.2
12.7
166.0
EVALUATION
Reported
Value
(ug/D
81.5
91.8
72.4
102.6
348.0
10.1
24.4
12.2
31.2
78.0
Acceptance
Limits
(;ug/l)
6.0 - 8.9
34.0 - 51.0
13.6 - 20.4
65.4 - 98.2
119.0 - 179.0
74'. 6 - 112.0
3.8 - 5.7
40.7 - 61.1
9.8 - 14.3
129.0 - 193.0
STANDARDS3 WS008
Acceptance
Limits
(W/D
61.0 - 92.'0
73.0 - 110.0
57.0 - 85.0
79.0 - 120.0
270.0 - 410.0
8.2 - 12.0
18.0 - 27.0
9.4 - 14.0
26.0 - 39.0
62.0 - 93.0
Performance
Evaluation
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
MAY, 1981
Performance
Evaluation
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
71
-------
were based on a 99% confidence interval while Table 24 and 25 limits were
based on 20% of true value. The USEPA switched methods for calculating
acceptance limits to 20% because the 99% confidence interval was too large.
Five sets of samples were received from Ohio River Valley Water
Sanitation Commission (ORSANCO). These samples were part of the quality
assurance program for the Early Warning Organics Detection System (EWODS).
These results are contained in Tables 26 through 30. The acceptance limits
contained in the tables were calculated using the results from four to ten
participating laboratories. A one-way analysis of variance was performed to
determine if any of the participating laboratories were outside the popula-
tion mean. If a significant difference between the laboratories existed,
the site or sites were determined using the Duncan Multiple-Range Test and
the data not used in further evaluation. The 99% confidence interval from
the true value was then calculated using a T-distribution. The T-distribu-
tion was used due to the small number of participating laboratories. From
the tables, it can be seen that all 98 compounds present were properly
identified and in only two cases were the reported concentrations outside
the acceptance limits. Both compound concentrations outside the acceptance
limits were CHBr-Cl. The true values were 2.7 jag/1 and 2.1 ;jg/l and the
reported values 1.8 ;ag/L and l.S^ug/1 respectively.
Purgeable non-halogenated--Purgeable non-halogenated organics were
measured by the purge/trap analytical procedure. The analytical equipment
was standardized by the external standard method. If the average response
factor for the standard was not within 10% of the previous standard run, the
calibration run was not accepted and the standard was rerun. The quality
control discussion contained herein applies only to samples analyzed by CWW.
The integrity of the analytical procedure throughout the day and from
day to day was checked by performing one duplicate analysis every 10 samples
and one replicate analysis every 20 samples. The average percent relative
deviations for duplicate and replicate analyses can be found in Table 31.
It is apparent that of the compounds which were analyzed, only benzene was
found with any regularity. All concentrations were less than 1 ug/1. The
analytical procedure shows excellent precision with little variability from
analysis to analysis and from day to day.
Total organic carbon--The integrity of the analytical procedure through-
out the day was checked by performing a duplicate analysis on every fifth
sample. The average percent relative deviation was 2.19. This value shows
little variability from analysis to analysis.
The integrity of the analytical procedure from day to day and, there-
fore, from standardization to standardization was performed using the repli-
cate analyses procedure previously described. The average percent relative
deviation for TOC replicate analyses was 4.57. As expected, this value is
higher than the duplicate value but still well within acceptable limits.
In order to determine if any short term instrument drift or contamina-
tion factors occurred during an analysis day, an 1,800 jug/1 standard was run
72
-------
TABLE 26. USEPA PERFORMANCE EVALUATION STANDARDS3, MARCH, 1979
True Reported Acceptance Performance
Parameter Value Value Limits Evaluation
(ug/1) (ug/1) ' (ug/1)
Vial 1
Chloroform
Bromodichlorome thane
Dibromochlorome thane
Bromoform
Total Trihalomethane
1 , 2-Dichloroethane
1,1, 1-Trichloroethane
Carbon Tetrachloride
Trichloroethylene
Tetrachloroethylene
Vial 2
Chloroform
Bromodichlorome thane
Dibromochlorome thane
Bromoform
Total Trihalomethane
1 , 2-Dichloroethane
1,1, 1-Trichloroethane
Carbon Tetrachloride
Trichloroethylene
Tetrachloroethylene
22.8
2.4
2.1
2.8
30.1
2.1
1.7
4.2 ^
1.9
1.8
91.3
23.8
10.3
11.4
136.8
17.3
16.8
21.3
3.0
1.5
3.2
29.0
2.0
1.1
3.4
1.4
1.8
85.4
24.5
8.6
9.8
128.3
14.6
15.4
4.2 -
0
0.1 -
0
6.2 -
0
0
0
0
0
40.8 -
0
0
0
0
0
0
41.4
6.2
4.1
6.8
54.0
6.3
7.1
14.8
10.6
6.2
130.0
81.0
21.4
85.1
278.4
63.5
88.4
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
, Received from Orsanco EWODS .
Based on 99% confidence interval.
73
-------
TABLE 27. USEPA PERFORMANCE EVALUATION STANDARDS3, APRIL, 1979
True Reported Acceptance Performance
Parameter Value Value Limits Evaluation
(ug/1) (ug/1) (ug/1)
Vial 1
Chloroform
Bromodichlorome thane
Dibromochlorome thane
Bromoform
Total Trihalomethane
1 , 2-Dichloroethane
1,1, 1-Trichloroethane
Carbon Tetrachloride
Trichloroethylene
Tetrachloroethylene
Vial 2
Chloroform
Bromodichlorome thane
Dibromochlorome thane
Bromoform
Total Trihalomethane
1 , 2-Dichloroethane
1,1, 1-Trichloroethane
Carbon Tetrachloride
Trichloroethylene
Tetrachloroethylene
9.1
1.2
2.7
2.8
15.8
1.4
1.7
68.5
11.9
17.2
14.2
111.8
27.2
11.2
12.6
19.0
8.8
8.0
1.4
1.8
2.3
13.5
1.2
1.3
69.0
15.2
15.4
12.9
112.5
21.6
9.8
10.9
15.6
8.9
0.2 -
0
2.1 -
1.9 -
4.8 -
0
0.9 -
37.9 -
0
10.0 -
8.4 -
56.7 -
18.4 -
6.9 -
2.7 -
7.9 -
6.4 -
18.0
5.1
3.3
3.7
26.8
3.6
2.5
99.1
31.6
24.3
20.0
166.9
36.0
15.5
22.5
30.1
11.2
Acceptable
Acceptable
Not Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Received from Orsanco EWODS .
Based on 99% confidence interval.
74
-------
TABLE 28. USEPA PERFORMANCE EVALUATION STANDARDS3, NOVEMBER, 1979
True Reported Acceptance Performance
Parameter Value Value Limits Evaluation
(uq/1) (uq/1) (uq/1)
Vial 1
Chloroform
Bromodichlorome thane
Dibromochlorome thane
Bromoform
Total Trihalomethane
1 , 2-Dichloroethane
1,1, 1-Trichloroe thane
Carbon Tetrachloride
Trichloroethylene
Tetrachloroethylene
Vial 2
Chloroform
Bromodichloromethane
Dibromochlorome thane
Bromoform
Total Trihalomethane
1 , 2-Dichloroethane
1,1, 1-Trichloroethane
Carbon Tetrachloride
Trichloroethylene
Tetrachloroethylene
22.8
2.4
2.1
2.8
30.1
2.1
1.7
4.2
1.9
1.8
91.3
23.8
10.3
11.4
136.8
17.3
16.8
21.7
2.7
1.5
2.9
28.8
1.7
1.3
3.5
1.6
1.7
90.8
25.9
8.4
11.0
136.1
16.2
11.3
8.6 -
1.2 -
1.6 -
2.0 -
14.6 -
1.0 -
0.7 -
1.5 -
0.7 -
0.2 -
34.2 -
2.2 -
6.4 -
8.4 -
72.7 -
6.4 -
10.1 -
37.0
3.6
2.6
3.6
45.6
3.2
2.7
6.9
3.1
3.2
148.4
45.4
14.2
14.4
200.9
28.2
23.5
Acceptable
Acceptable
Not Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
k Received from Orsanco EWODS .
Based on 99% confidence interval.
75
-------
TABLE 29. USEPA PERFORMANCE EVALUATION STANDARDS3 , JUNE, 1980
True Reported Acceptance Performance
Parameter Value Value Limits Evaluation
Oig/1) (jug/1) (>ig/l)
Vial 1
Chloroform
Bromodlchlorome thane
Dibromochlorome thane
Bromoform
Total Trihalomethane
1 , 2-Dichloroe thane
1,1, 1-Trichloroethane
Carbon Tetrachloride
Trichloroethylene
Tetrachloroethylene
Vial 2
Chloroform
Bromodlchlorome thane
Dibromochlorome thane
Bromoform
Total Trihalomethane
1 , 2-Dichloroethane
1,1, 1-Trichloroethane
Carbon Tetrachloride
Trichloroethylene
Tetrachloroethylene
11.0
1.7
2.4
2.8
17.9
1.5
1.2
2.3
2.6
1.1
45.6
8.6
12.0
10.4
76.6
20.0
14.0
9.4
13.0
5.6
10.6
1.4
2.3
2.2
16.5
1.2
0.8
1.8
2.2
0.9
48.2
9.3
12.6
13.1
83.2
15.7
9.6
6.0
9.6
5.6
5.0 -
0.2 -
1.3 -
1.5 -
10.0 -
0.5 -
0.6 -
1.0 -
1.9 -
0.8 -
27.2 -
4.3 -
9.0 -
2.6 -
44.2 -
0
6.3 -
5.3 -
8.9 -
3.9 -
17.0
3.2
3.5
4.1
25.8
2.5
1.8
3.6
3.3
1.4
64.0
12.9
15.0
18.2
109.0
43.4
21.6
13.5
17.1
7.3
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Received from Orsanco EWODS.
Based on 99% confidence interval.
76
-------
TABLE 30. USEPA PERFORMANCE EVALUATION STANDARDS3 , FEBRUARY, 1981
True Reported Acceptance Performance
Parameter Value Value Limits Evaluation
(ug/1) Gag/1) (ug/1)
Vial 1
Chloroform
Bromodichlorome thane
Dibromochlorome thane
Bromoform
Total Trihalomethane
1 , 2-Dichloroe thane
1,1, 1-Trichloroe thane
Carbon Tetrachloride
Trichloroethylene
Tetrachloroethylene
Vial 2
Chloroform
Bromodichlorome thane
Dibromochlorome thane
Bromoform
Total Trihalomethane
1 , 2-Dichloroe thane
1,1, 1-Trichloroethane
Carbon Tetrachloride
Trichloroethylene
Tetrachloroethylene
11.0
1.7
2.4
2.8
17.9
1.5
1.1
2.3
2.6
1.1
45.6
8.6
12.0
10.4
76.6
20.0
14.0
9.4
13.0
5.6
9.6
1.4
2.4
2.1
15.5
1.3
0.8
2.3
2.6
0.9
47.8
10.5
14.2
11.5
84.0
16.5
9.0
7.2
10.4
5.1
6.8 -
0.9 -
1.8 -
1.2 -
13.6 -
0
0.3 -
1.3 -
1.6 -
0.5 -
31.1 -
5.9 -
6.2 -
3.2 -
62.5 -
11.6 -
7.8 -
6.3 -
0
4.0 -
15.2
2.5
3.0
4.4
22.2
3.8
1.9
3.3
3.6
1.7
60.1
11.3
17.7
17.6
90.7
28.4
20.2
12.5
38.1
7.2
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
Acceptable
, Received from Orsanco EWODS .
Based on 99% confidence interval.
77
-------
TABLE 31. AVERAGE PERCENT RELATIVE DEVIATION OF PURGEABLE NON-HALOGENATEDSa
b
Duplicate Samples
Average Percent
Parameter Relative Deviation
Hexane 8.3C
Benzene 2.4
Toluene ND
Ethyl Benzene 0C
o-Xylene ND
Tetralin ND
b
Replicate Samples
Average Percent
Parameter Relative Deviation
Hexane ND
Benzene 6.3
Toluene ND
Ethyl Benzene ND
o-Xylene ND
Tetralin ND
, Analysis performed by Cincinnati Water Works.
Based on samples with concentrations equal to or greater than 0.4 jag/1.
, Based only on one value.
Not detected, values were less than 0.4 ,ug/l.
at the end of the sample series and compared to the standardization run.
The average percent relative deviation was 2.57 which is well within accept-
able limits.
The results from the USEPA Performance Evaluation Study WP005 can be
found in Table 32. From the table, it can be seen that vial 1's result was
acceptable while vial 2's result was not acceptable. Upon investigation of
the result from vial 2, it was observed that although the 150 ul of persul-
fate reagent was sufficient for the normal concentration range observed, it
was not sufficient for the concentration level in vial 2. Further investi-
gation indicated that 150 jjl of persulfate reagent was sufficient for TOC
values less than 4.0 mg/1. The method of analysis was modified so that any
sample with a TOC value greater than 3.0 mg/1 would be rerun using 300 >il of
persulfate reagent. This provides sufficient reagent for TOC values up to
10.0 mg/1.
78
-------
TABLE 32. TOC RESULTS FROM USEPA PERFORMANCE EVALUATION STANDARDS WS005
TOO Concentration (mg/1)
Reported True
Value Value
Vial 1 3.2 3.2
Vial 2 8.6 107
79
-------
SECTION 6
RESULTS AND DISCUSSION
Discussion of the observations and recorded data resulting from the
operation of both pilot and full-scale GAC systems as well as the analytical
results of the many areas being investigated are presented in segments
addressed as specific objectives consistent with the goals of the grant
study. The format will follow the chronology of the three phase program.
PHASE 1. FULL-SCALE GAC FILTERS
Objective 1: To Compare the Relative Performance of Various Depths
and Types of GAC
Three sand filters (19A, 21A and 23A) were converted to GAC filters as
detailed in Section 4 with provisions for sampling at various depths as
indicated in Figure 15. The combination of sample probe locations and
differing GAC beds afforded the opportunity to evaluate the effect of varied
contact times and effectiveness of different GAC types.
Effect of Various Contact Times--
The GAC filters were designed so that samples could be collected at
various depths. Figure 15 shows the three GAC filters and their respective
intermediate sample depths. A comparison of the results from these inter-
mediate sample points would show the effects of longer contact time on
removing organics from the water. It would also be possible to watch the
progression of exhaustion through the bed.
Sample collection from the intermediate depths was accomplished by
placing stainless steel auger-type sample probes at the specified depths in
the filters. A pump was then used to draw the sample from the desired depth
for collection.
Collection of these intermediate sample points proved to be difficult.
Air had a tendency to collect in pockets around the sample probes which made
priming of the pump and sample line very difficult. In order to release the
bound air, the sample probe, or auger, was rotated in place. Rotating the
sample probe did not always solve the problem of sample collection.
Examination of the breakthrough curves showed inconsistent results from
what would have been expected when comparing similar bed depths. Figure 16
shows the THttttAX breakthrough curves for 19AE and 21A2 and is an example of
80
-------
13
CONCENTRATION,
0>
m
I
IS)
s
I
-------
the inconsistency observed. These two locations represented the same EBCT
and should, therefore, have identical curves. For the most part, the two
locations do parallel each other. However, on rundays 31 and 38, the inter-
mediate sample point (21A2) does not match 19AE. The concentrations for
21A2 on these two days approached and almost equaled the influent (FLIN)
THMMAX concentration. This inconsistency might have been due to the sample
collection technique since the rotation of the sample probe may have permit-
ted the channeling of the influent, (FLIN) down the sample probe. This
short circuiting would have resulted in little if any contact time with the
GAC, resulting in a higher THMMAX concentration.
Interpretation of the data from the intermediate sample points was
severely restricted because of this potential sample collection problem.
Examination of the breakthrough curves through the GAC filters does show
that longer contact time provides additional removal of organics. However,
determination of the benefits of this longer contact time cannot be made
with any real confidence by using the intermediate sample depths.
It is still possible to address the effect of contact time by comparing
the effluents from GAC Filters 19A and 21A which had 4.5 and 7.5 minutes
EBCT, respectively. Before comparing these two GAC filters, some clarifica-
tion of data interpretation is first necessary.
During Phase I, a review of the data revealed that the concentrations
from a large number of backlogged THM analyses were excessively high. This
covered the period from February 14, 1978 to May 15, 1978. Subsequent
investigation revealed that insufficient quench agent, sodium sulfite, was
being used. Therefore, free chlorine was available for further reaction
with THM precursors and additional THMs formed. Unfortunately, this invali-
dates much of the data collected during the early stage of the GAC filters'
life. The various types of THM analyses were affected as follows:
1. INSTTHM: Invalid, except for the GAC filter samples where the chlorine
had been adsorbed by the GAC.
2. THMMAX: Valid, little additional chlorine-precursor reaction occurred
after the one-week reaction time used for forming the THMI'IAX.
3. THMFP: Invalid, except for the GAC filter samples. The THMFP calcula-
tion involved the subtraction of INSTTHM from THMMAX.
4. THMSIMD1ST: Invalid, additional chlorine-precursor reaction was pos-
sible even after the one-week reaction time used for the THMSIMDIST
method.
The determination of the effect of different contact times between the
effluents from GAC Filters 19A and 21A was still possible even with the
limitations placed on data use. Breakthrough curves for TOC and THMMAX can
be found in Figures 17 and 18, respectively. Figure 19 shows INSTTHM con-
centrations for 19AE and 21AE. From these figures, it can readily be seen
that longer contact time provided improved effluent water quality. Selection
82
-------
a>
en
CONCENTRATION,
co c/i
CD CD
o
ro
v\
CD
CO
I
o
ro
i
-------
CO
300
0
INFLUENT CFUJO
LEGEND
6AC FILTER I8A EFF C18AE3
6AC FILTER 21A EFF C21AE5
10
20
60
30 40
TIME, days
Figure 18. Maximum THM (MTTT) breakthrough curves for GAC Filter 19A and
21A effluents, Phase 1-0.
70
-------
S8
ca
-i
o>
CONCENTRATION, >jg/ I
m
CD
fO
I
-------
of a treatment goal of 1,000 .ug/1 for TOC in the effluent (as represented by
the intersection of the treatment goal line and the breakthrough curve when
the average of any three-week set exceeded the goal) permitted further
evaluation of contact time. This effluent requirement resulted in an oper-
ating service time of about 32 days for GAC Filter 21A and 16 days for GAC
Filter 19A. Therefore, increasing the EBCT by a factor of 1.7 resulted in an
increase in the operating service time by a factor of 2.0.
The purgeable halogenated minimum, maximum and average results can be
found in Table 33. It is felt that the dichloromethane results represented
an artifact due to the use of dichlorome thane in the laboratory. The in-
fluent concentrations were very low and for most samples approached the
lower limits of the detector (approximately 0.1 ug/1). Both GAC filters
performed similarly when they received an elevated 1 to 2 ug/1 dose of carbon
tetrachloride, 1,1,1-trichloroethane, chlorobenzene and o-dichlorobenzene.
Except for 1,1,1-trichloroethane, they were very effective in the removal of
these compounds.
Purgeable non-halogenated samples were collected but very few analyses
were performed due to analytical problems encountered with the method.
The above data indicates that longer EBCT in GAC filters resulted in an
increase in the operating service time. However, the optimum EBCT for GAC
filters was not determined due to the limited number of sample points and
minimal bed depth.
Comparison of Different Grades of Bituminous-Based GAC--
As previously described, GAC Filter 21A contained 76.2 cm (30 in) of 12
x 40 WVG and GAC Filter 23A, contained 76.2 cm (30 in) of 20 x 50 WVW GAC.
This configuration permitted the comparison of two different grades of
bituminous-based GAC.
The TOC breakthrough curves for 21A and 23A can be found in Figure 20.
The breakthrough curves show 21A and 23A to be very similar for the first 20
days, after which 21A outperformed 23A. A treatment goal of 1,000 pg/1 for
TOC in the effluent was again used for further evaluation. This effluent
requirement would result in an operating service time of about 32 days for
21A and 24 days for 23A. Based on this criteria, 21A was only slightly
superior to 23A.
The THMMAX breakthrough curves for 21A and 23A can be found in
Figure 21. From the figure, it is apparent that 21A outperformed 23A. For
further evaluation, an exhaustion criterion of 100 un/1 THMMAX, similar to
the level specified in the Safe Drinking Water Act , in the effluent (as
represented by the intersection of the treatment goal line and the break-
through curve when the average of any three-week set exceeds the goal) was
selected. This exhaustion criterion would have resulted in an operating
service time of about seven days for GAC Filter 23A and 28 days for GAC
Filter 21A. Based on this exhaustion criteria, the configuration of 21A was
superior.
86
-------
00
TABLE 33. PURGEABLE HALOGENATED ORGAMICS FOR GAC FILTERS 19A AND 21A,
Influent 19A 21A
Min Max
Value Value
Parameter (ug/1) (ug/1)
Carbon tetrachloride
1 , 2-Dichloroe thane
1,1, 1-Trichloroe thane
Dichloromethane
Tetrachloroethylene
Trichloroethene
Chlorobenzene
0-Dichlorobenzene
a 1
a 0
a 1
a 21
a 0
a 0
a 2
a 1
.7
.7
.5
.9
.2
.4
.2
.3
Avg
Value
(ug/1)
0.1
0.0
0.1
0.8
0.1
0.1
0.1
0.2
PHASE
1-0
Min Max Avg Min Max Avg
Value Value Value Value Value Value
a 0.2 0.0 a 0.2 0.0
a a
a 1.
a 6.
a 0.
a 0.
a 0.
a 0.
2
0
2
2
2
2
a
0.
0.
0.
0.
0.
0.
1
6
0
0
0
0
a a
a 1.
a 5.
a 0.
a 0.
a 0.
a 0.
0
1
2
2
2
2
a
0.
0.
0.
0.
0.
0.
1
5
0
0
0
0
Not Detected ( 0.1 ;ig/l)
-------
00
CO
3080
0
INFLUENT CFLDO
LEGEND
SAC FILTER 2IA EFF C2IAE5
(SAC FILTER 23A EFF C23AE>
10
20
50
60
30 40
TIME, days
Figure 20. Total organic carbon (TOO breakthrough curves for GAC Filter 21A and
23A effluents, Phase 1-0.
70
-------
68
CONCENTRATION, jug/ I
on
en
ro
ro 3
oo Q
>• X
& 3
-h C
I* 3
C —I
S i
rt-
ffi /'N
O
V)
<&
i Q
O 7T
O
ro
-HO
H
rn
o o.
c Q
"i •<
< n
o
O
1
m
o
rn
9
««i
w
§
-------
The INSTTHM concentration in 21A and 23A effluents can be found in
Figure 22. The effluents from both GAC filters were almost identical.
However, the highest INSTTHM concentration observed was approximately
10 ug/1. This was a very low concentration probably due to low temperature
water. It is important to note that comparison of performance should always
take concentration into account.
The purgeable halogenated organics minimum, maximum and average results
for influent, 21A and 23A can be found in Table 34. As before, the dichloro-
methane results were likely artifacts. The influent concentrations were
very low and, for the most part, approached the lower limit of the detector.
The performance of 21A and 23A were very similar based on very low influent
concentrations.
Purgeable non-halogenated samples were collected but very few analyzed
due to analytical problems encountered with the method.
One set of base-neutral extractables was analyzed. The only reported
organics were phthalate plasticizers that seem to be laboratory artifacts,
since often blank concentrations were higher than those of the samples.
At the end of the run, 23A was found to have lost 34% of its GAC volume
compared to only 14% for 21A. Due to the smaller particle size and weight
of 23A GAC, utilization of backwash procedures acceptable for 12 x 40 GAC
filters likely resulted in the excessive losses for 20 x 50 GAC contained in
23A. Modifications to the procedure were initiated to minimize this loss
but backwash velocities sufficient to remove floe were still likely great
enough to continue the higher GAC attrition through the rest of the phase.
Given that shorter 23A filter runs occurred throughout this phase (discussed
in Objective 2), it is apparent that floe was not completely flushed out at
the reduced backwash rates.
In summary, the combined results from organic evaluations and backwash-
ing experiences tend to indicate that 20 x 50 GAC would not function properly
in our plant. Rather, the 12 x 40 GAC would be preferred.
Objective 2. To Compare the Relative Performance of Sand vs. GAC as a
Filter Media for Particulate Matter
GAC, as a filter media, performed the filtration function as well as, or
better than sand. Table 35 shows that the GAC filters with 12 x 40 GAC
consistently gave longer service between washes than the plant average,
including sand and GAC filters. It is possible that the plant average could
have been slightly higher than shown since sand filters were washed after
attaining 1.8 m (6 ft) of head loss or 60 hours of operation whichever
occurred first, while the GAC filters were washed on head loss alone.
During the period of the study, filter runs became increasingly longer.
The 76.2 cm (30 in), 12 x 40 GAC bed also experienced increased run times,
but at a rate less than the plant average. The run time of the 45.7 cm
(18 in), 12 x 40 bed tended to be stable, but still performed better than
90
-------
LEGEND
SAC FILTER 21A EFF C21AE5
SAC FILTER 23A EFF C23AE)
12
C9
8
1 6
2
UJ
O
2
O
O
10
28
30 4
TIME, days
50
68
78
Figure 22. Instantaneous THM CITTT) concentration curves for GAC Filter 2tA and 23A
effluents, Phase 1-0.
-------
TABLE 34. PURGEABLE HALOGENATED ORGANICS FOR GAC FILTERS 21A AND 23A, PHASE 1-0
Influent 21A 23A
Parameter
Carbon tetrachloride
1,2-Dichloroethane
1,1,1-Trichloroethane
Dichloromethane
Tetrachloroethylene
Trichloroethene
Chlorobenzene
0-Dichlorobenzene
Min
Value
(ug/l)
a
a
a
a
a
a
a
a
Max
Value
. (ug/l)
I
0
1
21
0
0
2
1
.7
.7
.5
.9
.2
.4
.2
.3
Avg
Value
(ug/D
0
0
0
0
0
0
0
0
.1
.0
.1
.6
.1
.1
.1
.1
Min Max
Value Value
(.ug/D (ug/D
a 0
a
a 1
a 5
a 0
a 0
a 0
a 0
.2
a
.0
.1
.2
.2
.2
.2
Avg
Value
(ug/l)
0
0
0
0
0
0
0
.0
a
.]
.5
.0
.0
.0
.0
Min Max Avg
Value Value Value
(ug/l) (ug/l) (ug/l
a 0.2 0.0
a a a
a 0.9 0.0
a 4.9 0.7
a a a
a 0.4 0.0
a 0.2 0.0
a 0.2 0.0
Not Detected ( 0.1 ug/1)
-------
TABLE 35. COMPARISON OF FILTER RUN BETWEEN GAC FILTERS AND PLANT AVERAGE
Yearly Average Hours Between Washes
Filter 1978 1979 1980
19A 61.9 58.7 59.0
21A 48.0 55.5 64.7
23A 21.0 36.8 31.8
Total Plant3 27.5 38.9 45.8
Includes GAC filters.
the plant average. This increase in the length of filter runs may be attrib-
uted to the improved river quality. Except for three brief episodes with a
slight rise in turbidity (all less than 1.0 NTU) on the 45.7 cm (18 in) bed
early in the run, the effluent clarity of GAC filters was as good as that of
the sand filters.
At no time were the length of filter runs for the 20 x 50 GAC filter as
long as the plant average, some runs being as short as 6.4 hours. One
likely explanation was that the backwash rates which were reduced to mini-
mize GAC losses did not completely remove filterable solids. During the
winter months, when water temperature ranged below 2°C (35°F), air binding
further shortened filter runs. The 12 x 40 GAC filters also had a tendency
to air bind more than sand filters but not as pronounced as the 20 x 50 GAC.
To eliminate the effect of the inclusion of the GAC filters in estab-
lishing the plant average as well as possible suppression due to the 60 hour
maximum run on a sand filter, direct comparison of sand filters was made
with GAC Filters (Table 36). The week of August 9 through 15, 1981 was
selected because the sand filter runs were consistently below 60 hours. The
results are fairly conclusive that the service life of the GAC filters was
approximately 150% of that for the sand filters.
TABLE 36. COMPARISON OF FILTER RUN TIMES DURING TYPICAL PERIOD
WHEN SAND FILTER RUNS WERE LESS THAN 60 HOURS9
Sand Filters GAC Filters
Number of Runs 141 8
Avg. Time in Service, Hours 29.2 45.0
Week of August 9 through 15, 1981.
93
-------
From the foregoing, it is evident that while GAC was generally effective
as a filter media, not all grades and types are equal in performance.
Specifically, the 20 x 50 GAC would require more frequent backwashes during
extended periods of increased turbidity and extremely cold weather.
Further, the dual-media [30.5 cm (12 in) sand, 45.7 cm (18 in) GAC] GAC
filter would be less practical than the full-depth GAC bed, since lower
backwash rates dictated by the use of GAC may not expand and clean the sand
portion of the bed. As an adsorption agent, the reduced contact time and
reduced bed volume would greatly reduce the bed life.
PHASE 2. PILOT GAC FILTERS AND PILOT CONTACTORS
Objective 3. To Compare the Relative Performance of Virgin and
Regenerated GAC Filters
In this aspect of Phase 2, bituminous-based GAC was utilized as a
direct filtration media in the 7.6 cm (3 in) pilot columns" described in
Section 4 and operated within the parameters set forth in Section 5. This
comparison extended through two laboratory regeneration cycles.
Figures 23 through 26 show a comparison of virgin versus once-regener-
ated GAC for various parameters. This and the subsequent regeneration were
performed by the Westvaco Corp. using a laboratory-scale, fluidized-bed
regenerator at 1500°F. These figures indicate that the percent removals for
TOC, INSTTHM, THMSIMDIST and THMFP were practically identical.
Figures 27 through 30 show a comparison of virgin versus twice-regener-
ated GAC for the same parameters. These figures again indicate that the
percent removals for TOC, INSTTHM, THMSIMDIST and THMFP were practically
identical.
Analyses performed included TOC, THMSIMDIST, INSTTHM, THMFP, non-THM
purgeable halogenated, purgeable non-halogenated and base/neutral extractable
compounds. Due to difficulties in obtaining consistent results from contract
laboratories, data concerning the latter two classes of compounds do not
provide for any useful evaluation. Purgeable halogenated compounds, other
then THMs, were seldom seen in sufficient quantities to be useful.
Based on the data, there appears to be no appreciable effect on the
GAC's adsorptive capabilities after two regeneration cycles.
Objective 4. To Compare the Relative Performance of Bituminous-
and Lignite-Based GAC in Pilot Contactors
A description of the physical facilities used to evaluate this object-
ive is contained in Section 4. Both pilot column systems contained essent-
ially the same volume of GAC and received essentially the same influent
water, loading rate and throughput over the same time frame (October 24, 1973
through May 31, 1979). It should be noted that during this time frame,
94
-------
LEGEND
Q H ONCE REGENERATED PILOT SAC FILTER CRV3)
VIRGIN PILOT GAC FILTER CRV55 X X
100
0
L
0>
a
o
Id
QL
60
40
20
0
-20
-40
0
20
40
100
120
60 80
TIME., days
Figure 23. Total organic carbon (TOO percent removal curves for virgin and once
regenerated pilot GAC filter effluents, Phase 2-1.
140
-------
LEGEND
Q D ONCE REGENERATED PILOT 6AC FILTER CRV3>
VIRGIN PILOT GAC FILTER CRV55
100
50 _
-«j
c
fl>
0
L
0
Q.
<
>
o
LJ
-300 .
-350 _
40
100
120
60 80
TIME, days
Figure 24. Instantaneous THM QTTD percent removal curves for virgin and
once regenerated pilot GAC filter effluents, Phase 2-1.
140
-------
LEGEND
a Q ONCE REGENERATED PILOT 6AC FILTER CRV3)
VIRGIN PILOT GAC FILTER CRV5) X X
-*>
c
©
u
L
0
Q.
o
z
Id
100
80
60
40
20
0
-20
0
20
40
60
TIME,
100
120
140
80
days
Figure 25. Seven-day simulated distribution syst. THM CSTT7) percent removal curves
for virgin and once regenerated pilot GAC filter effluents, Phase 2-1.
-------
LEGEND
0 m ONCE REGENERATED PILOT GAC FILTER CRV35
VIRGIN PILOT GAC FILTER CRV5)
ID
CD
.*»
c
0
u
1_
-------
-n
66
00
REMOVAL, percent
ro
en
o>
00
(D
o> o
Q
o> ""
"J O
Q ">
CT
<7> O
3> D
O
— O
•— O
3
« o
<0 C
f\> <
I (t>
ro w
o
ro
p
«"»
!
-------
LEGEND
Q H TWICE REGENERATED PILOT GAC FILTER CRV3D
VIRGIN PILOT GAC FILTER CRV5> X X
o
o
0
L
0
a
o
2:
UJ
a:
-80 .
-100
0
40
60
TIME, days
80
100
120
Figure 28. Instantaneous THM (ITTT) percent removal curves for virgin and
twice regenerated pilot GAC filter effluents., Phase 2-2.
-------
LEGEND
B Q TWICE REGENERATED PILOT GAC FILTER CRV3D
VIRGIN PILOT GAC FILTER CRV5>
u
L
0>
a
A
_J
<
>
o
2:
kl
108
98
80
70
60
50
40
30
20
0
20
40 60
TIME, days
80
100
120
Figure 29. Seven-day simulated distribution syst. THM (STT7) percent removal curves
for virgin and twice regenerated pilot GAC filter effluents, Phase 2-2.
-------
LEGEND
D H TWICE REGENERATED PILOT 6AC FILTER CRV3!)
VIRGIN PILOT GAC FILTER CRV55
o
100
90
80
"c 70
0
o 60
Q.
J 50
O
I 40
30
20
10
0
20
40 60
TIME, days
80
100
120
Figure 30. THM formation potential CFTTT) percent removal curves for virgin and
twice regenerated pilot GAC filter effluents, Phase 2-2.
-------
water temperatures were lower than during the summer months when shorter bed
lives relative to THMs would be expected. Differences in the two systems
include the GAC characteristics contained in Table 37 and the GAG costs as
indicated in Table 38. Analyses performed included TOC, THMSIMDIST,
INSTTHM, THMFP, non-THM purgeable halogenated, purgeable non-halogenated and
base/neutral extractable compounds. Due to difficulties in obtaining consis-
tent results from contract laboratories, data concerning the latter two
classes of compounds do not provide for any useful evaluation of
bituminous/lignite GAC comparative performance. Purgeable halogenated
compounds, other than THMs, were seldom seen in sufficient quantities to be
useful in comparing these two GACs.
The comparative results of the GACs are contained in Table 39. Only
TOC and THMSIMDIST results were compared and only for selected exhaustion
criteria. A discussion of the reasons for these selections is contained in
Objective 9. Since the preferred criteria identified in Objective 9
(TOC = 1.0 mg/1 or STT7 = 0.1 mg/1) did not provide sufficient comparative
data, two other related criteria also explained in Objective 9 were included
(TOC loading retained i 75% or STT7 loading retained > 75%).
At exhaustion, the preferred GAC will have: 1) lasted a greater number
of days, 2) adsorbed a greater weight of contaminants, and 3) cost less per
weight of contaminant removed.
As Figures 31 and 32, and Table 39 indicate, bituminous outperformed
lignite GAC in almost every case regardless of the exhaustion criterion or
parameter. Perhaps this was due to the differences in characteristics such
as the greater BET surface area and Iodine number of the bituminous GAC.
For example, using the exhaustion criterion of 1.0 mg/1 TOC, bituminous GAC
lasted 198 days, removed 785 grams of TOC and cost 4.1C per TOC gram adsorbed.
Lignite GAC lasted only 35 days, removed 248 grams of TOC and cost 6.4C per
TOC gram adsorbed. These cost figures were merely calculated by dividing
the cost of the GAC in the system by the weight of TOC adsorbed. However,
since lignite GAC has a lover apparent density and costs less per unit of
weight, the cost per gram of contaminant adsorbed needs to be examined more
fully.
It was important to consider the adsorbed weight of contaminants from
two different perspectives when considering cost: 1) per unit of weight of
GAC since GAC is purchased by weight and 2) per unit of volume of GAC since
the size and cost of facilities constructed are dependent upon the volume of
GAC applied in the treatment process.
The cost estimates in Table 39 are based solely on the cost of GAC and
not on facilities or regeneration costs. Given that the bituminous GAC
adsorbs more grams of contaminant to exhaustion per unit volume, the use
of lignite GAC would require larger tanks to hold proportionately greater
volumes of GAC to expect similar removals to that of bituminous GAC. There-
fore, although lignite is less expensive per pound, a larger volume of it
would have to be applied to match the adsorptive capability of bituminous.
Further, given the longer life of the bituminous GAC to exhaustion, lignite
103
-------
TABLE 37. BITUMINOUS AND LIGNITE GAC CHARACTERISTICS
Bituminous Lignite
Weight of GAC, kg (Ibs)
Apparent Density, gm/ml (Ib/cu ft)
Particle Size
Surface Area, BET,
Iodine No., mg/g
m /g
16.33 (36.00)
0.43-0.46 (27-29)
12x40
1100
1050
14.40 (31.75)
0.39 (24.3)
10x30
650
600
TABLE 38. BITUMINOUS AND LIGNITE GAC PRICES
Bituminous
Lignite
August, 1978
S.65/lb
$.37/lb
June, 1981
$.81/lb
$.55/lb
Based on prices quoted in June, 1981 by Wes±vaco Corp.
and by ICI Americas, Inc. for lignite GAC.
for bituminous GAC
24
TABLE 39. COMPARISON OF BITUMINOUS (BIT) AND LIGNITE (LIG) PILOT
CONTACTORS FOR SELECTED EXHAUSTION (EXH) CRITERIA
TOC STT7
TOC STT7 Loading Loading
1.0 mg/1 0.1 mg/1 Retained Retained
^75% -Z 75%
Run Length, Days
Effluent Cone., mg/1
Removal, %
Loading, g
Load. Retained, %
Load./GAC wt., g/kg
GAC Cost/g Load. ,
-------
SOT
-------
90T
CONCENTRATION, /Ug/
-------
GAC would have to be regenerated five times more often (applying the exhaus-
tion criteria of 1.0 mg/1 TOC) at considerable cost.
Due to the fact that the bed lives to exhaustion of the two carbons
varied, depending on exhaustion criteria used, and that grant objectives did
not include developing costs on the regeneration of lignite GAC, it would be
impossible to determine comparative annual costs of the two GACs. However,
it seems apparent, based on the available data, that bituminous GAC would be
the more cost effective GAC to use in water systems.
Objective 5. To Predict Phase 3 Full-Scale Performance Using Pilot Columns
Phase 2 pilot column data were used to predict the operation and the
adsorptive performance of the Phase 3 full-scale GAC systems. This informa-
tion was also used to predict the life expectancy and, therefore, was useful
in estimating the regeneration frequency. The data from this phase not only
added to our knowledge gained from Phase 1, but also filled the data voids
from Phase I.
Pilot GAC filters for Phase 2 included two runs which represented both
winter, Phase 2-1, and summer, Phase 2-2, conditions. Thus, the performance
of GAC under different influent concentrations and temperatures were
observed.
The breakthrough curves for TOC, INSTTHM, THMSIMDIST and THMFP from the
Phase 2 pilot GAC filter can be found in Figures 33 through 40. From the
figures, it is quite apparent that varying influent concentrations resulted
in varying breakthrough curves, but the same basic trends existed for both
runs. Initially, the GAC removed most of the influent TOC, THMSIMDIST and
THMFP concentrations. As the run progressed, the GAC removed a decreasing
amount of the influent concentration until a steady state was reached for
the remainder of the run. At steady state, the GAC removed a constant
percentage of the influent concentration. The trend observed in the break-
through curves for INSTTHM were similar to those detailed above except that
instead of reaching a steady state, the effluent concentration eventually
equalled the influent and then ultimately exceeded it. For the most part,
the Phase 3 GAC filters tended to mimic these results. In order to estimate
or predict the regeneration frequency for the Phase 3 GAC filters, TOC and
THMSIMDIST exhaustion criteria were established.
The TOC exhaustion criterion was defined as the point at which the
three-week running average for TOC exceeded 1,000 >ig/l (as represented by
the intersection of the treatment goal line and the breakthrough curve when
the average of any three-week set exceeds the goal). The pilot GAC filters
would have required regeneration after 24 days of operation for Phase 2-1
and 34 days for Phase 2-2. This meant that the full-scale system in Phase 3
would likely require regeneration approximately every 29 days. Applying
this TOC exhaustion criterion to the Phase 3 GAC filters resulted in an
average operating life of 24 days which approximated the predicted 29 days.
107
-------
INFLUENT CFLBO
LEGEND
VIRGIN PILOT 6AC FILTER EFF CRV5>
3500
3000
_ 2500
a
_> 2000
a:
i-
z
UJ
o
~z.
o
o
1500
1000
500
0
0
20
40
60 80
TIME, days
100
120
140
Figure 33. Total organic carbon (TOO breakthrough curve for virgin
pilot GAC filter, Phase 2-1.
-------
601
CONCENTRATION, JJQ/|
CO
o>
00
-------
OIT
(D
CONCENTRATION, >jg/
I I I I I I
rn
o
2
o
-------
LEGEND
INFLUENT CFLDO
VIRGIN PILOT SAC FILTER EFF CRV55
O
H
<
QL
t~
bJ
O
Z
O
O
408
350
300
250
200
150
100
50
0
0
20
40
60 80
TIME, days
100
120
140
Figure 36. THM formation potential CFTTT) breakthrough curve
for virgin pilot GAC filter, Phase 2-1.
-------
f\>
(D
CO
T3 —I
HI °
O Q~
CD O
>• -I
ooa
o
— O
rf
<& O
Jl Q
CT
T3 O
rr D
o
<& Q
o
ro o -i
i ^ H
rs> 2
o" n
rr « oo
T s>
o
(0
o
<
"J —
^. CO
0>
0
en
s>
SIT
CONCENTRATION,
G5 01 ^D
ro
en
oo
CO
en
e
m
CD
<
i
a
?
<^
XI
-------
LEGEND
INFLUENT CFLDO
VTOGIN PILOT SAC FILTER EFF CRV5)
60
o
H
h-
Z
LJ
O
z
o
o
30
20
10
0
0
28
40
80
60
TIME, days
Figure 38. Instantaneous THM (ITTT) breakthrough curve
for virgin pilot GAC filter, Phase 2-2.
100
120
-------
LEGEND
INFLUENT CFLBO
VIRGIN PILOT SAC FILTER EFF CRVSJ
•z.
o
H
o:
»-
ki
o
z
o
o
250
200
150
100
50
0
0
80
100
20 40 60
TIME, days
Figure 39. Seven-day simulated distribution system THM (STT7) breakthrough curve
for virgin pilot GAC filter, Phase 2-2.
120
-------
LEGEND
INFLUENT CFLDO
VIRGIN PILOT SAC FILTER EFF CRV55
400
350
300
N
° 250
z
H 200
z 150
UJ
o
z
8 100
50
0
0
28
40
80
60
TIME, days
Figure 40. THM formation potential (FTTT) breakthrough curve
for virgin pilot GAC filter, Phase 2-2.
100
120
-------
The THMSIMDIST treatment goal was not useful to this objective since
different storage times (seven days in Phase 1 and three days in Phase 3)
were used in the analytical process. Figures 35 and 39 are presented to
point out one problem with using THMSIMDIST as an exhaustion criterion. The
criterion was only valid during the summer months (Figure 39) because during
the winter months (Figure 35) the rate of THMSIMDIST formation was limited
by the cooler water temperatures and by lower concentrations of precursors.
The pilot contactor for Phase 2 involved only one run which occurred
under winter, or cold water temperature, conditions. The only operational
problem encountered during this run involved maintaining proper flow through
the columns. As mentioned earlier, this system consisted of four 10.2 cm
(4 in) I.D. glass columns in series containing 0.9, 1.2, 1.2 and 1.2 m (3,
4, 4 and 4 ft) of GAC, respectively, and a hydraulic loading of 302 Ipm/sq m
(7.0 gpm/sq ft). On rundays 91 and 141, it was necessary to backwash all
four pilot columns in order to maintain the 302 Ipm/sq m (7.0 gpm/sq ft)
loading rate. A backwash criterion was established for the full-scale
contactors which stated that if the drop across the bed exceeded 138 kPa
(20 psig), the contactor would be backwashed. However, this criterion was
never met during Phase 3 and, therefore, it was not necessary to backwash
any contactors.
The breakthrough curves for TOG, INSTTHM, THMSIMDIST and THMFP from the
pilot contactor can be found in Figures 41 through 44. The same trend of
breakthrough, gradual rise and steady state for TOC, THMSIMDIST and THMFP,
described earlier for pilot GAC filters, was observed here. Also, as before,
the INSTTHM in the effluent eventually equalled and then exceeded the
influent INSTTHM concentration.
The TOC exhaustion criterion, defined earlier, resulted in the predic-
tion of 180 days of operation for contactors before requiring regeneration.
Applying this exhaustion criterion to the Phase 3 contactors, resulted in an
average operating life of 89 days with a range of 36 to 168 days. It can be
seen that the predicted life of 180 days was at the upper range of the
observed values. This emphasizes the fact that caution should be employed
when using only one run upon which to base projections.
From Figure 43, it can be seen that the THMSIMDIST concentration in the
influent and effluent was increasing. This increase in THMSIMDIST concentra-
tion was due to the increasing water temperature and increasing concentra-
tions of precursors. However, the effluent THMSIMDIST concentration never
exceeded the 100 ug/1 exhaustion criterion.
For the most part, the pilot systems did give a good indication of what
to expect from full-scale systems. However, the main objective of this
grant was the operation of full-scale systems and not pilot systems. Anyone
wishing to use pilot systems to predict full-scale systems should perform
several runs covering all seasons of the year.
116
-------
LII
CONCENTRATION,
©
I
-------
INFLUENT CSFEF3
LEGEND
BITUMINOUS PILOT CONTACTOR EFF CPFAS)
180 150
TIME., days
200
250
Figure 42. Instantaneous THM CITTT) breakthrough curve
for bituminous pilot contactor, Phase 2-0.
-------
INFLUENT CSFEF3
LEGENP
BITUMINOUS PILOT CONTACTOR EFF CPFA55
0
0
200
100 150
TIME, days
Figure 43. Seven-day simulated distribution system THM CSTT7) breakthrough curve
for bituminous pilot contactor, Phase 2-0.
250
-------
021
CONCENTRATION,
01
o
en
en
o>
c
O IE
o~ -*> tn
— o
O O -I
O — M
3 3
**• *~* m
Q -n v
o —i
rf — I Q.
o — i a
m
00
CO 7C
O rt-
rs> -»
I O
S> C
o
1
ID
cn
ro
tn
-------
PHASE 3. FULL-SCALE GAC FILTERS AND CONTACTORS WITH ON-SITE REGENERATION
Objective 6. To Compare the Relative Performance of Full-Scale GAC
Filters and Contactors
As mentioned earlier. Phase 3 dealt with the study of GAC systems in
two operational modes. In the first mode, the systems were acting as GAC
filters or sand replacement systems. Therefore, the GAC acted both as a
filter for carryover solids and as an adsorbent material for dissolved
organics. In the second mode, the systems were used as adsorbers only since
they received water which already had the carryover solids removed by conven-
tional sand filters.
Comparison of Equal Contact Times—
In this section, the effluent from GAC filters was compared to the
effluent of a sample point in the contactor with a similar EBCT. This
comparison helped to show what effect, if- any, the filtering function of a
GAC filter had on its ability to adsorb dissolved organics. The EBCT for a
filter is 7.5 minutes which is within 5% of the 7.2 minutes EBCT for the
contactor sample point used.
Two comparison runs were performed during Phase 3. The first run
matched 15AE and D7 from Phase 3-0, while the second run matched 15AE and C7
from Phase 3-1. For the most part, the findings from 15AE and D7, Phase 3-0,
were similar to those from 15AE and C7, Phase 3-1. Since these two systems
paralleled, the discussion will be limited to 15AE and D7, Phase 3-0, unless
otherwise stated. The matching graphs for 15AE and C7, Phase 3-1, in addi-
tion to other graphs for both comparisons can be found in Volume Two.
Figures 45 through 49 are the breakthrough curves for TOC, THMSIMDIST,
INSTTHM and THMFP. These plots serve two purposes: 1) they enable the
comparison of similar EBCT in two different operational modes (GAC filters
and contactors); and 2) they present a comparison of what-the consumer would
receive with and without GAC treatment.
Figure 45 shows that the GAC filter effluent TOC concentration was
slightly higher than the similar EBCT contactor concentration. However, the
GAC filter TOC influent concentration was also slightly higher. Overall, it
appears that the two systems were performing in a similar manner.
Figure 46 shows that the THMSIMDIST curves for the two similar EBCTs
are identical. The two influent THMSIMDIST concentration curves are also
identical. It appears that the solids removed by the sand filters inter-
fered little with the GAC's ability to remove THMSIMDIST precursor.
Figure 47 shows that the GAC filter influent INSTTHM concentration is
slightly lower than the influent concentration for the contactor. This can
be explained by the fact that the influent to the contactor was the effluent
from the sand filter which provided a longer reaction time between the THM
precursors and chlorine. The average influent concentration throughout the
121
-------
rt-
— o_
ft>
3 O
<-»- -J
O
§ §
Q_ —
O
O
O O
3 Q
Q
O
rr-
O
"J
O
3
O O
O
— *-/
(t
< cr
» T
—
Q
?3
o c
CO
m
o>
-------
LEGEND
NJ
Co
z
o
H
Z
U
O
Z
O
o
140
128
_ 100
X
0»
3,
80
60
40
20
0
0
FILTER INFLUENT CFUDO
CONTACTOR INFLUENT CSFEF)
GAC FILTER ISA EFF CISAE)
CONTACTOR D LEVEL 7 CD7>
20
40
60
120
140
160
180
80 100
TIME, days
FIGURE 46. Three-day simulated distribution system THM (STT3) breakthrough curves
for GAC Filter ISA effluent and Contactor D level 1, Phase 3-0.
-------
LEGEND
5
z
o
H
hl
O
z
o
o
88
78
68
58
48
38
28
18
8
8
FILTER INFLUENT CFLINJ
CONTACTOR INFLUENT CSFEF)
GAC FILTER ISA EFF C15AE>
CONTACTOR D LEVEL 7 CD7)
48
68
128
148
168
88 188
TIME, days
FIGURE 47. Instantaneous THM CITTT) breakthrough curves for GAC Filter ISA
effluent and Contactor D level 7't Phase 3-8.
188
-------
LEGEND
tv)
Ul
H
458
400
350
308
2 250
o
H
< 200
I 150
2
O
° 100
50
0
0
FILTER INFLUENT CFLDO
CONTACTOR INFLUENT CSFEF>
GAC FILTER ISA EFF O5AE>
CONTACTOR D LEVEL 7 CD7>
20
40
60
120
140
160
80 100
TIME, days
FIGURE 48. THM formation potential CFTTT) breakthrough curves for GAC Filter ISA
effluent and Contactor D level 7, Phase 3-0.
180
-------
LEGEND
KJ
cn
108
FILTER INFLUENT CFLBO
CONTACTOR INFLUENT CSFEF)
GAC FILTER ISA EFF C15AE5
CONTACTOR C LEVEL 7 CC7)
0
10
15
35
40
45
20 25
TIME, days
FIGURE 49. Instantaneous THM CITTT) breakthrough curves for GAC Filter ISA effluent
and Contactor C level 1, Phase 3-1.
-------
run to the GAC filter was 21.1 ug/1 compared to 28.2 ug/1 average influent
concentration for the contactor. Although a rigid comparison was not pos-
sible, it appears that GAC systems with similar EBCTs followed the same
trends. Figure 48 is a plot of the INSTTHM breakthrough curves for 15AE and
C7, Phase 3-1. In this case, the two influents were closer in concentration
due to the warmer water temperature causing the rate of formation to
increase, thus reducing the effect of the longer contact time. The average
influent concentration throughout the run to the GAC filter was 81.0 ug/1
which was similar to the 89.4 >-ig/l average influent concentration for the
contactor. Figure 49 is a plot of the percent removal of influent INSTTHM
for 15AE and C7, Phase 3-0, which shows that the two systems performed
roughly the same.
Figure 50 shows that the GAC filter effluent THMFP concentration was
slightly higher thc.n the similar EBCT contactor concentration. This differ-
ence can be explained by the differences in INSTTHM concentration observed
in Figure 47. Figure 51 shows the two systems were very similar in their
removal of influent THMFP.
Tables 40 and 41 contain the minimum, maximum and average values for
the purgeable halogenateds (other than the INSTTHMs) for 15AE and D7, Phase
3-0, and 15AE and C7, Phase 3-1, respectively. In both tables, the influent
concentration, for the most part, approached the lower detection limit of
the instrument. The two similar EBCT systems performed approximately the
same. If the dichloromethane concentration represents a real value and not
a laboratory artifact, the GAC has little adsorptive capacity for it. GAC
was very effective in removing low concentrations of other purgeable halogen-
ated compounds.
Tables 42 and 43 contain the minimum, maximum and average values for
the purgeable non-halogenated organic compounds for 15AE and D7, Phase 3-0,
and 15AE and C7, Phase 3-1, respectively. For the most part, the influent
concentrations approached the detector's lower limit. Except for ethylben-
zene for 15AE and D7, Phase 3-0, the two similar EBCT systems performed the
same. Some of the results contained in Table 42 may be due to laboratory
artifacts since the analytical method was still being modified during this
run. It appears from Table 43 that at lower ug/1 influent concentrations,
GAC was ineffective to moderately effective in removing benzene and ethylben-
zene. Toluene and hexane were removed more effectively. Tetralin and
o-xylene were not detected, except for one case of o-xylene in the influent.
CAOX analyses were performed on samples from 15AE and D7, Phase 3-0.
Figure 52 is a plot of CAOX breakthrough curves for 15AE and D7, Phase 3-0.
Since the CAOX formation was dependent on chlorine contact time, the previous
discussion on differing influent concentrations for INSTTHM also applies
here. Generally, the two similar EBCT systems follow each other. On runday
113, it appears that either 15AE was desorbing CAOX or that the datum is
questionable. The datum on runday 85 for the contactor influent is also
questionable.
127
-------
6AC FILTER ISA EFF C15AE5
LEGEND
CONTACTOR C LEVEL 7 CC7)
K)
00
8
18
15
38
35
48
28 25
TIME, days
FIGURE 58. Instantaneous THM (ITTT) percent removal curves for GAC Filter ISA
effluent and Contactor C level 7, Phase 3-1.
45
-------
6AC FILTER ISA EFF C15AE5
LEGEND
CONTACTOR D LEVEL 7
100
to
<£>
0
t.
O
2:
kJ
20 .
10
20
40
80 100
TIME, days
120
140
(60
FIGURE 51. THM formation potential (FTTT) percent removal curves for
GAC Filter ISA effluent and Contactor D level 1, Phase 3-8
180
-------
LEGEND
FILTER INFLUENT CFLHO
CONTACTOR INFLUENT CSFEFJ
GAC FILTER ISA EFF CI5AE)
CONTACTOR D LEVEL 7 CD75
UJ
o
160
140
120
100
80
60
40
20
0
FIGURE
^
Z
o
H
Of.
y-
z
Id
O
Z
O
O
0
20
40
60
120
140
160
80 100
TIME, days
52. Carbon adsorbabIe organohaIides (CAOX) breakthrough curves for
GAC Filter ISA effluent and Contactor D level 7, Phase 3-0.
180
-------
TABLE 40. PURGEABLE HALOGENATED ORGANICS FOR CONTACTOR D LEVEL 7 AND GAC FILTER ISA EFFLUENT, PHASE 3-0
Influent 15AE Influent D7
Min
Value
Parameter
Carbon tetrachloride
1,2-Dichloroethane
1,1,1-Trichloroethane a
Dichloromethane
Tetrachloroethylene
Trichloroethene
Chlorobenzene
o-Dichlorobenzene
Min Max
Value Value
a 0.
a a
i a 0.
0.2 0.
a 0.
a 0.
a 0.
a 0.
2
2
6
2
2
2
2
Avg
Value
(ug/1)
0
0
0
0
0
0
0
.0
a
.1
.2
.1
.1
.0
.1
Min Max Avg Min
Value Value Value Value
(ug/1) (pg/1) (ug/1) (ug/1)
a a a a
a a a a
a 0.2 0.0 a
a 0.6 0.2 a
a a a a
a 0.2 0.0 a
a a a a
a a a a
Max
Value
:_ (ug/1)
0.
a
0.
0.
0.
0.
0.
0.
2
2
6
2
2
2
2
Avg Min Max Avg
Value Value Value Value
(ug/i) (ug/i) (pg/1) (ug/i
0.
a
0.
0.
0.
0.
0.
0.
0 a
a
0 a
2 a
1 a
1 a
0 a
0 a
a a
a a
0.2 0.0
0.5 0.2
a a
a a
a a
0.2 0.0
Not Detected.
-------
TABLE 41. PURGEABLE HALOGENATED ORGANICS FOR CONTACTOR C LEVEL 7 AND GAC FILTER ISA EFFLUENT, PHASE 3-1
Influent 15AE Influent C7
Min Max Avg Min Max Avg Min Max Avg Min Max Avg
Value Value Value Value Value Value Value Value Value Value Value Value
Parameter
Carbon tetrachloride a
1,2-Dichloroethane a
1,1,1-Trichloroethane a
Dichloromethane a
Tetrachloroethylene a
Trichloroethene a
Chlorobenzene a
o-Dichlorobenzene a
jg/1) (ug/1) (ug/1) (ug/1) (ug/1)
a a a a a
a a a a a
a a a a a
a a a a a
0.2 0.0 a a a
0.2 0.1 a a a
a a a a a
a a a a a
3/1) (ug/1) (ug/1) (ug/1)
0.2
0.2
a
a
0.0
0.1
a
a
a
a
a
a
a
a
a
a
0.2
a
a
a
a
a
a
a
0.0
Not Detected.
-------
Cd
U)
TABLE 42. PURGEABLE NON-HALOGENATED ORGANICS
PHASE 3-0
Influent
Parameter
Benzene
o-Xylene
Ethylbenzene
Toluene
Hexane
Tetralin
Min
Value
(ug/l)
0.2
a
a
a
a
a
Max
Value
(ug/1)
1.4
a
20.8
1.8
1.1
1.2
Avg
Value
(ug/1)
0.4
a
1.9
0.5
0.1
0.1
Min
Value
. (ug/1)
0.2
a
a
0.2
a
a
FOR CONTACTOR
15AE
Max
Value
(ug/D
0.9
a
2.4
2.2
1.2
1.2
D LEVEL 7 AND
GAC FILTER 15A EFFLUENT,
Influent
Avg
Value
(ug/1)
0.3
a
0.3
0.5
0.1
0.1
Min
Value
(ug/1)
a
a
a
a
a
a
Max
Value
. (ug/D
1.2
a
13.4
1.2
1.5
2.4
Avg
Value
(ug/1)
0.3
a
1.7
0.4
0.2
0.1
D7
Min Max
Value Value
(ug/l) (ug/l)
0.2 0.8
a a
a 0.8
0.2 2.8
a 1.0
a 0.9
Avg
Value
(ug/l)
0.3
a
0.1
0.4
0.1
0.0
Not Detected.
-------
TABLE 43. PURGEABLE NON-HALOGENATED ORGANICS FOR CONTACTOR C LEVEL 7 AND
PHASE 3-1
Influent 15AE Influent
Parameter
Benzene
o-Xylene
Ethylbenzene
Toluene
Hexane
Tetralin
Min Max Avg Min Max Avg Min Max
Value Value Value Value Value Value Value Value
(ug/1) (ug/1) (ug/1) (ug/1) (ug/1) (ug/1) (ug/1) (yig/1)
0.2 0.7 0.4 0.2 0.4 0.2 0.2 1.0
a a a a a a a 0.2
a 0.6 0.1 a 0.2 0.0 a 0.2
a 0.4 0.2 a a a a 0.9
a 0.2 0.1 a a a a 0.6
aaaaaaaa
GAC FILTER ISA EFFLUENT,
C7
Avg Min
Value Value
(ug/1) (ug/1)
0.6 0.2
0.0 a
0.0 a
0.2 a
0.2 a
a a
Max
Value
(ug/1)
0.6
0.2
0.2
0.2
0.7
a
Avg
Value
Qig/1)
0.4
0.0
0.1
0.0
0.1
a
a = Not Detected.
-------
Acid extract GC/FID profiles were performed on samples from 15AE and
D7, Phase 3-0. Figures 53 through 56 are the profiles after one week, three
months, four months and five months of operation, respectively. These
profiles enable a comparison of the two similar EBCT systems and also a
comparison of what the consumer would receive with and without GAC treatment.
Figure 53 shows that the GAC in both systems was effectively removing
organics after one week of operation. This is evidenced by the lesser
number of peaks and by lesser peak heights at the same retention time. It
appears that D7 was adsorbing more effectively than 15AE during the first
half of the chromatogram. This difference may be due to slight variations
in detector response. The last third of the chromatogram for D7 shows peaks
of greater magnitude than 15AE. Three possible explanations for this are:
1) the peaks were real since they are present to some extent in the influent;
2) they were picked up during the concentration procedure since they are
present to some extent in the solvent blank; or 3) the contactor liner was
leaching organics. Therefore, 15AE and D7 were similar in their performance
after one week of operation.
The third month of operation (Figure 54) shows an improvement in the
quality of the influent water when compared to the first week (Figure 53).
The GAC effluents also appear to be improved when compared to the effluents
after one week of operation. The GAC was not effective in removing organics
present in the first third of the chromatogram. After the first third of
the chromatogram, the GAC became effective in removing organics. The last
third of the D7 chromatogram shows more peaks of greater magnitude similar
to those observed in Figure 53. Except for this section of the chromatogram,
15AE and D7 were similar in their performance after three months of opera-
tion.
The performance of the two GAC systems after four months of operation
is presented in Figure 55. The figure shows a significant difference between
15AE and D7 during the first third of the chromatogram. It appears that
while D7 was still removing organics, 15AE was starting to desorb organics.
During the rest of the chromatogram, both GAC systems continued to remove
peaks observed in the influent. The last third of the chromatogram for D7
does not show the peaks observed earlier. This tends to support the hypoth-
esis that the contactor liner was leaching organics. Overall, the perform-
ance of 15AE and D7 no longer appear to be similar after four months of
operation.
The performance of the two GAC systems after five months of operation
is presented in Figure 56. During the first third of the chromatogram, some
desorption was occurring in both GAC systems. Both GAC systems were still
removing organics after the first third of the chromatogram. Once again,
15AE and D7 were very similar in their performance after five months of
operation.
Attempting to use acid extract GC/FID profiles to compare the two
similar EBCTs from two different GAC systems has pointed out the fact that,
in this case, a rigid comparison is not really possible. For example, it is
135
-------
BLANK
... - P7
FLIN
-u
DT
to
CD
cr
/O
rn
en
co
o
Q-
T (D
CO X
I — H-
cn i
<• J» O
o
c -»»**"
3 ~*7*O
o c" o
-< (!> -h
(D
-*J
O
Q —
O rt-
'C' 3
—
(D C
d> D
—^-j^— ^i*^
\ •
V
*^J
" CD
O
SEI
1
CO
—i
o
CO H
-H 2:
Li
fe?
-------
. ir
j
it
liiLiM
D -
INT,
STD,
LU
LO
uL.
CO
INT,
STD,
INT,
STD,
INT,
STD,
FIGURE 54. Acid extract profiles for filter influent,
GAC Filter ISA effluent and Contactor D
level 7'f Phase 3-0, runday 85.
137
-------
Mil it1
T(
I ' 1
JJuUtA1
I
INT,
STD,
INT
STD
INT,
STD.
FIGURE 55. Acid extract profiles for filter influent, GAC
Filter ISA effluent and Contactor D level 1,
Phase 3-0, rundoy 113.
138
-------
ra
137
LU
or
f «
'* *
CQ
tnj
• 4
• i
- INT.
STD.
_LJU—L
INT.
STD.
INT,
STD,
FIGURE 56. Acid extract profiles for filter influent, GAC
Filter ISA effluent and Contactor D level 7,
Phase 3-0, runday 141.
139
-------
possible to have desorption occuring in one system and not in the other at
the time of sample collection as observed in Figure 55. This would cause
one to conclude that the two systems were not performing equally. However,
the next samples, collected one month later, showed the two systems to be
equal. In order to obtain a more accurate comparison, it would be necessary
to have more frequent sampling. Following this consideration and assuming
that the peaks observed in the first two D7 samples during the latter third
of the chromatogram were due to the leaching of contactor liner, it would
appear that 15AE and D7 performed similarly.
The percent of occurrence of GC/MS tentatively identified compounds for
the acid extracts from GAC Filter ISA can be found in Table 44. The results
from D7 will be discussed later in this objective with the Contactor D
results. The results in the table deal only with the percent of occurrence
of the compounds and not their concentrations. Accordingly, use of these
data in discussing the ability of GAC to remove these compounds is difficult.
The following points should be taken into consideration when examining these
data with respect to removal by GAC:
1. Only compounds with a high percent of occurrence should be examined.
In this case a high percent of occurrence will be defined as 30% or
greater.
2. Compounds which have a high percent of occurrence in the influent and
not in the effluent are not necessarily well adsorbed by GAC. Given
the limits of detection of the GC/MS, it is possible that the concentra-
tion of the influent was just sufficient for the GC/MS to identify
while the effluent concentration was not.
3. Compounds with a higher percent of occurrence in the effluent than in
the influent could be due to either the limit of detection explained
above or desorption of the compound from the GAC.
Table 44 shows that of the 46 compounds identified, only seven had a
high percent of occurrence. Of these, 1,3,5-trimethyl-l,3,5-triazine-
2,4,6-(lH,2H,3H)-trione, nitrobenzene and 3,3,3-trichloro-l-propene appeared
to be well adsorbed by the GAC, while 2,H-pyran-2-one,2-cycloheKene-l-one,
dibromochloromethane, 7-oxabicyclo[4.0.1]heptane and 2,H-pyran-2-one varied
from marginal to only slightly adsorbed. Compounds identified in the blank
which would negate their identification in the samples included diethyl
ester-1,2-benzenedicarboxylic acid, 2-cyclahexene-l-ol and 3,3-dimethyl-
hexane. The other 36 compounds occurred too infrequently to draw any con-
clusions as to the ability of GAC to adsorb them given the above considera-
tions.
Breakthrough and percent removal graphs show the two GAC systems with
similar EBCTs to be equivalent. Next, in order to compare the operating
condition of the two systems, it was necessary to select exhaustion criteria
for various parameters. Arbitrary exhaustion criteria were selected for
TOC, INST CHC1 , INST CHBrCl , THMFP and THMSIMDIST, thus covering
individual organic compounds, groups of specific and non-specific organic
compounds.
140
-------
TABLE 44. ACID EXTRACT COMPOUNDS TENTATIVELY IDENTIFIED
BY GC/MS FROM GAC FILTER 15A
Percent of Occurrence
Compound
l,3,5-Trimethyl-l,3,5-triazine-2,4,6-(lH,2H,3H)-trione
Dibromochloromethane
Nitrobenzene
3,3,3-Trichloro-l-propene
2-Cyclohexene-l-one
7-Oxabicyclo(4.0.1)heptane
2,H-Pyran-2-one
Diethyl ester-l,2-benzenedicarboxylic acid
3-Penten-2-one
1-Pentyne
Butyl-2-methyl propylester-1,2-benzenedicarboxylic acid
l-Chloro-2-butene
1,4-Dichlorobenzene
Benzoic acid
2-Butene
Carbon tetrachloride
6-Chloro-N-ethyl-(lH ethylmethyl)-l,3,5-triazine
Cyclohexane
2-Cyclohexene-l-ol
Dibutyl ester ethanoic acid
2,2-Dimethylbutanal
2,3,-Dimethyl-1-butene
4,4-Dimethyl-l-pentene
1,4-Dinitropentane
3,4-Epoxy-2-hexanol
3-Ethyl-4-methylfurandiene
3-Methyl-2-butadiene
2-Methyl-l,4-dinitrobenzene
2-Nitropropane
Tribromomethane
Tributyl ester phosphoric acid
TriethyIborane
4,4,5-Trimethyl-2-hexene
2,2,3-Trimethylpentane
Bromocyclohexane
1,3,5-Cycloheptatriene
Cyclohexene
1,4-Dichlorobutane
l-Ethyl-2-methylbenzene
2-Hexen-3-one
1,7-Octadiene
Pentadinitrile
2-Pentanone
TrimethyIborane
2,2,5-Trimethylhexane
3,3-Dimethylhexane
, Based on ten samples.
Not detected.
FLIN 15AE Blank
90
60
50
50
40
40
30
20
20
20
10
10
b
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
b
b
b
b
b
b
b
b
b
b
b
b
20
50
b
b
30
40
10
10
b
10
20
20
20
b
b
b
b
10
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
10
10
10
10
10
10
10
10
10
10
10
b
10
b
b
b
10
10
b
20
b
b
b
b
b
b
b
b
b
b
10
10
b
b
b
b
b
b
10
b
b
b
b
b
b
b
b
b
b
b
b
10
b
b
b
10
b
10
1-11
-------
The criteria and results for 15AE and D7 , Phase 3-0, can be found in
Table 45 and for 15AE and C7, Phase 3-1, in Table 46. The service times, or
length of operations, for all exhaustion criteria defined were identical
except for INST CHBrCl in Phase 3-1 where C7 would have been kept on-line
one week longer. Thus, the two systems performed identically based on a
single point comparison. The carbon use rates for the two GAC systems were
similar with the GAC filter performing slightly better. The mass loading
values were similar except for INST CHC1 and INST CHBrCl and to a certain
extent THMFP. These differences in loading were probably aue to the differ-
ences in INSTTHM concentrations previously discussed. The effect of varying
temperature and influent concentrations on service time and the carbon use
rate is readily apparent by comparing Tables 45 and 46. Table 45 represents
lower temperatures and, therefore, lower CHC1 CHBrCl and THMSIMDIST
concentrations which resulted in significantly longer service times and
lower carbon use rates. The influent TOC and THMFP concentrations in
Table 45 represent the lower concentrations and, therefore, the longer
service times and carbon use rates. Although the carbon use rates and
service times are significantly different between Tables 45 and 46, the mass
loadings were similar except where desorption was occurring.
It is apparent that there was no major difference in adsorptive behavior
between GAC filters and contactors with a similar EBCT. Thus, the GAC
filters acting as a filter for carryover solids had little apparent effect
on the GAC performance as an adsorbent material for dissolved organics.
Effect of Various Contact Times--
The improved guality of water achieved by increased EBCT is shown by
the EBCT breakthrough curves for Contactor D, Phase 3-0, for INST CHC1
INST CHBrCl TOC, THMFP, THMSIMDIST and organohalides in Figures 57 through
62, respectively. Further evaluation of the breakthrough curves is possible
by selecting various effluent exhaustion criteria and comparing the carbon
use rates for the various EBCTs. It should be noted that.only systems which
remained in service long enough to fulfill the exhaustion criteria are
included in the following discussion. Lower carbon use rates would have
resulted had this data been available.
Effluent exhaustion criteria of 5>jg/l for INST CHC1 l^wg/1 for INST
CHBrCl and 1,000 >ig/l for TOC were selected. Exhaustion occurred when the
effluent concentration exceeded a criterion and remained above it for at
least three weeks. The exhaustion runday was then selected from the inter-
section of the horizontal criterion line and the breakthrough curve at or
just before the first data point in the three-week set. Ideally, the carbon
use rate should be calculated using parameters which have a constant concen-
tration in the influent. However, this is not the case in full-scale systems
where the influent concentration tends to fluctuate and follow seasonal
trends. Thus, if the influent concentration is decreasing, the carbon use
rates for the various EBCTs would tend to favor the longer contact times.
The reverse of this is true when the influent concentration is increasing.
Figure 63 presents a graph of the INST CHC1 carbon use rate for various
142
-------
TABLE 45. EXHAUSTION CRITERIA FOR CONTACTOR D LEVEL 7 AND GAC FILTER 15A EFFLUENT, PHASE 3-0
Exhaustion Criteria
TOC
1000
Parameter 15AE
Service Time, days 57
GAC Use, kg/mil gal 153
Mass Load, gm/kg 24.2
Percent Removal 30.5
a Desorption is occurring.
ug/1
D7
57
164
22.9
19.6
CHC13
5 ug/1
15AE 07
64 64
136 146
0.07 0.16
- 1.4 51.5
CHBrC12
5 ug/1
15AE 07
148 148
59.2 63.0
0.04a 0.16
51.3 48.1
THMFP
100 u/gl
1 5AE D7
57 57
153 164
3.2 2.6
43.8 38.4
THMSIMDIST
100 ug/1
15AC D7
169 169
51.8 55.5
1.9 1.9
16.2 14.9
-------
it.
TABLE 46. EXHAUSTION CRITERIA FOR CONTACTOR C LEVEL 7 AND GAC FILTER ISA EFFLUENT
Exhaustion Criteria
TOC
Service
GAC Use
Parameter
Time , days
, kg/mil gal
Mass Load, gm/kg
Percent
Removal
1000
15AE
22
410
20.6
48.2
ug/1
C7
22
429
21.0
52.5
CHC13
5 ug/1
15AE C7
15 15
604 628
0.29 0.37
54.7 73.6
CHBrC12
5 ug/1
15AE C7
29 36
310 259
0.17 0.23
39.9 44.3
THMFP
100 u/gl
15AE C7
29 29
310 324
2.7 2.4
63.3 48.4
, PHASE 3-1
THMSIMDIST
100 ug/1
15AE C7
36 36
249 260
1.6 1.5
29.6 24.2
-------
LEGEND
80
70
60
® 50
z
o
H
<
Of.
h-
Z
hi
O
Z
O
o
INFLUENT CSFEF5
CONTACTOR D LEVEL 7 CD7>
CONTACTOR D LEVEL 4 CD4> X X
CONTACTOR D EFF CDE3 H h
40
30
20
10
0
0 50 100 150
TIME, days
Figure 57. Instantaneous chloroform CICLR) breakthrough curves
for Contactor D, Phase 3-0.
200
250
-------
LEGEND
X
Q
O
H
z
UJ
O
z
O
O
B 0 INFLUENT C9FEF>
A A CONTACTOR D LEVEL 7 CD73
CONTACTOR D LEVEL 4 CD4)
CONTACTOR D EFF CDED
30
25 _
20 _
15
10 _
5 _
0
0
50
200
100 150
TIME, days
Figure 58. Instantaneous bromodichIoromethane CICL2) breakthrough curves
for Contactor D, Phase 3-0.
250
-------
Lvl
CONCENTRATION,
-n
-------
cs>
CONCENTRATION, >jg/
01 o
-------
-n
c
o>
CONCENTRATION, jug/
N) 4x O) 00 CS> !\)
O> 00
rn
o
m
-------
051
T
(D
CONCENTRATION, >jg/I
ro
CD
O)
00
CD
CS>
ro
CD
O)
S3
rn
I
o
f ^
s
I
P
vi
I
-------
LEGEND
£
Q
JK
LJ
UJ
PHASE 3-1
CONTACTOR D, PHASE 3-1
CONTACTOR C, PHASE 3-2 *—X
AVERAGE
8
2 4 6 8 18 12
EMPTY BED CONTACT TIME, minutes
Figure 63. Instantaneous chloroform CICLR) carbon use rates for
multiple runs of contactors, Phase 3.
14
16
-------
contactor runs against EBCT. From the figure, it is apparent that longer
EBCTs do provide better utilization of the GAC. In order to discuss the
effects of the changing influent concentrations, both increasing and decreas-
ing influent concentrations were examined. Contactor A, Phase 3-1, repre-
sents a system which had a decreasing influent concentration and Contactor C
Phase 3-2, represents a system with an increasing influent concentration.
While both systems show varying carbon use rates, they also show that a
longer EBCT provides a better GAC use rate.
The optimum EBCT for the majority of the systems appears to be between
7 and 15 minutes. However, the systems which came on line in August and
September, or during the warmer water temperature months, appear to have an
optimum EBCT somewhere beyond 15 minutes. Similar observations to the INST
CHC1_ results were found for INST CHBrCl_ and TOC presented in Figures 64
and 65, respectively.
The effects of longer contact time on compounds represented by the acid
profiles can be seen in Figures 66 and 67. Figure 66 presents a case where
the influent contains many peaks of a large magnitude and Figure 67 an
influent which contains few peaks of a large magnitude. In both cases, the
GAC adsorbed the substances contained in the influent. It appears that
longer contact time did not provide additional removal of the substances
contained in the influent. However, longer contact times may provide addi-
tional protection from higher influent concentrations.
The percent of occurrence of GC/MS tentatively identified compounds for
the acid extracts from Contactor D can be found in Table 47. The qualifiers
listed in the GAC Filter ISA discussion also apply here. Table 47 shows
that, of the 45 compounds identified, only seven had a high percent of
occurrence, 30% or greater. Of these, dlbromochloromethane, 3,3,3-trichloro-
1-propene, 1,3,5-trimethyl-l,3,5-triazine-2,4,6-(lH,2H,3H)trione, nitroben-
zene and tributyl ester phosphoric acid appeared to be well adsorbed while
2-cyclohexene-l-one and 7-oxabicyclo[4.0.2]heptane varied from marginal to
not well adsorbed. Compounds identified in the blank which would negate
their identification in the samples included dibutyl ester ethanoic acid and
2,2-proanyl chloride. The other 36 compounds occurred too infrequently to
draw any conclusions as to the ability of GAC to adsorb them.
Grob CLSAs were performed on Contactor A influent and effluent samples,
Phase 3-0, at startup and at approximately four week intervals thereafter
for a total of eleven samples. Approximately 225 compounds were identified
with an average of 106 on any one date and a range of 84 to 130 compounds.
Except for the four THMs all concentrations were in the very low part per
trillion or ng/1 range. The health significance of these concentrations is
unknown and is well beyond the scope and the intent of the project. In
order to provide a point of perspective for these concentrations, Table 48
presents the MCLs which have been established, to date, for organic
chemicals. The range of MCL concentrations is from 0.0002 to 0.1 mg/1. The
average concentration of the non-THM compounds observed was well below
0.00001 mg/1 (about 1/20 of the lowest MCL).
152
-------
LEGEND
UJ
580
13—H CONTACTOR A, PHASE 3-0
A—A CONTACTOR D, PHASE 3-0
O e> CONTACTOR A, PHASE 3-1
0—0 CONTACTOR C, PHASE 3-1
•CONTACTOR A, PHASE 3-2
CONTACTOR B, PHASE 3-0 X X
CONTACTOR BB, PHASE 3-0 H h
CONTACTOR BB, PHASE 3-1 O 0
CONTACTOR D, PHASE 3-1 X—Z
CONTACTOR C, PHASE 3-2 X %
450 .
£
O
\
hi
h-
hi
(0
Z
o
DQ
QL
<
O
AVERAGE
0
0
4 6 8 10
EMPTY BED CONTACT TIME, minutes
14
16
Figure 64. Instantaneous bromodichIoromethane (ICL2) carbon use rates for
multiple runs of contactors, Phase 3.
-------
LEGEND
Ln
E
O
LJ
H-
LJ
CO
O
£0
O
350
300
250
200
150
100
50
a m CONTACTOR A, PHASE 3-8
A—A CONTACTOR C, PHASE 3-0
<$ 0 CONTACTOR BB, PHASE 3-0
0 0 CONTACTOR BB, PHASE 3-1
CONTACTOR A, PHASE 3-2
0
CONTACTOR B, PHASE 3-8 X X
CONTACTOR D, PHASE 3-8 H h
CONTACTOR A, PHASE 3-1 0 ©
CONTACTOR C, PHASE 3-1 X X
CONTACTOR C, PHASE 3-2 * %
AVERAGE
0
24 6 8 10 12
EMPTY BED CONTACT TIME, minutes
Figure 65. Total organic carbon CTOC) carbon use rates for
multiple runs of contactors, Phase 3.
14
16
-------
K
CO
i rr
,l
jl
T r J
" [ 11
' ,
1
1
I
I
•
i
i
i
t t
j'-'lJu
i$/
1 t
UL
-
vil'v
U
Is
L
liu
INT.
STD.
^
-------
fe:
CO
I I
fe
UJ
II
• i
J-Jj—^
INT.
STD.
INT.
STD.
INT,
STD,
CO
INT,
STD,
FIGURE 67. Acid extract profiles for Contactor D,
Phase 3-0, runday 113.
156
-------
TABLE 47.
ACID EXTRACT COMPOUNDS TENTATIVELY IDENTIFIED
BY GC/MS FROM CONTACTOR D
Percent of Occurrence
Compound
Dibromochlorome thane
3,3, 3-Trichloro-l-propene
l,3,5-Trimethyl-l,3,5-triazine-2,4,6-
7-Oxabicyclo (4. 0.1) heptane
2-Cyclohexene-l-one
Nitrobenzene
Tributyl ester phosphoric acid
Diethyl ester-l,2-benzenedicarboxylic
2-Methyl- 1 , 4-dinitrobenzene
Tribromome thane
Benzoic acid
Bromodichlorome thane
l-Chloro-2-butene
Butane
Cyclohexene
Decycl ester nitric acid
1 , 4-Dichlorobutane
1 , 1-Dichloropropane
1 , l-Dichloro-2-propanone
Dimethyl-1 , 4-dioxalane
2 , 2-Dimethylpentane
Ethylbenzene
3-Ethyl-4-methylfurandiene
3-Methyl-2-butadiene
5-Methylnonane
Nitrocyclcpentane
2-Nitropentane
1-Pentyne
1-H-Pyrrole
1-H- 1,2, 4-Triazolediamine
2,2, 3-Trimethylbutane
2,2, 3-Trimethylpentane
Cyclobutanal
2-Cyclohexene-l-ol
Cyclopropane
2 , 3-Dimethyl-l-butene
3 ,3-Dimethylhexane
4-Ethylheptane
2-Methylnaphthalene
2,4, 8-Trimethylnonane
1 , 3-Dimethylbenzene
2-Hexene-3-one
2,2,5 Trimethylhexane
Dibutyl ester ethanoic acid
2,2 Proanyl chloride
, Based on ten samples.
Based on nine samples.
Not detected.
FLINa
70
50
(lH,2H,3H)-trione 40
30
30
30
30
acid 20
20
20
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10 •
10
10
c
c
c
c
c
c
c
c
c
c
c
c
c
D7b
11
22
c
33
22
c
c
11
c
c
c
c
c
c
11
c
11
c
c
c
c
c
c
c
c
c
c
c
11
c
11
c
11
11
11
11
11
11
11
11
c
c
c
c
c
DE3
c
20
c
40
10
c
c
20
c
c
c
c
c
c
c
c
10
c
c
c
c
10
c
c
c
c
c
c
c
10
c
c
c
c
c
c
c
c
10
c
10
10
10
c
c
Blank
c
c
c
10
c
c
c
10
c
c
c
c
c
c
10
c
c
c
c
c
c
c
c
10
c
c
c
c
c
c
c
c
c
10
c
c
c
c
c
c
c
c
c
10
10
157
-------
TABLE 48. ESTABLISHED HCLS FOR ORGANIC COMPOUNDS
Compound MCL (mg/1)
Endrin 0.0002
Lindane 0.004
Methoxychlor 0.1
Toxaphene 0.005
2,4-D 0.1
2,4,5-TP(Silvex) 0.01
THMsa 0.10
Four quarter running average of four sample points.
Discussion of all 225 compounds identified would be a nearly impossible
task. Tables 49 and 50 present a data summary of the percent removals by
GAC for 55 compounds. These compounds were selected because they were
identified most frequently (on 50% of the sample days) and occurred in
higher concentrations (greater than 0.00001 mg/1 on at least one of those
days).
From Tables 49 and 50 it is apparent that, for the most part, GAC
removed the identified substances.
However, the degree of removal and the length of effectiveness varies
from compound to compound. Some compounds such as tetrachloroethane,
1,2-dichlorobenzene and hexachloroethane were well adsorbed, 85 to 100%, by
GAC over the entire sample period. Some compounds were initially well-
adsorbed, 85 to 100%, but eventually desorbed with time such as diisopropyl-
ether, benzene and carbon tetrachloride. A few compounds were adsorbed only
to a limited extent by GAC such as toluene, ethylbenzene, 1,2,4-trimethylben-
zene and 1,2,3,4-tetramethylbenzene. Nonanal and decanal-were higher in the
effluent than in the influent. It is possible that these compounds either
leached from the contactor liner, resulted from laboratory contamination
or resulted from bacterial degradation of adsorbed organics.
Objective 7. To Compare the Relative Performance of Full-Scale GAC Filters
and Contactors after Successive Regenerations
Objective 3 showed that once-regenerated GAC and twice-regenerated GAC
were very similar in adsorbent performance to virgin GAC. Laboratory regen-
erated GAC and parallel pilot column GAC filters were utilized in performing
this determination. One objective of Phase 3 was to take this finding one
step further and determine if full-scale, on-site regeneration affects the
adsorbent performance of GAC in full-scale systems.
The ideal approach to determine the effects of regeneration on the
GAC' s adsorbent performance would have been to run regenerated systems in
parallel with virgin systems as performed in the Phase 2, Objective 3, pilot
158
-------
TABLE 49. PERCENT REMOVAL DATA SUMMARY OF GROB CLSA RESULTS
FOR CONTACTOR A, PHASE 3-0, RUNDAYS 1 THROUGH 134
Compound
Di isopropy lether
Chloroform
1,1, 1-Trichloroethane
Benzene
Carbon tetrachloride
Cyclohexane
1 , 2-Dichloropropane
Trichloroethene
Bromodichlorometbane
Methylcyclohexane
4-Methyl-2-pentanone
Toluene
Dibromochlorome thane
Hexanal
Tetrachloroethene
Dichloroiodomethane
Chlorobenzene
Ethylbenzene
1,3-and 1 , 4-Dimethy Ibenzene
Bromoform
Styreae
1 , 2-Dime thy Ibenzene
Isopropy Ibenzene
Propylbenzene
E thy 1-4-me thy Ibenzene
1 , 3 ,5-Trime thy Ibenzene
1,2, 4-Trimethylbenzene
Octanal
1 , 4-Dichlorobenzene
1,2, 3-Trimethylbenzene
1,2-Dichlorobenzene
1,3-Diethylbenzene
1 , 4-Diethy Ibenzene
5-Ethyl-l,3-dimethy Ibenzene
Hexachlo roe thane
2-Ethyl-l ,4-dimethy Ibenzene
4-Ethyl-l ,2-dime thy Ibenzene
Nonanal
3-Ethyl-l,2-dime thy Ibenzene
1,2,3, 5 -Tetramethylbenzene
l,3-Diethyl-5-methylbenzene
1,2,3, 4-Tetramethy Ibenzene
Decanal
Dodecanal
2,6-Bis(l,l-dimethylethyl)2,5-
cyclohexadiene-1 , 4-dione
Pentadecane
Diethylphthalate
2,2, 4-Tnraethylpenta- 1,3-
dioldisobutyrate
Runday
1
a
100
e
a
100
h
a
a
100
d
a
38
100
-100
100
100
a
25
38
100
100
40
37
33
43
33
27
b
100
33
100
40
d
60
100
30
50
-350
70
- 50
50
60
-150
a
67
g
a
a
34
100
98
95
100
94
21
100
98
100
a
g
72
100
9
100
100
100
c
51
100
100
58
50
36
48
c
43
d
100
46
100
d
-100
e
99
50
e
64
100
73
37
a
93
g
g
98
95
a
100
3
97
34
98
100
92
22
2
100
40
17
33
0
17
e
11
77
100
100
99.0
e
h
0
98
d
20
-330 -150
e
55
0
50
-680
b
e
a
c
d
0
e
0
50
120
h
70
a
h
c
113
100
79
94
100
100
100
g
100
98
100
100
59
100
e
100
100
100
62
20
100
100
59
67
62
32
e
46
39
100
100
100
67
c
57
100
b
50
- 9
b
h
100
83 .
- 17
100
100
62
a
c
134
d
70
98
99
100
c
g
100
90
g
d
100
100
18
100
100
100
99
98
97
100
99
99
98
98
e
99
d
d
99
98
99
e
99
99
e
99
-1300
99
99
100
99
d
d
e
b
b
d
2,5-Bis(l,l-dimethylproply)2,5-
cyclohexadiene-1 , 4-dione
1 , 1 , 3-Trime thy 1-3-pheay lindan
Heptadecane
Dibutylphthalate
Dioctylphthalate
Hexane
Methylcyclopentane
t Not detected.
Not detected in influent.
Not quantified.
Influent not quantified.
, Effluent not quantified.
Not scanned.
? Influent not quantified and
b
33
33
a
a
a
a
effluent
d -200
c
d
c
a
a
a
not detected
10
180
c
g
a
a
-460
h
e
c
100
a
a
a
d
d
h
h
a
a
Influent not detected and effluent not quantified.
159
-------
TABLE 50. PERCENT REMOVAL DATA SUMMARY OF GROB CLSA RESULTS FOR
CONTACTOR A,
Compound
Diisopropylether
Chloroform
1,1, 1-Trichloroethane
Benzene
Carbon tetrachloride
Cyclohexane
1 , 2-Dichloropropane
Trichloroethene
Bromodichlorome thane
Me thy 1 eye lohexane
4-Methyl-2-pentanone
Toluene
Dibromochlorome thane
Hexanal
Tetrachloroethene
Dichloroiodome thane
Chlorobenzene
Ethylbenzene
1,3-and 1,4-Dimethy Ibenzene
Broraoform
Styrene
1,2-Dimethy Ibenzene
Isopropylbenzene
Propylbenzene
E thy 1 - 4-me thy Ibenzene
1, 3, 5-Trimethy Ibenzene
1,2, 4-Trimethy Ibenzene
Octanal
1 , 4-Dichlorobenzene
1,2, 3-Trime thy Ibenzene
1 ,2-Dichlorobenzene
1 , 3-Diethy Ibenzene
1 , 4-Diethy Ibenzene
5-Ethyl-l,3-dime thy Ibenzene
Hexachloroe thane
2-Ethyl-l,4-dimethy Ibenzene
4-Ethyl-l ,2-dimethy Ibenzene
Nonanal
3-E thy 1-1, 2-dime thy Ibenzene
1,2,3 , 5-Tetrame thy Ibenzene
l,3-Diethyl-5-me thy Ibenzene
1,2,3, 4-Tetramethy Ibenzene
Decanal
Dodecanal
2,6-Bis(l,l-dimethylethyl)2,5-
cyclohexadiene- 1 , 4-dione
Pentadecane
Die thy Iphtha late
2,2,4-Trimethylpenta-l,3-
dioldisobutyrate
PHASE
3-0, RUNDAYS 162
THROUGH
302
Runday
162
67
-340
56
100
100
90
a
100
0
100
88
90
68
63
ICO
100
100
e
71
75
95
84
96
83
e
c
68
68
99
80
99
g
83
e
99
c
73
54
b
e
a
87
65
68
94
h
29
65
190
-114
- 55
70
100
a
- 46
100
100
14
d
-166
- 83
71
- 15
100
24
a
-286
-220
83
h
- 75
0
- 33
- 25.
d
0
- 60
100
0
98
h
h
0
100
8
0
- 80
b
0
a
e
- 76
- 4
100
b
48
100
218
f
f
f
a
100
d
a
a
- 87
g
b
42
62
e
100
100
a
e
33
90
a
e
b
a
56
c
36
e
100
d
100
c
c
d
100
d
a
- 23
b
d
a
b
9
- 5
100
- 45
g
a
246
g
75
c
40
e
c
a
100
35
c
a
15
-714
a
e
a
h
c
a
c
a
e
g
g
100
e
57
h
e
e
100
g
g
100
a
e
100
- 48
g
e
a
g
- 22
h
c
- 44
42
a
274
b
- 8
- 35
- 4
e
c
h
a
29
g
g
- 21
73
0
86
100
c
-3900
a
100
e
-2000
d
c
- 43
e
-100
44
89
e
100
c
S
e
100
c
e
23
c
e
g
e
19
6
100
42
e
91
302
g
-1200
d
- 46
- 96
c
h
17
-580
g
a
- 75
- 93
g
92
100
g
65
a
50
e
23
c
c
-350
16
- 24
d
36
- 12
100
c
c
e
100
e
-220
- 15
g
0
g
e
- 10
100
- 52
75
35
- 2
2,5-Bis(l,l-di
-------
column work. This approach would eliminate the problem of varying influent
concentrations when two systems were brought on line at different times.
However, another more important objective of Phase 3 was to determine the
GAC losses across the regeneration system. Construction delays for the
regeneration system and for the contactors severely reduced the amount of
time available for GAC loss determination. In order to generate sufficient
data to determine the GAC loss across the regeneration system it was neces-
sary to obtain maximum utilization of the regeneration system. This involved
bringing all the GAC systems on line as soon as possible. It also forced
the regeneration of several GAC systems before they had reached complete
exhaustion as defined by the exhaustion criteria discussed in Objective 9.
Thus, it was not possible to compare the adsorbent performance of virgin and
regenerated GAC using parallel runs.
The above considerations made it difficult to compare the adsorptive
capacity of virgin and regenerated GAC. Since it was not possible to perform
parallel runs, the two systems received influent waters with varying concen-
trations. The systems also were started at different times of the year
which introduced temperature variations. It was hoped that these differences
would not be significant and that percent removal curves could be used to
compare the virgin and regenerated GAC.
Figures 68 and 69 present percent removal graphs for TOC and THMSIMDIST
from Contactor D, Phase 3-0, 3-1 and 3-2, respectively. Figures 70 and 71
present percent removal graphs for TOC and INST CHC1_ from GAC Filter ISA,
Phase 3-0, 3-1, 3-2 and 3-3, respectively. The wide variation in the
influent concentration between phases for both systems is presented in
Table 51. From the figures, it is apparent that varying influent concentra-
tions resulted in varying percent removal graphs. This is especially true
in the INST CHC1 percent removal graph for GAC Filter ISA (Figure 71).
Contactor D (Figure 69) gives a good graphic presentation of the effects
that three different THMSIMDIST average influent concentrations, (59.0 ug/1,
87.4/ug/l and 124 ug/1) have on the percent removal curves. Initially,
while the GAC had many adsorptive sites there was very little difference
between the curves. As the run progressed, fewer adsorptive sites were
available for the GAC systems that received the larger influent concentration
in Phases 3-0 and 3-1, causing them to adsorb a smaller percentage of the
influent concentration.
It appeared that in order to determine if regeneration had any effect
on the GAC's adsorbent performance, two systems with similar influent concen-
trations must be compared. In Figure 68, the average TOC influent concentra-
tion for virgin GAC, Phase 3-0, and twice-regenerated GAC, Phase 3-2, were
very similar. In Figure 70, the average TOC influent concentration for
virgin GAC, Phase 3-0, and GAC regenerated three times, Phase 3-3, were very
similar. If the GAC adsorbent performance was reduced due to regeneration,
the curves for the regenerated systems would be below those for the virgin
GAC. This is not evident in the figures. From this discussion, it is
apparent that only percent removal curves from systems with similar influent
concentrations can be used to study the effect of regeneration on the
161
-------
LEGEND
c
0
0
L
0
Q.
O
LU
95
98
85
80
75
70
65
60
55
50
45
40
0
VIRGIN C3-8)
ONCE REGENERATED C3-O
TWICE REGENERATED C3-2)
20
40
80
100
60
I, days
FIGURE 68. Total organic carbon (TOO percent removal curves for
multiple runs of Contactor D effluent, Phase 3.
120
-------
LEGEND
VIRGIN C3-e:>
0
L
9
OL
O
Ui
Of.
ONCE REGENERATED C3-l>
0
TWICE REGENERATED C3-2)
48
60
TIME, day
100
120
FIGURE 69. Three-day simulated distribution system THM CSTT3) percent removal
curves for multiple runs of Contactor D effluent, Phase 3.
-------
LEGEND
VIRGIN C3-4)
TWICE REGENERATED C3-2)
ONCE REGENERATED C3-O
THREE-TIMES REGENERATED C3-3)
-O
C
0
0
L
0
Q.
O
UJ
80 J
70 .
60 .
30 .
20
0
20
50
60
30 40
TIME, days
FIGURE 70. Total organic carbon (TOO percent removal curves for
multiple runs of GAC Filter ISA effluent, Phase 3.
70
-------
LEGEND
(Tv
Ul
JJ
c
0
0
i_
0
Q.
o
Ul
o:
180
80
60
40
20
0
-20
0
VIRGIN C9-83
TWICE REGENERATED C3-23
ONCE REGENERATED C3-D
THREE-TIMES REKNERATED C3-3)
20
50
60
30 40
TIME, days
FIGURE 71. Instantaneous chloroform CICLR) percent removal curves for
multiple runs of GAC Filter ISA effluent. Phase 3.
70
-------
TABLE 51. AVERAGE INFLUENT CONCENTRATION FOR CONTACTOR D AND GAC FILTER ISA
Contactor D GAC Filter ISA
Influent Influent
Phase Phase Phase Phase Phase Phase Phase
Parameter 3-0 3-1 3-2 3-0 3-1 3-2 3-3
TOC
THMFP
THMSIMDIST
Chloroform
Bromodichlorome thane
1914
184
87.4
25.0
8.7
2627
222
124
31.4
18.0
1818
149
59
11
8
.0
.8
.9
1870
216
74
14
4
.3
.3
.8
3275
290
179
54
19
.6
.7
2330
217
95
17
16
.5
.0
.4
1988
184
57.6
8.9
6.7
adsorbent performance. For systems with similar influent concentrations, it
appears that reactivation had restored the GAC to its virgin adsorptive
capacity.
In order to further determine the effect of regeneration on the adsorb-
ent performance, it was necessary to first normalize the varying influent
concentration. One way of eliminating the varying influent concentration
was by using the cumulative summation concentration of the various parameters
applied per weight of GAC rather than the actual influent concentration.
These data were further normalized by using the weight of parameter adsorbed
per weight of GAC. This eliminated the variations in GAC weight from system
to system. The cumulative summation concentration per weight of GAC was
plotted against the cumulative amount adsorbed per weight of GAC. Therefore,
a GAC System which was 100% effective in the removal of organics would
result in a 45° slope.
Figures 72 through 74 illustrate this effect for INST CHC13/ INST
CHBrCl and TOC, respectively, from Contactor D for Phases 3-0, 3-1 and 3-2.
Adsorption of specific compounds and groups of compounds, were not signifi-
cantly affected by multiple regenerations. Normalization techniques enabled
the comparison of different runs.
Figures 75 through 77 present the same effects for INST CHC1-, TOC and
THMFP, respectively, from GAC Filter ISA for Phases 3-0, 3-1, 3-Z and 3-3.
Figures 76 and 77 represent groups of compounds which show a pattern similar
to that observed in the contactors except for the TOC curves. The TOC
curves separated as more substance passed through the GAC filter. The INST
CHCl curves in Figure 75 are only identical up to a point, after which they
separate quite dramatically. Again Figure 75, Phase 3-0, desorption is
observed when the influent concentration decreases, with adsorption occuring
when the influent concentration increases. As expected, the effects of
varying influent concentration on equilibrium were minimized by the normali-
zation techniques, but not completely eliminated.
166
-------
0.7
CD
0
LEGEND
VIRGIN <3-0>
DICE REGENERATED C3-O
TWICE REGENERATED C3-23
0.1
0.2
0.7
0.8
0.3 0.4 0.5 0.6
SUM INFLUENT/SAC WT, G/l<0
FIGURE 72. Instantaneous chloroform CICLR) adsorbed per GAC weight for
multiple runs of Contactor D effluent, Phase 3.
0.9
-------
LEGEND
03
0)
JK
X
CD
0.4
0.35
0.3
h- 0.25
o
x
0.2
8 0.15
o
(0
Q
< 0.1
0.05
0
0
VIRGIN C3-0)
ONCE REGENERATED <3-t)
TWICE REGENERATED C3-2)
0.05
0.3
0.35
0.1 0.15 0.2 0.25
SUM INFLUENT/GAC WT, g/kg
FIGURE 73. Instantaneous bromodichloromethane CICL2) adsorbed per 6AC weight for
multiple runs of Contactor D effluent, Phase 3.
0.4
-------
LEGEND
D B VIRGIN C3-0>
ONCE REGENERATED <3-O
TWICE REGENERATED C3-2)
O
A:
N
O
40
o
o 30
X
o
UJ
-------
LEGEND
o
0.4
0
0
VIR8IN C3-a>
TWICE REGENERATED <3-2>
ONCE REGENERATED <3-O
TH?EE-TIMES REGENERATED C3-3)
0.2
0.4 0.6 0.8
SUM INFLUENT/GAC WT, g/kg
FIGURE 75. Instantaneous chloroform CICLR) adsorbed per GAC weight for
multiple runs of GAC Filter ISA effluent, Phase 3.
1.2
-------
LEGEND
VIRGIN C3-C>
TWICE REGENERATED C3-25
ONCE REGENERATED C3-15
THREE-T»C3 REGENERATED C3-35
Id
CQ
a
o
CO
2:
D
CO
25
a
o
S 15
10
0
0
FIGURE
10 15 20 25 30 35
SUM INFLUENT/GAC WT, g/kg
45
50
76. Total organic carbon (TOO adsorbed per GAC weight for
multiple runs of GAC Filter ISA effluent, Phase 3.
-------
LEGEND
O
x
X
Q
H-
3:
O
<
(D
X
Ct
LU
ffi
a
o
CO
2:
D
CO
VIRGIN C3-0)
TWICE REGENERATED C3-25
ONCE REGENERATED C3-O
THREE-TIMES REGENERATED C3-35
6 8 18 12
SUM INFLUENT/SAC WT, g/kg
14
16
18
FIGURE 77. THM formation potential CFTTD adsorbed per GAC weight for
multiple runs of GAC Filter ISA effluent, Phase 3.
-------
Graphs of the cumulative summation concentraion per weight of GAG of
the various parameters applied against the cumulative amount adsorbed per
weight of GAC was the best method available for comparing the adsorbent
performance of virgin and regenerated GAC. Full-Scale, on-site regeneration
restores the GAC to its virgin adsorptive capacity. This is further supp
orted by conclusions found in Objective 10 which used iodine number, molasses
number, BET surface area determination and modified phenol value to compare
the effects of regeneration on adsorptive capacity.
Objective 8. Correlation Between Pilot and Full-Scale GAC Systems.
The primary purpose of this objective was to determine the ability of
pilot systems to predict full-scale performance. Percent removal graphs
were studied in this evaluation for the following parameters.- INSTTHM,
THMSIMDIST, TOC, CAOX, and acid extract GC/FID profiles. Although other
parameters were considered, they were not significant for this evaluation.
Data concerning these parameters can be found in Volumes 2 and 3.
The evaluations conducted under this objective utilized three criteria
for comparing the pilot and full-scale systems: EBCT, percent removal and
exhaustion (the preferred criteria stated in Objective 9).
GAC Filters--
Data discussed below are contained in Table 52. During Phase 3-0, the
EBCTs were equal for both the pilot and full-scale GAC filters based on
throughput for the length of the run. Figures 78 and 79 and show the
percent removal of the two systems for THMSIMDIST and TOC to be identical.
TABEL 52. COMPARISON OF PILOT AND FULL-SCALE GAC FILTER SYSTEMS
Hydraulic Exhaustion
GAC Filter Phase Loading Rate TOC THMSIMDIST
System Run 1pm/sq m gpm/sq ft (runday) (runday)
Pilot 3-0 102 2.5 57 176
Fullscale 102 2.5 57 176
Pilot 3-1 97 2.4 22 36
Fullscale 102 2.5 22 36
Pilot 3-2 94 2.3 • 29 a
Fullscale 102 2.5 21 a
Pilot 3-3 94 2.3 50 a
Fullscale 102 2.5 46 a
Exhaustion did not occur during life of this run.
173
-------
LEGEND
O a PILOT GAC FILTER EFF CRV5>
GAC FILTER ISA EFF CISAE) X X
C
0
0
L
0
Q.
O
X
hi
20 _
10 .
0
20
40
60
140
160
180
80 100 120
TIME, days
Figure 78. Three-day simulated distribution system THM CSTT3) percent removal
curves for pilot GAC filter and GAC Filter ISA effluents, Phase 3-0.
200
-------
LEGEND
PILOT 9AC FILTER EFF CRVS)
6AC FILTER 15A EFF C15AED X X
•-J
U1
0
0
L
O
111
Of.
90
80
70
60
50
40
30
20
10
0
-10
0
20
40
60
140
160
180
80 180 120
TIME, days
F igure 79. Total organic carbon (TOO percent removal curves for pi lot
GAC filter and GAC Filter ISA effluents, Phase 3-0.
200
-------
Breakthrough curves (Figures 80 and 81) for the same parameters, when evalu-
ated for exhaustion criteria, verified these results.
Throughput calculations for Phase 3-1 showed the EBCTs to be the same.
Figure 82 demonstrates that the pilot column was about 5% more efficient
than the GAC filter for THMSIMDIST removal. Figure 83 shows about a 5%
better TOC removal for the pilot GAC filter for 25 days but both systems
were equal thereafter. Exhaustion for TOC and THMSIMDIST on the same runday
(Figures 84 and 85).
Throughput calculations for Phase 3-2 showed a 10% lesser volume of
water passed through the pilot system causing a longer EBCT. This effect is
reflected in Figure 86 where the pilot column is removing 20% to 30% more
THMSIMDIST. This was further evidenced for TOC in Figure 87, but only by
about 10%. THMSIMDIST exhaustion did not occur during the run (Figure 88),
however, the exhaustion criterion was met for TOC on runday 29 for the pilot
system and on runday 21 for the GAC filter (Figure 89). This represented up
to a 30% greater efficiency in the pilot system.
In Phase 3-3, the EBCT was 5% longer for the pilot column which explains
the approximate 5% greater removal of THMSIMDIST (Figure 90). TOC percent
removal curves were identical (Figure 91). THMSIMDIST exhaustion did not
occur (Figure 92). The exhaustion criterion for TOC was reached in 50 days
in the pilot system and in 46 days in the GAC filter, indicating up to a 9%
greater efficiency in the pilot system (Figure 93).
A review of the acid extract GC/FID profiles, comparing the two systems,
indicated that the organics removal efficiency was essentially identical as
shown in the Figure 94.
Data available for CAOX showed essentially no difference in removals
for the two systems (Figure 95).
Contactors--
Data discussed below are contained in Table 53.
Phase 3-0 EBCTs were practically the same for the two systems.
Figures 96 and 97 showed a 5% greater efficiency of the contactor over the
pilot system based on THMSIMDIST and TOC. Exhaustion criteria verify this
5% better removal (Figures 98 and 99).
The EBCTs observed during Phase 3-1 resulted in a 4% longer contact
time in the pilot system. However, this advantage for the pilot system was
not apparent in THMSIMDIST and TOC removal graphs (Figures 100 and 101).
These data showed a 10% better efficiency for the contactor system. Exhaus-
tion did not occur during this run as shown in Figures 102 and 103.
An 8% longer EBCT was experienced for the pilot system during Phase
3-2. Although this longer EBCT should have improved the efficiency of the
pilot system, this was not evidenced in the percent removal graphs
176
-------
INFLUENT CFi_IN>
LEGEND
PILOT GAC FILTER EFF CRVS5
GAC FILTER ISA EFF CISAE>
F i gure
(00
150
200
250
TIME, day*
Three-day simulated distribution system THM (STT3) breakthrough curves
for pilot GAC filter and GAC Filter ISA effluents, Phase 3-0.
-------
3800
2500
2000
H 1500
QL
O
O
O
1000
500
0
0
INFLUENT CFLJJO
50
LEGEND
PILOT GAC FILTER EFF CRV5)
6AC FILTER ISA EFF <15AE>
100 150
TIME, days
200
250
Figure 81. Total organic carbon (TOO breakthrough curves for pilot
GAC filter and GAC Filter ISA effluents, Phase 3-0.
-------
LEGEND
PILOT 6AC FILTER EFF CRV35
6AC FILTER ISA EFF C15AE)
20
0
10
15
30
35
20 25
TIME, days
Figure 82. Three-day simulated distribution system THM CSTT3) percent removal
curves for pilot GAC filter and GAC Filter ISA effluents, Phase 3-1
-------
LEGEND
PILOT 6AC FILTER EFF
-------
0
INFLUEKT CFLirO
LEGEND
PILOT 6AC FILTER EFF CRV3) X K
8AC FILTER ISA EFF C1SAE5
15
20 25
TIME, days
30
35
40
45
Figure 84. Three-day simulated distribution system THM CSTT3) breakthrough curves
for pilot GAC filter and GAC Filter ISA effluents, Phase 3-1.
-------
00
4000
3500 .
0
0
INFLUENT CFLIN>
LEGEND
PILOT 6AC FILTER EFF CRV3)
GAC FILTER ISA EFF C1SAED
10
15
35
40
20 25
TIME, days
Figure 85. Total organic carbon (TOO breakthrough curves for pi lot
GAC filter and GAC Filter ISA effluents, Phase 3-1.
45
-------
LEGEND
B a PILOT 6AC FILTER EFF CRV3)
6AC FILTER ISA EFF C1SAE)
CO
U)
c
0
0
L
0
Q.
O
21
Id
o:
100
90 _
80 .
70 .
60 _
50 .
40 _
30
10
15
20 25
TIME, days
30
35
40
45
Figure 86. Three-day simulated distribution system THM CSTT3) percent removal
curves for pilot GAC filter and GAC Filter ISA effluents, Phase 3-2.
-------
PILOT 6AC FILTER EFF CRV33
LEGEND
6AC FILTER ISA EFF C16AE3
CO
20
10
15
30
35
40
20 25
TIME, days
Figure 87. Total organic carbon (TOO percent removal curves for pilot
GAC filter and GAC filter ISA effluents, Phase 3-2.
45
-------
CD
in
X
0»
o
H
1-
J-
Z
LJ
O
O
O
Figure
a a INFLUENT CFLDO
120
100 .
80 _
60 _
40 .
0
0
LEGEND
PILOT 6AC FILTER EFF CRV35 X X
GAC FILTER ISA EFF CI5AE5
10
15
20
TIME,
25
days
30
35
40
45
Three-day simulated distribution system THM (STT3) breakthrough curves
for pilot GAC filter and GAC Filter ISA effluents., Phase 3-2.
-------
a a INFLUENT CFLJN)
3000
2500
2000
S H 1500
en r—
O
O
1000
500
0
0
10
LEGEND
PHOT 8AC FILTER EFF CRV35
GAC FILTER ISA EFF CtSAE)
20 25
TIME, days
30
35
40
Figure 89. Total organic carbon (TOO breakthrough curves for pi lot
GAC filter and GAC Filter ISA effluents, Phase 3-2.
45
-------
PILOT SAC FILTER EFF CRV3>
LEGEND
8AC FILTER 15A EFF CI5AE> X X
03
C
0
«
a
o
LJ
20
20
30 40
TIME, days
50
60
70
Figure 90. Three-day simulated distribution system THM CSTT3) percent removal
curves for pilot GAC filter and GAC Filter ISA effluents, Phase 3-3.
-------
LEGEND
PILOT 6AC FILTER EFF CRV33
6AC FILTER ISA EFF <1SAE)
00
00
18
10
20
30 40
TIME, days
50
60
70
Figure 91. Total organic carbon CTOC) percent removal curves for
pilot GAC filter and GAC Filter ISA effluents, Phase 3-3
-------
CD
oo co
-------
061
• O
— O
rt- 1
O Q
j> i
o cr
o
©
©
ro
©
o to
CO -b
i o
CO -5
cn
©
— ©
o
©
CONCENTRATION,
cn © cn
r\>
o
©
ro
cn
©
©
CO
©
©
©
2
?
2
o
I
2
1
rn
o
rn
G)
O
0
51
-------
M
Lu
LU
I
'!-
: U1.
\S>
cm
r
CD
H
t i
< T
r> t r
U :
LJLL
J
•Lx
JUJ-
Ji
INT.
STD,
INT,
STD,
INT,
STD,
INT.
STD.
FIGURE 94. Acfd extroct profiles for GAC Filter ISA and
pilot GAC filter., Phase 3-8, rundoy 57.
191
-------
ID
to
140
120
- 100
X
z
o
O
8
80
60
40
20
0
INFLUENT CFLIN3
LEGEND
PILOT OAC FILTER EFF CRV53
8AC FILTER ISA EFF C15AE>
0 20 40 60 80 100
TIME, day.
120
140
160
Figure 95. Carbon adsorbable organohaIides CCAOX) breakihrough curves for
pilot 6AC filter and 6AC Filter ISA effluents, Phaee 3-0.
180
-------
TABLE 53. COMPARISON OF PILOT AND FULL-SCALE CONTACTOR SYSTEMS
Hydraulic Exhaustion
GAC Filter Phase- Loading Rate
System
Pilot
Full-scale
Pilot
Full-scale
Pilot
Full-scale
Exhaustion did not
Run 1pm/ sq m
3-0 296
301
3-1 289
301
3-2 281
301
occur during life of
gpm/sq ft
7.3
7.4
7.1
7.4
6.9
7.4
this run.
TOC THMSIHDIST
(runday) (runday)
155 197
169
85
a
a
a
204
a
a
a
a
(Figures 104 and 105). These graphs show the contactor to be 10% more
efficient in removals. As shown in Figures 106 and 107, exhaustion did not
occur during this phase.
Based on acid extract GC/FID profiles, it was apparent that there was
only a small difference between the two systems in organic removals
(Figure 108).
The CAOX data showed that the contactor had a 30% better removal than
the pilot system (Figure 109).
The following observations could account for the difference in effic-
iency between pilot and full-scale systems:
1. The variations in the volume measuring device of the pilot systems.
2. The accuracy of full-scale flow controllers.
3. The pilot GAC filters had flow adjustment problems due to floe buildup
and insufficient head to permit accurate flow regulation.
4. The difficulty in measuring weight and volume of GAC in all systems.
Overall, although there were occasionally minor deviations in the
results from the pilot and full-scale systems, it was concluded that pilot
systems would be sufficiently predictive of full-scale performance to be
used for plant design purposes.
Objective 9: GAC Exhaustion Criteria in GAC Filters and Contactors
Identification of a specific indicator of GAC exhaustion turned out to
be an impossible task. Several different criteria were evaluated including
193
-------
LEGEND
BITUMINOUS PILOT CONTACTOR EFT CPFA5)
CONTACTOR D EFF CDE5
-4J
c
9
0
L
O
2:
UJ
01
100
90 .
30 _
20
0
50
100 (50
TIME, days
200
250
Figure 96. Three-day simulated distribution syst. THM CSTT3) percent removal curves
for bituminous pilot contactor and Contactor D effluents, Phase 3-0,
-------
LEGEND
BITUMINOUS PILOT CONTACTOR EFF
CONTACTOR D EFF X X
u
L
9
0.
U
or
tee
90
88
70
60
50
40
30
0
50
100 150
TIME, days
200
250
Figure 97. Total organic carbon (TOO percent removal curves for bituminous
pilot contactor and Contactor D effluents., Phase 3-0.
-------
CQ
961
CONCENTRATION,
-------
3008
2500
2000
z
H 1500
Of
I-
z
y 1000
Z
O
O
500
0
0
INFLUENT CSFEF)
50
LEGEND
BITUMINOUS PILOT CONTACTOR EFF CPFA5> X X
CONTACTOR D EFF
200
100 150
TIME, days
Figure 99. Total organic carbon (TOO breakthrough curves for bituminous pilot
contactor and Contactor D ©ffluents, Phase 3-0.
250
-------
LEGEND
BITUMINOUS PILOT CONTACTOR EFF CPFA5)
CONTACTOR D EFF CDE5
100
00
C
9
0
C
0
Q.
O
80
70
60
^ 50
40
30
0
20
60
TIME, days
80
100
120
Figure 100. Three-day simulated distribution syst. THM CSTT3) percent removal curves
for bituminous pilot contactor and Contactor D effluents, Phase 3-1.
-------
LEGEND
BITUMINOUS PILOT CONTACTOR EFF
-------
to
O
O
H Q INFLUENT
280
180
160
- 140
c»
* 120
\
z
H 100
H-
80
60
40
20
0
O
O
0
LEGEND
BITUMINOUS PILOT CONTACTOR EFF CPFA5) X X
CONTACTOR D EFF
20
40
60
TIME, days
100
120
Figure 102. Three-day simulated distribution system THM CSTT3) breakthrough curves
for bituminous pilot contactor and Contactor D effluents, Phase 3-1.
-------
102
Q.
Q
•<
-------
LEGEND
BITUHINOUS PILOT CONTACTOR OFF CPFA55
CONTACTOR D EFF CDED
NJ
O
55
20
40
60 80
TIME, days
100
120
140
Figure 104. Three-day simulated distribution syst. THM (STT3) percent removal curves
for bituminous pilot contactor and Contactor D effluents, Phase 3-2.
-------
LEGEND
n—B BITUHINOUS PILOT CONTACTOR EFF
CONTACTOR D EFF
tv)
O
-p
fl>
0
L
0
Q.
Id
o:
95
90
85
80
75
70
65
60
55
50
45
0
20
40
60 80
TIME, days
100
120
140
Figure 105. Total organic carbon CTOC3 percent removal curves for bituminous
pilot contactor and Contactor D effluents. Phase 3-2.
-------
to
a
O
O
H
I-
<
ft:
h-
ld
O
Z
O
O
70
68
50
40
30
20
10
0
0
INFLUENT CSFEF>
LEGEND
BITUMINOUS PILOT CONTACTOR EFF CPFA55
CONTACTOR D EFF CDE>
20
40
60
TIME,
100
80
days
Figure 106. Three-day simulated distribution system THM CSTT3) breakthrough curves
for bituminous pilot contactor and Contactor D effluents, Phase 3-2.
-------
CO
soz
CONCENTRATION,
01
PO
cs>
O —I
o o
3 n-
o
r«- O
O 1
-J <
I A
ro w
3
I
V)
6>
o>
m
0.
O
0> 00
O
Q
-------
u_
CO
.
jf,
b^
LU
O
Jl
£
ex.
CD
INT,
STD
INT,
STD,
INT.
STD,
INT.
STD,
FIGURE 188. Acid extract profiles for Contactor D and
pilot contactor, Phase 3-0, runday 113.
206
-------
LOZ
CONCENTRATION,
(S)
O> CO
CD CD
ro
(5*
o>
<0
o
°
ro
©
o
O Q
3 O.
«•»- to
Q O
2-3-
O Q
T CT
o>
D D
rr- O
Q 3T
O Q
T CL
»
o to
O r>
-h O
~o
ex:
_
rn
O.
O
ro
8
D-
* o
I
(D
(A
o>
CD
00
o
rn
CD
8
o
m
*
S
5
/N
TJ
S
-------
the three specified in the original GAC treatment requirement of the
February 9, 1978 proposed amendments to the Safe Drinking Water Act (SDWA) .
A discussion of each of the criteria considered under this objective
along with the merits and shortcomings of each will be presented. Major
shortcomings of all criteria are that health effects information and more
specifically, maximum contaminant levels (MCLs) for harmful substances are
practically non-existent. With such MCLs, the GAC would be regenerated when
an effluent concentration, for a substance known to be adsorbable by GAC,
approached the MCL for that substance.
The best summary solution seemed to be the regeneration of the GAC when
an established MCL is approached or when a three-week running average efflu-
ent concentration exceeds 1.0 mg/1 TOC (1,000 pg/1) or 0.1 mg/1 (100 ug/1)
THMSIMDIST, whichever occurs first.
Although INSTTHM and THMFP were also considered for exhaustion criter-
ia, they did not appear to be as useful as the TOC and THMSIMDIST criteria.
INSTTHM continues to form beyond the treatment process. THMFP represents
the reaction of all precursor with chlorine resulting in THM concentrations
far greater than those experienced in the distribution system and, therefore,
was not an appropriate criterion. THMSIMDIST would more accurately reflect
the distribution system THM concentrations.
There are some full-scale practical matters to consider with any ex-
haustion criterion. In a large plant, such as the CWW with 47 filters or an
equivalent amount of contactors, one would not regenerate all GAC systems at
one time. Further, it would be practically impossible to handle the sampling
load required from each filter. Therefore, banks of filters or contactors
would have to be treated as units and whenever the exhaustion criterion was
approached the entire unit would be regenerated.
During the research project, the CWW selected a three-month cycle for
contactor and a six-week cycle for filter regeneration based on USEPA data
and Phases 1 and 2 data. Although this seemed to be the optimum regenera-
tion frequency, special attention was given at the close of these time
frames to the organic data, so that temperature, seasonal influent concentra-
tion levels, and other considerations could be used to determine the best
time to remove each system for regeneration. Often, in attempting to evalu-
ate various exhaustion criteria and also due to the unavailability of the
regeneration furnace, systems were left on line past exhaustion.
Occasionally in the later stages of the grant, systems were taken off early
to gain regeneration experience and GAC loss data.
Exhaustion Criteria--
Tables 54 through 56 show the results of applying each criterion
discussed below. Criteria 1 through 3 are contained in the proposed rules for
the GAC treatment technique as published in the Federal Register (as
"design" criteria, not necessarily as "exhaustion" criteria). However, they
were evaluated herein as "exhaustion" criteria.
208
-------
TABLE 54. APPLICATION OF VARIOUS EXHAUSTION (EXH) CRITERIA TO GAC FILTER ISA AND CONTACTOR D, PHASE 3-0
Data Description
GAC Filter ISA
Time to EXH, days
Effluent Cone., mg/1
Removal, %
Loading, kg
Loading Retained, %
Loading/GAC Wt., g/kg
TOC Cone. @ STT3, EXH, mg/1
STT3 Cone. @ TOC EXH, mg/1
o Contactor D
vO
Time to EXH, days
Effluent Cone., mg/1
Removal, %
Loading, kg
Loading Retained, %
Loading/GAC Wt ., g/kg
TOC Cone. @ STT3 EXH, mg/1
STT3 Cone. @ TOC EXH, mg/1
TOC
0.5
mg/1
57
1.5
30
1,045
55
24.2
-
0.03
134
0.8
60
602
72
30.4
-
0.02
TOC
1.0
mg/1
57
1.5
30
1,045
55
24.2
-
0.03
176
1.1
57
824
69
41.6
-
0.06
TOC
1.2
mg/1
64
1.4
37
1,141
53
26.4
-
0.03
a
-
-
-
-
-
-
-
TOC
1.5
mg/1
127
1.6
13
1,567
37
36.3
-
0.05
a
-
-
-
-
-
-
-
TOC
% Removal
(Steady State)
29
0.7
52
642
67
14.9
-
0.01
50
0.3
83
259
85
13.1
-
0.00
STT3
0.1
mg/1
169
0.11
16
84
38
1.9
3
-
a
-
-
-
-
-
-
-
TOC Loading
Retained
75%
8
0.7
66
243
75
5.6
-
0.01
120
0.8
49
548
74
27.6
-
0.02
TOC
50%
Removal
50
0.9
40
961
59
22.2
-
0.02
120
0.8
49
548
74
27.6
-
0.02
Exhaustion criterion was not met during this run.
-------
TABLE 55. APPLICATION OF
Data Description
GAG Filter 15A
Time to EXH, days
Effluent Cone., mg/1
Removal, %
Loading, kg
Loading Retained, %
Loading/ GAC Wt., g/kg
TOC Cone. @ STT3 EXH, mg/1
STT3 Cone. @ TOC EXH, mg/1
KJ Contactor D
i — i
o
Time to EXH, days
Effluent Cone. , mg/1
Removal , %
Loading, kg
Loading Retained, %
Loading/ GAC Wt., g/kg
TOC Cone. @ STT3 EXH, mg/1
STT3 Cone. @ TOC EXH, mg/1
Exhaustion criterion was
VARIOUS EXHAUSTION (EXH) CRITERIA TO GAC FILTER ISA AND CONTACTOR D, PHASE 3-1
TOC
0.5
mg/1
22
1.6
48
919
68
20.6
-
0.09
70
1.0
65
632
81
32.0
-
0.03
not met during
TOC
1.0
mg/i
22
1.6
48
919
68
20.6
-
0.09
a
-
-
-
-
-
-
~
this
TOC
1.2
mg/1
22
1.6
48
919
68
20.6
-
0.09
a
-
-
-
-
-
-
~
run.
TOC
1.5
mg/1
29
1.7
38
1,088
62
24.4
-
0.12
a
-
-
-
-
-
-
~*
TOC
% Removal
(Steady State)
29
1.7
38
1,088
62
24.4
-
0.12
29
0.4
85
305
89
15.4
-
0.00
STT3
0.1
mg/1
36
0.12
30
72
60
27.6
2.1
—
a
-
-
-
-
-
-
~
TOC Loading
Retained
75%
15
1.2
60
694
74
15.6
-
0.06
a
-
-
-
-
-
-
_
TOC
50%
Removal
29
1.7
38
1,088
62
24.4
-
0.12
a
-
-
-
-
-
-
—
-------
TABLE 56. APPLICATION OF
Data Description
GAC Filter ISA
Time to EXH, days
Effluent Cone., mg/1
Removal , %
Loading, kg
Loading Retained, %
Loading/ GAC Wt., g/kg
TOC Cone. @ STT3 EXH, mg/1
STT3 Cone. @ TOC EXH, mg/1
Contactor D
Time to EXH, days
Effluent Cone., mg/1
Removal, %
Loading, kg
Loading Retained, %
Loading/GAC Wt. , g/kg
TOC Cone. @ STT3 EXH, mg/1
STT3 Cone. @ TOC EXH, mg/1
Exhaustion criterion was
VARIOUS EXHAUSTION (EXH)
TOC
0.5
mg/1
21
1.6
41
626
63
13.1
-
0.04
a
-
-
-
-
-
-
—
not met during
TOC
1.0
mg/1 m
25
1.3 I
46
CRITERIA TO
TOC TOC
1.2 1.5
g/1 mg/1
25 a
.3
46
710 710
60
14.9 14
-
0.04 0.
a
-
-
-
-
-
-
~
this run.
60
.9
-
04
a a
-
-
-
-
_
-
— —
GAC FILTER ISA
TOC
% Removal
(Steady State) _
21
1.6
41
626
63
13.1
-
0.04
43
0.4
82
255
84
12.8
-
0.00
AND CONTACTOR D, PHASE 3-2
STT3 TOC Loading
0.1 Retained
mg/1 75%
15
1.1
54
488
69
10.2
-
0.03
a a
-
-
-
-
-
-
_ _
TOC
50%
Removal
21
1.6
41
626
63
13.1
-
0.04
a
-
-
-
-
-
-
~"
-------
1. The concentration in the effluent of any of the volatile halogenated
organic compounds (except the THMs) shall not exceed 0.5 ug/1.
The discussion in the proposed rules supporting this criterion con-
cluded that the presence of these chemicals would also be indicative of the
presence of other potentially hazardous substances which would be more diffi-
cult to detect. The validity of this surrogate is open to debate.
This criterion has the advantage of being, in some sense, an MCL and
is one that the CWW found no problem with meeting even without GAC. The con-
centration of volatile halogenated organic compounds in excess of this limit
were seldom detected even in the raw water before any treatment. The short-
comings of this criterion include the fact that some volatile halogenated
organic compounds may be more hazardous than others and, therefore, could
cause considerable expense, associated with regenerating the GAC, when one
of the less harmful organics exceeds 0.5 /ig/1. Further, to our knowledge,
the health effects of ingesting O.S^ug/l of any of the volatile halogenated
organic compounds has not been substantiated.
2. Removal of influent TOC with fresh GAC shall be at least 50%.
The selection of any percent removal criterion suggests that it is
equally as safe to drink the water that contains 2,000 ug/1 of TOC (50% of
4,000;ig/l of TOC) as it is for one that contains 250 ug/1 of TOC (50% of
500 ,ug/l of TOC). It further places a considerable significance on a "50%"
level. For example, why not select 10%, 60%, 80%, or some other percentage?
Therefore, the actual effluent concentration being ingested across the
country would vary considerably due to the level of organics in raw water
supplies, with the resulting health effect advantages unknown.
As indicated in the proposed rules, the use of TOC as a measurement
tends to give excessive weight to naturally occurring, high molecular weight
compounds which are not known to be hazardous. This being the case, more
hazardous TOC constituents could exist in larger concentrations than accept-
able, but be disguised in the TOC levels.
An advantage of this criterion and all TOC-based criteria is that the
TOC analysis is relatively easy to run.
3. The effluent TOC may not exceed the value obtained with fresh GAC by
more than 0,5 mg/1.
As in the case of 50% removal, this criterion has no health effects
basis, allows considerable variability between the level of TOC in water
consumed across the country, and could disguise higher levels of harmful
organics. The further disadvantage is that it tends to unfairly condemn a
GAC that is extremely effective when first placed in service. Under
Objective 4, the use of bituminous and lignite GACs were evaluated and it
was noted that, although the influents were common, the initial effluent
concentrations of TOC in the two GACs were different. Therefore, by using
this criterion, the GAC that v/as most effective when fresh might have to be
212
-------
removed for regeneration earlier and yet the effluent concentrations at the
time of exhaustion would be lower than the other GAC whose initial effluent
concentration was higher.
4. The effluent concentration shall not exceed 0.10 mg/1 on an annual
running average of THMSIMDIST effluent concentration.
This concentration is an existing MCL for distribution system samples
and, therefore, would be an appropriate exhaustion criteria. Three-day
THMSIMDIST serves as a surrogate for predicting the distribution system
extremes for the CWW service area. One minor disadvantage is that it takes
approximately four days to obtain results from this analysis. This criterion
also ignores the hazard of other organics which are not detected under this
analytical technique. The major advantages of this criterion are that it
reflects an established MCL and further is representative of the quality of
water being ingested.
5. The TOC concentration shall not consistently exceed 1.0 mg/1.
Although this criterion has the disadvantage that hazardous substances
might be disguised by the non-harmful substances within TOC, at least there
would be consistent year-round, city-to-city" levels. Neither would the
problems inherent in using percentages exist. Unfortunately there were no
health effect relationships used to determine the 1.0 mg/1 exhaustion criter-
ion. Basically, it was developed because fresh GAC effluent TOC concentra-
tions ranged from about 0.25 to 0.50 mg/1 and when the recommended criter-
ion of 0.5 mg/1 TOC is added, approximately 1.0 mg/1 TOC results. Further,
this is a reasonable level to achieve considering the raw water source.
6. The cumulative percent of TOC or THMSIMDIST loading shall not be less
than 75%.
Under this criterion, the consumer will ingest water which contains,
on an average, 75% fewer GAC-adsorbable organics. At exhaustion, the percent
removal could be a much lower percentage than 75%, but the cumulative weight
of organics retained, as a percent of the cumulative weight of contaminants
contained in the influent, would never be less than 75% on a running average
basis. An advantage is that a specific degree of removals could be selected
and guaranteed to the consumer. The problem again with TOC or THMSIMDIST is
that more harmful organics are not separately detectable by these two
analytical techniques. Another disadvantage is that there are no health
effect information substantiating the value of the 75% running average.
7. The slope of the TOC or THMSIMDIST percent removal graph tends to
level off or plateau for at least three weekly samples. This condi-
tion, termed steady-state, would indicate exhaustion.
The primary disadvantage with this approach is that it is very diffi-
cult to detect a leveling off or at least where the leveling off occurs. The
actual attainment of steady-state may not be discerned until the system is
well past the point of exhaustion. Further, there is no health effects
213
-------
basis for this criterion. Also, steady-state could be reached at an unaccep-
table effluent concentration or percent removal level. The major advantage
is that the most efficient use of the GAC would be assured.
8. The effluent TOC or THMSIMDIST concentration equals that of the
influent (zero percent removal).
Keeping the GAC on line until this condition occurred would present
unacceptable effluent concentrations even though the average purity of the
water would be improved over sand filters alone. This approach would
certainly give maximum life to the GAC since this condition did not occur in
the bituminous GAC Systems studied.
Preferred Criteria—Based on the above considerations, the best cri-
terion for exhaustion and, therefore, regeneration is likely when an estab-
lished MCL is approached or when a three-week running average effluent con-
centration exceeds 1.0 mg/1 TOC or 0.1 mg/1 THMSIMDIST, whichever occurs
first. The reasons are as follows:
1. These criteria combine fixed MCLs for a gross organic indicator (TOC)
and for THMs. Therefore, a plant could be designed with minimum var-
iability.
2. The 1.0 mg/1 TOC concentration is approximately the level that would
have resulted for CWW had the proposed GAC regulations
been adopted. This regulation called for criteria of 0.5 mg/1 TOC
above fresh GAC (which for CWW is about 0.75 to 0.8 mg/1) and 50%
removal of influent TOC. At 1.0 mg/1 TOC (about 50% of the average
annual plant effluent TOC of 2.0 mg/1), both recommendations would be
approximately met.
3. Although TOC could hide specific organic contaminants, CWW seldom saw
any specific organics at quantities greater than 0.4 ug/1 (lower
detection limit of many EPA methods-) during this
study.
Tables 54 through 56 indicate the bed lives for GAC Filter 15A and
Contactor D when the preferred exhaustion critera were used. In all cases
the TOC criterion was more limiting than the THMSIMDIST criterion for GAC
Filter 15A, the average bed life was 34 days, ranging from 22 to 57 days for
the first three runs.
During Phase 3, attention was focused on maintaining, as nearly as
possible, the original plan for a three-month regeneration cycle for
Contactor D, the "control" contactor. As a result, only the longer virgin
run of Contactor D reached the preferred exhaustion criterion at day 176.
The two subsequent runs did not reach any of the criteria before being taken
off line in runs of 106 and 126 days. A review of data from other Phase 3
contactors (9 runs) indicated that exhaustion was reached in eight of the
runs for an average bed life of 93 days (range 50 to 148 days). Again, TOC
was the exhaustion determinant.
214
-------
Objective 10: Regenerated GAG Characteristics
To provide analytical data on regenerated GAC characteristics through-
out the project CWW contracted with the Georgia Institute of Technology
(GIT), Engineering Experimental Station. Analytical tests were conducted on
selected samples of virgin, spent and regenerated GAC to determine the
effectiveness of the regeneration process. Representative samples were
taken from GAC Filter 23A (WVW 20 x 50), GAC Filter ISA (WVG 12 x 40), and
Contactor D (WVG 12 x 40). These samples were dried, if necessary, at 100°C
(212°F), mixed and reduced in size by the coning, quartering and riffling
technique or by riffling alone. These samples were then sent to the GIT for
GAC quality tests.
The GAC analyses data (Table 57) from GAC Filter ISA indicated that
three of four regenerations proved successful and returned the GAC to the
approximate original characteristics. The fourth regeneration did not
return the GAC to the start-up qualities.
These data showed that three regenerations of GAC from Contactor D
weresuccessful in returning the GAC to nearly original qualities. The data
from GAC Filter 23A (WVW 20 x 50) indicated poor recovery which may have been
dueto long exposure of this GAC and the fact that the regeneration system was
calibrated for a different grain size GAC.
In-house GAC evaluations were also performed. Split GAC samples were
evaluated by GIT and CWW. These comparison tests were: iodine number,
sieve analysis (dry), apparent density and percent total ash. All of these
test methods were identical to those used by GIT. As indicated in Table 58,
a good correlation exists between the GIT and CWW results.
Objective 11. GAC Regeneration and Transport Losses
The bed-to-bed GAC losses included both regeneration and transport
losses (Table 59). Losses for the contactors ranged from-9.8% to 18.9%; the
average for nine regeneration cycles being 15.3%. GAC filter losses ranged
from 13.6% to 23.7%, the average being 18.5% for six Phase 3 regeneration
cycles. A detailed explanation of the technique used to obtain these figures
is contained in Section 5.
The losses on GAC filters were greater than contactors due to the
length and configuration of transport piping [maximum length 132.6 m (435 ft)
including a number of short radius ells] , the handling of the GAC with a
shovel and the sandtable inefficiency. The contactor system, which was
immediately adjacent to the regenerator, was specifially designed for mini-
mum transfer losses with short runs and long radius ells.
Attempts were made to measure the regenerator battery losses dynamic-
ally from the spent GAC storage tank to the regenerated GAC storage tank.
Grids were placed on top of the tanks and the contour of the GAC was measured
at 30.5 cm (1 ft) increments using a leveling rod. An effort was made to
complete the measurements of each tank within equal elapsed times. The
215
-------
n> 01
0,3
It
£-£•
o
o 3
T)
SV'iltH bFAlfc HUN
I5A Viigiu 3-0
I'»A Ketji-n 3-0
>- "«»*• -
,(leil
l')A Mai tup* 3-2
ISA K.yfn 1-2
K) !'.» 11. II up* J-J
H
ISA He (jcli J- J
2 JA Lipenl 1 -U
2JA He yen l-O
1* 'jpent 3-0
& Stdi tupd 3 1
p«-nl
•«, « „
IhfiLL 57 a m/g iy B/g
1184 572 0 1101 7 532 i
1011 48' 0 1022 4 49? 8
1 103 523 0 9/9 0 464 0
1083 504 0 1036 0 482 0
1104 510 0 1041 9 4t)4 0
9B6 SI 2 0 880 0 457 0
32/ 226 0 609 I WO 9
685 J90 5 721 3 4111
&24 4V* 8 1030 3 535 B
1096 b47 0 948 0 473 0
KHMYbtb
PHfNOl PHKNOl HOI A^bfcS
PPM iTl Mq/7 i7Tc/l DllT~~niU/inl ~% yiTTflOnT '%"" ~lfn/lliUmi \
16 2 33 b 1 84 J 81 10 50 5 07 1 4d
21 7 45 0 2 47 5 12 10 80 5 21 0 42
18 2 38 4 2 O/ 4 37 IS 00 7 11 2 52
20 0 43 0 2 28 4 90 12 60 b 86 4 02
IB 8 40 7 2 14 4 b3 Ib 00 fa 93 3 fcfi
18 9 36 4 2 Ib 4 Ib 11 30 5 H6 11 70
4 30 i 00 5 29
2? 9 49 0 3 20 5 bO it /O 5 00 2 39
30 2 54 1 3 44 i 16 6 80 3 J4
0 899
1 023
5bB
563
I 110
1 043
1 190 t,/ J
1 080 69 S
i no ij it
0 980 /j /
0 a JO 69 L
0 BSD '3 4
0 9/0 /'J /
540 /'j /
545 /O 8
1 OfiO (S 2
I 000 dO 4
Hill I AN
/.' i
/i 1
M <,
/•J ',
// 1
It, 5
Bl -1
t 1 1
/B 9
7 J tl
8L, 0
-------
TABLE
GAC
System
15A
ISA
15A
15A
15A
ISA
ISA
ISA
ISA
ISA
ISA
ISA
23A
23A
23A
D
D
D
D
D
D
58. COMPARISON OF CWW AND GIT ANALYSES OF
GAC
State
Virgin
Spent
Regenerated
Startup
Spent
Regenerated
Startup
Spent
Regenerated
Startup
Spent
Regenerated
Virgin
Spent
Regenerated
Virgin
Spent
Regenerated
Startup
Spent
Regenerated
Phase-
Run
3-0
3-0
3-0
3-1
3-1
3-1
3-2
3-2
3-2
3-3
3-3
3-3
1-0
1-0
1-0
3-0
3-0
3-0
3-1
3-1
3-1
Start-up GAC consisted of the
b virgin
Not run
GAC makeup as
•
Lab
CWW
GIT
CWW
GIT
CWW
GIT
CWW
GIT
CWW
GIT
CWW
GIT ,
CWW
GIT
CWW
GIT
CWW
GIT
CWW
GIT
CWW
GIT
CWW
GIT
CWW
GIT
CWW
GIT
CWW
GIT
CWW
GIT
CWW
GIT
CWW
GIT
CWW
GIT
CWW
GIT
CWW
GIT
same
Iodine
No.
mg/gni
1181
1184
805
711
1021
1011
1084
1128
902
1093
1061
1111
1093
1103
914
916
1060
1083
1067
1104
881
896
923
986
858
903
325
327
669
721
1077
1128
885
824
1008
1030
1064
1096
849
880
1075
1097
App.
Density
gm/ml
0.47
0.48
0.58
0.60
0.48
0.48
0.49
0.47
0.51
0.50
0.47
0.47
0.49
0.47
0.53
0.49
0.48
0.47
0.48
0.46
0.62
0.53
0.60
0.52
0.59
0.57
0.72
0.69
0.58
0.57
0.51
0.51
0.55
0.58
0.49
0.51
0.49
0.50
0.55
0.52
0.50
0.50
proportions of
were contained in full-scale
GAC CHARACTERISTICS
Sieve Analysis
Total
Ash
%
6.0
7.1
27.6
19.1
7.6
8.3
8.4
8.3
10.9
10.5
8.7
8.4
8.1
7.8
10.2
8.6
8.8
8.1
9.1
7.8
9.1
8.6
8.2
7.5
8.1
8.6
36.0
17.8
15.2
14.2
8.2
8.1
10.6
9.5
9.6
9.0
8.1
7.7
8.1
7.7
8.1
7.9
Eff.
Size
mm
0.78
0.73
0.90
0.75
0.70
0.69
0.83
0.72
0.62
0.52
0.60
0.61
b
0.63
0.62
0.58
0.62
0.58
0.62
0.58
0.71
0.73
b
0.66
0.42
0.40
0.38
0.40
0.40
0.42
0.68
0.64
0.76
0.60
0.70
0.58
0.62
0.62
0.84
0.68
0.62
0.61
regenerated GAC
for next
run.
Unif.
Coef.
1.56
1.74
1.62
1.91
1.63
1.70
1.50
1.67
1.60
1.92
1.67
1.72
b
1.44
1.51
1.41
1.45
1.31
1.48
1.35
1.60
1.44
b
1.36
1.45
1.47
1.53
1.43
1.45
1.36
1.82
1.80
1.60
2.00
1.68
1.90
1.74
1.49
1.67
1.52
1.71
1.45
and
217
-------
TABLE 59. TOTAL GAC LOSSES
Contactor GAC Filter
Location Phase
C
D
A
BB
C
D
A
D
BB
3-0
3-0
3-0
3-0
3-1
3-1
3-1
3-2
3-1
Average Loss, %
Percent
Loss
IS.8
14.9
13.8
18.0
15.6
18.9
14.2
16.4
9.8
15.3
Location
15A
21A
ISA
21A
ISA
ISA
Phase
3-0
3-1
3-1
3-2
3-2
3-3
Percent
Loss
22.6
22.6
23.7
13.7
13.6
14.8
Average Loss, %
18.5
Total losses calculated by volume measurements in each GAC adsorption bed.
volume of GAC was determined according to the "borrow-pit" method. The
losses determined by this method (Table 60) were inconsistent and not reflec-
tive of bed-to-bed losses. These losses, therefore, were considered invalid
due to the following shortcomings in the process: small sample size, vari-
ances in elapsed time of measuring each tank, erratic furnace operations
TABLE 60. GAC LOSSES ACROSS FURNACE3
Contactor GAC Filter
Location Phase
D
A
BB
C
D
A
D
3-0
3-0
3-0
3-1
3-1
3-1
3-2
Average Loss, %
Percent
Loss
15.9
12.0
14.6
16.7
3.4
22.8
27.9
16.2
Location
21A
ISA
21A
15A
Phase
3-1
3-1
3-2
3-2
Percent
Loss
31.2
16.4
16.6
8.7
Average Loss, %
18.2
Furnace losses calculated by volume measurements in spent and regenerated
GAC tanks.
218
-------
during tests and the difficulty associated with measuring an irregular
surface obscured by many feet of carbon-black water.
Three transport and regenerator battery loss tests were conducted on
the contactors (Table 61) to assess performance under the regenerator con-
struction contract guarantees. The average transport loss of 3.1% was
greater than anticipated, based on current literature, although it is compar-
able to losses reported from other GAC facilities. The regeneratory battery
losses fluctuated considerably from 12% to 7% to 14% in three tests. The
average 11% loss was much greater than that predicted by furnace manufac-
turers.
On several occasions, attempts were made to determine the efficiency of
the sand separator during GAC filter runs. Percent sand-in-GAC analyses
were run during GAC furnace loss tests, transport loss tests, and at other
times during runs. Samples were collected from the sand and GAC discharge
ends of the sand separator. The results contained in Table 62 indicate that
the sand separator, although removing most of the sand, did not completely
prevent sand from entering the furnace. As described in Objective 12, sand
which entered the furnace hampered furnace operations.
TABLE 61. CONTACTOR GAC TRANSPORT LOSSES
Contactor
C
BB
D
Phase-
Run
3-1
3-2
3-2
Average Loss, %
TABLE 62.
Sample
Date
Transport
Loss, %
4.0
3.0
2.4
3.1
Regenerator
Battery
Loss, %
11.6
6.8
14.0
10.8
Bed to Bed
Loss, %
15.6
9.8
16.4
13.9
Sep.
Feb.
Mar.
June
June
June
June
June
June
26,
20
25,
1,
1,
1,
2,
2,
1980
1981
1981
1981 (1st
1981 (2nd
1981 (3rd
1981 (1st
1981 (2nd
2, 1981 (3rd
shift)
shift)
shift)
shift)
shift)
shift)
SAND SEPARATOR, PERCENT SAND IN GAC
Percent Percent
Sand Separator GAC Sand by GAC by
Sample Location Filter Volume Volume
Sand Discharge 21A 84.6 15.4
Sand Discharge 15A 69.1 30.9
GAC Discharge 15A 4.6 95.4
Sand Discharge 21A 81.5 18.5
GAC Discharge 21A 1.5 98.5
GAC Discharge 21A 0.0 100.0
GAC Discharge 21A 5.2 94.8
GAC Discharge 21A 0.0 100.0
GAC Discharge 21A 2.5 97.5
Sand Discharge Avg. 78.4 21.6
GAC Discharge Avg. 2.3 97.7
219
-------
More important to the consideration of GAC losses is that an average of
22% (by weight) of the discharge from the sand end of the separator was GAC.
This is not as significant as it may seem since the total volume of discharge
off of the sand end of the separator averaged less than 3% of the total
volume of materials regenerated in any one run. Therefore, the GAC losses
from the sand end of the separator were less than 1% of the volume of mater-
ials regenerated. Further, it should be noted that the sample collection
process (core sampling or grab sampling) may have introduced considerable
error and, since separator adjustments were made continually throughout a
run, the results obtained may not be representative of that total run.
Objective 12. To Develop Parameters to be Used in the Design of Full-Scale
Systems
Sand Replacement GAC Filters—
Design criteria--
Empty bed contact time--Probably the most significant parameter in
designing a GAC system, EBCT, is somewhat fixed for sand replacement GAC
filters. Most existing rapid sand filters are designed for 1.4 to
2.8 Ips/sq m (2 to 4 gpm/sq ft) and bed depths of 61 to 76.2 cm (24 to
30 in). Conversion of these facilities for GAC use results in EBCTs of 3.7
to 9.4 minutes. Existing sand filters at CWW are designed for 1.7 Ips/sq m
(2.5 gpm/sq ft) and contain 76.2 cm (30 in) of media, resulting in an EBCT
of 7.5 minutes. Comparisons between filters with 30.5 cm (12 in) of filter
sand and 45.7 cm (18 in) of GAC and filters with 76.2 cm (30 in) of GAC have
established that the sand layer is unnecessary from the standpoint of filtra_
tion (Objective 2). Elimination of the sand layer results in a 40% increase
in EBCT.
Within the range of EBCTs available in converted sand filters, bed
service life, carbon use rate, and regeneration requirements are all improved
as the EBCT increases. Optimization of available EBCT should, therefore, be
evaluated when filters are converted from sand to GAC.
Surface loading—Full-scale performance at CWW indicated that WVG
12 x 40 GAC functioned satisfactorily at a surface loading of 1.7 Ips/sq m
(2.5 gpm/sq ft). Turbidity reduction was equivalent to that of sand filters
operated in parallel with GAC filters (Objective 2).
Carbon use rate and carbon service life are functions of EBCT, the
criteria selected to indicate GAC exhaustion, and the concentration of
contaminants in the influent to the GAC filters. Seasonal variations in the
concentrations of certain organic parameters must also be considered in
design. Plots of TOC and THMSIMDIST in the raw water, sand filter influent
and sand filter effluent indicate seasonal trends due to temperature, runoff,
etc. (Figures 110 and 111).
Full-scale adsorption systems must be designed for the shortest, or
critical, expected bed service life and maximum water production during the
220
-------
IZZ
CONCENTRATION,
CD
!D
D O
rt- T
CD
c-8
O —
O O
en
CD
CD
O —
-h
(D
»
O-
M
M
m
CD
CLCD
5
en
CD
o_
-*7
CD
CD
CO
en
CD
CD
CD
CD
CD CD CD CD
CD
CD
CD
CD
CD
CD
CD
CD
i
r\
1
v
2
§
P
m
m
u
2
?
i
-------
zzz
CONCENTRATION,
w co
01
cs>
C
O C/>
T —I
•H
CO
S>
CO
cn
(S
CD
I
3
m
1
y
?
-------
summer months. Operating costs, especially those for GAG consumption, will
be reflective of longer service life and reduced water production during
winter months.
Figure 112 and Tables 63 and 64 present seasonal TOC and THMSIMDIST
concentration data for GAC filters. "Summer critical" curves reflect the
shortest service life to achieve a given effluent goal, based on four summer
runs. "Winter average" curves present the average service life to achieve a
given effluent goal, based on four winter runs and is presented to emphasize
the critical nature of summer data.
These exhaustion criteria (e.g. 1,000 p.q/1 TOC) should not be applied
to any single GAC filter, but rather to total plant effluent. Actual plant
operation will include multiple GAC filters operating at the same time in
parallel. Individual GAC filters will be placed into and taken out of serv-
ice for GAC replacement on a "staggered schedule".
To meet a desired treatment goal in the combined flow from all filters,
some units may produce an effluent exceeding the treatment goal as long as
other units produce water equally below the treatment goal. Due to the fact
that the quality of the total plant output would be the average of the
quality of all filters in service, it follows that the bed life of each
filter would be approximately twice that of an individual filter operating
toward the same exhaustion goal.
For example, from Figure 112's TOC summer critical line, if 1,000 _ug/l
TOC were the plant average treatment goal, then a combined effluent from a
fresh GAC filter passing about 500 ug/1 TOC would permit an older GAC filter
to pass l,500jug/l TOC. The figure shows that it took 18 days to reach the
l,500jug/l treatment goal. Had the GAC filter been taken off line at a goal
of l,000_ug/l TOC, the service life would have been 8 days. Therefore, the
average life of a single filter would be approximately twice those indicated
in Tables 63 and 64. Since carbon use rate is inversely proportional to
service life, carbon use rates applicable to staggered-operation would be
one-half those of any single filter.
Therefore, carbon use rates and service life data are useful in the
design of GAC application and regeneration facilities and in the projection
of costs for building and operating them. It should be noted that TOC was
always the determinant of exhaustion rather than THMSIMDIST.
The "summer critical" carbon use rate is an important factor for design,
since facilities must be sized to handle this critical period. Thus for
1,000 ug/1 TOC the carbon use rate of 160 kg/mil 1 (1,334 Ib/mil gal) (calcu-
lated from Table 63's summer critical service life) and the service life of
16 days (2 times the Table 63 summer critical value) were used.
Due to significant variability in the influent concentration of contam-
inants, the annual average carbon use rate was used for determining operating
and maintenance factors and costs. Thus for 1,000/ig/1 TOC, the carbon use
223
-------
LEGEND
to
t\J
1600
B—B TOC - SUMMER CRITICAL
h—A STT3 - SUMMER CRITICAL
TOC - WINTER AVERAGE X—X
STT3 - WINTER AVERAGE -I 1-
1400 .
200
0
0
1 • | If 1
20 40 60 80 100
SERVICE TIME, days
i 1
120 140 160
160
140
120
100
80
60
a
3
CO
r
CO
i-
ll.
._ IL.
40 LJ
20
0
Figure 112. Total organic carbon CTOC) and three-day simulated distribution
system THM CSTT3) seasonal service time curves for 6AC filters.
-------
TABLE 63. GAC
TOC
GAC
Filter
Phase-
Run
Start
Month
FILTER SERVICE LIVES FOR SELECTED
TREATMENT GOALS, IN DAYS
250
Treatment Goals, >i
500 750 1,000 1
q/1
,250
1,500
Summer
ISA
ISA
21A
21A
Summer
3-1
3-2
3-1
3-2
Critical
Aug. '
Oct. '
Jul. '
Oct. '
80
80
80
80
0
0
0
0
0
049
2 7 11
148
2 7 12
048
14
16
12
21
12
20
42
18
30
18
Winter
15A
ISA
ISA
21A
Winter
GAC
Filter
3-0
3-3
3-4
3-3
Average
TABLE
Phase-
Run
Jan. '
Dec. '
Mar. '
Feb. '
80
80
81
81
0
3
0
0
1
64. GAC FILTER
THMSIMDIST
Start
Month
25
8 32 47
11 23 32
16 26 42
0 10 19
9 23 35
55
42
48
60
51
115
48
50
90
76
SERVICE LIVES FOR SELECTED
TREATMENT GOALS, IN DAYS
Treatment Goals , ju
50 75
g/1
100
125
Summer
ISA
ISA
21A
21A
Summer
3-1
3-2
3-1
3-2
Critical
Aug. '
Oct. '
Jul. '
Oct. '
80
80
80
80
4
10
2
7
2
10 17
42
8 14
21
8 14
24
22
22
34
33
33
Winter
ISA
15A
15A
21A
Winter
3-0
3-3
3-4
3-3
Average
Jan. '
Dec. '
Mar . '
Feb. '
80
80
81
81
52
35
36
38
103 131
55
60
73 131
160
160
-
-
225
-------
rate of 96 kg/mil 1 (800 Ib/mil gal) (one half of Figure 113 value) and the
service life of 38 days (two times Figure 114 value) were used.
Backwash requirements—A study was conducted to determine the appropri-
ate backwash rate for the GAC filters. A rate of 5.6 Ips/sq m (8 gpm/sq ft)
was found to expand the GAC bed without washing GAC out of the filter.
Backwashing was accomplished at a rate of 2 Ips/sq m (3 gpm/sq ft) for
4 minutes, 4.1 Ips/sq m (6 gpm/sq ft) for 4 minutes and then 5.4 Ips/sq m
(8 gpm/sq ft) for 4 minutes. GAC filters exhibited average service times,
between backwashings, which were approximately 50% longer than conventional
sand filters (Objective 2).
Regeneration requirements--GAC regeneration requirements are dictated
by summer operation. Raw water TOC, THHSIMDIST and water consumption are
all maximum during this period. This results in minimum GAC filter service
life and maximum required filtration capacity at the same time. Regeneration
capacity must, therefore, be sufficient to handle the spent GAC generated
daily at the maximum monthly average flow and the carbon use rate resulting
from the critical summer service life.
Standby GAC filter capacity--Removal and replacement of GAC requires
taking GAC filters out of service for a period of time for each regeneration
cycle. During this study, spent GAC was manually shovelled to an eductor
box. Regenerated GAC was returned to the filters as a slurry delivered by a
large diameter hose. The GAC bed was then backwashed and allowed to settle.
Additional GAC was added and this procedure repeated until the desired bed
depth was achieved. This process required approximately five days per
filter.
Full-scale GAC filters will require improved methods of spent GAC
removal. Alternate methods of GAC removal are site-specific and beyond the
scope of this study. It seems reasonable, however, to assume that GAC
filter downtime for removal and replacement of spent GAG can be reduced to
two days per regeneration cycle. To maintain equal plant capacity after
conversion from sand filters to GAC filters, additional filters must be
constructed to compensate for those out of service for GAC replacement. The
required standby GAC filter capacity is the maximum capacity which would
become exhausted daily multiplied by the expected duration of downtime to
remove and replace the spent GAC. The maximum capacity to be exhausted
daily is that resulting from maximum monthly average water production and
critical summer service life. As indicated in Objective 13, an additional
six filters would be required at an exhaustion criterion of 1,000 ug/1 TOC.
Design consideration--
GAC handling and transport—During this study, spent GAC was removed by
shovelling to an eductor box. This labor intensive removal method is imprac-
tical for full-scale facilities. Other methods should be evaluated prior to
design. This evaluation should consider minimizing the quantity of transport
water required as well as reducing operating labor.
226
-------
LZZ
CARBON USE RATE, kg/mI I I
en
©
6)
cs>
en
ro
§
co
CD
CO
1^
O TJ
o r
T >
D~ O
o rn
f* i^j
(0 Z
0>
•» m
0 37
<•«- <
« M
W O
m
co
S
B
O
TO
G?
H
O
2
xi
§
05
m
o
m
r~
3
o
H
O
O
«^>
4Q
-------
UJ
CD
LU
§
*^m
3
o
2
g
<
5
iI I I I ]i I r] r
j i
O.
LU
U
H %
> OJ
fi
H
e
o
9
^o. w
*=Z —
O —
o
-a:
CD
1 i
o
CO
CM —
Q
O>
U)
Q
CM —
228
-------
Less labor-intensive methods of GAC replacement should also receive
careful evaluation. A permanent installation of transport piping to each
GAC filter should be considered. As a minimum, consideration should be
given to an overhead crane for moving GAC handling equipment between filters.
GAC transport piping used in this study was black steel pipe with
forged steel fittings. Although proven to be satisfactory, other more
economical pipe materials should be evaluated. Special consideration must
be given to pipe sizing, bend radii and slurry consistency to minimize
abrasion and velocity to maintain suspension.
Related plant modifications—Several plant modifications are required
in support of filter conversion for GAC treatment. Additional filter
capacity will likely be required to compensate for filters down for GAC
removal and replacement. Washwater storage and pumping facilities may need
to be expanded to meet backwash and GAC transport requirements.
When converting sand filters, consideration should be given to optimiz-
ing the available EBCT. Methods of supporting the GAC bed, which will
prevent intermixing of GAC and sand, should be evaluated.
GAC treatment removes all free chlorine from the water resulting in
bacterial growth. Post-chlorination facilities will be required for any GAC
treatment system. Additional facilities may be required to ensure sufficient
chlorine contact time prior to distribution.
Post-filtration GAC Contactors
Design criteria--
Empty bed contact time—Figure 115 contains plots of carbon use rate
versus optimum EBCT for each contactor run during the study period. Carbon
use rates were calculated as the weight of GAC required per volume of water
throughput until the treatment goal of 1,000 /ig/1 TOC was reached based on a
three-week running average. (Each three-week set along the curve was
averaged and a new curve plotted from these values. The point at which this
curve intersected the treatment goal became the point of exhaustion.) As
shown by the average of all contactor runs, carbon utilization efficiency
improved (carbon use rate decreased) substantially as EBCT was increased to
15 minutes. Optimum carbon utilization apparently occurred at an EBCT
greater than 15 minutes, possibly 20 minutes, for the critical late summer
months. Determining which of the two EBCTs is the most cost-effective,
however, requires a look at the costs and benefits of increasing EBCT
(Objective 13).
Surface loading—Pressure contactors utilized during this study operated
at a surface loading of 5 Ips/sq m (7.4 gpm/sq ft). No significant reduction
in pressure was noted during contactor runs probably because the water had
previously been filtered. Therefore, backwashing was not required to maintain
a constant rate. Since filtration is not a factor in contactor design,
229
-------
400
350
- 300
£
250
H B CONTACTOR A, PHASE 3-8
A A CONTACTOR C, PHASE 3-0
0 $ CONTACTOR BB, PHASE 3-9
0 €) CONTACTOR BB, PHASE 3-1
'CONTACTOR A, PHASE 3-2
LEGEND
OJ
O
N
UJ
Of.
200
150
§ 100
03
Of
O 50
0
CONTACTOR B, PHASE 3-8
CONTACTOR D, PHASE 3-8
CONTACTOR A, PHASE 3-1
CONTACTOR C, PHASE 3-1
CONTACTOR C, PHASE 3-2
X X
AVERAGE
0
6 8 10 12 14
EMPTY BED CONTACT TIME, minutes
16
18
20
Figure 115. Total organic carbon (TOO carbon use rates for multiple
runs of contactors, Phase 3 (TOC=I000 yug/l).
-------
surface loading is a function of allowable head loss and space availability.
Surface loadings commonly range from 3.4 to 6.8 Ips/sq m (5 to 10 gpm/sq ft).
High liquid velocities resulting from surface loadings greater than
6.8 Ips/sq m (10 gpm/sq ft) may inhibit mass transfer and reduce adsorption
in the contractor.
Carbon use rate—as previously mentioned under GAC filters, carbon use
rate and service life are functions of EBCT, the criteria selected to indi-
cate GAC exhaustion, and the concentration of contaminants in the influent
to the contactors. Therefore, the considerable variations in influent
concentrations must be also considered in the design of contactors.
Table 65 and Figures 116 and 117 present the results of these variations
for TOC and THMSIMDIST concentrations for the contactors. Figure 118 illus-
trates significant variations in carbon use rates for a treatment goal of
1,000 jug/1 TOC and 15 minute EBCT. As can be noted therein, "summer
critical" and "staggered scheduling" phenomena, as observed for GAC filters,
also exist relative to the contactors, except that carbon use rates and
service lives are considerably improved. Therefore, carbon use rates and"
service life data are useful in the design of carbon application and regener-
ation facilities and in the projection of costs for building and operating
them.
The "summer critical" carbon use rate is an important factor for design,
since facilities must be sized to handle this critical period. Thus for
1,000 ;ug/l TOC and a 15 minute EBCT, the carbon use rate of 68 kg/mil 1
(562 Ib/mil gal) (one-half of the Figure 115 critical C (3-1) value) and the
service life of 78 days (two times the Figure 116 summer critical value)
were used.
Due to significant variability in the influent concentration of contam-
inants, the annual average carbon use rate was used for determining operating
and maintenance costs. Thus for 1,000 /ig/1 TOC and a 15 minute EBCT, a
carbon use rate of 34 kg/mil 1 (285 Ib/mil gal) (one-half of the Figure 118
average annual carbon use rate) and a service life of 186 days (two times
Figure 119's value) were used.
Although data were not collected at a 20-minute EBCT, service life may
TABLE 65. CONTACTOR SERVICE LIFE, IN DAYS
Treatment
Goal EBCT EBCT EBCT
Parameter ug/1 Season 2.8 min 7.2 min 15.2 min
THMSIMDIST 100 Summer (Critical) 9 23 70
THMSIMDIST 25 Summer (Critical) 2 9 29
TOC 1,000 Summer (Critical) 3 13 39
TOC 1,000 Winter (Average) 13 37 133
231
-------
1800
1600
1400
1200
o 1000
o
LEGEND
H—B SUMMER CRITICAL • 2.8 MIN. EBCT SUMMER CRITICAL » 7.2 MIN.
A—A SUMMER CRITICAL t 15.2 MIN. EBCT WINTER AVERAGE • 2.8 MIN.
WINTER AVERAGE • 7.2 MIN. EBCT WINTER AVERAGE t 15.2 MIN.
X—X
to
U)
to
Z
UJ
D
U.
U.
Id
800
600
400
200
0
50
200
100 150
SERVICE TIME, days
Figure 116. Total organic carbon (TOO seasonal service time curves for
contactors, Phase 3.
258
-------
to
U)
U)
160
LEGEND
H—Q SUMMER CRITICAL t 2.8 MIN. EBCT SUMMER CRITICAL 9 7.2 MIN. EBCT
A—A SUMMER CRITICAL • 15.2 MIN. EBCT SUMMER AVERAGE 0 2.8 MIN. EBCT H h
$—e> SUMMER AVERAGE t 7.2 MIN. EBCT SUMMER AVERAGE • 15.2 MIN. EBCT ©—0
X—X
140 .
10
20
50
60
30 40
SERVICE TIME, days
Figure 117. Three-day simulated distribution system THM CSTT3) seasonal
service time curves for contactors,, Phase 3.
70
-------
CARBON USE RATE., kg/ml I
ro
CO
o>
00
C5>
CO
CO
CO
0>
CO
c_
oo
§>
r»-
Q 2
O O
n- 2
O H
O TJ
O I-
-J >
O~ O
o m
•» m
o /o
»" H
- O
m
CO
oo
CO
rn
-------
to
CO
Ul
200
LEGEND
Q Q TOTAL ORGANIC CARBON C1880 jug/1)
JAN
MAR
MAY JUL SEP
MONTH PLACED IN SERVICE
NOV
Figure 119. Contactor service life., 1980.
-------
be approximated by referring to Figure 115. It can be assumed that the
maximum reduction in GAC utilization which might occur between EBCTs of 15
and 20 minutes would exist if carbon use rate improved following the same
slope exhibited between 7.2 and 15.2 minutes EBCT. Extrapolation of the
critical carbon use rate curve in Figure 115, Contactor C (3-1) indicates
that GAC utilization might be reduced to as low as 102 kg/mil 1 (850
Ib/mil gal) during the summer. This corresponds to a single contactor
service life of approximately 67 days (calculated from Figure 115's carbon
use rate) or 134 days for contactors in staggered operation.
Since average annual carbon use rate data was not generated for a
20 minute EBCT, an interpolation of the average of all Phase 3 runs contained
in Figure 115 was used as a surrogate. Accordingly, a carbon use rate of
20 kg/mil 1 (165 Ib/mil gal) (one-half of the Figure 115 value) was obtained.
Backwash reguirement--As indicated under "surface loading", backwashing
was not required during the study period. Backwash facilities are necessary,
however, to remove GAC fines after initial filling and to fluidize the bed
for GAC removal.
A backwash rate of 6.8 Ips/sq m (10 gpm/sq ft) was utilized for contac-
tors and proved satisfactory. Backwashing was initiated at 2.0 Ips/sq m
(3 gpm/sq ft) for five minutes, then 4.1 Ips/sq m (6 gpm/sq ft) for five
minutes, followed by two periods at 6.8 Ips/sq m (10 gpm/sq ft) of 20 and
15 minutes each with five minutes settling between the two.
Regeneration requirements--GAC regeneration capacity is controlled by
summer operation when maximum water production is required and GAC service
life is at a minimum. Regeneration capacity must, therefore, be sufficient
to handle the daily production of spent GAC at the maximum monthly average
flow and the carbon use rate resulting from critical summer contactor service
life.
Standby contactor capacity--Removal of spent GAC and replacement with
regenerated or virgin GAC necessitates that the contactor be out of service
for a period of time each cycle. Sufficient standby contactor capacity must
be provided to compensate for this lost capacity. The maximum capacity out
of service at one time will be that resulting from maximum monthly average
water production and critical summer contactor service life, multiplied by
the expected duration of downtime for GAC removal and replacement. As
indicated in Objective 13, one additional contactor would be required.
Design considerations--
GAC handling and transport--GAC was removed from the contactors by
fluidizing the GAC bed and utilizing hydraulic pressure to force out the GAC
slurry. An additional point of water injection was installed at the GAC
outlet to adequately fluidize the GAC so that the slurry would flow. The
contactor vessels had dished bottoms which were below the level of the
236
-------
treated effluent header. GAC in the dished portion could not be adequately
fluidized, which resulted in difficulties.
GAC transport piping was extra-strong black steel pipe with forged
steel flanges and fittings. All valves in the transport system were full-
port stainless steel ball valves with Teflon seats. This appears to have
been a satisfactory application.
Process pumps and piping—Ordinary water plant pumps, piping, valves
and rate controllers were used with contactors. This project yielded the
following observations which might be considered in future designs:
1. Contactor backwash discharge ports should be located above the expanded
bed when backwashing to eliminate GAC loss. Screens did not function
properly because of clogging problems.
2. The contactor effluent header should be lower than the bottom of the
contactor to allow draining and removal of transport water. If this is
not desirable from a design standpoint, a separate line must be instal-
led.
3. Air and vacuum relief valves must be adequately sized for the capacities
of air required.
4. Turbulence in the piping downstream from the rate-of-flow controller
caused cavitation problems with the piping. This was due to abrupt
changes in flow direction both upstream and downstream from the valve,
the position of the axis of the valve in relation to the flow, and the
sizing of the valve. In designing rate controllers and piping, attention
should be given to all factors that are normally considered for a valve
used to throttle flows. Proper sizing and piping arrangement should
eliminate these problems.
5. The piping layout did not permit adequate isolation"for working on all
sections of piping and valves without taking the entire installation
out of service.
Materials of construction—All equipment that will not be in contact
with GAC can be constructed of materials normally used for water treatment
plants. The contactors must be designed using materials or liners that do
not impart objectionable organics to the water and are resistant to abrasion
and chemical attack by GAC.
Related plant modifications—The installation of contactors will require
additional modifications to existing facilities. The extent of these modifi-
cations are site specific. New pumping facilities may be required to relay
filtered water to the contactor facilities. Additional washwater or GAC
transport water facilities may be required. Post-chlorination will be
necessary, possibly requiring additional chlorine contact time.
237
-------
On-site GAC Regeneration--
Design criteria--
Regeneration capacity--Regeneration capacity must be sufficient to meet
the demands of summer operation. Furnace downtime must be factored into the
design capacity. During this study, the regeneration system was operable
approximately 60% of the time. Experience at other installations has indi-
cated continued improvement in operating efficiency with time. Taking this
potential improvement into consideration, 70% operation appears reasonable
for design.
Make-up GAC storage—Transport and regeneration of GAC results in
losses due to incineration and carry-over of fines in motive water. Adequate
virgin GAC must be stored on site to replace regeneration losses. Transport
and regeneration losses for GAC filters averaged 18.5%. Losses for contac
tors averaged 15.3%. The higher losses from GAC filters were due to longer
transfer piping runs, shovel handling of GAC and sand separator inefficiency.
Sand removal—A vibrating sand separator table was utilized to separate
sand from GAC prior to filter regeneration. The separator was adjusted by
regulating the flow of water to the table and the slope of the table top.
Difficulties were encountered in this operation since sand accumulated in
the dryer and regeneration sections of the furnace which necessitated shut-
ting the furnace down for manual removal of sand. Future designs of full-
scale facilities should evaluate alternative methods of sand removal.
Design considerations--
Process considerations—Several observations were made during the
course of this study which should be considered in future designs.
1. Off-gases from the dryer section of the regeneration furnace passed
through a cyclone separator and were then reintroduced into the incin-
eration zone of the furnace prior to discharge to the atmosphere through
a venturi scrubber. Incineration of off-gases was utilized to ensure
that organics were not discharged to the atmosphere. Analysis of the
dryer off-gases indicated insufficient low temperature volatiles to
justify the required secondary burner ("Other Observations" section).
Fuel consumption would be substantially reduced by eliminating this
feature of furnace design. Results from this study should be compared
to those obtained from other installations.
2. A number of regeneration system operating problems were related to the
computerized control system. The sophisticated control package made
manual control difficult and resulted in substantial down-time. Future
designs should consider various control options.
3. Regardless of regenerator size, a minimum of three furnaces would be
recommended for a full-scale installation. This would increase
238
-------
flexibility of operation and would provide standby capacity during
maintenance.
4. The slurry pump used to transport GAC from the regeneration system
required injection of additional water to create the proper fluidity.
Backpressure created by thick GAC slurry resulted in separation of the
connecting hoses and in a long section of piping becoming clogged with
thick GAC paste.
Ancillary facilities—The regeneration system requires adequate power,
fuel, water, and air to support the installation. A source of softened
water is necessary for generation of steam which is used for bed fluidiza-
tion. Consideration must also be given to handling the quantities of motive
water which must be returned to the treatment process. Disposal of GAC
fines removed by the cyclone and scrubber gas streams must also be evaluated.
An overhead crane and large access doors are required to aid installation
and maintenance of the regeneration system.
Objective 13. To Determine Costs of Research Experience and Project Costs
of Full-Scale Plant Conversion
Costs Developed Under Research Conditions--
Although the costs developed under this section were real costs, they
can be misleading and should not be considered as indicative of the costs to
convert to full-scale operations. Factors which must be considered in
qualifying the costs developed in this section include:
1. The facilities were built primarily for research purposes adding
significantly to their cost.
2. The plant evaluated under this section compared a 0.44 cu m/s (10 mgd)
GAC filter system with a 0.17 cu m/s (4 mgd) contactor system, the
actual configuration operated during Phase 3. Differences of scale
were compensated to some extent by calculating costs per 3,785 1 (1,000
gal).
3. The application of GAC in filters would require the addition of filters
and GAC storage to compensate for the reduction in plant capacity due
to frequent filter changeovers for regeneration.
4. The furnace built for this grant was oversized relative to 0.44 cu m/s
(10 mgd) and 0.17 cu m/s (4 mgd) adsorption systems. Had there been
additional GAC to regenerate, the furnace capacity would have allowed
about 272 metric tons (300 short tons)/yr more to be regenerated over
the optimum levels presented in this section. Therefore, the optimum
costs for GAC regeneration would have been about ll<:/kg (5<:/lb) less.
5. Maintenance costs were practically non-existent during the grant since
most systems were new and covered by warrantees. This was hardly a
real life situation.
239
-------
The effect of these other factors will be addressed in the next section
of this report.
Tables 66 through 68 show some of the pertinent background data used in
the cost calculations that will be subsequently discussed. The useful life
cycles shown on Table 66 were estimated by CWW personnel based on normal
operating experiences gained to date. These were used in the calculation of
"actual" costs. Costs were also calculated using a straight 20-year life
and were used in the calculation of "optimum" costs. Costs incurred during
the years 1979 and 1980 were summarized for use in the "research condition"
part of this objective.
"Actual" costs are defined as those costs actually incurred during a
time frame most representative of "normal" operations under this grant.
These costs reflect the fact that the filters and contactors were not
regenerated or otherwise operated exactly according to plan (for reasons
indicated in other sections of this report). "Optimum" costs reflect costs
that would have been incurred if the operating and regeneration plan (primar-
ily a six week-cycle for a filter and a three-month cycle for a contactor)
had been rigidly adhered to.
No capital costs or electrical operating costs were included for GAC
filters because no additional costs were incurred above what was normally
required for a sand filter. Thus the costs shown reflect only the additional
costs required by the use of either GAC system. The sand replacement trans-
port system was manually fed and contributes to GAC losses through excessive
handling. In a full-plant conversion, capital expenditures for a more
efficient transport system would be a wise investment.
TABLE 66. USEFUL LIFE ESTIMATES FOR CAPITAL COST DATA
GAC Filters
Initial GAC Inventory - 20 year life
Contactors
Initial GAC Inventory - 20 year life
Contactors and Instrumentation - 8 year life
Building - 25 year life
Regenerator
Furnace, Tanks & Controls - 15 years
Building - 25 years
All unit capital costs are amortized at 10'
over the useful life span indicated.
240
-------
Power
Natural Gas
Water
GAC
Labor
Contactor & Furnace
General Labor
Maintenance
Natural Gas Fuel Value
Reactivation Fuel Use
TABLE 67. COST FACTORS
Metric
2.19- 2.91
-------
TABLE 68. DETAILED CAPITAL COSTS
Building Cost in Dollars
Foundation & Tunnel $ 70,700
Building Heat., Vent. & Light 175,190
Floor & Drain System 52,100
Potable Water Piping 3,080
Final Connections 8,160
Misc. 4,210
Engineering, Design 44,520
Engineering, Resident 20,970
Total Building $ 378,940
Contactors
Pressure Vessels (4) $ 355,550
Sample Troughs 24,310
Influent/Effluent Piping 156,380
Process Pumps 37,370
Backwash Piping 77,490
GAC Transport Piping 60,100
Compressed Air Equip. 38,130
Switchgear 53,500
Painting 12,100
Engineering, Design 82,920
Engineering, Resident 16,540
GAC Installation Labor 920
Initial GAC 85,290
Total Contactors $1,000,600
GAC Filters (2)
Sand Removal Labor 2,000
GAC Installation Labor 1,400
Sample Units 11,190
Turbidimeters 2,030
Initial GAC • 96,250
Total GAC Filters $ 112,870
Regenerator
Furnace 248,230
Instrumentation 224,500
Storage Tanks (2) 53,650
Steam Generator and Piping 14,900
Sand Separator 13,850
Motor Control Center & Wiring 17,010
Water Piping 9,810
Drain Piping 7,220
GAC Transport Piping 38,900
Engineering, Design 84,780
Engineering, Resident 7,410
Final Connections 38,630
Total Regenerator $758,890
Total Capital Costs $2,251,300
242
-------
TABLE 69. CAPITAL COST SUMMARY
GAC Filters
0.44 cu m/s
(10 mgd)
Contactors
0.17 cu m/s
(4 mgd)
Reactivation
230 kg/hr
(500 Ib/hr)
Actual Optimum Actual Optimum Actual Optimum
Construction $ 16,620 $ 16,620 $ 915,310 $ 915,310 $758,890 $758,890
Building
Cost
Initial GAC
96,250 96,250
189,470 189,470 189,470 189,470
85,290 85,290
Total
Capital Cost 112,870 112,870 1,190,070 1,190,070 948,360 948,360
Prorated
Reactivation 531,080 663,850 417,280 284,510 - -
Total Costs $643,950 $776,720 $1,607,350 $1,474,580
Land, legal, fiscal and administrative and insurance costs were not
included.
because of the eight-year life estimated for the entire contactor system.
The total annualized operating and capital costs are shown in Table 72 in
dollars and in Table 73 in C/1000 gal of water throughput or C/lb of GAC
throughput. The criterion of a six-week or a three-month bed life must be
kept in mind when considering the optimum costs shown.- Relaxed criteria
would result in lower costs, with reduced water quality being delivered.'
The actual and optimum costs including prorated on-site reactivation, were
3C/cu m (13C/1000 gal) and 4<:/cu m (14<:/1000 gal), respectively, for the GAC
filter system and 10<:/cu m (36C/1000 gal) and 6C/cu m (23C/100Q gal), respec-
tively, for the contactor system. The actual and optimum GAC reactivation
costs were 21<:/lb and 19
-------
TABLE 70. PRELIMINARY ANNUAL OPERATING AND MAINTENANCE COSTS
Water
GAC Transport
Process
Electric
Pumping
Building
Furnace
Nat'l. Gas
Operating Labor
GAC Filters
Contactors
Furnace
Operating Mat11.
Mtce. Labor
Mtce. Mat'l.
Subtotal
Related
React. Costs
Prorata
React. Costs
Totals
GAC Filters
0.44 cu m/s
(10 mgd)
Contactors
0.17 cu m/s
(4 mgd)
Reactivation
230 kg/hr
(500 Ib/hr)
Actual Optimum Actual Optimum Actual Optimum
$ 1,380 $ 1,930 $ 980 $ 930
14,800 17,000
12,670
3,810
14,790
3,810
3,810
4,440
35,520
97,400 136,360 53,920 51,220
3,810
5,100
40,900
29,400 41,160 - - 28,000 39,200
18,110 17,680 22,850 21,700
12,500 12,500
66,850 66,850
1,540 1,540
$128,180 $179,450 $ 89,490 $ 88,430 $190,310 $208,600
28,000 39,200 22,850 21,700
78,100 103,390 61,360 44,310 - -
$234,280 $322,040 $173,700 $154,400
filters. Also, as indicated in Table 73, the greater annual throughput
rendered further advantage to the GAC filters wherein even the operating
cost per 1000 gal was lower. An attempt to compensate for the disclaimers
was made in the next section.
244
-------
TABLE 71. COMPARISON OF ANNUAL AMORTIZED CAPITAL COSTS
Actual
Optimum
GAC Filters
Contactors
Reactivation System
Prorated Reactivation System
GAC Filters
Contactor
Total Costs
GAC Filters & Reactivation
Contactors & Reactivation
TABLE 72. TOTAL
GAC Filters
0.44 cu m/s
(10 mgd)
Based on
Useful Life 2
$ 13,260
202,460
120,650
67,560
53,090
$ 80,820
$255,550
ANNUAL COSTS RECAP
Contactors
0.17 cu m/s
(4 mgd)
Based on
0 Year Life
$ 13,260
139,790
111,390
77,970
33,420
$ 91,230
$173,210
Reactivation
230 kg/hr
(500 Ib/hr)
Actual Optimum Actual Optimum Actual
Optimum
Operating
Capital
Subtotal
Prorated React.
Operating
Capital
Subtotal
Total
Annual Costs
$128,180 $179,450
13,260 13,260
$141,440 $192,710
$106,100 $142,590
67,560 77,970
$173,660 $220,560
$ 89,490 $ 88,430 $190,310 $208,600
202,460 139,790 120,650 111,390
$291,950 $228,220 $310,960 $319,990
$ 84,210 $ 66,010
53,090 33,420
$137,300 $ 99,430
$315,100 $413,270 $429,250 $327,650
Proration of reactivation costs determined by adding prorata shares recorded
in Table 62.
This "optimum" is higher than "actual" since under optimum operating
conditions much greater amounts of makeup GAC and regenerated GAC would
have been used.
245
-------
TABLE 73. UNIT COSTS RECAP
C/1000 gal water throughput
GAC Filters
0.44 cu m/s
(10 mgd)
Actual
5.3
0.6
Operating
Capital
Subtotal
Prorated React.
Operating
Capital
Subtotal
Total
, To determine C/IOQQ 1 divide figures by 3.8.
To determine <:/kg multiply figures by 2.2.
5.9
4.5
2.8
7.3
13.2
Optimum
6.1
0.4
6.5
4.8
2.7
7.5
14.0
Contactor
0.17 cu m/s
(4 mgd
C/lb GAC throughput
Reactivation
230 kg/hr
(500 Ib/hr)
Actual
7.5
16.9
24.4
7.0
4.4
11.4
35.8
Optimum Actual
6.3 12.9
10.0 8.1
16.3
4.7
2.4
7.1
23.4
21.0
Optimum
12.3
6.5
18.8
Cost projections for full-plant conversion--
General—Conceptual cost estimates have been prepared for CWW in
order to determine the magnitude of costs of GAC treatment at a major water
treatment facility. The plant has a design capacity of 10'.3 cu m/s (235 mgd)
utilizing 47 rapid sand filters rated at 0.2 cu m/s (5 mgd) each. Design
average flow (maximum monthly average) is 8.8 cu m/s (200 mgd).
Estimates or probable costs have been developed for GAC filters and
contactors, each with on-site regeneration facilities. Where applicable,
data collected during this study have been utilized for sizing the required
facilities and estimating costs. Cost data from a previous USEPA study
were also used extensively. Operating and maintenance cost increases were
developed using the factors presented in Table 74 and current annual average
water production of 5.7 cu m/s (140 mgd). Labor costs were developed by
estimating the number and classification of additional employees required to
operate and maintain the expanded facilities. All costs developed in the
"full plant conversion" part of this objective were based on September, 1981
dollars and on preliminary conceptual designs and assumptions (Objective 12).
GAC filters--Two alternatives were considered in evaluating GAC
filter systems. One alternative was based on controlling TOC at less than
246
-------
TABLE 74. OPERATING AND MAINTENANCE COST FACTORS
Electricity $ 0.04 kWh
Natural Gas3 $ 0.12/cu m ($0.34/100 cf)
Service Water $ 0.13/cu m ($0.50/1,000 gal)
GAC $ 1.87 kg ($0.85/lb)
Average Labor Rate (inc. fringes) $ 12.50/hr
Chlorine $264 /mt ($240.00/ton)
GAC Transport Water 8.34 I/kg GAC (1 gal/lb GAC)
Regeneration Electricity 0.13 kWh/kg GAC (0.06 kWh/lb GAC)
Regeneration Natural Gas 4.2 joule/kg (2,000 Btu/lb GAC)
GAC Regeneration Loss 18.5 %
a At local utility quote of 1,000 Btu/cu ft.
1,000 jug/1 and the other at TOC less than 1,500 ;ug/l. Both alternatives
ensure continued control of THMSIMDIST. GAC filters would operate at
1.7 Ips/sq m (2.5 gpm/sq ft), providing 7.5 minutes EBCT.
GAC filter service lives were developed in Objective 12. However,
because the study included a limited number of runs, it would seem prudent
to incorporate a safety factor in a full-scale design. Summer critical GAC
filter service lives used in this evaluation have been reduced by approxi-
mately 15% to allow for this uncertainty.
A system to reduce TOC to less than 1,000 ,ug/l could be designed for
service lives of 14 days in the summer (85% of twice the summer critical
values shown in Table 63). Maximum monthly average water production of
4.4 cu m/s (200 mgd) requires that 40 filters be in service. A 14-day
summer service life would, therefore, necessitate removing three GAC filters
from service each day for replacement of spent GAC. • Assuming two days
downtime for GAC replacement, six additional filters would be required to
maintain the same plant capacity. Daily regeneration capacity must be
sufficient to handle spent GAC from three filters, each containing 48,500 kg
(106,750 Ib). Assuming a 70% regenerator design service factor (uptime),
required regeneration capacity would be 204,000 kg/day (450,000 lb/day).
A system designed to reduce TOC to less than 1,500 >ig/l could be
designed for service lives of 31 days in the summer (85% of twice the summer
critical values shown in Table 63). Operation of this system would require
that three GAC filters be removed from service every two days during the
summer for replacement of spent GAC. Assuming two days downtime per GAC
filter for servicing, three additional filters would be required to maintain
the same plant capacity. Regeneration capacity would be needed to handle
spent GAC from 1.5 filters [72,700 kg (160,125 Ib)] daily. A design capac-
ity of 104,400 kg/day (230,000 lb/day) would provide the required firm
capacity at 70% uptime.
247
-------
Table 75 presents a design summary of the two treatment alternatives.
Each alternative includes conversion of the existing 47 rapid sand filters
by replacing the filter sand with GAC and installing GAC handling facilities.
Additional GAC filters would be constructed for either alternative, would be
the same configuration as the modified units and would include surface
washing facilities. Preliminary calculations indicate that additional
backwashing facilities would not be required. Handling facilities would
include spent GAC removal troughs at the support gravel/GAC interface, spent
GAC transport piping and pumps and fresh GAC return piping. Regeneration
facilities would include multiple fluid bed furnaces housed in a separate
building along with spent and regenerated GAC storage, associated support
facilities, and controls. Virgin make-up GAC would be stored in cone-
bottomed vessels housed with the regeneration facilities. Post-chlorination
requirements would include new chlorine storage and feed facilities ahead of
the clearwells. Baffling would be installed in the clearwells to prevent
short-circuiting and to make full use of available contact time. Also, high
service pumping facilities for two small service areas must be relocated
from the filter effluent flume to the clearwells to ensure adequate chlorine
contact.
Estimated costs for the two GAC filter treatment alternatives are pre-
sented in Table 76. Capital costs were amortized at an interest rate of
10 percent over 20 years. This cost summary indicates that treatment to
reduce TOC to less than 1,000 jug/1 will result in an increased cost of
approximately 6C/cu m (24C/1.000 gal). Providing treated water with TOC
less than 1,500 /Kj/1 will result in increased costs of approximately 3C/cu m
(13
-------
TABLE 75. CONCEPTUAL DESIGN SUMMARY FOR GAC FILTERS
Treatment Goal
GAC Filters:
Surface Area
Hydraulic Loading
Empty Bed Contact Time
Critical Summer Serv. Life
Annual Avg. Carbon Use Rate
Number Units Req'd. -
Existing
New
GAC per Unit
GAC Regeneration:
Max. Spent GAC
Regenerator Capacity
Number Furnaces Req'd.
Furnace Capacity
Spent GAC Storage Capacity
Regen. GAC Storage Cap.
Virgin GAC Storage:
Assumed Regeneration
& Transport Loss
Max. Make-up GAC
Annual Avg. Make-up GAC
Storage Capacity
Post-Chlorination:
Cl Storage & Feed Cap.
Contact Time9
1,000 >ug/l TOG
130 sq m (1,400 sq ft)
1.7 Ips/sq m (2.5 gpm/sq ft)
7.5 min
14 days
96 kg/mil 1 (800 Ibs/mil gal)
1,500 jjg/l TOG
130 sq m (1,400 sq ft)
1.7 Ips/sq m (2.5 gpm/sq ft)
7.5 min
31 days
42 kg/mil 1 (350 Ibs/mil gal)
47
6
48,500
145,400
204,000
9
22,700
204,000
204,000
kg (106,750 Ib)
kg/day (320,250 Ib/day)
kg/day (450,000 Ib/day)
kg/day (50,000 Ib/day)
kg (450,000 Ib)
kg (450,000 Ib)
18.5 %
28,900 kg/day (59,250 Ib/day)
11,360 kg/day (25,000 Ib/day)
113,500 kg (250,000 Ib)
4,450 kg/day (9,800 Ib/day)
2.5 hr
47
3
48,500
72,700
113,500
5
22,700
113,500
113,500
kg (106,750 Ib)
kg/day (160,125 Ib/day)
kg/day (250,000 Ib/day)
kg/day (50,000 Ib/day)
kg (250,000 Ib)
kg (250,000 Ib)
18.5 %
13,450 kg/day (29,620 Ib/day)
5,000 kg/day (11,000 Ib/day)
50,000 kg (110,000 Ib)
4,450 kg/day (9,800 Ib/day)
2.5 hr
At 893 kcu m (235 mgd).
-------
TABLE 76. COST ESTIMATES FOR FULL-SCALE CONVERSION OF PLANT TO GAC FILTERS
Treatment Goal
1,000 ;ig/l TOC 1,500 jug/I TOC
Capital Costs
Convert Existing Filters
Additional GAC Filters
Regeneration System
Make-up GAC Storage
Chlorine Storage and Feed
Chlorine Contact Modifications
Miscellaneous Site Work
Contractor Overhead & Profit
Engineering, Legal &
Administrative
Initial GAC Inventory
Total Capital Cost
Annual Amortized Capital Cost
Annual 0 & M Cost
Natural Gas
Electricity
Maintenance Material
Service Water
Chlorine
Make-up GAC
Labor
Total Annual 0 & M
Total Annual Cost
Cost Increase per cu m
Cost Increase per 1,000 gal
$ 3,800,000
3,500,000
17,000,000
250,000
450,000
1,500,000
1,300,000
2,400,000
2,300,000
4,800,000
$37,300,000
$ 4,381,260
$ 300,000
180,000
380,000
50,000
100,000
6,450,000
570,000
$ 8,030,000
$12,411,260
$ 3,800,000
2,200,000
9,500,000
150,000
450,000
1,500,000
900,000
1,700,000
1,600,000
4,500,000
$26,300,000
$ 3,089,200
$ 130,000
90,000
200,000
30,000
100,000
2,830,000
360,000
$ 3,740,000
$ 6,829,200
$
$
0.063 $
0.240 $
0.036
0.135
, Amortized at 10% for 20 years.
Based on current average of 5.7 cu m/s (140 mgd).
(10 mgd) each having a GAC bed depth of 3m (10 ft) and a surface area of
39.6 sq m (1,400 sq ft). Design capacity flow of 10.3 cu m/s (235 mgd)
would require 24 contactors in service with maximum average production of
88 cu m/s (200 mgd) requiring 20 units. At the critical service life of 66
days, one contactor would be taken out of service approximately every third
day for GAC removal and replacement. Assuming that removal of spent GAC and
250
-------
replacement with fresh requires that the unit be out of service no more than
three days, one spare contactor would be sufficient. Regeneration capacity
must handle one-third of a contactor each day. Assuming a 70% uptime, the
required capacity would be 90,800 kg/day (200,000 Ib/day).
The 20-minute EBCT alternative was extrapolated (Objective 12) to have
a summer service life of 114 days. This system also includes 25 contactors
rated at 0.44 cu m/s (10 mgd) each having a GAC bed depth of 4 m (13.5 ft)
and a surface area of 130 sq m (1,400 sq ft). Peak design flow of
10.3 cu m/s (235 mgd) would require 24 contactors on line with maximum
monthly average flow of 8.8 cu m/s (200 mgd) requiring 20 units. At a
service life of 114 days, one contactor would be taken out of service approx-
imately every six days. Assuming that removal and replacement of spent GAC
would require that the unit be out of service no more than four days, a
single spare contactor would be needed. Regenerator capacity of
63,500 kg/day (140,000 Ib/day) would provide sufficient capacity to regener-
ate one-sixth of a contactor daily at 70% uptime.
Table 77 presents a design summary of the two contactor alternatives.
Each includes intermediate pumping to transfer filter effluent to the contac-
tor facility. The contactors would be similar in configuration to the exis-
ting sand filters but would be equipped with GAC removal troughs at the
interface of the support gravel and the GAC bed. Facilities would be pro-
vided to transport spent GAC to the regeneration facilities and to return
fresh GAC to the contactors. The entire contactor complex would be housed.
On-site regeneration would include sufficient spent and regenerated GAC
storage to hold the contents of an entire contactor. Each alternative would
include four fluid bed furnaces and associated support facilities and con-
trols. Virgin make-up GAC would be stored in cone-bottomed vessels which
would be housed in a complex with the regeneration furnaces and spent and
regenerated GAC storage facilities. Backwashing and transporting spent GAC
from the contactors would require washwater in addition to the normal
requirement for the sand filters. Each alternative includes 3.8 mil 1
(1 mil gal) additional washwater storage capacity. ' Post-chlorination
facilities would include new chlorine storage and feed ahead of the existing
clearwells as well as installation of baffling in the clearwells to prevent
short-circuiting. High service pumping suction for two small service areas
would also require relocation from the sand filter effluent flume to the
clearwells to provide GAC-treated and chlorinated water to these areas.
Because the USEPA cost curves assumed a sand/gravel base under the contac-
tors, the assumed regeneration and transport loss of 18.5% for the GAC
filters was also used for the contactors. The sand/gravel base in the GAC
filters was believed responsible for the higher GAC filter losses.
Relative costs for the two contactor alternatives are presented in
Table 78. Capital costs were amortized at ten percent over 20 years.
Annual operation and maintenance expenses were based on current water produc-
tion averaging 6.2 cu m/s (140 mgd).
The outcome of the economic analysis presented in Table 78 indicates
that both 15 and 20-minute EBCT alternatives will result in increased costs
251
-------
TABLE 77. CONCEPTUAL DESIGN SUMMARY FOR CONTACTORS
IV)
tn
rsi
Intermediate Pumping:
Design Capacity
Contactors:
Surface Area
Hydraulic Loading
GAC Bed Depth
Critical Summer Serv. Life
Annual Avg. Carbon Use Rate
Number Units Req'd.
GAC per Unit
GAC Regeneration:
Max. spent GAC
Regenerator Capacity
Number Furnaces Reg'd.
Furnace Capacity
Spent GAC Storage Capacity
Regen. GAC Storage Cap.
Virgin GAC Storage:
Assumed Regenerated
& Transport Loss
Max. Make-up GAC
Annual Avg. Make-up GAC
Storage Capacity
Additional Backwash Storage:
Storage Capacity
Post-Chlorination:
Cl Storage & Feed Cap.
Contact Time3
66
34
25
193,860
EBCT, 15 min
10.3 cu m/s (235 mgd)
130 sq m (1,400 sq ft)
3.4 Ips/sq m (5 gpm/sq ft)
3 m (10 ft)
days
kg/mil 1 (285 Ibs/mil gal)
64,620
90,800
4
22,700
195,220
195,220
kg (427,000 Ib)
kg/day (142,330 Ib/day)
kg/day (200,000 Ib/day)
kg/day (50,000 Ib/day)
kg (430,000 Ib)
kg (430,000 Ib)
18.5 %
12,000 kg/day (26,330 Ib/day)
1,076 kg/day (8,960 Ib/day)
10,808 kg (90,000 Ib)
3,785 cu m (1 mgd)
4,450 kg/day (9,800 Ib/day)
2.5 hr
EBCT, 20 min
10.3 cu m/s (235 mgd)
130 sq m (1,400 sq ft)
3.4 Ips/sq m (5 gpm/sq ft)
4.1 m (13.5 ft)
114 days
20 kg/mil 1 (165 Ibs/mil gal)
25
261,700 kg (576,450 Ib)
43,620
63,560
4
15,900
263,320
263,320
kg/day (96,075 Ib/day)
kg/day (140,000 Ib/day)
kg/day (35,000 Ib/day)
kg (580,000 Ib)
kg (580,000 Ib)
18.5 %
8,100 kg/day (17,770 Ib/day)
623 kg/day (5,190 Ib/day)
6,245 kg/day (52,000 Ib)
3,785 cu m (1 mgd)
4,450 kg/day (9,800 Ib/day)
2.5 hr
At 893 kcu m (235 mgd).
-------
TABLE 78. COST ESTIMATES FOR FULL-SCALE
Capital Costs
Intermediate Pumping
Contactor Facilities
Regeneration System
Make-up GAC Storage
Chlorine Storage and Feed
Chlorine Contact Modifications
Additional Washwater Storage
Miscellaneous Site Work
Contractor Overhead & Profit
Engineering, Legal &
Administrative
Initial GAC Inventory
Total Capital Cost
Annual Arnoritized Capital Cost
Annual 0 & M Cost
Natural Gas
Electricity
Maintenance Material
Service Water
Chlorine
Make-up GAC
Labor
Total Annual 0 & M Cost
Total Annual Cost
Cost Increase per cu m
y.
Cost Increase per 1,000 gal
, Amortized at 10% for 20 years.
Based on current average of 5.7 cu
CONVERSION OF
EBCT,
15 min
$ 1,700,000
10,800,000
7,900,000
150,000
450,000
1,500,000
1,100,000
1,200,000
2,200,000
2,400,000
9,100,000
$38,500,000
$ 4,522,210
$ 110,000
810,000
280,000
30,000
100,000
2,290,000
400,000
$ 4,020,000
$ 8,542,210
$ o
$ o
m/s (140 mgd)
PLANT TO CONTACTORS
EBCT,
20 min
$ 1,700,000
12,000,000
7,800,000
100,000
450,000
1,500,000
1,100,000
1,200,000
2,300,000
2,500,000
12,250,000
$42,900,000
$ 5,039,030
$ 60,000
840,000
310,000
20,000
100,000
1,330,000
400,000
$ 3,060,000
$ 8,099,030
.043 $ 0.042
.165 $ 0.160
of appronimately 4C/cu m (16C per 1,000 gal). Recalling that liberal extrap-
olations of GAC service life were used to evaluate the 20-minute system, it
can be concluded that 15-minute EBCT would likely be more cost effective in
actual operation.
253
-------
Summary--Estimates of probable cost were developed for four altern-
atives to reduce the presence of synthetic organics in potable water by GAC
treatment at CWW as an example of a large facility. These costs were based
on conceptual designs and were derived primarily using the USEPA cost
curves and operating data from this study.
A comparison of the increased cost to the consumer for the four alterna-
tives is shown in Table 79. This comparison indicates that for a given
treatment goal (TOC = 1,000 yug/1), contactors would be more cost effective
than GAC filters. Capital costs of the two alternatives would not be signifi-
cantly different, but operating costs for the GAC filter system would be
substantially higher due to less efficient GAC utilization.
As expected, it costs less to meet a less stringent treatment goal (TOC
(TOC = 1,500 jug/1 vs. TOC = l,000jug/l) due to longer GAC service life.
Treatment goals of TOC - 1,000 >ig/l and TOC = 1,500 jug/1 have no known
significance regarding the presence of certain undesirable compounds or
health risks. Either level of treatment might be acceptable at a given
location.
OTHER OBSERVATIONS
Bacterial, Physical, Chemical Observations
Bacterial--
Standard total coliform and total plate count tests were conducted on
various sample points throughout the three phases of the project.
Total coliform bacteria were never found in any GAC system influent nor
total plant effluent. High total coliform counts, observed in GAC system
effluents during Phase 1, are not believed to be true values. Laboratory
procedures used during this period were found to be incorrect. After proper
procedures were implemented, total coliform counts virtually did not recur
during the remainder of the project in any of the GAC systems.
Standard plate counts (SPC) in the influent were generally less than
TABLE 79. COST OF ALTERNATIVE TREATMENT GOALS
GAC System
GAC Filter
GAC Filter
Contactor
Contactor
EBCT,
min
7.5
7.5
15.2
20.0
Treatment
Goal,
;ug/l TOC
1,500
1,000
1,000
1,000
Water Throughput
Cost,3
<:/l,000 gal
13.5
24.0
16.5
16.0
To determine C/1,000 1, divide by 3.8.
254
-------
100/ml, but much higher counts were found in the GAC system effluents
(Table 80). These SPC results did not follow any observable trends and
support the necessity of post disinfection. As expected, following regener-
ation, the GAC systems were lower in SPCs for a few weeks.
Physical--
Nothing significant was observed from temperature measurements beyond
their expected effect on THM formation.
Compared to chlorinated sand filter effluents, the odor obtained from
GAC system effluents in Phases 1 and 2 was better. However, the odor from
the GAC system effluents was not better than dechlorinated sand filter
effluent samples. In many cases, the odors of both sand and GAC system
effluents were musty. The average threshold odor number for the systems
tested was around one. Since no real odor problems were detected in the GAC
filters of Phase 1 or pilot GAC filters and contactors of Phase 2, this
analysis was suspended.
Turbidity values in all phases were as good or better than those of the
sand filters and well below the established MCL of 1.0 NTU.
Chemical--
The dissolved oxygen concentration, for the most part, from carbon
system effluents followed the influent concentrations. However, the initial
dissolved oxygen concentration for both virgin and regenerated GAC in the
contactors was very low. Generally, the dissolved oxygen concentration in
TABLE 80. STANDARD PLATE COUNTS PER ML FOR PHASE 3-0
GAC Filter Influent
Min. I
Max. 92
Avg. 15
GAC Filter ISA Effluent
Min. 1
Max. 4,700
Avg. 292
Contactor Influent
Min. 1
Max. 52
Avg. 3
Contactor D Effluent
Min. 6.0
Max. 40,800
Avg. 4,840
255
-------
these systems increased over a two-week period until it approximated the
influent concentration.
The average pH of GAC system effluents was slightly lower (pH 8.0) than
those of the corresponding influents (pH 8.5). A similar reduction also
occurred through the sand filters. Therefore, pH had no significant effect
on research results.
The GAC filters removed all but marginal traces of free chlorine. Low
concentrations of chloramines passed through the filters with concentration
of the 45.7 cm (18 in) GAC filter being noticeably higher. The use of GAC
in any plant-wide configuration would require continuous post chlorination.
Several samples were analyzed for pesticides, herbicides and PCBs from
various effluents of the pilot and full-scale GAC systems. Raw river and
corresponding infuent samples were also analyzed. All results were below
the detectable limit of the instrument.
Regenerator Off-Gas Observations
In order to assess the effectiveness of the furnace configuration,
which incorporates an incinerator zone with a secondary burner, additional
analyses were also requested on samples taken from the dryer off-gas loop
(cyclone outlet).
The analytical results of the stack and dryer off-gas sampling are
shown in Table 81. After reviewing these data, the Southwest Ohio Air
Pollution Authority confirmed that the emission quality was well within the
limits established for a process plant.
The analysis of the dryer off-gas loop revealed that insufficient low
temperature volatiles existed to justify this feature which required the
secondary (incinerator) burner. An inordinate fuel demand was required to
maintain process temperatures due to the lack of sufficient combustibles and
the loss of heat resulting from the entrained moisture released in drying
the GAC.
Data shown on Table 81 indicates that most of the fines which pass
through the cyclone were incinerated in the furnace since stack particulate
was much lower. At the same time, it appeared that the 149°C (300°F) dryer
temperature was not high enough to drive off much of the organic loading,
resulting in very little fuel savings attributable to the dryer off-gas
loop.
The Work Environment for Fluidized Bed GAC Regenerator Operators
Some concern was expressed by the operators over the airborne GAC
particles. The National Institute of Occupational Safety and Health (NIOSH)
was contacted to perform a survey of the work environment.
256
-------
TABLE 81. REGENERATOR OFF-GAS ANALYSES RESULTS
Summary of Average Readings
Grams per Hour (Pounds per Hour)
First Sampling - Contactor
Particulate
Filterable
Condensable
Methane Equivalent
Nox
Stack
Particulate
Filterable
Condensable
Methane Equivalent ,
Nitrogen Oxide (Nox)
Second Sampling - GAG Filter
g/hr
4.5
19.5
131.7
(Ib/hr)
(0.01)
(0.043)
(0.29)
163.4 (0.36)
Stack
g/hr (Ib/hr)
4.5 (0.01)
13.6 (0.03)
81.7 (0.18)
202.5 (0.446)
Cyclone Outlet
g/hr (Ib/hr)
449.5 (.99)
35.0 (.077)
0.8 (.0017)
N/A
Cyclone Outlet
g/hr (Ib/hr)
276.9 (0.61)
13.6 (0.03)
13.6 (0.03)
N/A
Total gaseous non-methane organics (TGNMO) is expressed as methane
(CH ) equivalent.
b Avg of 12 readings.
Avg of 11 readings.
Since GAC is not classed as a hazardous or toxic material, it was
considered a nuisance dust in the findings. Under normal operating condi-
tions, the highest level of dust present was 0.29 mg/cu m, considered an
extremely low concentration in view of an allowable 5.0 mg/cu m. Analytical
results from organic vapor samples showed no appreciable amount of contamin-
ation. Additionally, noise levels within the control room were below the
allowable limits set by NIOSH, however, some areas of the building outside
of the control room had noise levels which exceeded the standards for contin-
uous (eight-hour) exposure. In view of the fact that operators spend only
brief periods outside of the control room, NIOSH determined that this was
not a hazard.
Scrubber and Quench Water Analyses
In order to assess the impact of the regeneration process waste water
upon any receiving stream, samples were collected from the quench tank and
off-gas scrubber water discharge lines on May 13 and 15, 1981. Normally,
the source water for the quench tank and the off-gas scrubber was the con-
tactor battery effluent. However, during this period, the source water
apparently included a considerable portion of sand filter effluent (SFEF)
possibly due to a partially-opened valve. This conclusion was supported by
257
-------
TOC, INSTTHM and chlorine residual concentrations which were considerably
higher than those of contactor effluent water for this period.
Tables 82 and 83 present this physical, chemical and some of the organic
data from the samples collected. The tables show a dramatic reduction in
the TOC and the INSTTHM concentrations and a slight reduction in the purgeable
non-halogenated and CAOX concentrations in the quench water relative to the
source water. These reductions were undoubtedly the result of the newly
regenerated GAC adsorbing the organics during the short contact time of the
quenching operation. The off-gas scrubber water showed a dramatic reduction
in INSTTHM concentrations with a slight reduction in the TOC, CAOX and
purgeable non-halogenated concentrations. These reductions, along with a
reduction in dissolved oxygen were likely due to the elevated temperatures
(49.0 and 50.0°C) of the scrubber discharge water. The off-gas analysis
(Table 81) did not contain any measurable methane equivalent which would
tend to indicate that the reduction in INSTTHM in the off-gas scrubber water
does not have a measurable effect on the off-gases. Both the quench and the
TABLE 82.
Parameter
PHYSICAL, CHEMICAL AND ORGANIC DATA FROM
MAY 13, 1981 QUENCH AND SCRUBBER SAMPLES
Temperature, °C
PH
Dissolved Oxygen, mg/1
Turbidity, NTU
Chlorine Residual
Free, mg/1
Total, mg/1
Hardness, mg/1
TOC, ;ug/l
CAOX, ^ug/1
Hexane, ug/1
Benzene, jug/1
Toluene, ,ug/l
Ethylbenzene, ug/1
o-Xylene, ,ug/l
Tetralin, ug/1
Inst THM
CHC1 jug/1
CHBrCl >ig/l
CHBr CI, >ig/l
, Jig/1
Source
Watera
17.0
7.5
9.4
0.2
b
b
143.0
,682
66
c
0.2
c
c
c
c
11.1
5.9
2.0
c
Quench
Tank
21.5
7.9
6.7
4.6
b
b
153.0
925
40
c
0.2
c
c
c
c
8.4
2.1
0.2
c
Off-gas
Scrubber
49.0
5.9
0.9
6.2
b
b
156.0
1,327
b
c
c
c
c
c
c
0.6
0.2
c
c
Mixture of SFEF and combined contactor effluents.
Data not available.
Mot detected.
258
-------
TABLE 83. PHYSICAL, CHEMICAL AND ORGANIC DATA FROM
MAY 15, 1981 QUENCH AND SCRUBBER SAMPLES
Parameter
Temperature, °C
pH
Dissolved Oxygen, mg/1
Turbidity, NTU
Chlorine Residual
Free, mg/1
Total, mg/1
Hardness, mg/1
TOC, jug/1
CAOX, .ug/1
Hexane, jug/1
Benzene, ug/1
Toluene, .ug/1
Ethylbenzene, jug/1
o-Xylene, jag/1
Tetralin, ug/1
Inst THM
CHC1 jug/1
CHBrCl jig/I
CHBr9Cl, jug/1
;, jag/1
Source
Water9
20.5
9.1
9.1
0.2
0.6
0.7
144.0
,746
66
0.2
0.2
b
b
b
b
16.0
11.8
7.3
0.4
Quench
Tank
21.0
7.8
6.2
5.6
0.0
0.0
152.0
773
40
b
b
b
b
b
b
10.1
2.6
0.2
b
Off-gas
Scrubber
50.0
6.6
0.7
11.8
0.0
0.0
155.0
1,445
41
b
b
b
b
b
b
1.3
0.4
0.2
b
Mixture of SFEF and combined contactor effluents.
Not detected.
scrubber water samples contained GAC fines as evidenced by the increase in
turbidity.
The THMMAX samples from May 13th were lost due to the absence of free
chlorine after the seven-day storage period. The May 15th quench and
scrubber samples were also devoid of chlorine residual, but in this case,
the remainder of the sample was dosed with additional chlorine. Both
samples were again devoid of chlorine after another seven days. Once again,
chlorine solution was added, but this time only to the scrubber water sample.
After another seven days storage, the scrubber water sample was devoid of
free chlorine. The quench and scrubber water THMMAX samples were very high
in CHBr3, CHBr2Cl and CHBrCl2 concentrations with the CHBrS concentrations
being the highest. For this reason, the results in Table 84 are presented
in micromoles (umoles) which permit comparison to the source water. The
scrubber water sample after 21 days is close in concentration to the source
water, which would indicate that the contact time with regenerated GAC was
not long enough to remove any of the precursor. The THMMAX concentration in
the quench tank is almost double that of the source water. It appears that
259
-------
TABLE 84. THMMAX DATA3 FROM MAY 13 AND 15, 1981 QUENCH AND SCRUBBER SAMPLES
Source Scrubber Scrubber Scrubber
Water 7 Day 14 Day 21 day
CHC1 1.00 0.02 0.03 0.03
CHBrCl 0.15 0.00 0.01 0.02
CHBr CI 0.04 0.00 0.04 0.10
CHBr^ 0.00 0.00 0.29 1.09
THMMAX 1.19 0.02 0.37 1.24
Source Quench
Water8 7 Day
CHC1 1.00 0.09
CHBrCl 0.15 0.06
CHBr CI 0.04 0.14
CHBr^ 0.00 0.67
THMMAX 1.19 0.96
, Concentrations expressed in/imoles/1.
Mixture of SFEF and combined contactor effluents.
either additional THM precursor was being introduced (perhaps from the ash
which is fractured off the GAC in the quench tank) or that there was a
possible sample error. It is difficult to believe that any THM precursor
could survive the regeneration temperatures. In any event, the THMMAX
concentration observed in the quench and scrubber waters should not present
any difficulty to a receiving stream.
The acid extract GC/FID profiles are presented in Figures 120 and 121.
For the most part, the scrubber and quench water samples showed a reduction
in the number and the magnitude of peaks. The May 15th scrubber sample
(Figure 121) shows the presence of two peaks in the latter third of the
chromatogram. These peaks have been observed in other acid extract GC/FID
profiles so that their presence here was not alarming.
Other than unidentified and unquantified high chlorine demand, the
parameters examined do not indicate the need for special disposal methods.
THMSIMDIST as a Surrogate for Distribution System THM Analyses
As previously described in Section 5, THMSIMDIST analyses of plant
effluent water were performed as a surrogate for actual distribution system
analyses. This afforded the opportunity to measure the effect of various
treatment alternatives at the distal end of the distribution system. In
Phase 1, the maximum travel time from treatment to consumption was estimated
to be seven days. As a result of tracking fluoride through our distribution
260
-------
LU
O
ex.
to
O
LU
O
LU
I
§
CO
••• i
Ji'L
INT.
STD.
!-:r -: JL
j*.,-»jj. ^ it ,,f -
INT.
STD.
ii i
INT
STD
M
CO
INT.
STD.
FIGURE 120. Acid extract profiles for quench and scrubber
samples. May 13, 1981.
261
-------
\kx
d
id
y
•I'-- (tl :•
INT.
STD.
ii.i! , i1
.'Li
o
z
LU
o
J
QQ
I
CO
CQ
'ii'lAi
INT.
STD.
INT,
STD,
INT,
STD,
FIGURE 121. Acid extract profiles for quench and scrubber
samples, May 15, 1981.
262
-------
system in January, 1979, a more precise estimate was made of three days for
this maximum travel time.
Figure 122 compares INSTTHM results from actual distribution system
samples with three-day THMSIMDIST results from plant clearwell samples over
the same time period. The figure demonstrates that three-day THMSIMDIST is
a fairly reliable surrogate for distribution system THM analyses.
263
-------
LEGEND
288
180
160
_ 140
H H ACTUAL SYSTEM CDS8O
SIMULATED SYSTEM CSFEF5 X—X
\
o>
120
Z
H 100
UJ
o
o
o
80
60
40
20
0
JAN
MAR
MAY
JUL
MONTH
SEP
NOV
JAN
Figure 122. THM comparison of actual distribution system (ITTT) vs three-day
simulated distribution system (SITS), 1980.
-------
REFERENCES
1. U.S. Environmental Protection Agency, National Interim Primary Drinking
Water Regulations. Federal Register, Vol. 40, No. 248, December 24,
1975.
2. U. S. Corps of Engineers. Ohio River Navigation, Past, Present, and
Future. Department of Army, October, 1979.
3. Kinman, R. N. and J. Rickabaugh. In-Plant Modifications for Removal
of Trace Organics from Cincinnati Drinking Water. University of
Cincinnati, April 1, 1976.
4. Great Lakes—Upper Mississippi River Board of State Sanitary Engineers.
Recommended Standards for Water Works. Health Education Service,
Albany, N. Y., 1976.
5. U.S. Environmental Protection Agency, Process Design Manual for Carbon
Adsorption. U.S. Environmental Protection Agency Technology Transfer,
October, 1973. EPA 625/l-71-002a.
6. Symons, J. M. Interim Guide for the Control of Chloroform and Other
Trihalomethanes. U.S. Environmental Protection Agency, Cincinnati,
Ohio, 1976.
7. Public Law 93-523. Safe Drinking Water Act. 93rd Congress, S. 433,
December 16, 1974.
8. DC-54 Ultralow Level Total Organic Carbon Analyzer System. Dohrmann,
Division of Envirotech Corp., Santa Clara, Calif. 1st ed. , May, 1977.
9. Stevens, A. A. and J. M. Symons. Measurement of Trihalome thane and
Precursor Concentration Changes. Journal AWWA, October, 1977.
10. The Analysis of Trihalomethanes in Finished Water by the Purge and
Trap Method. U.S. Environmental Protection Agency. Method 601,
Environmental Monitoring and Support Laboratory, Cincinnati, Ohio,
September 9, 1977.
11. Eichelberger, J. W. and L. Littlefield. An Evaluation of the Finnigan
Organics in Water Analyzer, U.S. Environmental Protection Agency, Cin-
cinnati, Ohio.
265
-------
12. Dressman, R. C., B. A. Najor and R. Redzikowski. The Analysis of
Organohalides (OX) in Water as a Group Parameter. Proceedings AWWA
Water Quality Technology Conference, Vol. II, Philadelphia, Pa, 1979.
American Water Works Association, Denver, Colorado, 1980.
13. Grob, K. and G. Grob. Organic Substances in Potable Water and in its
Precursor, Part II; Applications in the Area of Zurich. Journal
Chromatography, 1974.
14. Grob, K. and F. Zurcher. Stripping of Trace Organic Substances from
Water Equipment and Procedure. Journal Chromatography, Vol. 117,
1976.
15. Coleman, W. E., R. G. Melton, R. W. Slater, R. C. Kopfler, S. J. Voto,
W. K. Allen and T. A. Aurand. Determination or Organic Contaminants
by the Grob Closed-Loop-Stripping Technique. Proceedings of AWWA
Water Quality Technology Conference, Vol. 2, Philadelphia, Pa, 1979.
American Water Works Association, Denver, Colorado, 1980.
16. Sampling and Analyses Procedures for Screening of Industrial Effluents
for Priority Pollutants. U.S. Environmental Protection Agency.
17. Methods for Organic Pesticides in Water and Wastewater. U.S. Environ-
mental Protection Agency, 1971.
18. Standard Methods for the Examination of Water and Wastewater. 14th
ed., 1976.
19. Federal Register, Vo. 42, No. 16, August 18, 1977.
20. Brunauer, Emmett, and Teller. Journal American Chemical Society, Vol.
60, 1938.
21. AWWA Standard for Granular Activated Carbon. American Water Works
Association. AWWA B-604-74, 1st ed., January 28, 1974.
22. Modified Phenol Value Method. Westvaco Corp., November 1969.
23. Personal Communication with Mr. Boyd Jenson, Westvaco Corp.
24. Personnel communication with Mr. Rick Weatherly, Westvaco Corp., June,
1980.
25. Personal communication with ICI Americas, Inc., June, 1980.
26. Control of Organic Chemical Contaminants in Drinking Water. Federal
Register, Vol. 43, No. 28, February 9, 1978.
27. Gulp, Wesner and Gulp. Estimating Water Treatment Costs, Vol. 2 - Cost
Curves Applicable to I to 200 mgd Treatment Plants. EPA-60/2-79-162b,
August, 1979.
266
-------
APPENDIX A
COMPUTERIZED DATA DICTIONARY
F
-------
F
-------
F<*> FIELD NflME
CflTEGORIES
ft-4
C-7
C-9
C-EFF
D-4
D-7
D-9
D-li
D-EFF
C-4
15-fl-EFF
flIR
REGOV
SCRUB
QUENCH
SPENT
DISTR81
DISTR82
DISTR03
DISTR84
DISTR65
DISTR86
DISTR87
DISTR88
DISTR89
DISTR10
DISTRil
DISTR12
DISTR13
DISTR14
DISTR15
DISTR16
DISTR17
DISTR18
DISTR19
DISTR28
DISTR21
DISTR22
DISTR23
DISTR24
flBBR
(CONTINUED)
fl4
C7
C3
CE
D4
D7
D9
Dli
DE
C4
15RE
flIR
ROV
SCRB
QNCH
SPNT
DS81
DS02
DS83
DS04
DS05
DS86
DS07
DS88
DS09
DS10
DSii
DS12
DS13
DS14
DS15
DS16
DS17
DS18
DS19
DS20
DS21
DS22
DS23
DS24
VftLUE
41
50
51
52
53
54
55
56
57
59
68
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
30
81
82
83
84
85
36
37
38
39
DESCRIPTION
4 FOOT DEPTH IN 15 FOOT GflC
CONTftCTOR "ft"
7 FOOT DEPTH IN 15 FOOT GflC
CONTftCTOR "C"
9 FOOT DEPTH IN 15 FOOT GflC
CONTftCTOR 'C'
EFFLUENT OF 15 FOOT GfiC CONTflCTOR
•c-
4 FOOT DEPTH IN 15 FOOT GflC
CONTflCTOR "D"
7 FOOT DEPTH IN 15 FOOT GflC
CONTftCTOR "D"
9 FOOT DEPTH IN 15 FOOT GflC
CONTftCTOR "D"
11 FOOT DEPTH IN 15 FOOT GfiC
CONTflCTOR BD"
EFFLUENT OF 15 FOOT GflC CONTflCTOR
"D"
4 FOOT DEPTH IN 15 FOOT GflC
CONTftCTOR "C1
EFFLUENT OF 38 INCH 12X40 MESH GflC
FILTER
flIR REGENERfiTOR OFF GflS
REGENERflTOR OUTLET VftLVE
REGENERflTOR flIR SCRUBBER EFFLUENT
REGENERfiTOR QUENCH TflNK EFFLUENT
REGENERflTOR SPENT TflNK EFFLUENT
PfiRKLfiND 4 TWflIN SflYLOR PflRK
EPft ST. CLfllR
SPRINGFIELD 4 MflRION
EBENEZER 4 CLEYES-WfiRSftH
GEST 4 STflTE
BEECHMONT flT ONTflRIO
QUEEN CITY 4 GRftND
CHERRY GROVE TftNK
RIVERFRONT STfiDIUM
WOOSTER PIKE
GEST ST. SEWftGE
5TH 4 CftNTRfiL
4316 RIVER RCfiD
BRIDGETOWN 4 EBENEZER
MfilN 4 CHURCH ST.
W. 8TH 4 PEDRETTI
KENWOOD 4 HUNT
LUNKEN flIRPCRT
HRRRISQN ftVE.
WINTON 4 COMPTON
ItflDISON 4 BROTHERTON
ICMICKEN 4 VINE
MONTGOMERY 4 MITCHELL FfiRfl
NGRTHWQOD DR. 4 REftDING
269
-------
F«.'#> FIELD NfiflE
CflTEGORIES
DISTR25
DISTR26
DISTR27
DISTR28
DISTR29
DISTR39
OISTR31
DISTR22
DISTR33
OISTR34
DISTR35
DISTR36
DISTR37
OISTR38
OISTR39
DISTR43
OISTR41
OISTR42
DISTR44
DISTR45
DISTR46
DISTR47
DISTR48
DISTR49
DISTR58
DISTR51
15-fHNF
B8-4
BB-7
B8-EFF
DISTR52
DISTR53
DISTR54
DISTR55
DISTR56
DISTR57
15FIE-P
ftBBR
(CONTINUED)
DS25
0526
DS27
DS23
DS29
DS38
DS31
DS32
0533
D534
OS35
OSS6
OS37
0538
0533
0549
0541
0542
0544
0545
OS46
0547
0548
0549
0558
DS51
ISfll
BB4
B87
6BE
OS52
0553
OS54
0555
0556
DS57
ISflP
VflLlE
98
91
92
93
94
95
96
97
98
99
108
101
102
183
184
185
186
187
188
189
118
111
112
113
114
115
116
117
118
119
128
121
122
123
124
125
126
DESCRIPTION
KE31PER (FOREST PfiRK)
PERRY 4 KENWDQD
MfiRKBREIT & EDWfiROS
MONTGOMERY i GflLBRfllTH
ROUTE 8
CLOUGH 4 8 MILE
1643 MHRL014
2729 ERIE
MELISH i REfiOING
fllDUflY 4 GLENWflY
MfllN STflTIOf^
VINE i MfiPLE
VINE 4 SKEEHFiN
TENNYSON STflTION
VICTORY PKUY i TfiFT
WRRISON 4 JOW60N
SPRING GROVE i BfiRNfiRD
LOVELftND ttfiDEIRfl ROW)
2814 VINE ST.
6885 MONTGOMERY
4833 GLENWflY
3668 BANNING
9763 CCLERfllN
8218 PIPPIN
18835 MONTGOMERY
4788 BLUE ROCK
15 fl INF
4 FOOT DEPTH IN 15 FOOT
CQNTflCTOR '88"
7 FOOT DEPTH IN 15 FOOT
CONTflCTOR '88"
EFFLUENT OF 15 FOOT GftC
•DO"
7286 HfiRRISON
6926 Hf^RISON
6597 SPRINGDftl
6958 RIPPLE
285 CflLHXJN
3188 REflDING
GRC
GfiC
CONTfiCTCR
FILTER 15fl-£FF C«80N S«1PLE
REPRESENTflTIVE OF CflRBCN
CONDITION PRIOR TO BEING PLflCED
INTO SERVICE
15fiE-S 15RS 127 FILTER 15fi-£FF CflRBON SfiMPLE
REPRESENTflTIVE OF CflRBON
CONDITION flFTER BEING TflKEN OUT
OF SERVICE (SPENT).
15RE-R 15fiR 123 FILTER 15fl-EFF CflRBON SflMPLE
REPRESQffflTIVE OF CflRBON
CONDITION HFTER REGBOfiTION
21RE-P 21flP 129 FILTER 21fl-EFF CflRBCN SfiHPLE
270
-------
F(») FIELD NftHE fiBBR VftLUE DESCRIPTION
CATEGORIES (CONTINUED)
2if£-S 21flS 138
2UE-R 21ft? 01
fl-£FF-P flP 132
fl-EF-S RS 133
fl-EFF-R ftR 134
BB-EF-P B8P 135
B8-EFF-S BBS 136
BB-EF-R BBR 13?
C-EF-P CP 138
C-EFF-S CS 139
C-EFF-fi CR 148
D-EFF-P DP 141
D-EF-S DS 142
REPRESENTflTIVE OF CfiRBCN
CONDITION PRIOR TO BEING PLftCED
INTO SERVICE.
FILTER 2ifl-£F CflRBON SfiHPLE
REPRESENTflTIVE OF CflRBON
CONDITION flFTER KING TflKEN OUT
OF SERVICE (SPENT).
FILTER 21fl-£FF CfiRBON SfiMPLE
REPRESENTflTIVE OF CflRBON
CONDITION fiFTER REGENERflTION.
CONTflCTOR fKFF CfiftBON SfiMPLE
REPRESENTflTIVE OF CflRBON
CONDITION PRIOR TO BEING PLflCED
INTO SERVICE.
CONTflCTOR fl-EFF CflRBON SfiMPLE
REPRESENTflTIVE OF CflRBON
CONDITION flFTER BIENG TftKEN OUT
OF SERVICE (SPENT).
CONTflCTOR fl-EFF CflRBON SfiMPLE
REPRESENTfiTIVE OF CflRBON
CONDITION flFTER REGENERATION.
CONTflCTOR BB-EF CflRBON SflMPLE
REPRESENWIVE OF CflRBON
CONDITION PRIOR TO BEING PLflCED
INTO SERVICE.
CONTftCTOR BB-EFF CflRBON SftMPLE
REPRESENTflTIVE OF CfiRBON
CONDITION flFTER BEING TfiKEN OUT
OF SERVICE (SPENT).
CONTflCTOR BB-EF CftRBON SflMPLE
REPRESENTflTIVE OF CflRBON
CONDITION fiFTER REGENERflTION.
CONTfiCTOR C-EF CfiRBON SflMPLE
REPRESENTflTIVE OF CftRBON
CONDITION PRIOR TO BEING PLflCED
INTO SERVICE.
CONTflCTOR C-EF CflRBON SfiMPLE
REPRESENTflTIVE OF CflRBON
CONDITION ftFTER BEING TflKEN OUT
OF SERVICE (SPENT).
CONTflCTOR C-EF CflRBON SfiMPLE
REPRESENTflTIVE OF CftRBON
CONDITION ftFTER REGENERflTION.
CONTflCTOR D-EF CftRBON SfiMPLE
REPRESENTflTIVE OF CflRBON
CONDITION PRIOR TO BEING PLflCED
INTO SERVICE.
CONTflCTOR D-EF CflRBON SftMPLE
REPRESENTfiTIVE OF CflRBON
CONDITION fiFTER BEING TflKEN OUT
OF SERVICE (SPENT).
271
-------
F<*> FIELD NfiME
HB8R
VfLUE DESCRIPTION
CATEGORIES (CONTINUED)
D-EFF-R DR 143 CONTRCTOR D-EFF CfiRBON SflMPLE
REPRESENTflTIVE OF CfiRBGN
CONDITION flFTER REGENERflTlON.
CONEFF CNEF 144 THE COMBINED EFFLUENT FROtt THE
CflRBON CONTRCTORS WHICH fiRE
ON-LINE fiT THE TIME OF THE SRMPLE
COLLECTION
OISTR58 DS58 145 LIBERTY flND LINN
DISTR53 DS59 146 SflLEN flND SUTTON
DISTR68 DS68 147 LINWOOD flNO OaTfl
CCNTFILT COFL 148 MIXTURE OF SflN FILTER EFFLUENT fiND
VWIW) CONTftCTOR EFFLUENTS FROM
THE CONTfiCTORS WICH HERE ON-LIfE
flT THE TIME OF SfiMPLE COLLECTION.
149 BftSE (CUTRflL fETHOO BLfiNK.
158 BffiE NEUTRflL SOLVENT BLflfttC.
RECWIRED PfiRfiflETER REPORTED
-1 UNKNOW
1
-------
F(#) FIELD NftHE
CflTEGORIES
CLFREEft
CLTOTfiLfl
SPC
PH
PURGX
TX
TEMPSfiMP
TOT. COLI
TOTftLSOL
TFILSOL
TURB
NPX
D02
TC7TTHH
TC7CL3F
TC7BR3F
TC7BR1F
fi68R '
, PH=9. 5
JJG/L CHLOROFORN TERMINflL REflDING,
STORED 7 DfiVS, FREE CHLORINE
TE«P=85F(28. 5C), PH=9.5
JJG/L BROflOFORM TERMINflL REftDING,
STORED 7 DfiYS, FREE CHLORINE
TEMP=85F<28. 5C>, PH=9. 5
,UG/L BROtlCDICHLOROtCTHflNE
TERMINflL REflDING, STORED 7 DfiVS,
FREE CHLORINE TEff =85F(2a 5C),
TC7BR2F
TC7CCL4F
TC7CL3EF
TCTDCMfiF
TC7a4E
TC7TRC.E
MC82
ICL4
MDIC
ttC4E
«ET3
26
27
25
3d
31
32
JJGA, CHLORODIBROMOMETHflNE
TERMINflL REflDING, STORED 7 DfiVS,
FREE CHLORINE TEMP=S5F<28. 5C>,
PH=9 5
JX3/L CRRBOM TETRflCHLORIDE
TERMINflL REftDING, STORED 7 DflVS,
FREE CHLORINE TEMP=85F<28. 5C),
PH=5.5
jfi/L TRICHLOROETHflNE-LLi
TERMINfiL REftDING, STORED 7 DfiVS,
FREE CHLORINE TEKP-85FC29-38C),
JJG/L DICHLOROMETHflNE TERMINfiL
REflDING, STORED 7 DftYS, FREE
CHLORINE TEHP=85F<29-38C), PH=5. 5
JJG/L TETRftCHLOROETHftNE-Li,i2
TERNINflL REftDING, STORED 7 DfiVS,
CHLORINE TENP=85F(29-38C>, PH=9.5
UG/L TRICHLOROETHEHE TERfllNflL
273
-------
F, PH=9. 5
36 JJG/L TRIHft.OI€THflNL TERHINflL
REfiDING, STORED 3 DftVS, FREE
CHLORINE, TEWMWBIENL PH=8. 2,
flLL PW&S BUT i-fl, 2-a, 2-1 fiND
2-2
37 JJG/L CHLOROFORH, TERMINflL
R£ft)IN 2-2
41 UG^L CftRBOM TETRflCHLCRIDE,
TERMINfiL REfiDING, STORED 7 DftVS,
FREE CHLORINE, TEMP=flM8IENT,
PH=flnBIENT, PHfiSES 1-9 fiND 2-9
ONLV
43 UG^L TRICHLOROETHfiNE-LLl
TERMINflL REftDING STORED 7 DflVS,
FREE CHLORINE, TEWP=flMBIENT,
PH=ftHBIENL PHftSES 1-9 fiND 2-9
ONLV
44 jUG/L DICHLOROMETHfiNE TERMINFL
REflDING, STORED 7 DftVS, FREE
CHLORINE, TEMP=fi«8IENT,
PH=fl«BIENT, PHftSES 1-9 ftHD 2-9
ONLV
45 UG/L TETRflCHLOROETHftNE-l,L2,2
' TERMINflL REftOING, STORED 7 DftVS,
FREE CHLORINE T£MP=fiMBIENT,
274
-------
F<*> FIELD NfiME
VflLUE DESCRIPTION
CflTEGORIES (CONTINUED)
TR7TRCEF SET3 46
TftTODCSF SBZ2 47
TR7CLB2F SBNZ 49
FPTTTHfF FTTT 58
FP7CUF FCLR 51
FP7BR3F FCH3 52
FP7BR1F FCL2 53
FP7BR2F FC82 54
FP7CCL4F FCL4 55
FP7CL3EF F3C3 57
FPTDCMflF FDIC 58
FP7CL4EF FC4E 59
PTRCEEN FET3 68
PH=fiM8IENT, PHfiSES 1-9 ftND 2-9
ONLV
JJG/L TRICHLOROETHENE TERMINflL
RERDING, STORED 7 DftVS, FREE
CHLORINE, TENP=flM8IENL
PH=fi«BIENT, PHfiSES 1-9 RND 2-9
ONLV
UG/L DICHLOROBEfCENE-ORTHO
TERMINAL REflDINQ, STORED 7 DflVS,
FREE CilORIffc TEfP=fiHBIBff,
PH=fiMBIENL PHflSES 1-9 fiND 2-8
ONLV
^G/L CHLOROBEHZENE WtHlNPL
REfiDING, STORED 7 DfiVS, FREE
CHLORINE, TBP=fflBIENT,
PH=flMBIENT, PHfiSES 1-9 fiND 2-9
ONLV
JUQ/L TRIHfLOMflTHfiNE FORMflTION
POTEMTIfiL STORED 7 DftVS, FREE
CHLORINE TEfP=85F<29-38C), PH=9. 5
JJG/L CHLOROFORM FORMflTION
POTENTIflL STORED 7 DftVS, FREE
CHLORIfE TEfff>=85F(29-28C), PH=9. 5
jUO'L, BROMOFORH FORMfiTIOH
POTENTIfiL STORED 7 DflVS, FREE
CHLORINE, TEMP=S5F(23-26C>, PH=9. 5
UGA, BRCMODICHLOROI^THnrt
FCRflfiTION POTEHTIBL STORED 7
DftVS, FREE CHLORINE,
TEHP=85F(29-38C>, PH=9. 5
JJGA^ OtORODIBROMOflETHflNE
FORMflTION POTENTIflL STORED 7
DfiVS, FREE CHLORIC '
TEMP=85F(29-28C), PH=9. 5
JJG/L CfiRBOH TETRflCHLORIDE
FORMflTION POTENTIfiL STORED 7
DftVS, FREE CHLORINE,
TEMP=65F<29-38C), PH=9. 5
UG/L TRICHLOROETHflNE-Li,!
' FORMflTION POTENTIflL STORED 7
DfiVS, FREE CHLORINE,
TEMP=85F<29-26C), PH=9. 5
JJG/t, DICHLOROMETHfiNE FORflflTION
POTENTIfiL STORED 7 DfiVS, FREE
CHLORINE, TEMP=S5F<29-38C), PH=9. 5
JUG/L TETRfiCHLOROETHfiNE-l,L2,2
FORMflTION POTENTIfiL STORED 7
DftYS, FREE CHLORINE,
TEMP=85F(29-29C), PH=9. 5
UG/L TRICHLOROETMENE FORMflTION
' POTENTIfiL STORED 7 DfiVS, FREE
275
-------
F(t) FIELD WE
VfiLUE DESCRIPTION
CflTEGORIES (CONTINUED)
FP700C8F FBZ2
FP7CLBNZ FBNZ
TTHM
ODICLBN2
CLRBNZ
SUPROTG
DELFLOTG
BED. VOL
CfiRBWTPD
TR3PH
TC7PH
CLSDF
FP3TTH1F
ITTT
CHLROFRH
8ROHOFRN
BRDCUCT
CLDBRMET
CflRBTET
DCETfiN
TRCETRN
ICLR
IOC
ICL2
IC82
ICL4
ICC2
I3C3
DCUIETfiN IOIC
TCLRETfiN IC4E
TRCETEN IET3
IB22
IBN2
SFLO
DFLO
BVOL
C
SPH
FPPH
SOF
FCLF
QDDC
61
63
68
63
70
71
72
73
74
75
76
77
78
79
81
91
92
93
94
82
33
84
35
35
CHLORINE TeiP=85F<23-38C>, PH=S.5
^UG/L DICHLOROBEHZENE-ORTHO
FOMWTION POTENTIflL STORED 7
DfiYS, FREE CHORINE,
TEW=83:<23-18C)I1 PH=9. 5
JJG/L CHORCiBEHZEtE FORilfiTION
POTENTIfiL STORED 7 DflVS, FREE
CHQRIffc TE11P=85F<23-38C). PH=9. 5
.UG/'L TRIHfiLOMETHflNE INSTfiNTflNEOUS
REflDINQ, SUM OF FOUR
TRIHflLOIO*»€ REflDINGS
UG/L CHLOROFORM INSTfiNTfiNEGUS
'RERDING
jXi/L BROflOFORH INSTflMTWCOUS
REfOING
^VL BROnODICW-OROPtETHflNE
INSTflNTfiNEOUS REH)ING
/JJG/L CHOROOIBRCWDMETHflNE
JJG/L CflRBOM TETRftCHORIDE
INSTflNTfiNEOUS REfiDING
UG/L DICHLOROETHflNE-L2
INSTfiNTfiNEOUS REflDING
JJG/L TRICH.OROET1WC-LL1 G/L DICHLOR08ENZENE-CRTHO
INSTfiNTfiNEOUS REfiDING
UG/L CH.OROBENZENE INSTfiNTflNEOUS
' REflDING
SUtWTION FLOW IN 1689 US QflLLONS
DELTft FLCM IN 1888 US GALLONS,
FILTER BED VOLUCE IN US GfiLLONS
CfiRBON 1€IGHT IN THE FILTER, POUNDS
PH OF SflMPLE STORED FOR THM
ff^LVSIS, STORED 3 DfiVS
PH OR SflflPLE STORED FOR TKil
RNfiLVSIS, 'STORED 7 DflVS
CH.ORIHE-FREE, FROM SIMULflTED
DISTRIBUTION SYSTEM
JUG/L TRIHRLOfETVifiNE, FORilfiTION
POTEHTIfiL STORED 3 Dfi«, FREE
CHLORINE, TE>P*85F<29-38C>i PH=9.5
TOH-UNITS-THRESHCLD ODOR
276
-------
F
-------
z
o
i<:
X
iSiil^
liJ O ® CD O 'j
t
;3
t fe ^ £
R^^ -,'Mffi
ffi
rO
"* §
"!*
C.J
s
a. ui -rt ix
CD fe o &
Qi
gs^ §s^
S S *~
s
,&.
ffi
I &si ?
°°2:§§S£§
fl i^j «S IjJ
^S CO C5 Ct G* -^H O
•j_ ui. r— «— '_»_
|Ees&|
3 2
S
&**.
•«•
a
l?iS@2
s fc^fj
Q H w T
S!
A 5
T* Lu
ffi
C-d
if I
3
I ED S
^l01^^
'fe
Q S _
00 ?££ m
1$wS
f
pi^pi^li^p^i
a
s s s
s
00
r-
oo
z oi
-------
F<*> FIELD NftHE
flBBR
VfiLUE DESCRIPTION
CflTEGORIES (CONTINUED)
TflTBRIF SB17
Tft7BR2F
DECftNflL
UNDECfitt.
METYNflPH
12DPVHVD
fiNflPTHYL
flPICLINE
BNEUNK
SB27
TflTPH
TRMETYPN
DMETBZ34
DKETBZ2
PRPVLBN2
METVEE4
METVB2E2
TRMETBNZ
14DICLS2
12DICL8Z
HCLRETflN
NONflNflL
THETYBNZ
SPH7
TRHP
DM34
NG2
PP82
MBE4
HBE2
T»B
DC84
DC82
HOE
NNftL
TM82
143
144
145
146
147
148
149
158
151
152
153
154
155
TRCLRBN2 TRC8
DECL
UDCL
HTNP
DPHV
fiNflP
flPIC
KM
1-8,2-9, 2-L fiND 2-2
141 UG/L SROMOOICHLOROMETHflNE,
' TERMINflL REflDING, STORED 7 DfiYS,
FREE CHORINE, TEHP=fiflBIENT,
PH=fflfiIENT.. fWSES 1-9, 2-6, 2-1
flND 2-2
142 JX5/L OtORODIBROnonETHfiNL
TERHINflL REfiDING, STORED 7 DflYS,
FREE CHLORINE, TEMP=flflBIENT,
PH=flH8IENT.. PHASES i-8, 2-9, 2-1
flND 2-2
PH OF SflMPLE STORED FOR THM
flNfiLVSIS, STORED 7 DftVS
NG/L TRIfCTHVL - 1 - PENTENE -
2,2,4 INSTFtffTflNEOUS REflDING
NG/L DIMETH-TtKNZENE - L3 fiND 1,4
INSTflNTflNEOUS REflDING
NG/L DIMETHVUeGENE - L 2
INSTflNTflNEOUS REflDING
NG/L PROPYLBENZENE INSTflNTfllCOUS
REflDING
NG/L fETHVLBENZENE - 1 - ETHVL - 4
INSTflNTflNEfX/S REfiDING
f«VL fETHVLBENZENE - 1 - ETHYL - 2
INSTflNTfiNEGUS REflDING
NG/L TRIMETHYLBENZENE 1,2,4
INSTflNTflNEiDUS REflDING —
NG/L DICHOROBENZENE - L4
INSTflNTflNEOUS REfiDING
NG/L DICHOROSENZENE - 1,2
INSTflNTflNEOUS REflDING
NG/L HEXflCHLOROETHflNE .
INSTflNTflNEiJUS REflDING
NGA NONflNflL INSTflNTfi>£OLlS REflDING
NG/L TETRftHETHYLBENZENE - 1,2,3,5
INSTflNTflNEOUS REflDING
156 NG/L TRICHLGR08ENZENE
INSTflNTflNEiJUS REflDING
157 NG/L DECfiNflL INSTfiNTfiNEOUS REflDING
158 NG/L UNDECflNftL INSTfiNTflNEOUS
REflDItW
159 NG^. M£THYU-WPHTflLE>JE
INSTfiNTRNEOUS REflDING
168 NG/L DIPtefTlHYDRfiZINE
INSTfiNTfiNEOUS REftDING
161 /UG^L, NfiPHThYLflMINE-flLPHfl,
INSTflNTflNEiDUS REflDING.
162 JJG/L PICOLINE-fiLPHft,
INSTfiNTflNBDUS REflDINQ
163 JJG/b KSE NEUTRflL EXTRfiCTfiBLE
' HNflLYSIS, COMPOUND
279
-------
F<« FIELD ME
VflLUE DESCRIPTION
F<4) REflDING ftHT
CATEGORIES
+ +
WWW UNK
F<5) NOTE N
CATEGORIES
NONE NONE
188 188
288 288
288 288
488 488
568 588
608 688
788 788
358 958
238 228
231 231
232 232
REfiDING
-1 NOT ENTERED
-i NOT ENTERED
STftNDftRD NOTE CODE
-1 NONE
108 DUPUCflTE fiNftLYSIS (TWO SUCCESSIVE
RNfiLVSIS OF SflflPLE ON S»€ DfiV)
200 REPLICflTE flNfiLVSIS (fiNflLVSIS OF
SflHPLE ON TWO DIFFERENT DflVS)
388 SflflPLE OR fiNfiLYSIS ERROR, DflTfi MflV
NOT BE VfiLID. IF NO CONCENTRTION
OBTfllNED EhffER 99S?99 IN REfiDIfW
FIELD.
488 PROBfiBLE BOLTON PLfiNT WfiTER
588 CONTROL SflMPLE (TWO SEPflRftTE
FORMfiTION POTENTIflLS SET UP 4
flNflLVZED ON SfiME DflV)
688 SWLE DECHLORIHflTED BEFORE ODOR
ftNflLVSIS
788 VfiLUE RECORDED IS fl NEGATIVE
WLUE-SHOIJLD BE PRECEDED BV MI/JUS
SIGN
888 SflMPLE NOT TflKEN
988 ftRTIFflCT
950 1 LITER PURGE
238 THIS DflTUM WfiS DEVELOPED BV
OUTSIDE QMWX LfiBORfiTORIES.
1>£ INCONSIS7ENCV OF RESULTS
TENDS TO INDICfiTE TWT CflUTION
SHOULD BE USED IN INTERPRETflTION,
231 THIS DIFLICftTE DflTUH WfiS DEVELOPED
BV OUTSIDE CONTRfiCT LflBORflTORIES.
THE INCOHSISTENCV OF RESULTS
TENDS TO INDICflTE THfiT CaiTION
SHOULD BE USED IN INTERFRETflTION.
332 THIS REPLICflTE DflTUM WflS DEVELOPED
BY OUTSIDE CONTRftCT LflBORflTORIES.
280
-------
F<*> FIELD NflME
fiBBR
VfiLUE DESCRIPTION
CATEGORIES (CONTINUED)
F(e) NOTE2 N2
CATEGORIES (ONLV)
NONE NONE
661 961
603
864
005
862
063
064
066
-1
•<
867
8
036
931
832
833
834
835
836
037
038
633
840
041
042
058
851
053
052
060
087
0
036
031
832
833
834
035
036
037
838
839
046
041
042
050
951
053
852
066
7
-1
38
31
32
33
34
35
36
37
38
39
40
41
42
56
51
53
52
66
055
055
55
THE INCONSISTENCY OF RESULTS
TENDS TO INDICflTE THfiT CfiUTION
SHOULD BE USED IN INTERPRETfiTION.
MOTE2
NONE
PRESENT BELOW 0.1 PPB=NR ENTER
8. 8 IN REflDING FIELD.
NOT DETECTED. ENTER 0. 8 IN REfiDING
FIELD.
PRESENT IN THE RfiNGE OF 6.1 TO 0. 4
PPB. ENTER 0.2 IN REflDING FIELD.
LESS THflN 1 «1J. INTENDED FOR
BfiCTERIfl ONLV. ENTER 8. 8 IN
REfiDING FIELD.
HIGH BflCKGRGUND (GftEfiTER THAN 200
COLONIES). INTENDED FOR TOTfl
COLIFORM ONLY. ENTER 300 IN NOTE
FIELD flND COUNT IN REfiDING FIELD.
LESS THflN 1 «i> TOTflL COLIFORfl
HIGH BfiCKGRCUND (GREflTER THAN 208
COLONIES). ENTER 380 IN NOTE
FIELD flND 399999 IN REfiDING FIELD.
YflLUE RECORDED IS fl NE&flTIVE VftLUE
NONE
BRLSflMIC (FLOWERY) SWEETISH
CHEMICft
CHEMICa HYDROCfiRBON
CHEMICflL CHLDRINOUS
flROfflTIC (SPICEY)
DISfiGREEfiBLE
DISfiGREEfiBLE (FISHY)
DISfiGREEfiBLE (SEPTIC)
EfiRTHY
QRftSSY
MUSTY
MUSTY, MOLDY
VEGETfleLE
BftSED ON flCTUfiL DELIVERED WEIGHTS,
NOT ON VOLUME X fiPPflRENT DENSITY
BflSED ON VOLUME X ftPPflRENT DENSITY
OF 36.5 LSS. /CU. FT.
FLOWS BflSED ON INTERPOLflTICNS CF
METERED FLOWS
FL016 BflSED ON flVERFGE OF 5 MGD
PHftSE 1-0 FILTERS WERE DCWN FOR
FOUR DftYS. DflTE IN THIS i
-------
F(i) FIELD W£
CflTEGORIES
870
888
F<7) UNK. RET
F<8) PHftSE
CflTEGORIES
TONE
1-8
1-1
2-0
2-i
2-2
3-0
3-1
3-2
3-3
3-4
3-5
FO) DflTE
fiBBR
(CONTINUED)
79
88
U.R
RUN
(ONLY)
NONE
1-8
1-1
2-9
2-1
2-2
3-8
3-1
3-2
3-3
3-4
3-5
DTE
vauE
79
88
-i
1
2
3
4
5
6
7
8
9
10
11
DESCRIPTION
ENGINEERING EXPERIMENTfiL STflTIOR
PRESENT BELOW 3 PPB. BITER 1 5 IN
REflDING FIELD.
COMPOUND NOT IDENTIFIED.
CONCENTRfiTION NOT OBTfilNfiBLE.
ENTER 399959 IN REflDING FIELD.
UNK. RET
PHfiSE OR RUN
NO FWSE
PHftSE i, RUN 0. GflC SYSTEM
CONTfllNS VIRGIN GflC
PHflSE L RUN 1 GfiC SYSTEM
CONTfllNS ONCE REGENERfiTED GflC
PHfiSE 2* RUN 8. GflC SYSTEM
CONTfllNS VIRGIN GflC
PHfiSE 2, RUN i. GflC SYSTEM
CONTfllNS ONCE REGENERATED GflC
PHftSE 2, RUN 2. GfiC SYSTEM
CONTfllNS TklCE REGENERfiTED GflC
PHftSE 1, RUN 3. GfiC SYSTEM
CONTfllNS VIRGIN GflC
PHflSE i RUN 1 GftC SYSTEM
CONTfilNS ONCE REGENERflTED GflC
PHfiSE 3, RUN 2 GflC SYSTEM
CONTfilNS TWICE REGENERflTED GflC
PHftSE l> RUN 3. GflC SYSTEM
CQNTQINS THREE-TIt€S REGENERflTED
GflC
PHflSE 3, RUN 4. GfiC SYSTEM
CONTfllNS FOUR-TIMES REG£f£RflT£D
GflC
PHftSE 3. RUN 5. GflC SYSTEM
CONTfllNS FIVE-TIMES REGENERflTED
GfiC
DflTE OF SftMPLE
282
-------
|