Granular Activated Carbon Installations
Culp/Wesner/Culp
Cameron Park, CA
Prepared for
Municipal Environmental Research Lab
Cincinnati, OH
Sep 81
U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
-------
EPA-600/2-81-177
September 1981
PBS2-1024'??
GRANULAR ACTIVATED CARBON INSTALLATIONS
by
Russell L. Gulp
and
Justine A. Faisst
Culp/Wesner/Culp
Clean Water Consultants
Cameron Park, California
and
U.S. EPA-NEIC LBRARY
E. smith Denver Federal Center
Rubel and Hager, Inc. Building 25, Ent E-3
Tucson, Arizona P.O. BOX 25227
Denver, CO 80225-0227
Contract No. CI-76-0288
Project Officer
Robert M. Clark
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. DEPARTMENT OF COMMERCE
SPRINGFIELD, VA 22161
-------
TECHNICAL REPORT DATA
(Please read Inaructioru on the reverse before completing)
1. REPORT NO.
-Bi- 177
ORD Report
3. RECIPIENT'S ACCESSION NO.
PBB2 1 0249 2
4. TITLE AND SUBTITLE
Granular Activated Carbon Installations
5. REPORT DATE
September 1981
6. PERFORMING ORGANIZATION CODE
7. AUTHORIS)
Russell L. Gulp, Justine A. Faisst, and
Clinton E. Smith
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Culp/Wesner/Culp, Consulting Engineers
Cameron Park, California
10. PROGRAM ELEMENT NO.
PE# BNC1A
11. CONTRACT/GRANT NO.
CI-76-0288
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 COVERE D
Final and Interim
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
Project Officer - Robert M. Clark. (513) 684-7488
16. ABSTRACT
This report complies and summarizes design criteria, performance data, and
cost information from twenty-two operating granular activated carbon (GAC)
installations. These plants produce municipal water for drinking, treat munici-
"ol and industrial wastewaters, and process food and beverage products. It pro-
vides guidance in estimating the cost of GAC treatment for public water supplies.
The manual is intended for use in connection with a previous series of EPA reports
on "Estimating Water Treatment Costs" to obtain project-specific cost estimates.
This manual is not a design manual. Rather, it describes how to obtain design
criteria for water systems.
This report was submitted in fulfillment of Contract No. CI-76-0288 by
Culp/Wesner/Culp under the sponsorship of the U.S. Environmental Protection
Agency, and work was completed as a January, 1981.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTOR
b.IDENTIFIERS/OPEN ENDED TERMS c. COS AT I Field/Group
Economic Analysis;. Environmental
Engineering; Operating Costs; Water
Treatment; Water Supply; Cost Estimates
Safe Drinking Water
Act; Unit Processes;
Treatment Efficiency;
Granular Activated
Carbon
13B
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (Thit Report)
Unclassified
20. SECURITY CLASS (Thilpagfj
Unclassified •
22. PRICE
EPA Form 2220.1 (Rev. 4-77)
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental Research Labora-
tory, U.S. Environmental Protection Agency, and approved for publication.
Approval does not signify that the contents necessarily reflect the views and
policies of the U.S. Environmental Protection Agency, nor does mention of trade
names or commercial products constitute endorsement or recommendation for use.
11
-------
FOREWORD
The U.S. Environmental Protection Agency was created because of increasing
public and government concern about the dangers of pollution to the health and
welfare of the American people. Noxious air, foul water, and spoiled land are
tragic testimonies to the deterioration of our natural environment. The com-
plexity of that environment and the interplay of its components require a con-
centrated 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 solu-
tions. 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
communications link between the researcher and the user community.
This report presents data on the cost and operation of existing granular
activated carbon installations for water and wastewater treatment which may be
useful -in the planning of future similar projects.
Francis T. Mayo, Director
Municipal Environmental Research
Laboratory
111
-------
PREFACE
From 1977 to 1979, the U.S. Environmental Protection Agency's, Municipal
Environmental Research Laboratory contracted with Culp/Wesner/Culp, Clean Water
Consultants (CWC) for the preparation of a series of reports estimating the
costs for municipal water treatment. During the course of this work, the costs
of granular activated carbon (GAC) adsorption and reactivation in municipal
water treatment as they relate to the removal of organics from drinking water
became a subject of great national interest. Because of this, in 1978, the
original project was expanded to include a special study of the unit process
costs of GAC adsorption and reactivation in potable water treatment. This
special study was directed at visiting as many existing GAC installations as
possible to gather and publish'data on actual operating experiences, partic-
ularly on the costs of building, operating, and maintaining GAC plants. This
report presents the.findings of this special study of GAC installations and the
com'dilation of the information available on the use of GAC in water treatment.
IV
-------
ABSTRACT
This report compiles and summarizes design criteria, performance data, and
cost information from twenty-two operating granular activated carbon (GAC)
installations. These plants produce municipal water for drinking, treat munici-
pal and industrial wastewaters, and process food and beverage products. It pro-
vides guidance in estimating the cost of GAC treatment for public water sup-
plies. The manual is intended for use in connection with a previous series of
EPA reports on "Estimating Water Treatment Costs" to obtain project-specific
cost estimates. This manual is not a design manual. Rather, it describes how to
obtain design criteria for water systems.
This report was submitted in fulfillment of Contract No. CI-76-0288 by
Culp/Wesner/Culp under the sponsorship of the U.S. Environmental Protection
Agency, and work was completed as of January, 1981.
-------
CONTENTS
Disclaimer ii
Foreword iii
Preface .•.. . iv
Abstract v
Figures , .. •»..«. x
Tables xi
Abbreviations •,. xii
Conversion to Metric , xiv
Acknowledgement. xv
1. Introduction 1
2. Historical Use of Activated Caxbon in Water Purification 27
Powdered Carbon 27
Granular Carbon ..-, 27
3. Use of GAC in Food and Drink Processing and in Other Industries 29
4. Extrapolating GAC Use in Municipal Wastewater Reclamation to
Water Treatment • 31
South Lake Tahoe 31
Other Wastewater Installations 32
Operational.Problems • 32
Extending Wastewater Treatment Experience to Water
Treatment Systems 33
5. Designing GAC Adsorption & Reactivation Systems for
Water Treatment . • '•• 36
GAC System Components 36
Pilot Plant Tests 36
Design of Pilot GAC Columns 36
Similarities and Differences in the Use of GAC in
Water and Wastewater Treatment 37
Frequency of Reactivation 37
GAC Contactors 37
Carbon Contactor Underdrains 39
GAC Reactivation or Replacement 39
Thermal Reactivation Equipment 39
Required Furnace Capacity 40
Carbon Transport and GAC Process Auxiliaries 40
6. The Cost of GAC in Water Treatment 41
Developing Cost Curves 41
Design Cost Information 41
Derivation of Cost Curves 41
Example Cost Curves 42
Updating Cost Curves 46
GAC Use in Water Treatment 48
Sensitivity Analysis 49
Operating and Maintenance 56
Preceding page blank
vii
-------
TABLE OF CONTENTS (Continued)
Capital Cost Effect 56
O&M and Capital Cost Effect 56
Key Variables 56
Summary and Conclusions 69
7. Report Summary 70
Conclusions • 70
Bibliography 72
Vlll
-------
Appendix
Case Histories of GAC Installations in the United
No. Case History
1. South Tahoe (CA) Public Utility District
2. Tahoe-Truckee (CA) Sanitation Agency
3. Upper Occoquan (VA) Sewer Authority
4. American Cyanamid Corp., Boundbrook (NJ)
5. Vallejo (CA) Sewage Treatment Plant
6. Orange County (CA) Water District
7. Niagara Falls (NY)
8. Fitchburg (MA)
9. Arco Refinery, Carson (CA)
10. Rhone Poulenc, Portland (OR)
11. Reichhold Chemical, Tuscaloosa (.AL)
12. Stephan Chemical Co., Cordentown (NJ)
13. Republic Steel Corp., Cleveland (OH)
14. Village of LeRoy (NY)
15. City of Manchester (NH)
16. Passaic Valley Water Commission, Little
Falls Filtration Plant (NJ)
17. City of Colorado Springs (CO)
18. Hercules Inc., Hattiesburg (MI)
19. Industrial Sugar, St. Louis (MO)
20. Hopewell (VA)
21. Davenport (IA)
22. Spreckles Sugar, Woodland (CA)
.81
States
Installation, Type
Municipal Wastewater.... 1.81
Municipal Wastewater.......95
Municipal Wastewater..:.,. .104
Municipal Wastewater ...... 113
Municipal Wastewater .. „ .. .123
Municipal Wastewater.... .136
Municipal & Industrial...144
Wastewater
Municipal & Industrial...150
Wastewater
Refinery Wastewater 159
Industrial Wastewater....172
Industrial Wastewater....182
Industrial Wastewater....193
Industrial Wastewater....198
Municipal Physical- 207
Chemical Wastewater
Public Water Supply 212
Public Water Supply 221
Municipal Wastewater 230
Industrial Wastewater.... 242
Food Processing 251
Public Water Supply 258
Public Water Supply 267
Food Processing 274
IX
-------
FIGURES
Page
Construction cost for pressure carbon contactors 43
2 Typical activated carbon column installation 44
3 Operation and maintenance requirements for pressure carbon
contactors - labor and total cost 45
4 Total production cost versus plant capacity for post-filter
adsorption and sand replacement systems 52
5 Capital cost versus plant capacity for post-filter adsorption
and sand replacement systems « 53
6 O&M cost versus plant capacity for post-filter adsorber and
sand replacement systems 54
7 Total production cost versus reactivation frequency in months
for post-filter adsorption and sand replacement systems 55
(100 mgd)
8 O&M cost versus wage rate for 1, 5, 10, 100, and 150 mgd sand
replacement systems 57
9 O&M cost versus carbon loss per reactivation cycle
for 1, 5, 10, 100, and 150 mgd sand replacement systems 58
10 O&M cost versus fuel cost in $/therm for 1, 5, 10, 100, and
150 mgd sand replacement systems 59
11 O&M cost versus producers price index for 1, 5, 10, 100,
and 150 mgd sand replacement systems 60
12 O&M cost versus power cost for 1, 5, 10, 100, 150 mgd and
sand replacement systems 61
13 Capital cost versus construction cost index for 1, 5, 10,
100, and 150 mgd sand replacement systems 62
14 Capital cost versus interest rate for 1, 5, 10, 100, and
150 mgd sand replacement systems 63
15 Capital cost versus amortization for 1, 5, 10, 100, and
150 mgd sand replacement systems 64
-------
FIGURES (continued)
Number Page
16 Capital cost versus carbon cost for 1, 5, 10, 100, and
150 mgd sand replacement systems 65
17 O&M cost versus carbon cost for 1, 5, 10, 100, and
150 mgd sand replacement systems ..« 66
18 Capital cost versus reactivation frequency for 1, 5, TO, 100,
and 150 mgd sand replacement systems 67
19 O&M cost versus reactivation frequency for 1, 5, 10, .100, •
and 150 mgd sand replacement systems .. 68
TABLES
1 Summary of GAC System Characteristics 3
2 GAC in Municipal Water Plants in the United States 28
3 Carbon Reactivation Furnace Installations 30
4 Cost Calculation and Operation and Maintenance Requirements
for a 100 mgd Pressure Granular Activated Carbon Plant 47
5 Annual Cost for 110 mgd Gravity, Steel GAC Plant 48
6 Design Parameters for Granular Activated Carbon 50
7 Assumptions for Separate Post-Filtration Systems 50
8 Amortized Capital and Operating and Maintenance Cost
for GAC Systems (it/1,000 gal) 51
9 Contactor and Reactivation Costs for GAC Systems 51
XI
-------
ABBREVIATIONS
avg average
AWT advanced wastewater treatment
BLS Bureau of Labor Statistics
BOD Biological Oxygen Demand
BSP Bartlett-Snow-Pacific Inc.
Btu British thermal unit
CCI Construction Cost Index
cm centimeter(s)
COD Chemical Oxygen Demand
cu cubic
d day(s)
dia diameter
EBCT empty bed contact time
ENR Engineering News Record
GAC granular activated carbon
gal gallon(s)
gptn gallon(s) per minute
hr hour(s)
I.D. Inside diameter
in. inch(es)
JTU Jackson turbidity units
kg . kilogram(s) . : •
kwh " kilowatt hour(s)
1 liter(s)
Ib pound(s)
m meter(s)
max Maximum
MBAS Methylene blue active substances
mgd million gallon(s) per day
mg/L milligram(s) per liter
mhr manho ur(s)
mil million
min minute(s) or minimum
O&M operation and maintenance
PAC powdered activated carbon
psi pound(s) per square inch
scf standard cubic feet
sq square
SS suspended solids
SWD side wall depth
THM trihalomethane
TOC Total Organic Carbon
TON threshold odor number
XII
-------
Abbreviations (Continued)
TN total nitrogen
TP total phosphorus
TTSA Tahoe-Truckee Sanitation Agency
w/ with
yr year
304SS stainless steel
316SS stainless steel
XI11
-------
CONVERSION TO METRIC
English units have been used throughout this report. For the readers
convenience the following table can be used to convert to metric units .
English Units
in.
ft
sq ft or ft2
acre
cu ft of ft3
gal
gpm
mgd
lb
op
(fc/1000 gal
Multiplier
2.54
0.305
0.093
0.405
028
79
0
3
227
3785
0.454
5/9(°F-32)
0.26
Metric Units
cm
m
sq m or m2
hectares
cu m or m3
L
L/hr
cu m/d or
m3/d
kg
°C
-------
ACKNOWLEDGEMENT
The authors hereby acknowledge the valuable assistance of Stephen W. Work,
Denver Water Department, Robert L. Chapman, CH2M/H111 Engineers, Dr. Robert M.
Clark, U.S. EPA, and Marion Curry, editor, in the review and completion of this
report.
xv
-------
On
posed t)
ing wat
in th-
ing ;
EPA'
est
at
1
1,
be
cai
Pli
"Est
This
as rt
systei
detent
Po
rem
-------
effects, but the use of granular activated carbon (GAC) in the treatment of
municipal water in the U.S. is limited to a few facilities. In most cases,
GAC is used to remove taste and odor Lrom drinking, water. Its use may become
more common in light of new information on the occurrence of trace organics in
water, the recent regulations limiting the concentration of trihalomethanes in
public water supplies, and the possibility of requirements for carbon treat-
ment. European water works have had considerable experience over a long period
of time with- GAC installations. • ;
Unlike the limited use of GAC in municipal water treatment, GAC has been
used in industrial and municipal wastewater treatment and in various industrial
process applications. Although the specific uses of GAC are somewhat different
with these applications than with water treatment, much of the-information on
design and operations will prove useful to water purveyors. In general, the
application of GAC adsorption to drinking water is simpler than to wastewater.
A series of case studies has been prepared as a result of site visits and
literature -searches. Twenty-two GAC installations were visited; 18 of these
were industrial process or municipal wastewater facilities, whereas 4 were
municipal water treatment plants. Case histories presenting design, operating,
performance, and cost information are presented in the Appendix; the pertinent
information has been summarized (Table 1). Single page fact sheets for each of
these case histories follow Table 1.
A principal 'function of the site visits was to collect construction and
operation and maintenance (O&M) cost.information on GAC installations. Plant
records were used to obtain available construction. costs,' the dates for these
costs, and the most recent O&M costs. The basic data are presented in the indi-
vidual case study reports, but no attempt was made to update construction or
O&M costs- to present day prices or to convert the costs to water treatment
plants. This i.report does, however refer to procedures whereby data from exist-
ing GAC projects can be adjusted or modified (based on the results of pilot
plant test results of the water to be treated) so as to be useful to experi-
enced professionals in the field in making preliminary estimates of costs for
future potable water projects involving GAC adsorption and reactivation or
replacement.
-------
TABLE 1. SUMMARY OF GAC SYSTEM CHARACTERISTICS
Carbon contactors
Case
NO.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Owner
South Tahoe
Tahoe-Truckee
Upper Occo-
quan
American
Cyan amid
Vallejo
Orange Co .
Niagara Falls
Fitchburg'i'
Arco Petro-
leum
Rhone-
Poulenic
Reichhold
Chemicals
Stepan Chemi-
cals
Republic
Steel §
LeRoy
Manchester
Passaic Valleytt
Colorado
Springs
Hercules
Industrial
sugar
Hopewell
Davenport
Spreckles
Sugar
Type of
facility
Municipal
wastewater
Municipal
wastewater
Municipal
wastewater
Chemprocess
Municipal
wastewater
Secondary
effluent
Municipal w/
' significant •
industrial
Municipal w/
significant
industrial
Process wastes
w/significant
industrial
Herbicide
production
wastes
Chemical pro-
duction wastes
Surfactant pro-
duction wastes
Coke process
wastes
Municipal
wastewater
Water supply
Water supply
Secondary
effluent
Chemical pro-
duction wastes
Decoloring
sugar
Water supply
Water supply
Sugar thick
juice
Flow, mgd
7. 5 (max)
4. 83 (max)
15.0(avg)
20.0(avg)
13.0(avg)
15.0 (max)
48.0(avg)
15.0(avg)
4. 32 (max)
0.15(max)
1.0 (max)
0.015 (max)
0.95 (max)
1.0 (max)
40.0 (max)
2. 2 (max)
2.0 (max)
3.25(max)
-
3.0(mg)
30.0 (max)
-
Pretreatment
Extensive
Extensive
Extensive
Extensive
Moderate
Extensive
Moderate
Moderate
Minimal
None
Moderate
None
Minimal
Extensive
Moderate
Extensive
Extensive
Moderate
Minimal
Moderate
Moderate
None
Contact
time,
min
17
20
22
30
25
34
40
15
56
87
100
500
116/58
12
14
8
17
48
1080
-
7.5
20
Hydraulic
loading ,
6.5
8.4
8.0
6.0
5.8
1.67
8.00
1.74
2.00
1.55
-
2.3/4.6
-
-
-
4.5
6.6
-
2.0
2.0
-
•Not available
^Facility under construction
§Two separate treatment trains
tFluidized bed
ttGAC test facility
§§kw/lb carbon
-------
TABLE 1. (Continued)
Furnace
Rated
capacity,
Ib carbon/flay
6,000
3,840
12,000
122,000
29,000
12,000
.
Carbon
loss, Fuel
% type
8
5
10
9
7.5
6
-
Gas
Propane
LPG
-
. -. '
Natural
gas
-
Use,
Btu/lb
carbon
2,900
3,840
2,750
-
-
5,600
-
Carbon Dose
Costs
Capital
Ib/mil Ib organic/
qal Ib carbon
207 0.38
0.33
250 0.33
-
1,410 0.50
1.70
-
Capital ,
$1000 (year)
849(1969)
1,569(1976)
3,880(N/A)
N/A
4,359(1974)
3,307(1972)
. .N/A
5/lb day
capacity
141.50
408.59
323.33
-
150.31
275.58
- '
O&M, $/mil
gal (year)
36.07(1979)
N/A*
50.00(1979)
N/A
N/A
90.49(1979)
N/A
N/A
N/A
8,500
Natural
gas '
1,000
0.26
1,000(1971) 117.65 490.00(1973)
6,500 " 8.8' Natural ' 6',500 ' 43/000
gas
0.38"
300(1969) 35.29 • 319.00(1973)
32,500
6,480
68,000
12 , 000
12,OOCt
2,400
1,800
33,600
12,000
None
None
15,000
5 Natural 6,200 79,000 0.26
gas
6 LPG 6,000 500,000
8 Natural - 35,000
gas
Fuel oil -
10 Fuel oil -
10 Electricity 0.7§§
7.5 - - 160 0.50
5 Natural 7,000 9,500 0.44
gas
2.5 Natural -
gas
-
-
4.8 Natural 1,785
N/A
225(1973)
N/A
2,500(1975)
N/A
N/A
N/A
1,622(1973)
N/A
N/A
N/A
238(1958)
N/A
34.72 25,000(19'
N/A
208.33 N/A
N/A
N/A
89.59
48.27 1,470.00
N/A
30.00(191
20.00(191
15.87 N/A
gas
-------
MUNICIPAL WASTEWATER TREATMENT
FACT SHEET
CASE HISTORY NO. 1
SOUTH TAHOE (CALIFORNIA) PUBLIC UTILITY DISTRICT
Type of Wastewater: Municipal
Size of Treatment Facility: 7.5 mgd (maximum)
Date of GAG Startup: 1966
Pretreatment: Primary, activated sludge/ lime clarification,
ammonia stripping, recarbonation, filtration
purpose of GAC: Dissolved organics removal, effluent discharged
to recreational impoundment
GAC Design Features:
Type of contactors - upflow, packed bed, countercurrent
Number, size, and bed depth - eight; 12-ft dia x 12-ft
straight sidewall; 14.5 ft
Minimum contact time - 17 min
Hydraulic (surface) loading - 6.5 gpm/ft2
Carbon size - 8 x 30 mesh
GAC Reactivation Furnace:
Type and size - six hearth; 54-iri. dia
Rated capacity - 6,000 Ib carbon/day
Regeneration temperature - 1650° to 1750°F
Type of fuel - gas with propane standby
Fuel demand - 2,900 Btu/lb carbon .
Carbon loss - 8%
System Performance:
COD MBAS Color
Concentration in, mg/L 20.3 0.6 11 (color units)
Concentration out, mg/L 10.0 0.1 5-6 (color units)
Removal efficiency, % 51 77 50
Carbon dose rate - 0.38 Ib COD removed/lb carbon
Carbon exhaustion - 207 Ib carbon/mil gal treated
Capital Costs (1969):
GAC units with carbon- $656,000
Reactivation system - 193,000
$849,000
O&M Costs (1970):
Adsorption - $n.74/mil gal
Reactivation - 2J.33/mil gal
$36.07/mil gal
-------
MUNICIPAL WASTEWATER TREATMENT
FACT SHEET
CASE HISTORY NO. 2
TAHOE-TRUCKEE (CALIFORNIA) SANITATION DISTRICT
Type of Wastewater: Municipal
Size of Treatment Facility: 4.83 mgd (maximum)
Date of GAC Startup: 1979
Pretreatment: Primary, pure oxygen activated sludge, lime
clarification, two-stage recarbonation, mixed media
filtration
Purpose of GAC: Organics removal, effluent discharged to
Truckee River
GAC Design Features:
Type of contactors - upflow, packed bed, countercurrent
Number, size, and bed depth - six; 12-ft dia x 20-ft straight
sidewall; 20 ft •• • '• '
Minimum contact time - 20 min
Carbon size - 8 x 30 mesh
GAC Reactivation Furnace:
Type and size - six hearth; 81-in. dia
Rated capacity - 3,840 Ib carbon/day plus-1,980 Ib
adsorbate/day
Hearth loading - 45 lb/day/ft2 ' •
. Type of fuel - propane-with fuel oil. standby .
Fuel demand- 3,780 Btu/lb carbon plus adsorbate
Carbon loss - 5%
System Performance:
Carbon dose rate - 0.33 Ib COD removed/lb carbon
Capital Costs (1976):
GAC units with carbon - $ 619,000
Reactivation system - 950,000
$1,569,000
O&M Costs: Not available
-------
MUNICIPAL WASTEWATER TREATMENT
FACT SHEET
CASE HISTORY NO. 3
UPPER OCCOQUAN (VIRGINIA) SEWAGE AUTHORITY
Type of Wastewater: Municipal
Size of Treatment Facility: 15 mgd (average)
Date of GAC Startup: 1978
pretreatment: Primary, activated sludge, lime clarification,
two-stage recarbonation, mixed media filtration
purpose of GAC: Soluble organics removal, effluent disposal to
domestic water supply reservoir
GAC Design Features:
Type of contactors - upflow, countercurrent
Number, size, and bed depth -•sixteen; 12-ft dia x 24-ft
straight sidewall; 24.3.ft
Minimum contact time - 22 min
Hydraulic (surface) loading -'8.40 gpm/ft
Carbon size T 8 x 30 mesh
GAC Reactivation Furnace:
Type and size - two at seven-hearth; 129-in. OD
Rated capacity - 12,000 Ib/day
Hearth loading - 45 lb/day/ft^
Type of fuel - liquid propane gas
Fuel demand - 2,750 Ib/day ;.....
Carbon loss - 10%
System Performance: Operation based on column effluent upper
and lower COD of 10 mg/L
Carbon dose rate - 0.18 to 0.33 Ib COD/lb carbon
Carbon exhaustion - 250 Ib carbon/mil gal treated
Capital Costs (N/A*):
GAC units with carbon - $2,250,000
Reactivation system - 1,630,000
$3,880,000
O&M Costs (1979): $50.00/mil gal
*Date not available
-------
.INDUSTRIAL WASTEWATER TREATMENT
FACT SHEET
CASE HISTORY NO. 4
AMERICAN CYANAMID CO., ORGANIC CHEMICALS DIVISION
BOUND BROOK, NEW JERSEY
Type of Wastewater: Chemical wastes, sanitary sewage, cooling
water, stormwater runoff
Size of Treatment Facility: 20 mgd (average)
Date of GAC Startup: 1977
Pretreatment: Primary with lime neutralization, activated
sludge, mixed media filtration
purpose of GAC: Color, odor, toxicity and suspended solids
(refractory organics) removal; effluent disposal comprises
up to 25 percent Raritan River flow
GAC Design Features:
Type of contactors - expanded upflow, countercurrent
Number, size, and bed depth - ten; 16-ft dia x 40-ft straight
sidewall; 30 ft
Minimum contact time - 30 min
Hydraulic (surface) loading - 8.0 gpm/ft
Carbon size -8x30 mesh
GAC Reactivation Furnace:
Type and size - ejght-hearth; 309-in. dia
Rated capacity - 122,000 to 152,000 lb/day
Regeneration temperature - 1750°F
Carbon loss - 9%
System Performance: TOC removal in pilot studies averaged 68%
Capital Costs: Not available
O&M Costs: Not available
-------
MUNICIPAL WASTEWATER TREATMENT
FACT SHEET
CASE HISTORY NO. 5
VALLEJO (CALIFORNIA) SANITATION AND FLOOD CONTROL DISTRICT
Type of Wastewater: Municipal
Size of Treatment Facility: 13 mgd (average)
Date of GAC Startup: 1979
pretreatment: Lime sedimentation, single—stage recarbonation
Purpose of GAC: Organic and suspended solids removal; effluent discharge to
Carquinez Straight
GAC Design Features:
Number, size, and bed depth - six; 18-ft square; 16 ft
Minimum contact time - 25 min
Hydraulic (surface) loading - 6 gpm/ft
Carbon size - 12 x 40 mesh
GAC Reactivation Furnace:
Type and size - six-hearth; 12 ft ID
Rated capacity - 29,000 Ib/day
Hearth loading - 3 Ib/hr ft2
Regeneration temperature - 1500°F
Carbon loss - 7.5%
System Performance:
COD MBAS SS
Concentration in, mg/L 175 104 66
Concentration out, mg/L 90 45 22
Removal efficiency, % 49 57 67
Carbon dose rate - 0.5 Ib COD removed/Ib carbon
Carbon exhaustion - 1,410 Ib carbon/mil gal treated
Capital Costs (1974):
GAC units with carbon - $2,929,000
1,430,000
$4,359,000
O&M Costs: Not available
-------
WASTEWATER RECLAMATION
FACT SHEET
CASE HISTORY NO. 6
ORANGE COUNTY (CALIFORNIA) WATER DISTRICT
Type of Wastewater: Secondary effluent ' ~
Size of Facility: 15 mgd
Date of GAC Startup: 1976
pretreatment: Lime clarification, air stripping,
recarbonation, chlorination, filtration
purpose of GAC: Organics removal, effluent injection to
prevent sea water intrusion
GAC Design Features:
Type of contactors - downflow
Number and size - seventeen; 12-ft dia x 24-ft straight
sidewall
Minimum contact time - 34 min
2
Hydraulic (surface) loading - 5.8 gpm/ft
Carbon size1 - 8 x 30 mesh
GAC Reactivation Furnace:
Type - six-hearth
Rated capacity - 12,000 Ib/day
, Regeneration temperature - 1650°F
Type of fuel - natural gas
Fuel demand - 5,600 Btu/lb carbon
Carbon loss- 6%
System Performance:
COD TOC
Concentration in, mg/L 42.0 14.4
Concentration out, mg/L 16.6 7.0
Removal efficiency, % 60 51
Carbon dose rate - 1.7 Ib COD removed/lb carbon
Capital Costs (1972): $3,307,000
0 & M Costs (1979):
Adsorption - $29.80/mil gal
Reactivation - 60.69/mil gal
$90.49/mil gal
10
-------
MUNICIPAL WASTEWATER TREATMENT
FACT SHEET
CASE HISTORY NO. 7
CITY OF NIAGARA FALLS, NEW YORK
Type of Wastewater: Municipal with significant industrial
contribution
Size of Treatment Facility: 48 mgd (average)
Date of GAC Startup: 1977
Pretreatment: Chemical clarification, acid addition
Purpose of GAC: Dissolved organics removal, effluent discharge
to Niagara River
GAC Design Features:
Type of contactors - open-top, downflow
Number, size, and bed depth - twenty-eight; 17.3-ft x 42-ft
rectangular; 8.75 ft
Minimum contact time - 40 .min
Hydraulic (surface) loading - 1.67 gpm/ft
GAC Reactivation Furnace:
Type and size - six-hearth; 14.25-ft dia
System Performance: Not available
Capital Costs: Not available
O&M Costs: Not available
11
-------
MUNICIPAL WASTEWATER TREATMENT
FACT SHEET
CASE HISTORY NO. 8
CITY OF FITCHBURG, MASSACHUSETTS
Type of Wastewater: Municipal and industrial (significant
paper processing wastes)
Size of Treatment Facility: 15 mgd (average)
Date of GAC Startup: Under construction
Pretreatment: Primary, lime clarification, alum addition as
required
Purpose of GAC Removal: BOD, COD, color removal, portion of
effluent to be used by paper processer
GAC Design Features:
Type of contactors - downflow, series
Number, size and bed depth - twelve; 20-ft dia, 30-ft height;
. 15.5 ft •
Minimum contact time - 15 min
Hydraulic (surface) loading - 8 gpm/ft^
GAC Regeneration Furnace:
Type - multiple hearth
Hearth loading - 4.17 lb/hr/ft2 .
Carbon loss- 5%
System Performance (pilot study):
. -. ..... COD , BOD ;..
Concentration in, mg/L 53.4 13.2
Concentration out, mg/L 6.8 1.7
Removal efficiency, % 87 87
Capital Costs: Not available
O&M Costs: Not available
12
-------
INDUSTRIAL WASTEWATER TREATMENT
FACT SHEET
CASE HISTORY NO. 9
ARCO PETROLEUM PRODUCTS COMPANY, WATSON REFINERY
CARSON, CALIFORNIA
Type of Wastewater: Process wastes and storm water runoff
(stored prior to treatment)
Size of Treatment Facility: 3,000 gpm (maximum)
Date of GAC Startup: 1971
Pretreatment: Reservoir storage
purpose of GAC: Soluble COD and phenols removal, effluent
discharged to Dominquez Channel
GAC Design Features:
Type of contactors - downflow, gravity, parallel
Number, size, and bed depth - twelve; 12-ft x 12-ft x 26-ft
deep; 13 ft
Minimum contact time - 56 min
Hydraulic (surface) loading - 1.74 gpm/ft2
Carbon size - 8 x 30 mesh
GAC Reactivation Furnace:
Type and size - six hearth; 56-in. ID
Rated capacity - 8,500 Ib/day carbon
Regeneration temperature - 1600° to 1750°F
Type of fuel - natural gas
Fuel demand - 3,000 SCFH
Carbon loss - 5%
System Performance:
COD Oil
Concentration in, mg/L 233 28
Concentration out, mg/L 48 12
Removal efficiency, % 79 57
Carbon dose rate - 0.26 Ib COD removed/lb carbon
Carbon exhaustion- 1,000 Ib/carbon/mil gal treated
Capital Costs (1971): Approximately $1 million
O&M Costs (1971-73): $490.00/mil gal
13
-------
INDUSTRIAL WASTEWATER TREATMENT
FACT SHEET
CASE HISTORY NO. 10
RHONE-POULENIC INC.
PORTLAND, OREGON
Type of Wastewater: Herbicide production wastewater
Size of Treatment Facility: 0.15 mgd (maximum)
Date of GAC Startup: 1969
Pretreatment: None
Purpose of GAC: Phenol removal, effluent discharged to City
sewer system
GAC Design Features:
Type of contactors - upflow series, no backwash
Number, size, and bed depth - two; 8-ft ID x 35-ft height;
12 ft
Minimum contact time - 87 min
Hydraulic (surface) loading - 2.0 gpm/ft2
Carbon size - 12 x 40 mesh
GAC Reactivation Furnace (reactivation off-site since 1977):
Type and size - six-hearth; 54=in. ID
Rated capacity - 8,500 Ib/day
Regeneration temperature - 1600° to 1800°F
Type of fuel - natural gas . •
Fuel demand -• 6,500 Btu/lb carbon
Carbon loss - 8.8%
System Performance:
Ortho-cresol Total Phenol
Concentration in, mg/L 16.5 52.9
Concentration out,* mg/L 0.62 0.69
Removal efficiency, % 96 99
Carbon dose rate - 0.38 Ib organics removed/lb carbon
Carbon exhaustion - 43,000 Ib carbon/mil gal treated
Capital Costs (1969): $300,000
O&M Costs (1973): $319.00/mil gal
*Two-stage series
14
-------
INDUSTRIAL WASTEWATER TREATMENT
FACT SHEET
CASE HISTORY NO. 11
REICHHOLD CHEMICALS INC.
TUSCALOOSA, ALABAMA
Type of Wastewater: Chemical production wastewater
Size of Treatment Facility: 10 mgd (maximum)
Date of GAC Startup: 1973
pretreatment: Neutralization, clarification with polymer
Purpose of GAC: Dissolved organics (including phenol) removal,
effluent discharged to Black Warrior River
GAC Design Features:
Type contactors - upflow, countercurrent, parallel
Number, size, and bed depth - two; 12-ft dia x 40-ft depth
Minimum contact time - 100 min
*\
Hydraulic (surface) loading - 1.55 gpm/ft
Carbon size - 8 x 30 mesh
GAC Reactivation Furnace:
Type and size - six-hearth; 13.5-ft ID
Rated capacity - 32,500 Ib/day
Regeneration temperature - 1700°F
Type of fuel - natural gas, liquid propane gas standby
Fuel demand - 6,200 Btu/lb carbon
Carbon loss - 5%
System Performance:
COD BOD
Quantity in, Ib/day 9750 3640
Quantity out, Ib/day 1868 1078
Removal efficienty, % 81 70
Carbon dose rate - 0.26 Ib COD removed/lb carbon
Carbon exhaustion - 79,000 Ib carbon/mil gal treated
Capital Cost (N/A*): $1.3 million (complete plant)
O&M Cost (1976): $793.00/mil gal (all treatment)
*Date not available.
15
-------
INDUSTRIAL WASTEWATER TREATMENT
PACT SHEET
CASE HISTORY NO. 12
STEPAN CHEMICAL COMPANY
BORDENTOWN, NEW JERSEY
Type of Wastewater: Surfactant process wastewater
i Size of Treatment Facility: 15,000 gpd
Date of GAC Startup: 1972
Pretreatment: None
Purpose of GAC: Organics removal, effluent discharged to
| Delaware River
| GAC Design Features:
Type of contactors - downflow, pressure, series
Number, size and bed depth - three; 6-ft dia x 10-ft height;
' 7.5 ft
Minimum contact time - 500 min
Carbon size - 8 x 30 mesh
'. GAC Reactivation Furnace:
Type and size - six-hearth; 54-in. ID
Rated capacity - 6,480 Ib/day
' Regeneration temperature - 1600° to 1800°F
Type of fuel - liquid propane gas
Fuel demand - 6,000 Btu/lb carbon
Carbon loss - 6%
| System Performance:
COD BOD TOC MBAS
: Effluent concentration, mg/L 117 16 19 0.12
: Carbon exhaustion - 500,000 Ib carbon/mil gal treated
• Capital Costs (1973): $225,000
O&M Costs (1978): $25,000/mil gal
16
-------
INDUSTRIAL WASTEWATER TREATMENT
FACT SHEET
CASE HISTORY NO. 13
REPUBLIC STEEL CORPORATION
CLEVELAND, OHIO
Type of Wastewater: Coke process wastewater
Size of Treatment Facility: 220 gpm (System A), 440 gpm (System
B)
Date of GAC Startup: 1977
Pretreatment: Chemical addition, dissolved gas flotation,
media filtration
Purpose of GAC: Phenol removal, effluent discharge to surface
water
GAC Design Features:
Type contactors - two separate systems; downflow, series, no.
backwash
Number, size, and bed depth .(.each system) - 2; 11-ft dia x
24.5-ft height; 18 ft
Minimum contact time - 116 -min (A), 58 min (B)
Hydraulic (surface) loading - 2.3 gpm/ft (A), 4.6
gpm/ft2 (B) . . :
GAC Reactivation Furnace:
Type and size - eight-hearth; 16-ft OD
Rated capacity - 68,000 lb/day
Regeneration temperature - 1600° to 1800°F
Type of fuel - natural gas, coke oven gas standby
Carbon loss - 5 to 8%
System Performance (A) Phenol
Cencentration in, mg/L 30.0
Concentration out, mg/L 0.025
Removal efficiency, % 99+
Carbon exhaustion- 35,000 Ib/mil gal (A), 25,000 Ib/mil gal
(B)
Capital Costs (N/A*): $10 million (complete system)
O&M Costs (N/A*): $8,000/day (complete treatment)
*Date not available.
17
-------
MUNICIPAL WASTEWATER TREATMENT
FACT SHEET
CASE HISTORY NO. 14
VILLAGE OF LE ROY, NEW YORK
Type of Wastewater: Municipal
Size of Treatment Facility: 1 mgd (maximum)
Date of GAC Startup: 1978
pretreatment: Clariflocculation, rapid sand filtration
Purpose of GAC: Residual organics removal/ effluent discharge
to Oatka Creek
GAC Design Features:
Number and size - four; 11-ft dia
Minimum Contact Time - 12 min
Gas Reactivation Furnace:
Type and size - five-hearth; 9.25-ft OD
Rated Capacity - 12,000 Ib/day
Type of fuel - No. 2 fuel oil
Fuel demand - 30,000 gal/yr
Carbon loss - 8.5%
System Performance: Not available
Capital Costs (1975): $2,500,000
O&K Costs: Not available . . •
18
-------
MUNICIPAL WATER PURIFICATION
FACT SHEET
CASE HISTORY NO. 15
CITY OF MANCHESTER, NEW HAMPSHIRE
Type of Water: Raw water
Size of Treatment Facilities: 40 mgd
Date of GAC Startup: 1974
pretreatment: Chemical clarification, sand filtration
Purpose of GAC: Taste, odor and volatile organics removal from
potable water supply
GAC Design Features:
Type of contactors - series with sand filtration, downflow
Number and bed depth r four; 4-ft
Minimum contact time - 14 to 22 min
Hydraulic (surface) loading-.- 4.0 gpm/ft^
Carbon size - 8 x 30 mesh
GAC Reactivation Furnace:
Type - fluidized bed
Rated capacity - 12,000 Ib/day
Type of fuel - fuel oil
Fuel demand - 240-gal/day
Carbon loss - 10%
System Performance (complete processing):
Color (color units) TOC
Concentration in, mg/L 15 4.0
Concentration out, mg/L 1 1.8
Removal efficiency, % 93 55
Capital Costs (1979): $882,406 (reactivation only)
O&M Costs: Not available
19
-------
MUNICIPAL WATER PURIFICATION
FACT SHEET
CASE HISTORY NO. 16
PASSAIC VALLEY (NEW JERSEY) WATER COMMISSION
Type of Water: Raw water
Size of Treatment Facility: 2.2 mgd GAG Test Facility
Date of GAC Startup: 1978
pretreatment: Prechlorination, alum clarification/ dual-media
filtration
Purpose of GAC: Experimental testing of GAC regeneration
GAC Design Features:
Minimum contact time - 8 min, pressure filters, downflow
GAC Reactivation Furnace:
Type and size - infrared tunnel; 4 ft x 20 ft
Rated capacity - 2,400 Ib/day
Regeneration temperature - 1600°F
Type of fuel - electric power (100 kw)
Fuel demand - 0.7 kw/lb carbon
Carbon loss - 8 to 10%
System Performance: TOC removal 20 to 40%
Capital Costs: Not available '..'".."•.'
O&M CoiSts: Not available
20
-------
MUNICIPAL WASTEWATER TREATMENT
FACT SHEET
CASE HISTORY NO. 17
COLORADO SPRINGS, COLORADO
Type of Wastewater: Secondary effluent (overloaded
bio-filters)
Size of Treatment Facility: 2 mgd
Date of GAC Startup: 1970
Pretreatinent: Lime clarification, dual media filtration
Purpose of GAC: Organics removal, effluent recycled for
industrial use
GAC Design Features:
Type of contactors - downflow, series
Number, size, and bed depth - three; 20-ft dia x 14-ft
straight sidewall; 8 ft
Minimum contact time - 17 min
Hydraulic (surface) loading - 4.5 gpm/ft
Carbon size - 8 x 30 mesh
GAC Reactivation Furnace:
Type and size - one, six-hearth; 3-ft dia
Rated capacity - 1,800 Ib/day
Regeneration temperature - 1650°F
Carbon loss - 7.5%
System Performance:
COD BOD TOC
Concentration in, mg/L 57 27 24
Concentration out, mg/L 31 13 10
Removal efficiency, % 46 52 58
Carbon dose rate - 0.50 Ib COD removed/lb carbon
Carbon exhaustion - 160 Ib/mil gal treated
Capital Costs: Not available
O&M Costs (1972): $89.59/mil gal
*Color units.
21
-------
INDUSTRIAL WASTEWATER TREATMENT
FACT SHEET
CASE HISTORY NO. 18
HERCULES, INC.
HATTIESBURG, MISSISSIPPI
Type of Wastewater: Industrial chemical process wastewater
Size of Treatment Facilities: 3.25 mgd
Date of GAC Startup: 1973
Pretreatment: Impoundment, dissolved air flotation, mixed media
filtration
purpose of GAC: Phenol and TOC removal, effluent discharged to
Bowie River
GAC Design Features:
Type of contactors - upflow, countercurrent
Number and size - three; 12-ft dia x 40-ft height
Minimum contact time - 48 min
O
Hydraulic (surface) loading - 6.6 gpm/ft
Ca.rbon size - 12 x 40 mesh
GAC Reactivation Furnace (replaced with fluidized bed furnace
in 1977):
Type and size - five-hearth; 12-ft OD
Rated capacity - 33,600 Ib/day
Regeneration temperature - 1600° to 1800°F
Type of' fuel - natural gas, fuel oil standby ,
Fuel demand - 6,000-7,000 Btu/lb carbon
Carbon loss - 5%
System Performance:
TOC Phenolics
Concentration in, mg/L 292 -
Concentration out, mg/L 168 0.13
Removal efficiency, % 43 -
Carbon dose rate - 0.1 to 0.44 Ib TOC removed/lb carbon
Carbon exhaustion - 9,500 Ib carbon/mil gal treated
Capital Costs (1973): $1,622,000
O&M Costs (1979): $1,470/mil gal
22
-------
FOOD PROCESSING
FACT SHEET
CASE HISTORY NO. 19
INDUSTRIAL SUGAR, INC.
ST. LOUIS, MISSOURI
Type of Liquid: Liquid sugar
Size of Treatment Facility: Not applicable
Date of GAC Startup: 1958
processing: Filtration for particulate impurity removal
Purpose of GAC: Decoloring liquid sugar
GAC Design Features:
Type of contactors - downflow, series
Number, size, and bed depth .- six; 6-ft dia x 9-ft height;
four; 10-ft dia x 9-ft height; 9 ft
Minimum contact time - 18 .hr total
Carbon size - 12 x 40 mesh
Gas Reactivation Furnace:
Type and size - six-hearth;. 6-ft OD
Rated capacity - 12,000 lb/day
Regeneration temperature - 1600°F
Type of fuel - natural gas
Carbon loss - 2.5%
System Performance:
Carbon exhaustion - 2 Ib carbon/100 Ib raw sugar ,
Capital Costs (1965): $125,000 reactivation system only
O&M Costs: Not available
23
-------
MUNICIPAL WATER PURIFICATION
FACT SHEET
CASE HISTORY NO. 20
AMERICAN WATER WORKS SERVICE COMPANY
HOPEWELL, VIRGINIA
Type of Water: Raw water
Size of Treatment Facility: 3 mgd
Date of GAC Startup: 1960
Pretreatment: Chemical coagulation, settling, filtration
through spent carbon
purpose of GAC: Municipal water purification - taste and odor
removal; spent carbon used for filtration and
dechlorination
GAC Design Features;
Bed depth - 24 in.
Hydraulic (surface) loading - 2 gpm/ft2
GAC Regeneration Furnace: None
System Performance:
Turbi-
Color* Mn Pb dityt Chlorine
Concentration in, mg/L 4-14 0.15 0.37 1.4 2.8
Concentration out, mg/L 0-2 0.017 0.025 0.15 0.25
Removal Efficiency, % 86-100 89 89 89 91 .
Capital Costs: Not available
O&M Cost (1980): $30.00/mil gal - . . •
*Color Units
tJTU
24
-------
MUNICIPAL WATER PURIFICATION
FACT SHEET
CASE HISTORY NO. 21
AMERICAN WATER WORKS SERVICE COMPANY
DAVENPORT, IOWA
Type of Water: Raw water
Size of Treatment Facility: 18 to 20 mgd (avg), 30 mgd (max)
Date of GAC Startup: 1973 (conversion of gravity sand filters
to GAC)
Pretreatment: Prechlorination/-coagulation, sedimentation,
filtration
Purpose of GAC: Municipal water purification - taste, odor,
turbidity and trihalomethane (THM) removal
GAC Design Features:
Type of contactors - gravity (converted sand filters)
Number and bed depth - twenty; 24 in.
Minimum contact time - 7.5 min
Hydraulic (surface) loading - 2 gpm/ft2
GAC Reactivation Furnace: None
System Performance:
Turbidity* TOC THMt
Concentration in, mg/L 3 . 5.3 97
Concentration out, mg/L 0.32 4.2 69
Removal efficiency, % 89 21 .29
Capital Costs: Not available
O&M Costs (1980): $20.00/mil gal
25
-------
FOOD PROCESSING
CASE HISTORY NO. 22
SPRECKLES SUGAR
WOODLAND, CALIFORNIA
Type of Liquid: Sugar juice
Size of Treatment Facility: Not applicable
Date of GAC: Removal of color and impurities from beet sugar
thick juice
GAC Design Features:
Type of contactors - upflow expanded, continuous adsorption
process (CAP)
Number, size, and bed depth - three; 9.5-ft dia x 47.6-ft
height; 13 to 14 ft
Minimum contact time - 20 min
Carbon size - 12 x 40 mesh
GAC Reactivation Furnace:
Type and size - six-hearth; 8.33-ft dia
Rated capacity - 15,000 Ib/day
Regeneration temperature - 1600° to 1720°F
Type of fuel - natural gas
Fuel demand - 1,?85 Btu/lb carbon
Carbon loss - 4.8%
System Performance-- Not available
Capital Costs (1958):
GAC units, carbon, rotary kiln furnace - $238,000
O&M Costs: Not available
26
-------
SECTION 2
HISTORICAL USE OF ACTIVATED CARBON IN WATER PURIFICATION
POWDERED CABBON
Powdered activated carbon (PAC) has been used successfully for more than
50 years to remove taste and odor from public drinking water supplies. Except
for partical size, this PAC is often identical to the GAC used in wastewater
treatment. In 1930, Hassler reported that PAC was being used in more than 150
municipal water treatment plants. In this type of use, PAC dosages usually are
in the range of 1 to 5 mg/L, although dosages as high as 20 to 30 mg/L have
been used in some places for short periods of time when taste and odor problems
are severe. PAC is commonly used on a one-time, throwaway basis, with no
attempt at recovery or reuse .
PAC's principal benefit'in water treatment is to remove taste and odor. In
some waters, PAC may also remove color or organics that otherwise would inter-
fere with coagulation or filtration. During its widespread use by waterworks
for more than 50 years, no harmful effects have been reported.
GRANULAR CARBON
Altogether 46 water treatment plants in the United States are presently
using GAC (AWWA, 1979); a partial list is given in Table 2. In these plants,
GAC is used to control taste and odor in drinking water. At the Hopewell, Vir-
ginia, and Davenport, Iowa plants however, other benefits have been observed by
the American Water Works Service Co., especially the reduction of trihalome-
thanes in finished water.
Because there has been no need for on-site reactivation of carbon, few
useful cost data are available for estimating the costs involved in organics
removal, which necessitates more frequent renewal of the GAC.
Recently, Manchester, New Hampshire installed a fluidized bed in its water
plant for on-site GAC reactivation. Capital costs are available from EPA and
the city, but operating and maintenance costs will not be forthcoming until
more data are accumulated.
Philadelphia, Cincinnati, and Passaic (New Jersey) are operating pilot GAC
projects that will soon yield capital and O&M cost data for water
purification.
27
-------
TABLE 2. GAC INSTALLATIONS IN MUNICIPAL WATER PLANTS
IN THE UNITED STATES (partial list)
Location
E. St. Louis, IL*
Granite City, IL*
Peoria, IL*
Kokomo, IN*
Muncie, IN*
Richmond, IN*
Terre Haute, IN*
Davenport, IA*
Lexington, KY*
Ashtabula, OH*
Marion, OH*
New Castle, PA*
Pittsburgh, PA*
Chattanooga, TN*
Hopewell, VA*
•Huntington, WV*
Madison, WV*
Princeton, WV*
West on, WV*
Montecito Co. Water Dist.
Santa Barbara, CA
Del City, OK
Somerset, MA
Pawtucket, RI
Lawrence , MA
Pi qua, OH
Bartlesville, OK
Winchester, OH
Mt . Clemens , OH
Manchester, NH
passaic, NJ (pilot)
Cincinnati, OH (pilot)
Queensbury, MA
Amesbury, MA
Bed depth, in.
18
18
30
20
24
24
30
24
14
24
18
30
30
30
60
: 30
18
18
18
144
36
11
18
24
30
18
24
24
24
48
Capacity, mgd
35
13
30
11
8
8
6
30
20
4
7
8
86
72
3
24 - ' --'. ,'
1
3
1
1.5
5.25
4.5
24
10
8
4.5
1.5
7
15
2.2
*19 American Water Works Service Co. system locations presently using GAC,
Total capacity, 370 mgd.
28
-------
SECTION 3
USE OF GAC IN FOOD AND DRINK PROCESSING AND IN OTHER INDUSTRIES
A partial list of carbon reactivation furnace installations in the United
States contains many of the largest GAC users in the country (Table 3).
Included are seven large sugar refiners and nine large corn syrup refiners who
use GAC for decolorizing their food products. The list also includes 15 indus-
tries that use GAC to remove organics from their wastewater discharges and 19
cities that use GAC in the advanced treatment of their wastewaters.
GAC has also been used for many years in industrial water.purification and
in'the production of soft drinks, Pharmaceuticals, fats and oils, and alcoholic
beverages. From 1935 to 1960, the utilization of decolorizing grades of GAC
increased from 11,000 to 60,000 tons per year (API, 1969). An even greater use
of GAC since 1960 is due in .part to the introduction of GAC to municipalities'
and sewer districts' wastewater (see Section 4).
29
-------
TABLE 3. CARBON REACTIVATION FURNACE INSTALLATIONS (partial list)
Installation
Date
Use
Colorado Springs, CO
Rocky River, OH
Derry Township, PA
Vallejo, CA
Santa Clara V.W.D., Palo Alto, CA
Tahoe-Truckee San. Dist., CA
No. Tonawanda, NY
Nassau County, NY
South Tahoe P.U.D., CA
Orange County (CA) Water District
Fitchburg, MA
Arlington County, VA
Niagara Falls, NY
Lower Potomac Plant, VA
St. Charles, MO
San. Dist. of L.A. County (CA)
Courtland, NY
Le Roy, NY
Washington Suburban San. Comnu (DC.)
Occoguan, VA
Hollytex Carpet Mills, PA
BP Oil, NH
Stepan Chemical Co., NY
Hercules, MI
Amerada Hess, NJ
American Aniline, PA
American Cyanamid, NJ
Esso Research
DuPont Deepwater, NJ
Republic Steel Corporation .
Atlantic Richfield, Wilmington, CA
Mobay Chem., New Martinsville, W.VA
Mobay Chem., Baytown, TX
Reichold Chemical
TRA, Irving, TX
Spreckles Sugar, Mendota, CA
Spreckles Sugar, Woodland, CA
Florida Sugar Refining, FL
Supreme Sugar Co., LA
Spreckles Sugar, AZ
Spreckles Sugar, Manteca, CA
Corn Siweeteners, Cedar Rapids, IA
Industrial Sugar Co., St. Louis, MO
Pennick & Ford, Ltd., Cedar Rapids, IA
Rhodia Inc., Portland, OR
Phizer Inc., Groton, CT
South Coast Corp., L.A., CA
Anheuser Busch Inc., Lafayette, IN
American Maize Products, Hammond, P.*
Clinton Corn Proc. Co,, Montezuma, KV
Cargill Inc., Dayton, OH
PepsiCo, Long Island, NY
1969
1972
1974
1974
1975
1976
1976
1977
1965
1972
1972
1977
1977
1977
1977
1975
1975
1975
1971
1978
1969
1971
1972
1972
1973
1973
1977
1973
1974
1974
1970
1-9-72
1973
1974
1976
*
*
ft
*
*
*
*
*
*
*
*
*
*
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Dye wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater .
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Sugar refining
Sugar refining
Sugar refining
Sugar refining
Sugar refining
Sugar refining
Corn syrup refining
Sugar refining
Corn syrup refining
Corn syrup refining
Corn syrup refining
Corn syrup refining
Corn syrup refining
Corn syrup refining
Corn syrup refining
Corn syrup refining
Sugar refining
Municipal
Municipal
Municipal
Municipal
Municipal
Municipal
Municipal
Municipal
Municipal
Municipal
Municipal
Municipal
Municipal
Municipal
Municipal
Municipal
Municipal
Municipal
Municipal
Municipal
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
Industrial
*Date of installation not available.
30
-------
SECTION 4
EXTRAPOLATING GAG USE IN MUNICIPAL WASTEWATER RECLAMATION TO WATER TREATMENT
SOUTH LAKE TAHOE
The first plant-scale use of GAC in a municipal wastewater treatment plant
was at South Lake Tahoe, California, in 1965. This plant has operated continu-
ously since that time and now has 15 years of operating experience with GAC;
the GAC system has processed more than 12 billion gallons of pretreated munici-
pal wastewater. The reclaimed water COD ranges from 10 to 30 mg/L. The South
Tahoe installation was a United States Environmental Protection Agency (EPA)
Demonstration Plant. For three years, EPA funded the collection of very
detailed and complete plant operating data and cost information (Appendix, Case
History No. 1). • •
Except for a few months in 1977-1.978 when some carbon was reactivated off-
site during an emergency, all carbon has been reactivated at the plant with the
use of a multiple hearth furnace. Carbon losses in reactivation have averaged
about 8 percent. This means that all of the original carbon has now been
replaced by makeup carbon. More significantly, plant experience has now veri-
fied the results of bench-scale tests of reactivation (made in 1963)—tests
indicating that the GAC could be maintained at or near full adsorptive capacity
by means of thermal reactivation until the carbon was eventually all replaced
by making up losses from attrition and burning. It has been demonstrated that
the GAC has a service life of at least 12 cycles of reactivation under actual
full-scale plant operating conditions.
At South Tahoe, the carbon furnace refractories were all replaced after 10
years of service. In 1977-78, due to maladjustment of the burners in the react-
ivation furnace, excess oxygen was present during reactivation of a consider-
able volume of GAC. This carbon was overburned and suffered a dramatic loss
(over 50 percent) in adsorptive capacity. Fortunately, it was possible to fully
restore the adsorptive capacity of the carbon in one pass through the furnace,
once the burners were re-adjusted properly. During this emergency, some off-
site reactivation of GAC was necessary. This is the only major difficulty
encountered in 15 years of operation of the GAC system. Corrosion of the
exhaust gas scrubbe r has been somewhat of a problem, especially when breakpoint
chlorination is practiced before GAC adsorption. Exhaust gas scrubbers should
be fabricated from 316 SS (Stainless Steel) where such corrosive conditions
prevail. Also the 50 psi, 304 SS carbon column inlet and outlet screens are
being replaced by 100 psi, 316 SS over a period of years due to corrosion and
mechanical distortion. The average service life of the original screens will
probably exceed 15 years by the time all are replaced.
31
-------
OTHER WASTEWATER INSTALLATIONS
Water reclamation plants constructed at the Orange County (CA) Water Dis-
trict (Water Factory. 21), the Upper Occoguan (VA) Sewer Authority, and the
Tahoe-Truckee (CA) Sanitation Agency (Nos. 6, 3, and 2, respectively) have the
same configuration as the South Tahoe plant both in respect to the type of GAC
facilities provided and the high degree of pretreatment afforded. All of these
plants operate successfully with few GAC system problems. As might be expected
with second and .third generation designs, these later plants embody some
improvements over the original South Tahoe installation, although there are no
major changes or deviations.
Although there are many other successful applications of GAC in advanced
waste treatment (AWT) plants, the experience with GAC in AWT in some places
has, unfortunately been poor. These failures in AWT applications have not stem-
med from deficiencies in the basic GAC processes or in organics adsorption and
thermal reactivation, but, rather, from mechanical problems.
OPERATIONAL PROBLEMS
In discussing the operational problems encountered with GAC systems, those
problems associated specifically with sewage treatment must be distinguished
from general problems that might be encountered with any type of GAC system.
Many problems with GAC in wastewater treatment will not occur in water purifi-
cation. For example, in water treatment,. fewer or no problems could be
expected with excessive slime growths, hydrogen sulfide gas production, or
corroision from adsorbed organics released during carbon regeneration.
Some of the types of problems encountered with GAC systems in wastewater
treatment include:
a inadequate carbon .transfer and feed equipment,
» undersized slurry and transfer lines,
« failure to provide for venting air from backwash lines with destruc-
tion of filter bottoms and disruption of GAC,
© failure to house or otherwise protect automatic control systems, from
the weather,
a inadequate means for continuous, uniform feed to furnace, which
result in temperature fluctuations, inconsistent reactivation
efficiency, and wasted energy,
ID location of furnace and auxiliary drive motors in areas of very high
ambient temperature (e.g., above top of furnace), and
IB the use of nozzles in filter and carbon contactor bottoms, which
produced major failures in carbon systems just as they have for many
years in water filtration plants. Their use is risky.
32
-------
Common problems related to wastewater treatment have been growth of bio-
logical organisms in the carbon contactors and development of anaerobic condi-
tions with the production of corrosive hydrogen sulfide. Both of these problems
have been successfully circumvented by providing adequate flow through the col-
umns or frequent backwashing. Failure to provide adequate pretreatment has
caused column clogging and mud balls with the need for more frequent
backwashing.
Corrosion has been a proolem with some of the GAC systems. The furnace
system, transfer piping/ and storage tanks are susceptable components. Many
operations require frequent replacement of the rabble arms and teeth and
replacement of the hearths every few years. At one installation, titanium or
ceramic coated rabble teeth were no more resistant to corrosion than stainless
steel teeth. In one case, the. corrosion problem in the furnace was solved by
eliminating the use of auxiliary steam during reactivation. In another case,
corrosion was linked to fluctuating temperatures in the hearths caused by
irregular feed to the furnace and frequent startup and shutdown. These problems
can be partially remedied by better operation and avoided by better engineering
design.
In several industrial applications, the wastewater itself has been highly
corrosive. In these cases, the contactors have been subject to corrosion. At
Spreckles Sugar, the epoxy linings in the columns must be replaced every 3
years; Republic Steel also replaces its column linings on a regular basis.
Public water supply sources would not be expected to consist of corrosive
water.
By properly applying the best current engineering design knowledge and
practices for GAC systems, these rather serious problems might be avoided. When
water works engineers apply GAC to produce high quality drinking water, they
should make the most of the experiences of the consultants for industry and
wastewater agencies.
EXTENDING WASTEWATER TREATMENT EXPERIENCE TO WATER TREATMENT SYSTEMS
Caution must be observed in extrapolating GAC cost data from operating
industrial installations and municipal wastewater treatment plants to the
design of water works. The purpose for using GAC in each of these types of
applications is generally the same—to remove organics. There are, however,
important differences. In industry, the GAC serves to remove a rather narrow
band of organics—color molecules—from a viscous liquid. In wastewater treat-
ment, the GAC remoT es (with or without biological activity) a broad spectrum of
organic substances from water as measured by BOD, COD, and TOC. In water treat-
ment, the objectives of GAC treatment are not completely defined at this time.
For raw waters with color or taste and odor problems, using GAC unquestionably
improves drinking .water from an aesthetic standpoint. In many cases, the cost
of GAC may be warranted for either of these purposes alone. For the great
number of water systems without color or taste and odor problems, the only
concern with respect to organics is the possible health effects over long peri-
ods of time from ingesting trace quantities of organics that may cause cancer.
33
-------
Public health officials and water works managers still disagree as to
whether the health risks that may be involved in the presence of minute traces
of organics in drinking' water are sufficient to warrant the cost of GAC treat-
ment. A major problem is that the potentially harmful organics in drinking
water have not all been identified at this time, and many of those that have
been tagged as suspect have widely different adsorptive characteristics. Some
adsorb readily on GAC; others do not.
GAC loading rates at exhaustion of adsorptive capacity vary widely among
the different potentially hazardous organics. This affects the length of ser-
vice life of GAC before carbon reactivation or replacement is necessary—a
determining factor in GAC treatment costs. Similarly, the reactivation times
and temperatures for thermal reactivation of GAC saturated with different
organics also differ, and all are not known at this time. Again, this has an
important bearing on GAC treatment costs. Because of the widely varying adsorp-
tive and reactivation characteristics of trace organics on GAC in water
supplies, pilot plant tests of both adsorption and reactivation are mandatory
preludes to treatment system design at this time.
Over a period of years, general, average design parameters may emerge from
the results of pilot plant studies and demonstration projects, but this time is
not yet at hand. Once the GAC design parameters for water treatment have been
established from pilot tests for a particular water source, then the knowledge
and experience from other GAC installations in industry and wastewater plants
can be put to good use. Carbon dosages, GAC contact times,- and spent carbon
reactivation times and temperatures can be determined. Contactor sizes .can be
calculated and furnace sizes and fuel requirements can be determined. Transport
facilities -for. GAC in water treatment can be the same as for other types of GAC
installations provided the differences in quantities and possible differences
in the viscosities of carbon slurries due to any slime growths are taken into
account. Also, with the GAC design parameters pinpointed as a result of pilot
plant studies, construction costs can be accurately estimated based on the
costs of existing installations in AWT and industry. The estimates cannot,
however, be based on an MGD capacity basis; rather, they must be based on
adsorption and reactivation data applicable to each specific installation.
Selecting the most economical number of contactors for a water system of a
certain size involves the same principles as are used for other systems.
Because of shipping regulations, factory-fabricated contactor vessels are gen-
erally limited to about 12-ft maximum diameter. For large capacity installa-
tions, a smaller number of field-erected steel vessels or poured-in-place
concrete vessels may be less costly.
Because upflow contactors provide all of the advantages of countercurrent
operation with respect to carbon savings, they are favored for most types of
service. The exception is water treatment. In this case, downflow is used
because the discharge of carbon fines in the effluent (a characteristic of
upflow columns) is avoided.
Cost estimates must be evaluated on the same basis as all other estimates
of construction cost; there is no reason that they should be more or less
34
-------
accurate than estimates made for the rest of the treatment plant. Fifteen
percent is generally accepted as being an allowable difference between costs
estimated from construction plans and the best bid received from contractors.
With good pilot plant data and with proper application of cost data from exis-
ting GAC installations/ preliminary cost estimates for GAC treatment of public
water supplies should be accurate enough for planning purposes.
The extrapolation of wastewater treatment experience with GAC to the
design of water treatment systems is a task for trained, experienced, engineer-
ing professionals. Even then, the foregoing discussions are intended to be no
more than an introduction to the subject.
35
-------
SECTION 5
DESIGNING GAC ADSORPTION AND REACTIVATION SYSTEMS FOR WATER TREATMENT
The following discussion is devoted to some of the procedures and details
for developing the design basis and costs for GAC water treatment systems from
pilot plant test results and information from full-scale applications.
GAC SYSTEM COMPONENTS
Systems utilizing GAC are rather simple. In general, they provide for con-
tact between the carbon and water to be treated for the length of time required
to obtain the necessary removal of organics; reactivation or replacement of
spent carbon; and transport of makeup or reactivated carbon into the contactors
and transport of spent carbon from the contactors to reactivation or hauling
facilities.
PILOT PLANT TESTS
Despite the simplicity of GAC systems, laboratory and pilot plant tests
are needed to select the carbon and the most economical plant design for both
water and wastewater treatment projects. Pilot column tests make it possible to
determine treatabiiity; select the best carbon for the specific purpose based
on .performance; determine the required empty bed contact time; establish' the
required carbon dosage, which, together with'laboratory' tests of reactivation,
will determine the capacity of the carbon reactivation furnace or the necessary
carbon replacement costs; and determine the effects of influent water quality
variations on plant operation. During pilot testing, the influence of longer
carbon contact times on reactivation frequency can be measured; these measure-
ments allow costs to be minimized through a proper balance of these two design
factors.
DESIGN OF PILOT GAC COLUMNS
Detailed information, including a list of materials, on the design and
construction of pilot GAC columns is presented in Appendix C of EPA's "Interim
Treatment Guide For Controlling Organic Contaminants in Drinking Water Using
GAC." Appendix B of EPA's "Interim Treatment Guide" describes the analytic
methodology for monitoring pilot column tests. Also included are data on the
adsorbability of various organic compounds; the performance of GAC in their
removal; information on the use of multiple-hearth, infrared, fluidized bed,
ana rotary kiln furnaces for reactivation of spent GAC; and example calcula-
tions for balancing added costs of increased contact time versus savings (if
any) from less frequent reactivation.
36
-------
SIMILARITIES AND DIFFERENCES IN THE USE OF GAC IN WATER AND WASTEWATER
TREATMENT
FREQUENCY OF REACTIVATION
One of the principal differences in costs for GAC treatment between water
and wastewater is the more frequent reactivation required in water purification
due to earlier breakthrough of the organics of concern. In wastewater treat-
ment, GAC may be expected to adsorb 0.30 to 0.55 Ib COD/lb carbon before the
carbon is exhausted. From the limited amount of data available from research
studies and pilot plant tests (most of it unpublished), it appears -that some
organics of concern in water treatment may breakthrough at carbon loadings as
low as 0.05 to 0.25 Ib organic/lb carbon. The actual allowable carbon loading
or carbon dosage for a given case must be determined from pilot plant tests.
Costs taken from wastewater cost curves, which are plots of flow in mgd versus
cost (capital or operation and maintenance costs), cannot be applied directly
to water treatment. Allowance must be made in the capital costs for the differ-
ent reactivation capacity needed and in the operation and maintenance costs for
the actual amount of carbon to be reactivated or replaced.
Because the organics adsorbed from water are generally more volatile than
those adsorbed from wastewater, the increased reactivation frequency due to
lighter carbon loading may be partially offset, or more than offset, by the;
reduced reactivation requirements of the more volatile organics. The times and
temperatures required for reactivation may be reduced due to both the greater
volatility and to the lighter loading of organics on the carbon.
From the limited experimental reactivations to date, it appears that
reactivation temperatures may be less than the 1650° to 1750°F required for
wastewater carbons. The shorter reactivation times required for water purifica-
tion carbons may allow the number of hearths in a multiple hearth reactivation
furnace to be reduced. Also, less fuel may be required for reactivation. Theses
factors must be determined on a case-by-case basis.
GAC CONTACTORS
Selection of the general type of carbon contactor to be used for a partic-
ular water treatment plant application may be based on several consideration:;
including economics and the judgement and experience of the engineering
designer. The choice generally would be made from three types of downflow
vessels:
1. Deep-bei, factory-fabricated, steel pressure vessels of 12-ft maximum
diameter. These vessels might be used over a range of carbon volume:;
from 2,000 to 50,000 ft3.
2. Shallow-bed reinforced concrete, gravity-filter-type boxes may be
used for carbon volumes ranging from 1,000 to 200,000 ft3. Shallow
beds probably will be used only when short contact times are suffi-
cient or when long service cycles between carbon regenerations can be
expected from pilot plant test results.
37
-------
3. Deep-bed, site-fabricated, large (20- to 30-ft) diameter, open con-
crete or steel, gravity tanks may be used for carbon volumes ranging
from 6,000 to 200,000 ft3, or larger.
These ranges overlap, and the designer may very well make the final selec-
tion based.on local factors, other than total capacity, that affect efficiency
and cost.
The AWT experience with GAC contactors may be applied to water purifica-
tion if some differences in requirements are taken into account. The required
contact time must be determined from pilot plant test results. Although con-
tactors may be designed for a downflow or upflow mode of operation and upflow
packed beds or expanded beds provide maximum carbon efficiency through the use
of countercurrent flow principles, the leakage of some (1 to 5 mg/L) carbon
fines in upflow carbon column effluent make downflow carbon beds the preferred
choice; in most municipal water treatment applications. At the Orange County
Water Factory 21, upflow beds were converted to downflow beds to successfully
correct a problem with escaping carbon fines. This full-scale plant operating
experience indicates that leakage of carbon fines is not a problem in properly
operated downflow GAC contactors.
Single beds or two beds in series may be used. Open gravity beds or closed
pressure vessels may be used. Structures may be properly protected steel or
reinforced concrete. In general, small plants will use steel, and large plants
may use steel or reinforced concrete.
Sand in rapid filters has, in some instances, been replaced with GAC. In
situations • where contact times . are short and GAC regeneration or replacement
cycles are exceptionally long (several months or years) as may be the case in
taste and odor removal, this may be a solution. However, with the short cycles
anticipated for most organics, conventional concrete-box-style filter beds may
not be well suited to GAC contact. Deeper beds may be more economical in first
cost and provide more efficient use of GAC. In converted filter boxes, possible
corrosion effects of GAC on existing metals, such as surface wash equipment and
metal nozzles in filter bottoms, must be taken into account. Beds deeper than
conventional filter boxes, or contactors with greater aspect ratios of depth to
area, provide much greater economy in capital costs. The contactor cost for the
needed volume of carbon is much less. In water slurry, carbon can be moved from
contactors with conical bottoms easily and quickly and with virtually no labor.
Flat-bottomed filters of a type that require labor to move the carbon unnecces-
sarily add to carbon transport costs. The labor required to remove carbon from
flat-bottomed beds varies considerably in existing installations from a little
labor to a great deal, depending upon the design of the evacuation equipment.
For many GAC installations intended for precursor organic removal or syn-
thetic organic removal, specially designed GAC contactors should be installed.
Contactors should be equipped with flow measuring devices. Separate GAC con-
tactors are especially advantageous where GAC treatment is required only part
of the time during certain seasons because they then can be bypassed when not
needed, possibly saving unnecessary exhaustion and reactivation of GAC.
38
-------
Tremendous cost savings can be realized in GAC treatment of water through
proper selection and design of the carbon contactors. The design of carbor.
contactor underdrains requires experienced expert attention.
CARBON CONTACTOR UNDERDRAINS
Although good proven underdrain systems are available, often they have not
been used, and there have been numerous underdrain failures due to poor design.
Some of the designs used in the past have failed in many installations for con-
ventional filter service, and they continue to be misapplied to GAC contactors:
as well as filters.
GAC REACTIVATION OR REPLACEMENT
Spent carbon may be removed from contactors and replaced with virgin car-
bon, or it may be reactivated either on-site or off-site. The most economical
procedure depends on the quantities of GAC involved. For larger volumes, on-
site reactivation is the answer. For small quantities of carbon, replacement or
off-site reactivation will probably be most economical.
Carbon may be thermally reactivated to very near virgin activity. Carbon
burning losses may, however, be excessive under these conditions. Experience in
industrial and wastewater treatment indicates that carbon losses can be mini-
mized (held to 8 to 10 percent per cycle) if the activity of reactivated carbon
(as indicated by the iodine number) is held at about 90 percent of the virgin
activity. To remove certain organics, there may be no decrease in actual
organics removal despite a 10 percent drop in iodine number.
THERMAL REACTIVATION EQUIPMENT
GAC may be reactivated in a multiple-hearth furnace, a fluidized bed fur-
nace, a rotary kiln, or an electric infrared furnace. Spent GAC is drained dry
in a screen-equipped tank (40 percent moisture content) or in a dewatering
screw (40 to 50 percent moisture) before being introduced to the reactivation
furnace. Dewatered carbon is usually transported by a screw conveyor. Following
thermal reactivation, the GAC is cooled in a quench tank. The water-carbon
slurry may then be transported by means of diaphragm slurry pumps, eductors, or
a blow-tank. The reactivated carbon may contain fines produced during convey-
ance; these fines should be removed in a wash tank or in the contactor. Maximum
furnace temperatures and retention time in the furnace are determined by the
amount -(lb organics/lb carbon) and nature (molecular weight or volatility) of
the organics adsor ed.
Off-gases from carbon reactivation present no air pollution problems pro-
vided they are properly scrubbed. In some cases, an afterburner may also be
required for odor control.
Despite recent advances in the design of infrared and fluidized bed
reactivation furnaces, the multiple hearth furnace is still the simplest, most
reliable, and easiest to operate for GAC reactivation. The infrared and fluid
bed units still have problems. to be worked out; experience with the multiple
39
-------
hearth equipment has already solved these problems. Still, it is necessary with
all four types of furnaces to specify top quality materials to suit the condi-
tions of service and to see that these materials are properly installed. Cor-
rosion resistance is important in the furnace itself and especially in all
auxiliaries to the furnace.
REQUIRED FURNACE CAPACITY
The principal cost differences between GAC treatment of water and waste-
watesr may lie in the capital cost of the furnace and in the O&M cost for carbon
reactivation. As already explained, the two principal differences between car-
bon exhausted in wastewater treatment and carbon exhausted in water purifica-
tion are that water purification carbons are likely to be easier to regenerate
(less time in furnace and lower furnace temperatures) and more lightly loaded
(greater volume of carbon to be reactivated per pound of organics removed).
Accurate estimates of GAC costs require knowledge and consideration of these
two factors. To repeat, it is not possible to use AWT cost curves based on mgd
throughput or plant capacity to obtain costs for water treatment. Differences
in reactivation requirements must be taken into account.
CARBON TRANSPORT AND GAC PROCESS AUXILIARIES
There can be large differences in O&M costs for GAC systems depending on
the method selected for carbon transport. Hydraulic transport of GAC in water
slurry by gravity or use of water pressure is simple, easy, inexpensive, rapid,
and uses very little labor. Moving dry or dewatered carbon manually or with
mechanical means involving labor can be very difficult, time consuming, and
costly. . • .
40
-------
SECTION 6*
THE COST OF GAC IN WATER TREATMENT
DEVELOPING COST CURVES
Little information is available concerning the cost and performance of GAC
for drinking water treatment. As discussed in the previous sections, most of
the data available on GAC performance have been acquired from wastewater and
industrial applications. An attempt has been made, however, to extrapolate from
these existing systems and to develop standardized and flexible cost data that
can be used to prepare cost estimates for GAC systems that treat drinking
water.
Design Cost Information
Much of the analysis and cost information contained in this section are
based on a four-volume report prepared for EPA (Culp/Wesner/Culp, 1979).
Twelve cost curves, discussed .in detail in the above report, were developed
specifically for GAC applications.**
Derivation of Cost Curves
Relating Costs to Design Parameters—
The construction cost for each unit process was presented as a function of
the process design parameter. This parameter was determined to be the most use-
ful and flexible under varying conditions, such as loading rate, detention
time, or other conditions that vary because of designers preference or regula-
tory agency requirements. For example, the contactor construction cost curves
were presented in terms of cubic feet of contactor volume, an approach that
allows various empty bed contact times (EBCT) to be used. Contactor O&M curves
were presented in terms of square feet of surface area, since O&M requirements
are more appropriately related to surface area rather than contactor volume.
Reactivation facility cost curves were presented in terms of square feet of
hearth area for the multiple hearth furnance and pounds per day of reactivation
capacity for the other reactivation variables. This allows the loading per
square foot of hearth area to be varied for the multiple hearth furance and for
design of reactivation furnaces at carbon reactivation rates less than the
* Section 6 was pi spared and written by Dr. Robert M. Clark, EPA, Cincinnati,
Ohio.
**Cost curves were developed for 12 GAC processes (Culp/Wesner/Culp, 1979):
Large systems, 1 to 200 mgd, include gravity carbon contactors (concrete
construction); gravity carbon contactors (steel construction); pressure car-
bon contactors; conversion of sand filters to carbon contractors; granular
activated carbon (material cost); multiple hearth granular carbon reactiva-
tion; infrared granular carbon reactivation; fluid bed granular carbon
reactivation; atomized suspension powdered carbon reactivation; fluidized
bed powdered carbon reactivation; and off-site regional carbon reactivation.
Small systems, 2,500 to 1,000,000 gpd, include package granular activated
carbon columns.
41
-------
reactivation capacity. This approach provides greater flexibility in the use of
the cost curves than if the costs were related to water flow through the treat-
ment plant. .
Methods Used to Develop Cost Curves—-
The construction cost curves were developed by using equipment cost data
supplied by manufacturers, cost data from actual plant construction, unit take-
offs from actual and conceptual designs, and published data. The costs (devel-
oped by Culp/Wesner/Culp) were then check by a second consulting engineering
firm (Zurhe-ide-Herrmann, Inc.), using an approach similar to an approach used
by a general contractor in determining a construction bid. Any discrepancies
existing between the initial estimate and the second estimate were then
resolved.
Construction Cost Components—
The costs for eight principal construction components were developed and
then aggregated to give the construction cost for each unit process. These
eight principal components were: (1) excavation and sitework, (2) manufactured
equipment, (3) concrete, (4) steel, (5) labor, (6) pipe and valves, (7) elec-
trical and instrumentation, and (8) housing» These eight categories also pro-
vide enough detailed information to permit accurate future cost updating.
The construction cost curves are not the final capital cost for the unit
process. The construction cost curves do not include costs for general con-
tractor overhead and profit, engineering, legal, fiscal, and administration
interest during construction. These items are all more directly related to the
total cost of a project, rather than the cost of the individual unit processes.
They, therefore, are.most appropriately added following summation of the cost
of the individual unit processes, if more than one unit process is required.
Example calculations are included below to illustrate the recommended method
for these costs.
Operation and Maintenance Components—
O&M requirements were developed for (1) building related energy, (2)
processs energy, (3) maintenance material, and (4) labor. The separate determi-
nation of building energy makes it possible to allow for geographic variations
in this component. Energy requirements were presented in kilowatt-hours per
year for electricity, standard cubic feet per year for natural gas, and gallons
per year for diesel fuel. Labor was presented in hours per year, allowing local
cost variations to be incorporated into the O&M cost calculations. Maintenance
material cost was in dollars per year, but did not include the cost of chemi-
cals. Chemical costs were added separately, as will be shown in the example
presented later.
Example Cost Curves
For the majority of the unit processes, three separate figures were used
to present construction and O&M curves. The first graph presented the construc-
tion cost; the second graph, energy (electrical, natural gas, and diesel fuel)
and maintenance material requirements; and the third graph, labor requirements
and total O&M cost. Figures 1, 2, and 3 present the cost curves developed for
pressure carbon contactors.
42
-------
en
C
3
_i
o
o
I
I-
8
o
o
e
1.000.00C
8
§ 100,0
8
100 2 3456789 1000 2 34567 8 9 10,000 2 3 456789
INDIVIDUAL CONTACTOR VOLUME - ft 3
10 . 100
INDIVIDUAL CONTACTOR VOLUME - m 3
1000
Figure 1. Construction cost for pressure carbon contactors
(from Gulp, Wesner, Gulp, 19791..
43
-------
cr
LJ
in
o
\
til
9
8
7
6
5
4
3
2
100,01
I
7
6
5
4
3
2
10,0
9
8
7
6
9
4
-s
z
1..C
9
B
7
6
5
4
3
2
9
8
7
6
5
4,
3
2
)0 1.0C
" 9
8
7
6
5
4
3
- -| 2
J
DO I 10
o
ENERG\
4 * WO>HOX»^ N 01 * WfflHOW)
2
. 1,000
0,000
•»
0,000
— •
,000
.-=
^
/
^_
S
/
*"
X
x
S*
4?
^
/
X
x
s
'
X
^
/
s
/
/
_4
s
s
/
J* —
S
X
X
X
x
/
/
/
/
*
f
/
/
/
A
*
f
/
/
I
BUILDIN
MAINTE
MATERI
GENE
NANC
AL
ERG
E
V
_
PROCESS ENERGY
100 2 3456789 1000 2 34 5 6 7 8 9 io.OOO 2 * 456769
TOTAL CONTACTOR AREA - ft*
_l 1 1
10
_ . .
' TOTAL CONTACTOR AREA - m2
1000
Figure 2. Typical activated carbon column, installation Cfrom
Gulp, Wesner, Gulp, 1979).
44
-------
9
8
7
6
5
4
3
2
1.000.
9
8
' 7
6
5
4
3
2
100.
$ e9
£ 7
03 &
i 5
I 4
i *
o
< 2
K
O'
10.C
9
6
7
6
5
4
3
2
1.000
- 9
~ 8
7
6
- 4
2
300
9
7
6
5
4
2
300 100,
~ _ 6
10 | |
LABOR - 1
P N W •»
- I
6
- 5
- 4
- 1.000
DOO
^^
300
.
^-«
**'
— — •
-*"
. —
«*^
.— •
*
^
,—
^
^>
^
^
-•
«*
i
Ix''
^^T
^^
^^
s"
J?
X"^
x^
l
,x
s
/
/
/
t
/
-~ 7
x"
. TOTAL
y
LABOF
COST
(
— [~
I
100 2 34 567691.000 2 34 5678910.000 2 3 456769
TOTAL CONTACTOR AREA - tt2
10 100 1,000
TOTAL CONTACTOR AREA - m2
Figure 3. Operation and maintenance requirements for pressure carbon
contactors - labor and total cost.
45
-------
Updating Cost Curves
Use of Indices for Updating Construction Costs—•
For many engineering purposes, a single cost index, such as the Engineer-
ing News ^ Record's (ENR) Construction Cost Index (CCI) may be used to update
construction costs. When this approach provides sufficient accuracy, a single
index is certainly the simplest and easiest • approach. In the case of water
treatment construction costs, especially those for carbon treatment, use of the
CCI is not recommended because the construction <. mponents used in its
development are not related to those required for water treatment plants. For
example, the CCI does not include process equipment or electrical equipment for
instrumentation and control. For water treatment plants, the use of several
indices is recommended such as the Bureau of Labor Statistics (BLS) (Bureau of
Labor Statistics, 1979), indices in conjunction with ENR Building Cost Index
and Skilled Labor Wage Index.
The tabulation below shows the indices recommended for each of the eight
categories used to derive construction costs. This approach allows updating to
be done proportionally to changes in the cost of each component. The eight con-
struction cost components represent the major items of material and labor
affecting the cost of water treatment plant construction.
CONSTRUCTION COST COMPONENT INDEX
Excavation and Sitework ENR Wage Index (Skilled Labor)
Manufactured Equipment . BIS General Purpose
Machinery and Equipment
' Code '114
Concrete BLS Concrete Ingredients -
Code 132
Steel BLS Steel Mill Products
Code 1013
Labor ENR Wage Index (Skilled Labor)
Pipes and Valves BLS Valves and Fittings
Code 114901
Electrical and Instrumentation BLS Electrical Machinery and
Equipment - Code 117
Housing ENR Building Cost Index
The principal disadvantages of this approach are the lack of geographical
specificity of the BLS indices and the extra effort involved in using seven
rather than a single index.
46
-------
TABLE 4. COST CALCULATION AND OPERATION AND MAINTENANCE REQUIREMENTS FOR A 100. MGD PRESSURE GRANULAR
ACTIVATED CARBON PLANT (from Culp/Wesner/Culp, 1979)
System and design criteria
Gravity, Steel Contactor -
18 min. EBCT, 30' dia, !
gpm/ft2
Initial Carbon Charge -
26 lb/ft3
Backwash Pumping -
12 gpm/ft2
Multiple Hearth Carbon
Reactivation Furnace
Make-up Carbon, 6% regen.,
5 times/yr
Subtotal
Sitework', Interface Piping,
Roads @ 5%
Subsurface Consideration
Standby Power
Total Construction Cost
General Contractor's Over-
head and Profit
Subtotal
Engineering @ 10%
Subtotal
Land, 20 acres @ $2,000/acre
Legal, Fiscal, and Admin-
istrative
Interest During Construction -
8%
Design Construction Operating
parameter cost - $ parameter
11,489 ft3/ $ 4,503,440 183,824 ft3
Contactor
4,779,400 Ib 2,746,810
7,520 gpm 101,710
1,173 ft2 2,467,640 1,173 ft2
1,433,820
0 Ib/yr
$ 9,819,600
490,980
0
0
$10,310,580
927,950
$11,238,530
1,123,850
$12,362,380
40,000
79,070
1,108,240
Energy Natural gas Maintenance Labor
kwhr/yr scf/yr material-$/yr hr/yr
3,720,810 0 21,590 8,362
0 000
0 000
1,111,310 : 158,149,050 12,740 13,824
0 0 857,740 0
4,832,120 158,149,050 892,070 22,186
—
—
— — »— — — __
_— _ — _ _.
..
TOTAL CAPITAL COST
$13,589,690
-------
TABLE 5. ANNUAL COSTS FOR 110-MGD GRAVITY, STEEL GAC PLANT
: (from Culp/Wesner/Culp, 1979)
ilcein . Total cost/year
Amortized capital (§8%, 20 yrs ' $ 1,384,110
Labor, 22,186 hr @ $10/hr (total labor costs
including fringes and benefits) 221,860
Electricity, 4,832,120 kwhr @ $0.03/kwhr 144,960
Natural gas, 158,149,050 scf @ $0.00175/scf 276,760
Maintenance material 892 ,070
TOTAL ANNUAL COST* $ 2,919,760
$2,919,760 (100)
*Cents per 1,000 gal treated =
- 10.4
-------
The cost analysis contained in this section will include both situations—
the use of GAC as a replacement for existing filtration media (sand replace-
ment) and as a separate adsorption system (post-filter adsorption). On-site
multiple hearth reactivation will be assumed. Standard levels of key design
parameters will be set at predetermined levels and then one variable at a time
will be changed to determine its effect on system cost.
The need to consider the cost of separate GAC contactors is eliminated if
GAC is replacing sand in existing filters. For the purposes of the sand
replacement analysis, a water treatment plant is assumed to consist of an inte-
gral number of 1 mgd filters with an effective volume of 18.5 ft x 18.5 ft x
2.5 ft or 856 ft . Design parameters assumed for the sand replacement systems
are shown in Table 6, and design assumptions for post-filter adsorber systems
are shown in Tables 6 and 7.
Note that for sand replacement, a GAC loss of 10 percent per reactivation
cycle is assumed, but a GAC loss of only 6 percent per cycle is assumed for
post-filter adsorbers. These two assumptions are intended to reflect differ-
ences in the operation of the- two systems. Sand replacement systems are labor
intensive and increase the possibility of GAC loss because the activated carbon
is changed manually. In post-filtration systems, the activated carbon is
assumed to be changed hydraulically, leading to fewer possibilities for han-
dling losses. Table 8 shows the unit costs for both types of systems for 1, 5,
10, 100, and 150 mgd plants at 70 percent capacity.
The costs shown in Table 8 can be further divided into reactivation costs
for sand replacement systems and into reactivation and contactor costs for the
post-filter adsorbers. Table 9 contains amortized capital and O&M costs for
sand replacement and post-filter adsorption systems. As can be seen, a large
portion of the capital cost for both processes is associated with reactivation.
Post-filter adsorption is more capital intensive in general than is the sand
replacement mode.
Sensitivity Analysis
In this section, the effect of changing variable levels will be consid-
ered. Standard values shown in Table 6, which yield the unit costs shown in
Tables 8 and 9, will be assumed. The effect of changing the design parameter
around the standard values will be examined for GAC sand replacement and post-
filter adsorbers separately (Clark et al, 1977). A comparison will be made
between 100 mgd, sand replacement, and filter-adsorption systems.
Figure 4 shows the cost of both sand replacement and post-filter adsorp-
tion systems versu . plant capacity,* based on the assumptions in Tables 7 and
8. Figures 5 and 6 show amortized capital costs and the O&M for both system
types, respectively. Significant scale economies exist with respect to plant
capacity. Similar relationships hold with respect to loading rate as well. Unit
costs will rise dramatically for a plant of a given capacity as loading rate
decreases. Figure 7 shows the effect of reactivation frequencies on a total
*Note: The proposed regulation (Part B) was to apply to utilities serving
75,000 people or more (approximately 10 mgd) (Culp/Wesner/Culp, 1979).
49
-------
TABLE 6. DESIGN PARAMETERS FOR GRANULAR ACTIVATED CARBON
Design parameter
Unit of measure
Sand replacement:
Activated carbon cost
Activated carbon loss per reactivation cycle
Fuel cost
Electric power cost
Construction Cost Index
Producers Price Index
Direct hourly wage rate
Amortization rate
Amortization period
Volume per filter
Loss in adsorptive capacity
Design capacity
Empty bed contact time
Reactivation frequency
Post-filter adsorption:
Activated carbon cost
Activated carbon loss per reactivation cycle
Fuel cost
Electric power cost
Construction Cost Index . '
Producers Price Index
Direct hourly wage rate
Amortization rate
Amortization period
Loss in adsorptive capacity
Design capacity
Empty bed contact time
Reactivation frequency
?0.65/lb
10%
$ 1.75/million Btu
($1.66/th J)
$0.04/kwhr
325
243.8
$10/hr
8%
20 yrs
856 ft3
0%
70%
9 min
Every 1.2 months
$0.65/lb
6%
$1.75/million Btu
($1.66/th J)
$0.04/kwhr
325
$10/hr
8%
20 yrs
0%
70%
18 min
Every 2.4 months
TABLE 7. ASSUMPTIONS FOR SEPARATE POST-FILTRATION SYSTEMS
Design capacity, mgd
Item
Number of contactors
Diameter of contactors (ft)
Depth of contactors (ft)
Volume of GAG per
contactor (ft3)
Minimum empty bed contact
time ( min )
1
3
8
13
653.1
18
5
6
12
13
1469.5
18 .
10
12
12
13
1469.5
18
100
40
20
14
4396.0
18
150
60
20
14
4396.0
18
50
-------
TABLE 8. AMORTIZED CAPITAL AND OPERATING AND MAINTENANCE COST FOR GAC SYSTEM
(
-------
<
o
5 -
0
SAND REPLACEMENT
POST-FILTER ADSORPTION
20
40
60
~r~
120
T~
140
160
180
80 100
PLANT CAPACITY . mgd
Fiyure 4. Total production cost versus plant capacity for post-filter adsorption and sand replacement
systems.
-------
U1
SAND REPLACEMENT
POST-FILTER ADSORPTION
Figure 5.
PLANT CAPACITY, mgd
Capital cost versus plant capacity for post-filter adsorption and sand replacement systems
-------
30 ~
25-
o
01
- 15 -
\n
O
U
•6
O 10
20
T
40
SAND REPLACEMENT
POST-FILTER ADSORPTION
~] - 1 - 1 - 1 - 1 - 1
60 80 100 120 140 160
PLANT CAPACITY, mgd
180
Fi.qure f>. OSM cost versus plant capacity for post-filter adsorber and sand replacement systems.
-------
01
Ui
35 -,
30 -
25 -
0>
8
20 -
o
u
O
U
15-
10 -
POST-FILTER ADSORPTION
SAND REPLACEMENT
Fiqure 7.
Total production cost versus
and sand replacement systems
REACTIVATION FREQUENCY, month*
reactivation frequency in months for post-filter adsorption
(100 mqd)-
-------
production 'cost for both post-filter adsorption and sand replacement systems—
cost rises dramatically with increased reactivation.
Operating and Maintenance
Some operating variables influence only O&M cost, some only amortized cap-
ital cost, and some both O&M and capital cost. The first set of variables to be
examined are those that influence O&M cost only: hourly wage rate in $/nr;
activated;carbon loss per reactivation cycle in %; fuel cost in $/therm; Pro-
ducers Price Index; and electrical power cost in $/kwh. This analysis is per-
formed for 1, 5, 10, 100 and 150 mgd systems operating at 70 percent of
capacity.
Figure 8 shows that changes in hourly wage rate have a greater impact on
O&M cost in small plants than in larger plants. For example, it can be seen
that as the hourly wage rate increases from $6/hr to $10/hr, the O&M cost for a
1 mgd plant increases from slightly over 17^/1,000 gal to slightly less than
19.5ifc/1,000 gal. The same wage rate changes in a 10 mgd plant increases the O&M
cost from approximately 11
-------
35 -i
30 -
25 -
o
o
0>
g 20 -
in
-j
o
01
8
U
3
00
o
15 -
10 -
I
10
T
12
14
1 MGD
5 MGD
150 MGD
I
16
I
18
DIRECT HOURLY WAGE RATE. dollarsAour
l-'Jk)Ui:i' 8. nf,H cost: versus waqo rnt.o Oir 1, '">, in, 1(10, and 150 mqd sand rei^lacement systems.
-------
IJ\
o>
35_
30-
25-
o
01
8
c
0)
u
8
O
•s.
15-1
10-
5-
2.5
5.0
7.5
—r~
10.0
12.5
15.0
~r~
17.5
1 MGD
150 MGD
20.0
1
22.5
CARBON'LOSS PER REACTIVATION CYCLE. %
Fiqure 9- O&M cost versus carbon loss per reactivation cycle for 1., 5, 10, 100, and 150 mgd sand
replacement systems.
-------
in
vo
35-,
30 -
25-
o
en
I 20 -j
II
u
5 -
1 MGD
5MGD
150 MGD
1 1 1 1 1 1 1 1 I
U 05 10 15 20
FUEL COST, dollars/therm
P'iqure .10. ORM cost versus fuel cost in $/therm Tor .1, 5, JO, 100, and 150 mqd sand replacement
systems.
-------
o>
o
a
en
O
U
35 _,
30 -
25 -
20
15 -
10 -
5 ~
0 5
T~
1.0
I
3 5
1MGD
5MGD
10MGD
100 MOD
150 MGD
T-
4.0
Figure 1.1.
'-•- 20 25 3.0
PRODUCERS PRICE INDEX (x 100)
O&M cost versus producers price index for 1, 5, 10, 100 and 150 myd sand replacement
systems .
4.5
-------
35
30 -
25 -
20 -
H 15 -
O
U
Ofl
0
10 -I
5 -
.01
T
.02
"T"
06
07
1 MGD
5MGD
10 MGD
100 MGD
150 MGD
08
Fiqure 12.
.03 04 05
POWER COST, dolIorsAilowatt hour
OSM cost versus power cost for 1., 5, .1.0, 100, 150 mgcl sand replacement systems.
.09
-------
35 _,
30 -
25 -
o
o>
20 -
8 15
U
OL
U "'
5 -
1 MGD,
150 MGD
r
1.0
I
1.5
I
2 0
I
2.5
I
3.0
I
3.5
I
4.0
I
4,5
Figure 13.
CONSTRUCTION COST INDEX (x 100)
Capital cost versus construction cost index for 1, 5, 10, 100, and 150 mgcl sand
replacement systems. •
-------
35 -,
30 H
g 25
_o
~o
20 H
01
u
8 '5
<
51
U
10 H
5 -A
T
1MGD
to
14
5MGD
150 MGD
16
18
INTEREST RATE, %
F.i.cjure 14. Capital cost versus interest rate for I, 1>, 10, 100, and 150 mgd sand repl acement systems.
-------
35 _
30 -
25 -
c
_o
~o
en
o
8
c:
v
8'5
_J
<
H
CL
u
to -
5 -1
10
20
T
30
T
1 MGD
5MGD
10 MGD
100 MGD
150 MGD
40
AMORTIZATION PERIOD, years
Fiqure 15. Capital cost versus amortization for 1, 5, 10, 100, and 150 mcjd sand replacement system.
-------
in
o
o>
35 -
30 -
25 -
20 -.
4)
U
8
U 15
Q.
5
10 -
5 -
,2
1
,4
1
3 ,4 5 6
CARBON COST, dollars/lb
T~
,7
1 MGD
5MGD
10 MGD
100 MGD
150 MGD
1
.9
riqure 16. Capital cost versus carbon cost for I, 5, 1.0, 100, and 150 mgd sand replacement systems.
-------
01
30 -
25 -
in
c
JJ
~0
cr>
§20 H
u
10 -
5 -
T
.2
T
T
.345
CARBON COST, dollars/I b
.7
1 MGD
150 MGD
,9
K.icpire 1.7. os,M cost versus carbon cost for 1, 3, .10, 100, and 150 mgd sand replacement system.
-------
35 -,
30 -
25-1
171
8
o
O
O
S15
_l
<
QL
U 10
5 -
1 MGD
5MGD
I
REACTIVATION FREQUENCY, months
•'igure .1.8. Capital cost versus reactivation frequency for 1, 5, 10, 100, and 150 mgd sand replacement
systems.
-------
cr>
00
35
30 H
25 J
c
o
a
o>
o
8 20
o
to 15
O
U
10 -
5 -1
1 MGD
150 MGD
o
i
T
T
T
Fiqure 19.
23456789
REACTIVATION FREQUENCY, months
OSM cost versus reactivation frequency Cor .1., 5, 10, TOO, and .150 mcjd sand replacement
system.
-------
insensitive to power cost and Producers Price Index. Power cost has a greater
impact on smaller systems than do some of the other variables.
Capital Cost Effects—
Plant capacity, amortization period, Construction Cost Index, and interest
rate have a significant impact on system cost. System costs are relatively
insensitive to activated carbon cost. Capital costs are moderately sensitive to
reactivation frequency.
SUMMARY AND CONCLUSIONS
Extensive development of curves for various unit processes involved in GAC
operations was done in a previous report (Culp/Wesner/Culp, 1979). These curves
are designed to be flexible in application and easily updated. Based on these
data, an analysis of the cost of GAC used in water supply has been completed.
The sensitivity of GAC system costs to changes in certain operating variables
has been examined. Several variables have been identified as important. These
include choice of system configuration, loading rate and size of system, reac-
tivation frequency, interest rate and life of system, local construction and
operating costs, inflation, and carbon use rate.
69
-------
SECTION 7
REPORT SUMMARY
On January 9, 1978, the U.S. Environmental Protection Agency (EPA) pro-
posed the use of Granular Activated Carbon (GAC) as a means of treating drink-
ing water. Since that time much has been written both for and against using GAC
in this manner. Serious challenges and many questions have been raised regard-
ing EPA' s cost estimates for GAC use. To respond to some of these questions,
EPA's Drinking Water Research Division initiated a carefully designed study to
establish water supply unit process cost curves on a consistent and understand-
able basis.
A study performed by Culp/Wesner/Culp under EPA Contract No. 68-03-2516,
focused on processes capable of removing those contaminants included in the
National Interim Primary Drinking Water Regulations and developed construction
and operation and maintenance cost curves for these processes. Its final
report for this project contains cost curves for 99 different unit processes.
These cost curves were divided into categories: large water treatment systems
applicable to flows between 1 and 200 mgd and small water treatment systems
applicable to flows between 2,500 gpd and 1 mgd. A computer program for
retrieving, updating, and combining the cost data was also developed. This pre-
• sent report addresses that portion of the study related to cost curves associ-
ated with GAC applications. It also extends the cost study by presenting 22
case histories of operating municipal and industrial. GAC installations for
treating water and wastewater and for processing food and beverage products.
CONCLUSIONS
Some conclusions resulting from the inspection of 22 operating GAC absorp-
tion and reactivation systems and a study of the costs involved are:
1. A substantial number of GAC installations have been operating success-
fully for many years. These installations include facilities for
treating water and wastewater and for processing food and drink.
2. Although there has been widespread, long-term use of GAC and powdered
activated carbon (PAC) by municipal waterworks and industries, no
adverse health effects have been reported.
3. GAC treatment of public water supplies to remove trace organics is
fairly new. Only limited direct design data or cost information are
available.
4. GAC treatment of public water supplies for taste and odor control or
color removal is practiced in about 46 cities in the United-States. In
70
-------
this application of GAC, however, there is little need for reactiva-
tion and little experience with on-site reactivation.
5. EPA is presently sponsoring five reactivation studies that will soon
yield good GAC performance and cost data for water supply
installations.
6. Twenty to thirty GAC installations in food (corn syrup and beet sugar)
processing plants give insight concerning the safety of GAC but add
little to the available cost data. Some design ideas and information
can be obtained from these sources.
7. Twenty to thirty GAC reactivation furnaces in industrial water pollu-
tion control plants yield some information on operation and cost of.
GAC systems.
8. The best sources of detailed cost information and equipment design
data are the 20 or so operating municipal advanced waste treatment.
(AWT) plants using GAC.
9. Pilot plant testing, of GAC adsorptive and reactivation characteristics
in municipal water treatment plants can be combined properly with the
known costs of GAC systems in wastewater treatment to yield a desigr.
basis and preliminary cost estimates that are satisfactory for devel-
oping water treatment projects. Designers should take full advantage
of applicable GAC use and experience available from:
- municipal water treatment for taste and odor control,
- municipal AWT,
- industrial wastewater treatment, and
- food and pharmaceutical production.
71
-------
BIBLIOGRAPHY
American Petroleum Institute, "Adsorption as a Treatment of Refinery Efflu-
ents". Report for the CDP Subcommittee on Chemical Wastes, 1969.
Ancona, A.J., "SuCrest Stationary Bed Activated Granular Carbon System". SIT
Proceedings, 1970.
f
Argo, D.G., "Wastewater Reclamation Plant Helps Manufacture Fresh Water".
Water and Sewage Works, (Reference Number): R-160 1976.
AWWA, Proceedings of AWWA Seminar (June 24, 1979) on Controlling Organics in
Drinking Water, (San Francisco), American Water Works Association, 1979.
Berthouex, P.M., "Evaluating Economy of Scale"<> Journal Water Pollution Control
Federation, 2111, November, 1972.
Bishop, D.F., et al., "Studies on Activated Carbon Treatment".'Journal Water
Pollution Control Federation, February, 1967«
Bunch, R.L., et al., "Organic Materials in Secondary Effluents". Journal:Water
Pollution Control Federation, 33: 122-126, February, 1961.
Bureau of Labor Statistics, U.S. Department of Labor. Producer-Prices and Price
•Indices, October, 1979. ' ,
Calgon Corp., "Water and Waste Treatment with Granular Activated Carbon".
General Catalog, 1970.
Cheremisinoff, P.N., and Ellerbusch, F., "Carbon Adsorption Handbook". Ann
Arbor Science Publishers, Ann Arbor, Michigan, 1978.
Cheremisinoff, P.N., and A.C. Moressi, "Carbon Adsorption". Pollultion
Engineering, 6(8): 66-68, 1974.
Clark, R.M., D.L. Guttman, J.L. Crawford, and J.H. Machisko, "The Cost of
Removing Trihalomethanes from Drinking Water". EPA-600/2-77. U.S. Environmen-
tal Protection Agency, Cincinnati, Ohio, 1977.
Clark, R.M. and Paul Dorsey, "The Costs of Compliance: An EPA Estimate For
Organics Control", Journal American Water Works Association, Vol. 72, No. 8
August 1980, 450-457.
Cookson, J.T., Jr., "Design of Activated Carbon Beds". Journal Water Pollution
Control Association, 42(12):2124, 1970.
Cookscn, T.J., Jr., "Mechanism of Virus Adsorption on Activated Carbon".
Journal American Water Works Association, 61:52, 1969.
72
-------
Cooper, J.C., and Hager, D.G., "Water Reclamation with Activated Carbon".
Chemical Engineering Progress, 85, October, 1966.
Cover, A.E., and Pieroni, L.J., "Appraisal of Granular Carbon Contactors,
Phases I & II". Federal Water Pollution Control Association Report No. TWRC-11,
May, 1969.
Cover, A.E., and Wood, C.D., "Appraisal of Granular Carbon Contactors, Phase
III". Federal Water Pollution Control Association Report No. TWRC-12, May,
1969.
Cf^M, et al., "Process Design Manual for Carbon Adsorption". U.S. EPA, Tech-
nology Transfer, Cincinnati, 1973.
Gulp, G.L., and Gulp, R.L., "New Concepts in Water Purification". Van Nostrand
Reinhold Co., New York, 1974.
Gulp, G.L., and Slechta, A., "Plant Scale Regeneration of Granular Activated
Carbon". USPHS Demonstraton Grant Report, 84-01, 1966.
Gulp, G.L., and Slechta, A., "Plant Scale Reactivation and Reuse of Carbon in
Wastewater Reclamation". Water S Sewage Works, 425-431, November, 1966.
Gulp, G.L., and Shuckrow, A.J., "Physical-Chemical Techniques for Treatment of
Raw Wastewater". Public Works, July, 1972.
Gulp, R.L., "Wastewater Reclamation at South Tahoe PUD". Journal American
Water Works Association, 84, 1968.
Gulp, R.L., and Gulp, G.L., "Advanced Wastewater Treatment". Van Nostrand Rein-
hold Company, New York, 1971.
Gulp, R.L. Wesner, G.M., and Gulp, G.L., "Handbook of Advanced Wastewater
Treatment". Van Nostrand Reinhold Co., New York, 1978.
Gulp, R.L. et al., "Estimating Water Treatment Costs". Vol. I, Gumerman, R,C,,
Hansen, S.P., Gulp, R.L. Summary, EPA-600/2-79-162a; Vol. II, Gumerman, R.C.,
Hansen, EPA-600/2-79-162b; Vol. Ill, Hansen, S.P., Gumerman, R.C., Gulp, R.L.
Cost Curves Applicable to 2,500 gpd to 1 mgd Treatment Plants, EPA-600/2-79-
162c; Vol. IV, Lineck, T.S., Gumerman, R.C., and Gulp, R.L. Computer User's
Manual for Retrieving and Updating Cost Data, EPA-600/2-79-162d. MERL, EPA,
Cincinnati, Ohio, 1979.
Gulp, R.L., Wilson, J., and Evans, D., "Advanced Wastewater Treatment as Prac-
ticed at South Tahoe". EPA 17010ELQ08/71, August, 1971.
DeBoer, J.H., "The Dynamic Character of Adsorption". Oxford University, 1968.
De Marco, J., and Brodtmann, N., "Prediction of Full Scale Plant Performance
From Pilot Columns". U.S. EPA, MERL, Cincinnati, 1979.
73
-------
Dostal, K.A., Pierson, R.C., Hager, D.G., and Roebeck, C.C., "Carbon Bed Design
Criteria Study at Nitro, West Virginia". Journal American Water Works
Association, 57:663, May, 1965. ,y
English, J.N., et al., "Removal of Organics From Wastewater by Activated Car-
bon". (Water-1970), Chemical Engineering Symposium Series, 67:147-153, 1970.
Environmental Protection Agency. Refinery Effluent Water Treatment Plant Using
Activated Carbon. EPA 660/2-75-020, June, 1975.
Environmental Protection Agency. Process Design Manual for Carbon Adsorption.
Environmental Protection Agency, Office of Technology Transfer, October, 1973.
(Also see U.S. EPA)
Erskin, D.B., and Schuliger, W.G., "Graphical Method to Determine the Perfor-
mance of Activated Carbon Processes for Liquids". AIChE Symposium Series, No.
124, 68 ("Water-1971"):185-190, 1971.
Flentje, M.E., and Hager, D.C., "Advances in Taste and Odor Removal with Gran-
ular Carbon Filters". Water and Sewage Works, February, 1964.
Fornwalt, H.J., and Hutchins, R.A., "Purifying Liquids with Activated Carbon".
Chemical Engineering, April 11, 1966 and May 9, 1966.
Giusti, D.M., Conway, R.A., and Lawson, C.T., "Activated Carbon Adsorption of
Petrochemicals". Journal Water Pollution Control Federation, May, 1974..
Guirguis, W., Cooper, T., Harris, J., and Unger, A., "Improved Performance of
Activated Carbon by Pre-Ozonationd". 49th Annual' Conference, Water Pollution
Control Federation, Minneapolis, October, 1976.
Hager, D.G., "Industrial Wastewater Treatment by Granular Activated Carbon".
Industrial Water Engineering, January/February, 1974.
Hager, D.G., and Flentje, M.E., "Removal of Organic Contaminants by Granular
Carbon Filtration". Journal American Water Works Association, 1440, 1965.
Hager, D.C., and Fulker, R.D., "Adsorption and Filtration with Granular Activa-
ted Carbon". Journal of the Society for Water Treatment and Examination, 17:41,
1968.
Hager, D.G., and Rizzo, J., "Removal of Toxic Organics from Wastewater by
Adsorption with Granular Activated Carbon". Paper presented at EPA Technology
Transfer Session on Treatment of Toxic Chemicals, Atlanta, Georgia, April,
1974.
Hale, J.H., Myers, L.H., and Short, T.E., Jr., "Pilot-Plant Activated Carbon
Treatment of Petroleum Refinery Wastewater". In Open Forum on Management of
Peitroleum Refinery Wastewaters", proceedings of a conference sponsored by the
Environmental Protection Agency, the American Petroleum Institute, the National
74
-------
Petroleum Refiners Association and The University of Tulsa, Oklahoma, p. 375,
1976.
Hansen, R.E., "Granular Carbon Filters for Taste and Odor Removal". JournaI
American Water Works Association, 57:176, March, 1972.
Hassler, J.W., "Activated Carbon". Chemical Publishing Co., New York, New York,,
1963.
Hassler, J.W., "Purification With Activated Carbon". Chemical Publishing Com-
pany, Inc., New York, 1974.
Himmelstein, K.J. Fox, R.D., and Winter, T.H., "In-Place Regeneration of Acti-
vated Carbon Used in Industrial Waste Treatment". A1ChE Symposium Series, No.
144, 70:310-315, 1974.
Hoak, R.D., "Recovery and Identification of Organics in Water". In Advances in
Water Pollution Research, Vol. 1, Macmillan, New York, pp. 163-180, 1964.
Hopkins, C.B., Weber, W.J., Bloom, A., "Comparison of Expanded Bed and Packed
Bed Adsorption Systems". Report No. TWRC-2, Federal Water Pollution Control
Association, December, 1968.
Humphrey, M.F. et al., "Carbon Wastewater Treatment Process". Journal ASME,
74-ENAS-46, July, 1974.
Hutchins, R.A., "New Method Simplifies Design of Activated-Carbon Systems".
Chemical Engineering 80(19):133, August 20, 1973.
Hyndshaw, A.Y., "The Selection of Granular Versus Powdered Activated Carbon".
Water and Wastes Engineering, :49, February, 1970.
Jolley, R.L., "Chlorination Effects on Organic Constituents in Effluents From
Domestic Sanitary Sewage Treatment Plants". Oak Ridge National Lab., Publ. No.
565, ORNL-TM-4290, October, 1973.
Joyce, R. S., Allen, J.B., and Sukenik, V.A., "Treatment of Municipal Wastewater
by Packed Activated Carbon Beds". Journal Water Pollution Control Federation,
:813, 1966.
Juhola, A.J., "Optimization of the Regeneration Procedure for Granular
Activated Carbon". U.S. Environmental Protection Agency, Water Pollution
Control Research £ >jries, Project No. 17020 DAO, July, 1970.
Juhola, A.J., and Tupper, F., "Laboratory Investigation of the Regeneration of
Spent Granular Activated Carbon". Federal Water Pollution Control Association,
Report N.TWRC-7, February, 1969.
Juhola, A.J., and Tupper, F., "Regeneration of Spent Granular Activated Car-
bon". Department of the Interior, Report No. TWRC-7, 1969.
75
-------
Klaffke, K., "Granular Carbon System Purifies Water at Del City, Oklahoma".
Water and Wastes Engineering; March, 1968.
Klotz, M., Werner, P., and Schweisfurth, R., "Investigations Concerning the
Microbiology of Activated Carbon Filters". In Translation of Reports on Special
Problems of Water Technology, Vol. 9—Adsorption, H. Sontheimer, Ed., EPA
Report EPA600/9-76-030, pp. 312=330, December, 1976.
Kovach, J.L., "Activated Carbon Dechlorination". Industrial Water Engineering,
30, October/November, 1971.
Lawson, C.T., "Activated Carbon Adsorption for Tertiary Treatment of Activated
Sludge Effluents from Organic Chemicals and Plastics Manufacturing. Plants -
Application Studies and Concepts". Research and Development Dept., Union Car-
bide Corporation, South Charleston, West Virginia, August, 1975.
Linstedt, K.D., et al., "Trace Element Removals in Advanced Wastewater Treat-
ment Processes". Journal Water Pollution Control Federation, 43:1507-1513,
July, 1971.
Los Angeles Water Recycling Project. Los Angeles Department of Water and Power,
September, 1972.
Love, O.T., Jr., and Symons, J.M., "Operational Aspects of GAC Adsorption
Treatment". U.S. EPA, MERL, Cincinnati, 1978. ;
Loven, A.W., "Activated Carbon Regeneration Perspectives". AIChE Symposium
Series No. 144, 70:285-295, 1974.
Manka, J., Rubhum, M., Mandlebaum, A., and Bortinger, A., "Characterization of
Organics in Secondary Effluent". Environmental Science and Technology, 1017,
November, 1974.
Mantell, C.L., "Adsorption". McGraw-Hill Book Co., New York, 1945.
Mantell, C.I., "Carbon and Graphite Handbook". Wiley-Interscience, New York,
1968.
McCreary, J.J., and Snoeyink, V.L., "Granular Activated Carbon in Water Treat-
ment". Journal American Water Works Association, 69(8):437-444, 1977.
McCrodden, B.A., "Treatment of Refinery Wastewater Using Filtration and Carbon
Adsorption". Advanced Petroleum Refinery Short Course, Principles and Practice
in Refinery Wastewater Treatment, University of Tulsa, June, 1973.
McDonnell, G.M., "Water Reuse Through Advanced Wastewater Treatment". Water and
Sewage Works, (Reference Number): R-60-R-64, 1973.
Morris, J.C., and Weber, W.J., "Adsorption of Biochemically Resistant Materials
from Solution". USPHS AWTR Publication No. 9, 1964.
76
-------
Mattson, J.S., and Kennedy, F.W., "Evaluation Criteria for Granular Activated
Carbon". Journal Hater Pollution Control Federation, 43:2210-2217, November,
1971.
O'Farrell, T.P., Bishop, D.F., and Bennett, S.M., Advanced Waste Treatment at
Washington, D.C., Presented at the 65th Annual AICHE Meeting, Cleveland, Ohio,
May, 1969.
Parkhurst, J.D., bryden, F.D., McDermott, G.N., and English, J., "Pomona Acti-
vated Carbon Pilot Plant". Journal Water Pollution Control Federation, R70,
1967.
Perrotti, A.E., and Rodman, C.A., "Factors Involved with Biological Regenera-
tion of Activated Carbon". AIChE Symposium Series, No. 144, 70:316-325, 1974.
Person, W.D., and Truemper, J.To, "Decolorization of Various Foreign and Domes-
tic Sugars by Activated Carbons". SIT proceedings, 1970.
Phillips, J.D., and Shell, G.L., "Pilot Plant Studies of Effluent Reclamation".
Water and Wastes Engineering, 38-41, November, 1969.
Poggenburg, W., "Activated Carbon Filters in Water Treatment Plants. Process-
ing Techniques—Engineering—"Operation". In Translation of Reports on Special
Problems of Water Technology, Vol. 9—Adsorption, H. Sontheimer, ED., EPA
Report, EPA600/9-76-030, ppo. 74-97, December, 1976.
Rizzo, J.L., and Schade, R.E., "Secondary Treatment with Granular Activated
Carbon". Water S Sewage Works, 307,.August, 1969.
Rizzo, J.A., "Activated Carbon Clears Effluent". Oil and Gas Journal, 74(22)-
:52-56, 1976.
Rizzo, J.L., and Sheperd, A.R., "Treating Industrial Wastewater with Activated
Carbon". Chemical Engineering, January 3, 1977.
Rodman, C.A., and Shunney, E.L., "Bio-Regenerated Activated Carbon Treatment of
Textile Dye Wastewater". EPA Report 12090 DWM, Washington, D.C., January,
1971.
Rosen, A.A., et al., "The Determination of Stable Organic Compounds in Waste
Effluents at Microgram per Liter Levels by Automatic High Resolution Ion
Exchange Chromatog •aphy". Water Research, 6:1029, 1972.
Rosen, A.A., et al., "Relationship of River Water Odor to Specific Organic Con-
taminants". Journal Water Pollution Control Federation, 35:77-782, June, 1963.
Rubhun, M., and Manka, J., "Classification of Organics in Secondary Efflu-
ents". Environmental Science and Technology, 5:606-609, 1971.
Schalekamp, M., "Use of Activated Carbon in the Treatment of Lake Water". In
Translation of Reports of Special Problems of Water Technology, Vol.
77
-------
9—Adsorption, H. Sontheimer, Ed., EPA Report EPA 600/79-76-030, pp. 128-159,
December, 1976.
Schwartz, H.G., Jr., "Adsorption of Selected Pesticides on activated Carbon and
Mineral Surfaces". Environmental Science and Technology, 1:332-337, April,
1967.
Shell, G.L., et al., "Regeneration of Activated Carbon". In Applications of New
Concepts of Physical-Chemical Wastewater Treatment, Pergamon Press, Elmsford,
New York, pp. 167-198, 1972.
Shumaker, T.P., "Granular Carbon Process Removes 99.0-99.2% Phenols". Chemical
Proceedings, 36(5):45-47, 1973.
Sigworth, E.A., and Smith, S.B., "Adsorption of Inorganic Compounds by Activa-
ted Carbon". Journal American Water Works Association, 386-391, June, 1972.
Slechta, A.F., and Gulp, G.L., "Water Reclamation Studies at the South Tahoe
Public Utility District". Journal Water Pollution Control Federation, 787-814,
May, 1967.
Simsek, M., and Cerny, S., "Active Carbon". Elsevier Publishing Co., New York.
Smith, C.E., and Chapman, R.L., "Recovery of Coagulant, Nitrogen Removal, and
Carbon 'Regeneration in Wastewater Reclamation". Final Report to Federal Water
Pollution Control Federation, Demonstration Grat WPD-85, June, 1967.
Spencer,. E.G., and G.P. Meade, "Cane Sugar Handbook". 9th Edition, New York:
John Wiley & Sons, Inc.
Spink, G.F., "Activated Carbons—Systems and Economics". Sugar Industry
Technologists, Inc. Proceedings, 1969, Medford, Massachusetts.
Stander, G.J., and Funke, J.W., "Direct Cycle Water Reuse Provides Drinking
Water Supply in South Africa". Water and Wastes Engineering, 66-67, May, 1969.
Stander, G.J., .and Van Vuuren, L.R.J., "The Reclamation of Potable Water From
Wastewater". Journal Water Pollution Control Federation, 355-367, March, 1969.
Suhr, L.G., and Gulp. G.L., "State of the Art—Activated Carbon Treatment of
Wastewater". Water and Sewage Works, (Reference Number): R-104, 1974.
Symons, J.M., "Interim Treatment Guide for Controlling Organic Contaminants in
Drinking Water Using Granular Activated Carbon". U.S. EPA, MERL, Cincinnati,
1978.
Symons, J.M., and Robeck, G.C., "Treatment Processes for Coping with Variations
of Raw Water Quality". Presented at the 1973 American Water Works Association
Conference, Las Vegas, Nevada, May 16, 1973.
78
-------
Thiem, L.D., Badorek, and O'Conner, J.F., "Removal of Mercury from Drinking
Water Using Activated Carbon". Journal American Water Works Association, August
1976.
Truemper, J.T., "The Porosity of Activated Carbon and Its Relation to Cane
Sugar Refining". SIT proceedings, 1968.
U.S. Department of Interior. "Appraisal of Granular Carbon .Contacting: Phase
One: Evaluation of Literature". Report No. TWRC-11, U.S. Department of the
Interior, May, 1969.
U.S. Department of Labor. Bureau of Labor Statistics. Producer Prices and Price
Indices, Washington, D.C., October, 1979.
U.S. EPA-NATO. Proceedings - International Conference on Adsorpton Techniques,
Washington, D.C., 1979. EPA, Office of Drinking Water, Washington, D.C.
U.S. EPA. "Estimating Cost for Water Treatment as a Function of Size and Treat-
ment Plant Efficiency". EPA-600/2-78-182, MERL, Cincinnati, August, 1978.
Van Lier, W.C., Graveland, A., Rook, J.J., and Schultnik, J..J., "Experiences;
with Pilot Plant Activated Carbon Filters in Dutch Waterworks". In Translation
of Reports on Special Problems of Water Technology, Vol. 9—Adsorption, H.
Sontheimer, Ed. EPA Report, EPA600/9-76-030, pp. 160-181, December, 1976.
Van Vuuren, L.R.J., et al., "The Full-Scale Reclamation of Purified Sewage.
Effluent for the Augmentation of the Domestic Supplies of the city of Wind-
hoek" . Paper presented at the 5th International Water Pollution Research Con-
ference, July-August, 1970.
Von Dreusche, C., Jr., "Process Aspects of Regeneration in a f-'niltiple Hearth
Furnace". AIChE Proceedings, August, 1974.
Walker, P.L., Jr., "Chemistry and Physics of Carbon". Marcel Dekker, Inc., New
York, 1965.
Ward, T.M., and Getzen, F.W., "Influence of pH on tne Adsorptibn of-Aromatic
Acids on Activated Carbon". Environmental Science and Technology, 4:64-67, Jan-
uary, 1970.
Weber, W.J., Jr., "Physiochemical Processes for Water Quality Control". John
Wiley and Sons, Inc., New York, 1972.
Weber, W.J., Jr. and J.P. Gould, "Sorption of Organic Pesticides from Aqueous
Solution". Advances in Chemistry Series, No. 60, 280, R.F. Gould, Ed., American
Chemical Society, Washington, D.C., 1966.
Wesner, G.M., "Reuse is on Tap". Water and Wastes Engineering, :46-47, Novem-
ber, 1972.
79
-------
Wesner, G.M., and Argo, D.G., "Pilot Wastewater Reclamation Study May 1970-June
1971". Orange County Water District, Fountain Valley, CA July, 1973.
Westetrmark, M., "Kinetics of Activated Carbon Adsorption". Journal Water
Pollution Control Association 42(4):709, 1975.
Wherry, C.R., "Activated Carbon". In Chemical Economic Handbook, Stanford
Research Institute, Menlo Park, California.
Zogorski, J.S., and Faust, S.D., "Operational Parameters for Optimum Removal of
Phenolic Compounds from Polluted Waters by Columns of Activated Carbon". 69th
AIChE Meeting, November 28, 1976.
80
-------
APPENDIX
CASE HISTORIES OF MUNICIPAL AND INDUSTRIAL
GRANULAR ACTIVATED CARBON INSTALLATIONS IN THE UNITED STATES
CASE HISTORY NO. 1
SOUTH TAHOE (CALIFORNIA) PUBLIC UTILITY DISTRICT
MUNICIPAL WASTEWATER TREATMENT
HISTORY AND INTRODUCTION
The South Tahoe wastewater reclamation plant was the first full-scale
municipal wastewater treatment plant in the United States to use GAC, and
therefore, it has the longest record of performance to draw upon. During 14
years of GAC treatment, the character of the wastewater to be treated and the
standards to be met have changed so that operational data are available for
more than one set of conditions. To supplement the routine data collection
efforts, a special EPA funded 3-year study made it possible to collect and ana-
lyze much more complete and detailed data on plant performance and operational
costs than would be possible under normal circumstances.
In 1963, the South Tahoe plant was designed without the benefit of any
information from full-scale municipal wastewater treatment GAC operations,
because there was none at .the time. Rather, the design was based on: (1) the
knowledge and experience gained by industries in decolorizing sugar and refin-
ing corn syrup with GAC; (2) laboratory and pilot plant work done by EPA
researchers at Cincinnati; and (3) information from GAC suppliers and carbon
reactivation furnace manufacturers, all as modified by applying the results of.
laboratory experiments (carbon isotherms and muffle-furnace reactivation) and
pilot plant operations (at 1 and 25 gpm rates) at South Tahoe.
This transfer and modification of data from an indirectly related field
(food processing) to wastewater treatment, as modified by pilot plant experi-
ence, is somewhat parallel to the transfer of data that is now being attempted
from wastewater treatment to drinking water purification. Actually, these pre-
sent relationships are probably closer than these earlier ones. The transition
from food processing to wastewater treatment was successful. Therefore, it is
logical to expec4-. that the lessons learned and the knowledge gained in waste-
water treatment * 1th GAC may be applied even more directly and more accurately
to water purification than was done in the previous transition. Indeed, by
using accumulated experience in wastewater treatment as properly modified by
pilot plant tests of water purification, it should be possible to design, con-
struct, and operate GAC systems to improve drinking water quality without risk
of major overdesign or underdesign. Also, capital and O&M (operation and main-
tenance costs) should be predictable within a reasonable degree of accuracy.
81
-------
DESIGN DATA
Originally, the GAC facilities at South Tahoe were designed for a maximum
daily flow of 2.5 mgd. This plant was operated from 1966 until 1968, when plant
capacity was expanded to a maximum daily flow of 7.5 mgd. Because no major
changes in the basis of GAC system design, other than the capacity increase
were made in the plant expansion, the design data are presented for the 7.5 mgd
plant.
It is important to note that the design is based on maximum daily flow (in
accordance with water works practice) rather than on average dry weather flow
(wastewater treatment practice). This was done because of the strict effluent
quality standards that applied under maximum flow conditions,—standards that/
at the time, were out of the ordinary for wastewater treatment. In a year
during which the maximum daily flow is 7.5 mgd, the corresponding average daily
dry weather flow is 5 mgd.
Basic of GAC System Design
Flow
Maximum daily, 7.5 mgd
Peak hourly flows are equalized by
in-plant storage before GAC treatment
COD
Raw wastewater, average 280 mg/L
Carbon column influent, average 35 mg/L
Carbon column effluent, average 20 mg/L
GAC contact time, minimum
Granular activated carbon
mesh
Makup carbon unloading
17 mln " ••••.-..
Calgon Filtrasorb; coal base; 8 x 30
Bag dump (manual)
Bulk - hydraulic eductor
Carbon contactors (see typical section Figure A-1)
Type
Number
Inlet
Outlet
Column
Upflow packed bed; countercurrent
Eight; 12-ft diameter by 12-ft straight
sidewall steel shell; coal-tar epoxy
coated inside; 45° conical top and
bottom; 14.5-ft carbon bed depth
Eight; 12-in. diameter x 2.5-ft long
Johnson well screens; 304 SS; 100 psi;
0.020-in. slot opening; closed bail
bottom
Same as inlet
Equipped with screened vent—combination
pressure air and air-vacuum release
valve
82
-------
CARBON IN
TOP WAFER VALVE
OUTLET SCREENS (.8)
u_lo
PRESSURE VESSEL
12 FT DIAMETER
SURFACE OF
CARBON
u.
-
-------
Rate-of-flow control Dahl tube and butterfly valve on effluent
header
• ;-Backwash Flow can be reverse cleaned
Adding/removing Carbon can be added and removed while
column is in operation
Carbon slurry transfer piping Black steel? fittings extra heavy;
standard radius elbows and tees
Carbon wash tanks For carbon fines removal from reactivated
and makeup carbon; two pressure tanks
with Johnson-screened outlets; equipped
for flow reversal to clean screens;
each tank holds 5% of one carbon column
volume; wash tanks also serve as
blowcases (hydraulic) for carbon
transfer
GAC reactivation (see schematic, Figure A-2)
Number One; 54-in. diameter by six-hearth
BSP—Envirotech gas-fired furnace with
propane standby • : ' ;
Rated capacity 6,000 Ib/day • • v:••-" . .
Odor/particulate control Equipped with afterburner for odor
control (ordinarily not used); exhaust
gas scrubber for removal of
particulates
Spent carbon dewatering tanks Two; to 40% moisture in 10 min when
. . . screen drains are clean; each tank
• holds 5% of volume'of one carbon column
Carbon feed to furnace Two inclined 304 SS screw conveyors with
variable speed (10:1) drive
Quench tank location Single reactivated carbon quench tank
directly below furnace outlet and above
two diaphragm slurry pumps that
discharge to carbon wash tanks
CARBON SYSTEM OPERATION
Carbon Adsorption System
The eight carbon columns are operated in parallel, with the flow travel-
ing upward from top to bottom. The flow can be reversed to flush the top
screens, compact the carbon bed, or remove the carbon in the lower portion of
the column. This backwashing is usually done when the headloss across the
column reaches 12 ft of water. The regenerated carbon is added to the top of
the column.
Flow through each column is maintained at greater than 500 gpm to ensure
proper performance. Studies showed that at flows less than this value, mud
balls and odors became operational problems. Carbon losses from the tanks were
small—about 20 ft^ per column between regeneration periods.
84
-------
CXHa
I
H
•Spent carbon
drain and feed
tanks
Screw conveyors
T
Carbon
• lurry pumps
-Quench tank
Spent carbon
1 from carbon
columns
/'Scrubber and air
I pollution control
\ equipment
Carbon
regeneration
furnace i
Steam
supply
V.Regenerated carbon defining
and storage tanks
M
Regenerated
carbon to
carbon columns
Figure A-2. Carbon regeneration system, South Tahoe Public
Utility District, CA.
85
-------
Carbon Reactivation
Spent carbon is withdrawn from each column and transferred to a dewatering
bin. From there it is fed to a 6-hearth furnace for reactivation. Reactivation
is done at a temperature of 1650° to 1750°F in an oxygen-limited atmosphere.
Organic impurities are volatized, and the reactivated carbon is cooled and
de-fined before being reused. During the initial startup and intensive testing
periods, an attempt was made to return the reactivated carbon to the column
from which it came. Virgin carbon was added to the columns as needed to make up
for reactivation losses.
Major GAC System Repairs
Carbon reactivation equipment and carbon slurry conveying facilities are
subject to rather severe service conditions. The 14 years of experience at
South Tahoe indicates the nature and cost of major repairs that may be expected
in addition to routine maintenance (Table A-1).
TABLE A-1. MAJOR CHANGES AND REPAIRS, SOUTH TAHOE GAG SYSTEM
Year
1973
1976
1977 ,.
1978
Description
Rebuild furnace burners
Rebuild reactivation furnace
. Repair, top. hearth .
Add new facilities for bulk han-
Maintenance
labor , hr
480
1,440 .
480
160
Materials &
supervision,
cost
$ 3,000
8,000
2,000 • ..;
8,000
dling of GAC
1979 Rebuild exhaust gas system, repair 72,000
screw conveyor
1973-1979 Replace 128 50 psi carbon column 6,144 90,880
screens with 100 psi screens
In addition, plant maintenance crews are presently spending about 20
man-hours per week replacing 10 to 14 year old carbon slurry lines on a system-
atic basis; the aim is to replace all original lines. In the past (1965 to
1979), about 3 hours per week was needed to maintain and repair carbon transfer
lines.
Plant maintenance crews estimate the current GAC reactivation furnace
downtime for service, adjustment, and repair is only about 5 percent of the
time (1 1/2 days per month).
86
-------
CARBON SYSTEM PERFORMANCE
From 1965 through March 1979, during the 14 years of GAC system oper tion
at South Tahoe, all plant flows (a total of 14.5 bil gal of wastewater) have
been carbon treated. In evaluating the performance of the GAC system, three
time periods are significant because of three different sets of operating con-
ditions that prevailed.
• 1965-1974 - Initial period when design COD loading conformed to pilot
plant conditions, plant operation was normal, and data collection was
at a peak.
• 1974-1977 - Unusual operating conditions due to 80 percent increase in
design COD; carbon use increased to maintain discharge requirement;
high carbon use because spent carbon overburned during reactivation.
• 1978-1979 - Return to-.normal operation as a result of relaxed COD stan-
dard; high influent COD loading prevailed.
The first and third periods of plant performance represent excellent and
good conformance to discharge requirements, whereas the second period was one
during which COD requirements were slightly exceeded intermittently.
Detailed Performance Data
The information concerning plant operations for 1968-1970 is typical for
the initial period of plant operations from 1965-1974. The GAC system operated
up to the full expectations developed in the pilot plant tests. Discharge
requirements met fully and continuously from the system's first application in
1968 through 1974—a period of about 7 years. All GAC was reactivated on-site
with virtually no decrease in adsorption of wastewater organics despite a
slight drop in iodine number and other indirect indicators of carbon capacity.
The performance data from 1968 to 1971 are summarized from "Advanced Wastewatex
Treatment as Practiced at South Tahoe."21
Between November 1968 and January 1971, the carbon column flow was 3,341
mg and 23,050 ft3 or 691,000 Ib of activated carbon were regenerated; this
provided a carbon dosage rate of about 207 Ib regenerated carbon/ mil gal
carbon column flow (Table A-2). For the same period, an average of 0.8 Ib of
COD were removed per Ib of carbon. An average of 0.027 Ib of MBAS was applied
per Ib of regenerated carbon, and 0.021 Ib of MBAS was removed per Ib of
carbon.
The average furnace .feed rate for the batch regeneration periods was 176
Ib/hr, or about 6 ft3/hr. Fuel requirements per Ib of carbon averaged 2,900
Btu at 860 Btu/ft3 of natural gas at 18 to 20 psia. Numbers four and six fur-
nace hearth temperatures averaged 1650° and 1760°F, respectively. Carbon losses
in the furnace and in transport to and from the furnace averaged 8 percent that
period. In the batch regeneration used at Tahoe, it took approximately 1 year
for an entire column to pass through one generation cycle.
87
-------
TABLE A-2. AVERAGE CARBON EFFICIENCY FOR THE REGENERATION PERIOD
NOVEMBER 1968 THROUGH JANUARY 1971
Parameter/unit
Average
Maximum
Minimum
Carbon dosage
(Ib reg/mg treated)*
Iodine numbert
Spent carbon
Regenerated carbon
Apparent density (g/cc)
Spent carbon
Regenerated carbon
Percent asht
Spent carbon
Regenerated carbon
Chemical oxygen demand
Percent removal
Ib COD applied
Ib COD applied/mil gal
Ib COD removed/mil gal
Ib COD applied/lb carbon
regenerated*
Ib COD removed/lb carbon
regenerated*
207
583
802
0.571
0.487
6.4
6.8
0.39
418
633
852
0.618
0.491
7.0
7.2
0.71
*Based on ft? of carbon fed to furnace at 30 Ib/ftAT
tNovember 1968 through November 1970,
111
497
743
0.544
0.478
5.8
5.8
49.9
28,250
162
81
0.78
63.3
54,970
254
149
1.56
30.1
15,680
105
32
0.52
0.16
Methylene blue active substances
( MBAS )
Percent removal
Ib MBAS applied
Ib MBAS applied/mil gal
Ib MBAS removed/mil gal
Ib MBAS applied/lb carbon
regenerated*
Ib MBAS removed/lb
carbon regenerated*
77.0
995
5.7
4.4
0.027
0.021
93.0
1,675
10.7
8.2
0.045
0.039
58.0
457
2.6
1.6
0.012
0.007
88
-------
The virgin carbon, from the 1965 and 1968 purchases, had an average iodine
number of 935, an apparent density of 0.485 g/cc, and an ash content of 5.0
percent. The spent carbon before regeneration had an average iodine number of
583, an apparent density of 0.571 g/cc, and an ash content of 6.4 percent. Th«
regenerated carbon had an average iodine number of 802, an apparent density of.
0.487 g/cc, and ash content of 6.8 percent. The performance of virgin and
regenerated carbon in removing MBAS is illustrated by the isotherms shown in
Figure A-3. There is little difference in the performance of MBAS removal,
related to the number of regeneration cycles. The same is true for COD removal,
except for the considerable difference in performance between the virgin carbon
and the first regeneration; after that, the performance is quite consistent.
When the influent COD to the South Tahoe plant suddenly increased from 280
to 500 mg/L in 1976, the GAC columns were stressed far beyond their original
design conditions. The influent COD to the columns jumped from 20 mg/L to 45
mg/L. To meet discharge requirements of 20 mg/L, the carbon dosage was
increased from 210 Ib/mil gal of flow treated to 320 Ib/mil gal. At the same
time, the removal of COD dropped from 0.39 to 0.31 Ib/lb carbon reactivated. On
occasion, the discharge requirements were violated and operating costs sky-
rocketed. A study of the situation left two alternatives—either doubling the
number of carbon columns or. relaxing the discharge requirements. The latter
option was exercised after an extensive study concluded that no ill effects
would result from an effluent COD limitation of 30 mg/L. Recent operations data
show that this concentration can be easily met.
CAPITAL AND OPERATION AND MAINTENANCE COSTS
The most detailed and accurate cost data collected at South Tahoe resulted
from the 1968-1970 EPA Demonstration Grant study.
Capital Costs
The capital costs shown in Table A-3 include all equipment and construc-
tion costs, but not engineering. Since construction took place over many years
and under several contracts, these costs have been adjusted to 1969 using an
EPA Sewage Treatment Plant Constructon Cost Index of 127.1.
89
-------
10
x/mw MGS OF MBAS ADSORBED/MG CARBON
p _^
2 M (it •*> IB O> --IOKO 1. l\> W * W 0>-JOO«O CJ l\> W -8k W O-JOXD
^
^
x--
^
•
*•
&
+•
*
&
X
?^
^
~*
^\
<£
^
^
**
&
*
f
%
•
^,
CC-r8
J^><
**^s&£;£
*2ss*r
7/70
\
\
J
/
*'
VI
ia
n
i
RC
/6
/7
^71
,Ih
B
0
D
J
0.01 2 34 567690.1 2 34 567891.0 2 34 5678910
Co RESIDUAL MBAS»mg/l
VIRGIN AND THREE REGENERATION CYCLES
OF CC-5 (DECEMBER 1968, JANUARY 1970, AND NOVEMBER 1970)
AND FOURTH CYCLE OF CC-8 (JULY 1970)
Figure A-3. Isotherms for regenerated carbon.
90
-------
TABLE A-3. SOUTH TAHOE PUBLIC UTILITY DISTRICT, CALIFORNIA, CAPITAL COSTS FOR
CONVENTIONAL AND ADVANCED WASTE TREATMENT PLANT, 7.5 MGD DESIGN
CAPACITY
Treatment nhase
Actual contract
total const.
cost per phase*
Estimated
national average
replacement const.
cost for 1969
Estimated
replacement
costs per mcj
for 1969
Carbon treatment
Carbon adsorptiont
Carbon regeneration
Subtotal, carbon
treatment
$656,000
193,000
849,000
$632,000
199,000
831,000
$16.30
5.20
21.50
•* Construction .costs are taken from .District records of actual contracts awarded
for various phases. Contracts for construction were completed at various
periods between 1960 and 1968. These costs have not been adjusted to a common
year.
•(•Includes initial carbon costs"to fill all carbon columns.
Operation and Maintenance
The operating costs are based on the plant design capacity of 7.5 mgd from
February 1969 to December 1970. During this period, the actual average monthly
influent flows varied between 1.79 mgd and 3.15 mgd. As a result of recycling
the water from scrubber flows, backwashing filters, and backflowing carbon
columns, the filtration and carbon adsorpton flows during the same period var-
ied between 3.13 mgd and 5.22 mgd.
To compare operating costs on the common basis of the plant design
capacity, it was assumed the total cost for fuel, chemicals, make-up lime, emd
make-up carbon would increase in proportion to the flow. The same assumption,
however, could not be made for electricity, O&M labor costs, equipment repair,
and instrument maintenance. Costs per day for electricity were adjusted upwsird
to reflect the equipment characteristics at 7.5 mgd. The cost per day for O&M
labor, repair materials, and instrument maintenance were assumed to be the Seime
as present design flows.
Table A-4 summarizes the cost of operating the GAC units to treat 7.5 ngd
in 1970. The uni-. costs for labor, energy, and make-up carbon are given in
Table A-5.
91
-------
TABLE A-4. OPERATING AND UNIT CAPITAL COSTS FOR SOUTH TAHOE GAC COLUMNS AND
REACTIVATION SYSTEM (1970 DOLLARS)
Operating cost
Electricity
Natural gas
Make-up carbon
Operating labor
Maintenance labor
Repair material
Instrument maintenance
Total operating cost
Total cost
Operating
Capital t
Total
GAC
treatment
$/day
$47.34
24.53
3.27
1.15
4.24
$80.53
$/mil gal
$10.74*
16.30
$27.04
GAC
reactivation
$/day
$ 2.23
6.15
70.39
91.90
16.21
1.17
1.90
$189.95.
$/mil gal
$ 25.33
5.20
$ 30.53
Total
$/day
$ 49.57
6.15
70.39
116.43
19.48
2.32
6.14
$220.48
$/mil gal
$ 36.07
21.50
$ 57.57
*Total operating cost per million gallons of water treated would be 8.77/mil
gal. This figure includes 7«5 mgd plant influent/ plus recycle streams. ..
•(•Includes initial, carbon charge. " • ,
TABLE A-5. UNIT COSTS*.1969 AND 1970, AND 1979
1969/1970
1979
Labort
Operations
Maintenance
Electricity**
Fuel§
Activated carbon makeupSl
Off- site carbon reactivation
$ 6.11/hr
5.05/hr
12.10/1000 kwh
0.0 54 3 /therm
0.305/lb
"
$10.30/hr
44.00/1000 kwh
2.52/1000 ft3
0.66/lb
0.45/lb
*All appropriate unit costs are f.o.b. South Lake Tahoe and include a 5%
California sales tax.
tLabor costs include all direct and indirect monies paid upon the employees
behalf. The rates are averages for 1969 and 1970.
**Includes energy and demand charges.
§Natural gas at about 860 Btu ft3 at 6,200 ft elevation and billed on the
basis of interruptable service.
^Activated carbon at 30 lb/ft3.
92
-------
BIBLIOGRAPHY
Gulp, G.L. and Hansen S., "How to Clean Wastewater for Reuse". American City!
page 96, June, 1967.
Gulp, G.L., and Slechta, A.F., "Plant Scale Reactivation and Reuse of Carbon in
Wastewater Reclamation". Water '& Sewage Works, pages 425-431, November, 1966.
Gulp, G.L., and Gulp, R.L., "Reclamation of Wastewater at Lake Tahoe", Public:
Works, February, 1966.
Gulp, G.L., and Slechta, A.F., "Plant Scale Regeneration of Granular Activated
Carbon", Final Progress Report, USPHS Demonstration Grant 84-01, February,
1966.
Gulp, G.L., and Slechta, A.F., "Tertiary Treatment Practices", Presented at
38th Annual Conference of the California Water Pollution Control Assoc., Mon-
terey, California, April, 1966.
Gulp, G.L., and Slechta, A.F., "Water Reclamation Studies at the South Taho«
Public Utility District", JWPCF, pp. 787-814, May, 1967.
Gulp, R.L., "A New Process for Wastewater Reclamation", Proceedings of the
Fifth Texas Industrial Water and Waste Conference, June, 1965.
Gulp, R.L., and Gulp, G.L., "Advanced Wastewater Treatment", Van Nostrand Rein-
hold Co., New York, February, 1971.
Gulp, R.L., "Integration of Advanced Wastewater Treatment Processes", FWQA AW1:?
Seminar, San Francisco/ October, 1970.
Gulp, R.L., "Reclaimed Water As A Resource", Presented at Seminar on Tertiary
Treatment of Wastewater, University of Arizona, Tucson, January, 1969.
Gulp, R.L., and Roderick, R.E., "The Lake Tahoe Water Reclamation Plant",
JWPCF, p. 147, February, 1966.
Gulp, R.L., "The Operation of Wastewater Treatment Plants", Public Works (in
three parts-October, November, and December, 1970).
Gulp, R.L., "The World's Most Advanced Wastewater-Purification Plant", American
City, p. 77 August, 1968.
Gulp, R.L., and Moyer, H.E., "Wastewater Reclamation and Export at South
Tahoe", Civil Engineering, p. 38, June, 1959.
Gulp, R.L., "Wastewater Reclamation by Tertiary Treatment", JWPCF, p. 799,
June, 1963.
93
-------
Gulp, R.L., "Water Reclamation at South Tahoe", Water and Wastes Engineering,
p. 36, April, 1969.
Gulp, R.L., "Wastewater Reclamation at South Tahoe Public Utility District",
JAWWA, 60:84, January, 1968.
Gulp, R.L., Wesner, G.M., and Gulp, G.L., "Handbook of Advanced Wastewater
Treatment", Second Edition, Van Nostrand Reinhold Co., New York, 1978.
t
Dean, R.B. and Forsythe, S.L., "Estimating the Reliability of Advanced Waste
Treatment",, Water & Sewage Works, p. 87, June, 1976.
Evans, D.R., and Wilson, J.C., "Actual Capital and Operating Costs for Advanced
Waste Treatment", WPCF national meeting, Boston, October, 1970.
Evans, D.R., Wilson, J.C., and Gulp, R.L., "Advanced Wastewater Treatment as
Practiced at South Tahoe", EPA 17010/ELQ/08/71, August, 1971, Supt. of Docu-
ments, U.S. Printing Office, Washington, DC, 20402.
Gumerman, R.C., Gulp, R.L., and Hansen, S.P., "Estimating Costs for Water
Treatment as a Function of Size and Treatment Efficiency", U.S. EPA,
600/2-78-182, August, 1978, National Technical Information Service, Spring-
field, Virginia, 22161.
Moyer, Harlan E., "The South Tahoe Water Reclamation Project", Public Works, p.
7, December, 1968.
Priday, w., Moyer, H.E., and Gulp, R.L., "The Most Complete Wastewater Treat-
ment Plant in the World", American City, p. 123, September, 1964. •
Sebastian, F.P., and Sherwood, R.J., "Clean Water and Ultimate Disposal", Water
& Sewage Works, p. 297, August, 1969.
Sebastian, F.P., "Wastewater Reclamation and Reuse", Water and Wastes
Engineering, p. 46, July, 1970.
Stevens, L., "Breakthrough in Water Pollution", Readers Digest, June, 1971.
Smith, C.E., and Chapman, R.L., "Recovery of Coagulant, Nitrogen Removal, and
Carbon Regeneration in Wastewater Reclamation", Final Report of Project Opera-
tions, FWPCA Grant WPD-85, June, 1967.
Suhr, L.G., and Gulp, R.L., "Design and Operating Data for a 7.5 MGD Nutrient
Removal Plant", WPCF Annual Conference, Chicago, September, 1968.
94
-------
CASE HISTORY- NO. 2
TAHOE-TRUCKEE SANITATION AGENCY
MUNICIPAL WASTEWATER TREATMENT
HISTORY AND INTRODUCTION
The Tahoe-Truckee Sanitation Agency (TTSA) was formed .on May 1, 1972, to
carry out the mandate of the Porter Cologne Water Quality Control Act, a-Cali-
fornia law requiring the exportation of all sewage from the Lake Tahoe Basin.
The newly formed Agency's initial task was to consider the design and construc-
tion of a regional wastewater treatment facility that would replace existing
interim facilities operated by its member entities (the Tahoe City Public Util-
ity District, the North Tahoe Public Utility District (both located in the Lake
Tahoe drainage basin), the Alpine Springs County Water District, the Squaw
Valley County Water District, and the Truckee Sanitary District), which arcs
located in the Truckee River drainage basin. The Agency is governed by a Board!
of Directors composed of five appointed directors, one from each member
district..
It was decided that retaining the effluent in the Truckee River Basin was;
of primary importance so that the quantity of water available to downstream
users would not be diminished by this project; that treating the wastewater to
a high degree at a plant near. Truckee Was the most economical with the least
adverse impact on the environment; and that adhering to the existing high water
quality standards .required of the Truckee River could not be done by simply
upgrading the existing plants. The State of California's rigid waste discharge
requirements necessitated the implementation of the most highly sophisticated
treatment available.
The project, which was finally selected from many alternatives studied,.
required constructing an interceptor line from Tahoe City to Truckee, Califor-
nia; constructing a 4.83 mgd advanced waste treatment plant in Martis Valley
near Truckee, California; and installing an underground effluent disposal sys-
tem at a total cost of $32 million. Pull tertiary treatment, including maximum
removals of nitrogen and phosphorus and granular activated carbon adsorption,
was deemed necessary to protect the quality of the Truckee River, into which
the effluent would ultimately find its way after percolation into the permeable
glacial outwash soil near the plant. This would ensure the safety and integrity
of this primary water source for Reno, Nevada, 30 miles downstream.
Studies of the background concentration of pollutants in the Truckee River
revealed that the critical pollutants were total nitrogen and total phosphorus.
These were also the most difficult contaminants to remove; other pollutant!;
could easily be removed. The primary goal of process design was to ensure that
the treated effluent, when mixed with background concentrations existing in the
river, would not violate the river standards even if discharged directly into
the river. Percolation into the Martis Valley soils (where a residence time o::
about 150 days would be experienced) would provide an additional treatment
95
-------
factor. The effluent's mixing with natural groundwater moving through the soil
would provide additional dilution of any residual pollutants.
Briefly, the major sewage treatment processes employed are: primary clari-
fica.tion; pure oxygen activated sludge; lime treatment and chemical clarifica-
tion for phosphorus removal; two-stage recarbonation with intermediate calcium
carbonate settling; flow equalization, mixed media filtration; granular acti-
vated carbon adsorption; ion exchange ammonia removal using clinoptilolite; and
disinfection. Sidestream processes include sludge thickening; anaerobic diges-
tion of organic sludges; sludge dewatering using high pressure filtration of
chemical and organic sludges; thermal reactivation of activated carbon; and
clinoptilolite regenerant recovery using chemical clarification and ammonia
stripping. The process flow diagram is presented in Figure A-4.
Sewage originates from the communities on the north and west shores of
Lake Tahoe and along the Truckee River Corridor from Lake Tahoe to Truckee,
California. The area's major industry, tourism, has two primary seasons; the
winter ski season and the summer season. Heavy influx of tourists results in
extremes of sewage flows and loadings over holidays, weekends, and seasonal
peaks.
Effluent from the plant is indirectly discharged to the pristine receiving
waters of the Truckee River and, consequently, the discharge limits (Table A-6)
imposed on the Agency by the State Water Resources Control Board are stringent,
and necessitate a very high degree of treatment. Compliance with COD require-.
ments depends to a great extent on the performance of the GAC system. ; ..
' ' TABLE A-6. TAHOE-TRUCKEE SANITATION
Constituent/unit
COD , mg/L
Suspended solids , mg/L
Turbidity, NTU
Total nitrogen, mg/L
Total phosphorus, mg/L
MBAS, mg/L
Total dissolved solids, mg/L
Chloride, mg/L
Total coliform organisms, MPN/100 mL
AGENCY DISCHARGE
10- sample
average
15
2.0
2.0
2.0
0.15
0.15
440
110
"
'REQUIREMENTS
Daily
maximum
40
4.0
8.0
4.0
0.4
0.4
-
-
23
DESIGN DATA
GAC Adsorption System
Flow
Maximum daily = 4.83 mgd
Peak hourly flows are equalized
by in-plant storage before GAC
treatment
96
-------
TO LANDFILL
vo
CYCLONE SEPARATION
GRIT WASHER
CHLORINE
HYDROGEN PEROXIDE
GRIT DISPOSAL
BYPASS 8,
OVER-FLOW TO
EMERGENCY
GRIT REMOVAL FACILITY
RETENTION BASIN
PRIMARY "™"OG|EN- WASTE SLUDGE THICKENER
CLARIFIERS
ORGANIC SLUDGE
SECONDARY
CLARIFIERS
RAPID MIX
EXHAUST GAS
TO RECARB BASINS
LIME
FERRIC CHLORIDE
CHEMICAL
CLARIFIERS
CHEM SLUDGE
THICKENERS
RECARB
_ CLARIFIERS
SLUDGE DEWATERING
Figure A-4. Plant process diaqram, Tahoe-Truckee Sanitation Agency, CA.
-------
vO
CD
DIGESTED SLUDGE
READY TANK
THICKENED SLUDGE
REGENERANT
BASINS
OtL
PLANT EFFLUENT TO
DISPOSAL BASINS
Figure A-4. (Continued)
-------
COD
Raw wastewater, average
Carbon column influent, average
Carbon column effluent
10-sample average, maximum
GAC contact time, minimum
Granular activated carbon
Makeup carbon unloading
Carbon contractors
Type
Number
Size
- Inlet
Outlet
Column
Rate-of-flow co trol
Backwash
Carbon slurry transfer piping
400 mg/L
26 mg/L
15 mg/L
40 mg/L
20 minutes
400,000 Ib; Westvaco; coal base; 8 x 30
mesh; suitable for 20 cycles of reuse
after thermal reactivation; less than
10% of total loss per cycle; COD to be
reduced from 20 mg/L to 10 mg/L at
detention time of 20 min; carbon shall
attain max COD loading of 0.333 Ib
COD/lb carbon under these conditions
Bag dump, manual with hydraulic eductor
Upflow packed bed; countercurrent
Six
Vertical, cylindrical 12-ft diameter by
20-ft straight sidewall steel shell; 90
psi working pressure; coal-tar epoxy
coated inside; 45° conical top and
bottom; 20-ft carbon bed depth
Eight; 12-in. diameter x 2.5-ft long
Johnson well screens; 304 SS; 100 psi;
0.020-in. slot opening; closed bail
bottom
Same as inlet except screen slot size is
0.015 in.
Equipped with screened vent—combination
pressure air and air-vacuum release
valve
Propeller meter and butterfly valve
Flow can be reversed for cleaning at a
rate of 15 gpm/sf
Black steel; fittings extra heavy
standard radius elbows and tees
99
-------
GAG Reactivation System
The carbon reactivation system includes facilities for the following:
« de-fining and transferring makeup carbon to the carbon columns,
• receiving and dewatering spent carbon,
• feeding spent carbon at a controlled rate to the furnace and thermally
reactivating spent carbon,
• quenching the"carbon and conveying the reactivated carbon to de-fining
and transfer tanks,
• de-fining and transferring reactivated carbon to the carbon columns
and carbon storage tank,
® scrubbing and cooling all exhaust gases,
• storing spent, reactivated, or makeup carbon in the carbon storage
tank.
The major equipment items installed for the carbon reactivation system
include the following:
• two dewatering bins and conveyors,
• '•• two carbon'feeders,
• carbon reactivation furnace exhaust gas cooling and scrubbing
equipment,
» quench tank,
• two transfer and defining tanks,
• one storage tank, and
e" virgin carbon hydraulic eductor and dust collector. ".' :-• ' p"
The reactivation furnace has the following design characteristics:
Number/type
Rated capacity
Odor and particulate control
Hearth loading
Fuel consumption
Particulate emissions
Odor emissions
Steam
Carbon losses through furnace
One; 81-in. diameter by six-hearth BSP -
Envirotech propane fired furnace with
fuel oil standby
3,840 Ib/day of carbon plus 1,980 Ib/day
of adsorbate
Afterburner for odor control; exhaust gas
scrubber for removal of particulates
Less than 45 Ib/day reactivated carbon
per ft2 of hearth
Less than 22,000,000 Btu/day
Less than 0.1 grain/ft3 of gas at
standard conditions
None objectionable at 50 ft from point of
discharge
Up to 200 Ib/hr of steam at 15 psig
available for carbon reactivation
Less than 5 percent per single pass
through furnace
100
-------
Auxiliaries to the reactivation furnace include:
Transfer and defining tanks
Virgin carbon feeder
Dust collector exhaust fan
Storage tank for carbon
Dewatering bins
Carbon feed to furnace
Combustion air blower
Carbon quench tank
Quench tank eductors
Precooler
Scrubber for combustion gases
Induced draft fan
Dust scrubber
Filter
CARBON SYSTEM PERFORMANCE
Two; vertical cylindrical; 8-ft diameter,
7-ft 6-in. sidewalls; dome top; 45°
conical bottom; carbon steel; epoxy
lined; 100 psig working pressure; 304
SS screens with 0.020 in. slots
Hydraulic eductor to defining tanks
2,000 scfm at 10-ft of water
Vertical cylindrical; carbon steel; open
top; 45° cone bottom; 12-ft diameter,
15-ft straight sidewall; screened
overflow
Two; vertical; 7-ft square; 7-ft 6-in.
straight sidewalls; open top; hopper
bottom; carbon steel; drain screen, 304
SS, 4-in. diameter by 3 ft long with
0.020 slots
Two; inclined 304 SS screw conveyors with
variable speed (10:1) drive; drain
connection at bottom with 304 SS
screens with 0.020-in. slots
1,000 scfm at 22 ounces
Single reactivated carbon quench tank
directly below furnace outlet and above
two centrifugal type slurry pumps;
capacity 160 Ib/hr from 1,800° to
150°F; 1-ft 8-in. diameter by 2-ft
6-in. high, with screened overflow;
carbon steel
Two; abrasive-resistant metal; 15 gpm;
1/2 Ib carbon/gal water
316 SS, with spray assemblies
Three-stage impingement; plate-type
vertical unit; pressure drop 9-in. of
water; 50 gpm at 30 psig scrubber
water
Centrifugal type; 800 scfm at 170°F;
static pressure of 18-in. water; 304
SS
304 SS; 3-in. pressure drop; 30 gpm at 40
psig
Combustion air intake; dry type; replace-
able element; washable.
Operating the activated carbon adsorption column has been troublefree, and
the removal process has consistently worked well in removing COD and MBAS. From
May 1, 1979, to November 1, 1979, plant performance for removals of COD, nitro-
gen, and phosphorus have conformed to waste discharge requirements.
101
-------
Indications of COD breakthrough were initially attributed to exhausted
carbon, and the reactivation procedure has been initiated three times to date.
During each reactivation procedure, 20 percent of the carbon in each carbon
column was removed, reactivated, and replaced. This was done contrary .to the
design recommendation of 10 percent carbon withdrawal because of problems in
furnace operation during the first two reactivations that extended the reacti-
vation process over long periods. During the third reactivation procedure, 20
percent carbon reactivation was again undertaken because the peak summer tour-
ist season was approaching and because the reactivation process was performing
so well. In retrospect, it is now believed that the preliminary indications of
COD breakthrough in carbon column operation were the result of high pH water
that reduced carbon removal efficiency rather than the result of completely
exhausted carbon columns.
The scant data accumulated to date regarding carbon removal efficiency an
carbon attrition (Table A-7) may be susceptible to considerable errors because
twice as much carbon as recommended was reactivated each cycle and the actual
state of exhaustion was unknown before reactivation. The extensive carbon make
up during the first reactivation procedure (resulting from unknown losses asso
ciated with the de-fining of the original carbon), the bedding of the furnace,
and the significant burning of carbon during the initial reactivations detract
from the value of preliminary data.
TABLE A-7. ACTIVATED CARBON ADSORPTION, SYSTEM PERFORMANCE
Exhaustion
cycle .
" #2
#3
Average
contact
time,
. min ,
47"' :"'
45
Final
eff .-,
mg/L
11.'3
13.1
COD
Overall
•'removal,
mg/L
13.2
14.0
Carbon
'•mg/L .
52
57
usage"
Ib/mg
431
473
Performance
kg COD removed
"• kg
carbon
'0.26
0.25
spent
Average COD removals through the carbon columns for the first 6 months of
operation are presented in Table A-8. Location of the ion exchange beds with
their fine-mesh media following the carbon columns has proven effective in
reducing losses of carbon fines. Fines that represent both suspended solids
(SS) and COD are continuously washed out of the carbon columns in low concen-
trations and are washed out in high concentrations during the reactivation
transfer procedures.
CAPITAL AND OPERATION AND MAINTENANCE COSTS
Capital Costs
In the Tahoe-Truckee project, bid by contractors in December 1976, the GAC
system was bid as a part of the contract for the entire AWT plant, The cost
information for the carbon adsorption system and the carbon reactivation system
have been taken from the Contractor's bid breakdown, which was used for
102
-------
TABLE A-8. PLANT PERFORMANCE, (6 month averages)
Process stream
Raw
Primary effluent
Secondary effluent
Chemical clarifier effluent
Filter effluent
Carbon column effluent
Final effluent
Discharge requirement
Overall removal , %
SS,
mg/L
156
120
16
17
2.8
2.8
1.3
2.0
99.2
COD,
mg/L
338
-
57
-
26
15
12
15
96.4
T.N.
mg/L
32
34
28
-
-
25
3.7
2.0
88.4
T.P.,
mg/L
10.5
8.7
4.9
0.7
-
-
0.3
0.15
97.1
progress payments on the construction work. On this basis,, the capital costs
are as folows:
Carbon adsorption system (previously described in the text),
including the original charge of GAC = $619,000
Carbon reactivation system = $950,000
It should be noted that these capital costs do not include: (1) the cost
of buildings to house the equipment; (2) electrical costs; or (3) piping costs
for connecting the carbon adsorption system to the processes that precede or
follow adsorption. Piping and valves internal to the carbon system are included
in the capital cost figures given.
Operation and Maintenance Costs
No O&M cost records for the GAC system are available at this time (July
1979), because it is early in the operations phase of this project. The methods
and procedures for breaking-down and accumulating cost records for O&M are
still under organization and development.
Recently, to determine the efectiveness of land treatment (which follows
AWT) in removing COD, the regulatory agencies have temporarily waived the COD
requirement for AWT plant effluent. Therefore, it may be some time before
accurate records of O&M costs for the GAC system become available.
BIBLIOGRAPHY
CH2M/Hill, Consul ting Engineers, TTSA Project Specifications for Carbon
System.
Kennedy, T.J., Butterfield, O.S., and Woods, C.F. Start-up and Operation of the
Tahoe-Truckee Sanitation Agency Advanced Wastewater Treatment Plant 1979.
103
-------
CASE .HISTORY NO. 3
UPPER OCCOQOAN SEWAGE AUTHORITY, MANASSAS PARK, VIRGINIA
MUNICIPAL WASTEWATER TREATMENT
HISTORY AND INTRODUCTION
The Occoquan Reservoir is the principal water supply source in northern
Virginia, serving over 600,000 people. During the past decade, the increased
discharges of conventionally treated sewage caused the water quality of the
reservoir to deteriorate. To reverse this trend, the Virginia State Water Con-
trol Board adopted a comprehensive policy for waste treatment and water quality
management in the Occoquan Watershed. A principal element of this policy
requires a highly sophisticated regional AWT plant to replace existing inade-
quate treatment plants.
The task of meeting this State mandate required that political and juris-
dictional boundaries be set aside to provide a regional solution to a regional
problem. In response to the mandate, in 1971, the jurisdictions in the Occoquan
Basin formed the Upper Occoquan Sewage Authority (UOSA). The member jurisdic-
tions of the Authority include Fairfax County, Prince William County, the City
of Manassas, and the City of Manassas Park (Figure A-5).
The initial project of the Authority was construction of a regional water
reclamation system to eliminate 11 existing treatment plants and reclaim the
wastewater in .an AWT plant.. •';-• •.'•.-. " ' ' •
The treatment standards (Table A-9) are among the highest in the country.
TABLE A-9. ADVANCED TREATMENT REQUIRED TO PROTECT THE OCCOQUAN RESERVOIR*
Weekly average Approximate
Parameter concentration removal rate
BOD5
COD
Suspended solids
Unoxidized nitrogen
Total phosphorus
MB AS
Turbidity
Coliform bacteria
1 . 0 mg/L
10.0 mg/L
1,0 or less mg/L
1 . 0 mg/L
0 . 1 mg/L
0 . 1 mg/L
0.4 JTU
>2/100 mL
>99%
98%
>99%
96%
>99%
>99%
—
>99%
*The effluent limitations established by the State Water Control Board of the
Commonwealth of Virginia are among the most stringent in the country. The
plant effluent discharged to Bull Run, a tributary to Occoquan Reservoir, must
be within the limits listed above.
The project included construction of a 15 mgd wastewater reclamation
plant, 122,000 ft of gravity sewers, 31,000 ft of force mains, and 6 pump
104
-------
o
01
DULLES INTERNATIONAL AIRPORT
WARRENTON
LAKE
WARRENTON
Figure A-5. General location of Occoquan watershed and UOSA project.
-------
stations. In addition, the system has extensive back-up units and emergency
retention facilities to ensure reliability. The regional facilities became
operational in June, 1978. The total cost of the project was $82 million, which
was funded by a combination of authority revenue bonds and state and federal
grants.
Basis of GAC System Design
The purpose of the activated carbon adsorption process is to remove the
soluble organic compounds that are not readily removed by the upstream biologi-
cal, chemical clarification, or filtration processes. Concentrations of the
soluble organic compounds are expressed as mg/L of BOD, COD, and/or MBAS.
Activated carbon adsorption is the most practical method for reducing the con-
centrations of these constituents in the wastewater to meet the treatment
reguirments.
Activated Carbon Adsorption
Number of carbon columns
Type
Size of columns
Effective carbon depth
Carbon size
Sixteen
Upflow, countercurrent
12-ft dia x 24-ft SWD
24 ft 8 in.
8 x 30 mesh
Normal Carbon Column Operation
Columns - in service
••"••• Columns - spare ' ••• •'•'
Flow per column
Surface loading rate
Contact time
Maximum Loading Rate
Columns - in service
Columns - spare
Flow per column
Surface loading rate
Contact time
Eleven
''Five'- '.-••'••'' .•' .'; '•
1.0 mgd (694 gpm)
6.15 gpm/ft2
30 min
Eight
Eight
1.37 mgd (951 gpm)
8.40 gpm/ft2
22 min
Estimated Carbon Dosage
Average
Maximum
250 Ib/mil gal
450 Ib/mil gal
Reactivation (based on carbon dosage 450 Ib/mil gal)
Percent of carbon volume removed
per reactivation cycle
10%
Frequency of reactivation with
30 min contact time 18.7 days
106
-------
Frequency of reactivation
with 20 min contact time 12.4 days
Weight of carbon regenerated
per cycle per column 8,400 Ib
Estimated makeup carbon 10%
Weight of makeup carbon per
cycle per column 840 Ib
Backwash
Supply Filter effluent
Headloss to backwash 15 ft
Duration of backwash 15 min
Backwash rate 13.3 gpm/ft2
Backwash rate per column 1,500 gpm
Operation
The 16 carbon columns are operated in parallel as upflow, countercurrent
beds. They are arranged in four beds to provide maximum flexibility should one
of the influent or effluent lines fail. At the present flow rate of 10.9 mgd,
11 columns are on line and 5 serve as standby units. This high level of redun-
dancy was required since the plant effluent is discharged to a public water
supply.
The optimum flow range is 694 to 951 gpm; lower flows pose the threat of
anaerobic conditions developing in the columns, whereas higher flows cause
escaping carbon fines to plug the outlet screens. Manual controls on the
columns ensure the uninterrupted operation of the columns.
When the average effluent COD from the carbon bed reaches 10 mg/L, about
10 percent of the carbon is removed from the bottom of the bed; an equal amount
of regenerated carbon is replaced in the top of the column. The frequency of
withdrawal ranges between 12 and 34 days depending on the carbon dosage and
contact time.
Periodic backwashing of the columns is necessary to reduce headlosses
through them. This procedure is done when the headloss reaches 8 ft above the
normal clean-bed headloss or at least every 3 days.
Process Control
The carbon column operation is monitored by the plant computer system.
Performance is evaluated using data collected on a preestablished schedule. COD
is used as the basic parameter for determining bed efficiency; declining
efficiency is an indication that regeneration is necessary. Carbon loading is
another operational control. A decrease in loading (normal range is 0.18 to
0.33 Ib COD/lb carbon) is an indication that the regeneration process is not
effective.
107
-------
Basis of Reactivation Design
Annual operation (days)
@ 250 Ib/mil gal
@ 450 Ib/mil gal
83
149
Daily spent carbon production (Ib)
@ 250 Ib/mil gal 2,725
@ 450 Ib/mil gal 4,905
Dewatering bins
Number
Volume - each
Two
416
Feed conveyors (screw)
Number
Capacity - each Ib/hr
-' each Ib/hr •„•'
Reactivation furnace
•Units in service
Diameter
Number of hearths
Hearth loading
Capacity
Fuel requirements
LP gas
Steam requirements •
Makeup carbon requirements @ 10%
Regeneration loss
i 250 Ib/mil gal
@ 450 Ib/mil gal
Storage tanks
Units - in service - spent or
regenerated or virgin
Volume - ft3
Storage @ 250 Ib/mgd
Storage @ 450 Ib/mgd
Wet scrubber
Type
Stack gas
particulate emissions (max)
Gas outlet temperature
Quench tank
Units in service
Two
500 dry reactivated carbon
1,500 spent wet carbon
Two
10-ft 9-in. OD
Seven
45 Ib/ft2/day
12,000 Ib/day
2,750 Ib/day
12,00.0 Ib/day
99,600 Ib/yr
178,800 Ib/yr
One
6,540 ft3
72 days/tank
40 days/tank
Venturi
0.05 grains/SCF
120°F
One
108
-------
Quenched carbon transfer pumps
Units in service One
Units spare One
Type Diaphragm
Capacity 500 Ib/hr
De-fining, receiving, and transfer tanks
Units in service Two
Dry carbon volume 280 ft3 each
Transfer capacity 8,400 Ib/cycle each
SYSTEM PERFORMANCE
Reactivation Process
The plant includes a complete carbon reactivation system for: (1) receiv-
ing spent carbon from the carbon columns, {2) storing and dewatering the spent
carbon, (3) feeding it at a controlled rate to the carbon regeneration furnace,
(4) thermally reactivating the carbon, (5) quenching the carbon, (6) pumping
the reactivated carbon to the transfer and de-fining tanks, (7) storing the
reactivated and makeup carbon, (8) adding new makeup carbon, 9) conveying the
reactivated carbon to the carbon columns, and (10) disposing of the furnace
exhaust gases to the atmosphere (through a quencher and scrubber) without
nuisance from dust, smoke, or odor.
Carbon withdrawn from the carbon columns for reactivation is. discharged
into two dewatering bins. Each dewatering bin holds about 416 ft^ of carbon
(12,368 Ib dry weight), or a total of 832 ft3 in the two tanks (24,736 Ib dry
weight). Typically however, only about 280 ft3 (8,400 Ib) is transferred at
once, equivalent to 10 percent of the volume of one carbon column.
At least 10 minutes is allowed between filling a dewatering bin and with-
drawal to the furnace to allow free water to drain from the carbon, leaving
about 40 percent moisture on the carbon. Withdrawal of carbon from a dewatering
bin is stopped before the screw conveyor is entirely empty to provide an air
seal for the furnace. For continuous reactivation, the second dewatering bin is
filled before the bin in use is empty.
The dewatered carbon is fed to the carbon regeneration furnace by one of
two furnace dewatering feed conveyors. Each feed conveyor is driven by a 2-hp
motor with a mechanical adjustable speed drive.
The multiple hearth furnace has a capacity of 12,000 Ib carbon/day. The
carbon is moved across each hearth by rotating stainless steel rabble arms
equipped with vertically-oriented stainless steel rabble teeth. A modulating
burner control system is provided for each hearth so that the desired tempera-
ture on it can be closely and independently maintained. In the furnace, the
carbon is thermally reactivated in the presence of steam. Cooling air is pro-
vided to the central shaft by a shaft cooling air blower (M-1013) driven by a
15-hp motor. Combustion air is provided to six hearths in the furnace by the
109
-------
use of a combustion air blower. An afterburner blower provides air to the top
hearth of the furnace.
The furnace exhaust gases are discharged through a quencher to precool
them before scrubbing, and through a scrubber to capture particulate matter and
to reduce the exit gas temperature to 120°F. An induced draft fan discharges
the furnace gas to either the C02 compressor or to the atmosphere.
Furnace operation is monitored routinely by tests for the apparent density
of the carbon, supplemented at times by tests for Iodine Number, ash content,
adsorption isotherms, or other laboratory analyses to check the activity of the
regenerated" carbon.
Under normal operation (plant flow at 10.9 mgd and carbon dosage 250
Ib/mil gal), the regeneration furnace is in use about 23 percent of the time.
The carbon is discharged from the furnace by gravity into a quench tank
where the carbon is cooled and put into water slurry. The carbon slurry is then
pumped by either of two diaphragm slurry pumps that take suction from the
quench tank and discharge into the two transfer and de-fining tanks. The pumps
are operated by compressed air, and each delivers 13 to 20 gpm of slurry.
Each of the two transfer and de-fining tanks will hold about 280 ft3
(8,400 Ib dry weight) of carbon, or a total of 560 ft3 (16,800 Ib) in the two
tanks. A flow of water directed .upward through the tank flushes the fines to
the first-stage recarbonation basin with the wash water; the de-fined carbon
remains in the tank.' Carbon.is transferred from the de-fining tanks to the top
of 'the carbon columns by pressurizing the de^-fining tanks with water through a
pressure water connection in ' the top of the tanks. Water flow is established
through the drainline first before carbon is released from the bottom of the
tank. The water flow is continued until the tank and pipeline have been flushed
free of carbon.
• A temporary storage tank is provided for spent carbon in case the reacti-
vation furnace is out of operation. When the furnace is in operation, the car-
bon is transferred through the de-fining tank, to the dewatering tanks, and
processed through the reactivaton furnace as per normal operations. Virgin car-
bon can also be stored in this tank if necessary. Sufficient virgin carbon is
stored to operate the carbon adsorption process for 1 month without the benefit
of the reactivation system.
Carbon losses due to attrition and excess burn-off range from 5 to 10 per-
cent of the carbon regenerated.
Bags of virgin (makeup) carbon are added to the de-fining tanks by a vir-
gin carbon conveyor. The conveyor can discharge 40 Ib/min of granulated carbon.
It works in conjunction with a vibrating virgin-carbon shaker that feeds carbon
granules onto the conveyor. An exhaust fan directs the dust from the hopper to
the scrubber.
110
-------
A portable eductor can also be used to convey virgin carbon (bulk in
trucks) to the de-fining tanks. This unit is capable of supplying 36 gpm and
100 psi of water. The eductor pump capacity is 100 gpm.
A carbon dioxide compressor is provided to compress the stack gas for use
in recarbonation.
Instrumentation and Controls
A central control panel is provided to monitor the carbon regeneration
system operation and to permit the system to be manually started and stopped
and automatically controlled. The most important operating variables are indi-
cated and/or recorded on the panel as summarized below:
Alarms
High and low dewatering tank level (2)
Combustion low air pressure
Furnace high temperature
Flue exhaust - high temperature
I.D. fan - high inlet temperature
Draft loss
Cooling air stack - high temperature
Quench tank high level - water
Quench tank low level - water
Scrubber temperature - high
Transfer and de-fining tank - high level
Induced draft fan - vibration
Separator - high
All alarm functions are annunciated.
Indicators
Dewatering bin bottom valve (2)
Feed conveyor speed (2)
Hearth temperature (6)
Furnace draft and pressure
I.D. fan outlet temperature
Main exhaust stack temperature
Controllers and Switches
Furnace feed conveyors (2)
Dewatering bin bottom valves (2)
Purge
Burner controls
Afterburner turboblower
Combustion blower
Eductor water pump
I.D. fan
111
-------
Center shaft drive
Cooling air fan
Center shaft speed
CO2 compressor
A temperature recorder records the temperature of each of the seven fur-
nace hearths, .the scrubber inlet and outlet gas, the furnace gas outlet, and
center shaft outlet.
A multipoint draft gauge indicates the draft at critical points in the
combustion air and gas system.
The furnace is equipped with an automatic draft control to maintain preset
draft on the furnace. An automatic afterburner air injection system is also
included.
CAPITAL AND OPERATION AND MAINTENANCE
Capital Cost
The capital cost of the carbon system has not been precisely defined. The
carbon adsorption columns and the carbon regeneration furnace were portions of
3 lump sum contracts. The capital costs were approximated by using the contrac-
tors' break out for payment purposes and apportioning the miscellaneous costs.
Carbon adsorption columns $2,250,000
Carbon regeneration system 1,630,000
Total carbon system 3,880,000 . .. .
Operation and Maintenance Cost
The O&M costs for each million gallon for 1979 axe estimated at:
Personnel $13.46
Carbon 8.28
Fuel 7.00
Electricity 4.00
Replacements, repairs, and
miscellaneous 17.26
Total O&M cost $50.00/mil/gal
These estimates are based on very limited data. Carbon has been regener-
ated only twice, and the furnace has not yet been accepted from the supplier.
112
-------
CASE HISTORY NO. 4
AMERICAN CYANAMID CO.
ORGANIC CHEMICALS DIVISION
BOUND BROOK, NEW JERSEY
INDUSTRIAL WASTEWATER TREATMENT
HISTORY AND INTRODUCTION
The Bound Brook, New Jersey, plant of American Cyanamid Company employs
2,200 persons and produces over 1,000 different products including dyes, chem-
ical intermediates, organic pigments, plastic additives, Pharmaceuticals, fine-
chemicals, agricultural chemicals, elastomers, and rubber chemicals. The multi-
ple waste discharges from the production operations are combined with cooling
water, sanitary sewage, and stonnwater runoff from the plant area to form a
single effluent stream that is processed in an AWT facility. Typical ranges of
BOD, COD, Color, and TOC in the raw waste stream are shown in Table A-10.
TABLE A-10. INFLUENT WASTEWATER CHARACTERISTICS
Parameter Concentration range
BOD 270 - 400 mg/L
COD ' 1,000 - 1,400 mg/L
TOC 260 - 350 mg/L
Color 400 - 600 CDAPHA units
Wastewater treatment has been a major part of the Bound Brook operation
for over 30 years. As wastewater technology has progressed with time, the Bound
Brook plant has made significant improvements to its wastewater treatment:
facilities. In the late 1930's, a primary treatment plant was installed at a
cost of $500,000. This plant included a 19 mil gal equalization basin, a lim«
neutralization facility to neutralize 30 tons of sulfuric acid a day, and a 60
mil gal primary settling lagoon.
Beginning in 1949, extensive laboratory and pilot plant studies were ini-
tiated to procure design data for the construction of an activated sludge
treatment plant. This facility was built in 1957 at a cost of $4,500,000 an3
placed into operation in 1958. The secondary waste treatment facilities consist
of six aeration basins, each having a capacity of 3.3 mil gal, six secondary
clarifiers, and a chlorine contact chamber. Figure A-6 illustrates the secon-
dary treatment facilities.
The secondary waste treatment facilities provided more than a 90 percent
reduction in BOD and approximately a 65 percent reduction in TOC and COD.
113
-------
PLANT SEWERS
i i i
RAW WASTE PUMPS
INFLUENT
PUMPS
SOMERSET
RARITAN VALLEY
SEWERAGE
AUTHORITY
• CHLORINE
AIR BLOWERS
-€)
j'
t>
ra
AERATION BASINS
i
Xfl
o
L_.
L.
SECONDARY
(C
$
I
3^
/
CLARIFIERS O
5
o
ID
£
3
o
/f*
( /r*
I/O
7zh
^j-i
/
\<
Ct
UJ
tE
O
IT/
HI ,
III //^
TTTfH^
nil /
11 n /
J'!JJ-_^ WASTE
nTM I SLUDGE
HIM
Mm
inn
i
F'iqure A-G. Waste treatment pJant f]ow rliaqcam IwForo consl'.ruction of AWT plant; American Cyanamid
Company, Pound Brook, NJ.
-------
Unfortunately, many of the organic compounds present in the raw wastewater are
resistant to biodegradation, and significant quantities of nonbiodegradable
organics are present in the secondary effluent. In 1971, it became evident that
additional treatment was needed, particularly for removal of color, odor, toxi-
city, and suspended solids. At times during the year, because the Bound Brook
plant effluent constitutes up to 25 percent of the nearby Raritan River, it
must remain high in quality.
Two alternatives were considered for upgrading the Bound Brook plant
effluent. They were: (1) heavier emphasis on at-source treatment or control in
the manufacturing areas and, (2) additional end-of-pipe treatment (AWT) to
supplement biological treatment. Source control or treatment was not considered
feasible; therefore, AWT processes were studied.
• Bench Scale Studies were conducted to evaluate activated carbon
adsorption, high lime treatment, and ozonation. Of the three pro-
cesses, carbon adsorption was the most favorable for reducing color,
foam, and refractory organics.
• Pilot Plant Studies of granular and powdered activated carbon were
used to determine which was more effective and least costly. Froir.
these, the design criteria for a full-scale GAC plant were
determined.
The advanced treatment plant that was put into service in 1977 consists of
multimedia filtration, carbon adsorption, transport facilities, and carbon
reactivation (Figure A-7).
DESIGN DATA
The design data for the carbon adsorption system for Bound Brook as deter-
mined from.the extensive studies conducted is as follows:
Carbon contactors:
Type Upflow, expanded bed
Number 10 units piped in parallel
Size 16-ft diameter, 40-ft straight sidewall; 45°
cone-shaped columns
Flow 8 gpm/ft2
Detention time 30 min
Carbon 8 x 30 mesh GAC
30-ft depth in each column
Reactivation procedure 5 percent pulse of exhausted carbon removed from
bottom of each column daily; equivalent amount:
of fresh carbon replaced in top of column;
Carbon slurry transferred to equalization
vessel;
115
-------
Slurry dewatered by screw conveyors leading to
furnace;
Reactivated carbon'quenched to ambient tempera-
ture and transferred to storage tank.
Reactivation furnace
Type 8-hearth furnace with afterburner
Size 25-ft, 9-in. diameter
Temperature 1750°F
Capacity 122,000 Ib/day average; 152,000 Ib/day maximum
Carbon Iqss 9%
SYSTEM STARTUP AND OPERATION
Carbon Adsorption
Although the carbon adsorption towers were started up with a minimum of
problems, several problems are worthy of note.
As the start-up and operation of the facility progressed into the winter
months, the areas of the plant that had not been adequately heat traced and
insulated for cold weather operation became immediately apparent. Approximately
4 miles of electrical tracing were needed to adequately heat those areas.
During the start-up, • it was found that each tower has to be pulsed every
30 to 36 hours. Longer frequencies result in a buildup of suspended solids on
the carbon in the lower sections of the towers; the buildup causes the carbon
to bridge across the cone bottom .when an.-attempt is made to pulse the tower.
The bridging can be broken by reversing the pulse flow direction, but this
procedure creates excessive fines in the tower effluent. The upset can be
accommodated only by holding the tower on recycle for an extended time.
Another unexpected problem was the result of the free fall of water from
the adsorbers effluent launders into the main 36-in. final effluent line. Air
was entrained by the free fall (approximately 10 ft in the 36-in. line) and
caused a severe vibration problem. The 36-in. line had been trapped to provide
a sidestream supply to booster pumps feeding the regeneration furnace scrub-
bers. Removing the trap to provide venting solved the vibration problem but, in
turn, necessitated installing an air disengagement tank for the scrubber pump
water supply.
.In the upflow design, carbon fines are purged from a tower during pulsing.
During this operation, the tower is placed on recycle to prevent any carbon
fines and other suspended solids generated by the upset of the carbon bed from
leaving with the final effluent. The initial design returned the adsorber recy-
cle stream containing fines to the multimedia filter influent. The carbon
fines, which range in size from 10 to 100 microns, caused extremely rapid sur-
face blinding of the filter media. As a result, it was necessary to redirect
the fine-bearing recycle streams to the head of the secondary treatment plant.
The carbon fines are now purged from the secondary plant with the waste
sludge.
116
-------
CARBON SLURRY
EFFLUENT FEED FROM CLARIFIERS
—
i.
MULTIMEDIA FILTERS
GAS
GAS SCRUBBER
FRESH. CARBON
MAKEUP
REGEN
CARBON
SURGE
TANK
* f
V
REGEN
W> CARBON
PULSE
TANK
UJ
_l
u
u
UJ
Of.
r t
111 FEED
<|pTANK
DEWATER
SCREW , WATER
REGEN »RECYCLE
FURNACE
B QUENCH TANK
M FURNACE
W PULSE TANK
X)
rn
m
o
>
m
n
5
70
/ADSORBERS ^
>
o
ADSORBER
PULSE TANK
SPENT CARBON
SURGE TANK
SPENT CARBON
PULSE TANK
i'j'.'re A-"7. Advanced v:aste treatment flow diagram, American Cyanarnid Company, NJ.
-------
Carbon Handling
All carbon in the AWT plant is conveyed hydraulically. All of the carbon
movement into and out of the adsorbers storage tanks and reactivation furnace
is controlled automatically by a programmable controller utilizing 8K of mem-
ory. No major problems were encountered with the start-up of the programmer
itself; however, some hardware problems are worthy of discussion.
Standard pivot arm microswitches are used throughout the carbon handling
system to detect and verify valve positions. This type of switch proved very
troublesome in an outdoor installation, especially in freezing weather. Carbon
level sensing also proved to be a difficult proposition. Nuclear carbon level
and carbon-water density instruments were initially installed to detect high
carbon levels and empty pipeline points for the automatic carbon pulsing pro-
grams. Carbon levels proved difficult to detect by this system. After a period
of trial and error, it was found that ultrasonic probes would detect carbon
levels accurately and reliably.
The carbon handling systems utilize piping of 4 in. and 6 in. Pipeline
velocities are controlled at approximately 3.5 ft/sec. Slurry density during
the pulse period averages 0.35 Ib carbon/lb water. At these conditions, very
few line blockages occur and attrition losses are minimized. Some difficulties
were initially encountered when the carbon was moved through lengths of pipe
exceeding 200 equivalent ft and to elevations exceeding 80 ft. Modifications
had to be made to these flow conditions to transport the carbon reliably over
the.longer runs to the higher elevations in the plant. The basic carbon trans-
port loops utilizing .the water pulsing method are designed to permit reversal
of flow through'most -of ...the: system. This allows clearing of-carbon line block-
ages with relative ease.
Carbon Reactivation
The spent carbon is reactivated in a 25-ft 9-in. diameter multiple hearth
furnace. A number of start-up problems were encountered with this unit. One of
the first problems centered around the spent carbon feed system, which consists
of a holding tank, an automatic discharge valve, and a dewatering screw. The
mechanical problems were readily solved, but the balancing of the carbon flow
into and out of the dewatering screw and into the furnace was more difficult. A
very precise balancing was required to ensure a constant feed rate into the
furnace. Variations in the feed rate cause temperature fluctuations that are
reflected through the upper hearths. To control the furnace operation, a steady
temperature profile must be set and maintained. Good control is needed over the
carbon discharge flow rate. The carbon tends to discharge from the furnace in
slugs as the rabble teeth move the carbon to the discharge point. As the carbon
is quenched, steam is generated, and as the steam rises through the furnace, an
uncontrollable draft fluctuation occurs. It was necessary to throttle and bal-
ance the carbon discharge flow to maintain a controllable furnace internal-
pressure. Separation of the two steps of carbon dewatering and conveying by
installation of dewatering tanks ahead of the screw conveyors might have
avoided these problems.
118
-------
The last problem with the reactivation furnace concerned the very rapicl
and unexpected failure of two hearths after approximately 3 months of opera-
tion. Hearth No. 7 failed somewhat dramatically while the furnace was operating
at design temperature. Approximately 75 percent of the hearth inverted and then.
collapsed when the furnace cooled down. During the cool-down, Hearth No. 5
flattened to the point that it had to be removed also. Dependence on a single
reactivation unit necessitated resumption of operation with two hearths
removed. Carbon reactivated under these conditions was acceptable, with appar-
ent bulk densities in the 0.49 to 0.51 gm/cc range at feed rates of 90,000
Ib/day.
Reactivation performance has been difficult to control and optimize
because of the size of the furnace. It can take several days of operation to
reach a desired operating point. When an operating variable is altered, several
hours are required for the furnace to reach equilibrium and reflect the change*
The constantly changing nature of the primary effluent discharge from the Bound
Brook plant produces a varying adsorbate loading on the carbon that further
complicates determining the optimum furnace conditions. These problems have
also made it difficult to determine accurately the turndown capabilities of the
reactivation funace. Acceptable bulk densities can be obtained by relying on
operating experience to set reactivation conditions, but optimization of these
conditions has not yet .been achieved.
Most of the described start-up experiences and problems probably are typi-
cal for a 20 AWT facility. The problems, although somewhat peculiar to a chemi-
cals manufacturing plant, have been resolved. The overall plant has been in
continuous operation since December 1977, and treatment performance has been
basically as predicted by the 40 gpm pilot plant.
SYSTEM PERFORMANCE
The most meaningful performance data for the Bound Brook facility were
collected during the pilot plant studies. Figure A-8 shows the TOC breakthrough
curves developed from one run of the 5-in. diameter carbon columns. The four
columns were operated as downflow packed bed units piped in series.
Refinement of the pilot studies led to the operation of a 30-in. diameter
pilot plant testing program. Figure A-9 shows the performance of this system
relative to the removal of TOC. In this study, a single column was operated as
an upflow expanded bed unit. A pulse of exhausted carbon was removed from the
bottom of the column each day and an equivalent amount of fresh carbon (either
virgin or regenert ted) was added to the top of the unit. This unit was the pro-
totype (in terms of operation) for the full scale plant.
Laboratory analyses of dissolved solids and nitrogen were run on the
influent and effluent of the initial 5-in. pilot system. Average dissolved sol-
ids and ammonia nitrogen removals were minimal—3 and 2 percent, respectively.
On the other hand, the average organic nitrogen removal was 85 percent.
119
-------
N)
O
90
80
o> 70
D 60
OC
O
2 50
1
O 40
2 30
.20
to
FLOW RATE: 3.4 gom/sq ft
I J I
10
- 15 20
TIME - DAYS
25
30
Figure A-8. Breakthrough curve for total organic carbon; run '2, 5 in. diameter columns.
-------
Z
o
CO
a:
<
o
o
<
(5
tt
O
60
50
30
20
10
PULSE CONDITIONS: VIRGIN CARBON 2.5% per day
FLOW RATE: 8 gpm/ft 2
EFFLUENT
10
15
20
25
30
TIME - DAYS
Figure A-9. Total organic carbon removal; 30-in. diameter pilot carbon column.
-------
CAPITAL AND OPERATION AND MAINTENANCE COSTS
Detailed cost estimates for this facility are not available; however, the
entire Bound Brook AWT plant cost $23 million. The annual operating costs for
these facilities currently exceed $9 million.
BIBLIOGRAPHY
Askins, W., Marek, A. C., and Wilcox, D. R., Tertiary Treatment of an Organic
Chemicals Plant Effluent, Proceedings 9th Mid-Atlantic Industrial Waste Treat-
ment Conference, Bucknell University, August, 1977.
Cherry, A. B., Biological Treatment of a Complex Chemical Waste, presented at
Southwestern Regional Meeting of the American Chemical Society, Richmond, Vir-
ginia, November 5-7, 1959.
Grube, J. R., Procuring the Design Data in Order to Uprate and Optimize the
Operation of an Industrial Waste Treatment Facility, Proceedings 22nd Annual
Purdue Industrial Waste Conference, West Lafayette, Indiana, May, 1967.
Lamb, J. C., The Cyanamid Story at Bound Brook, presented at Air Stream Control
Seminar, Bound Brook, New Jersey, June 4, 1958.
Marek, A. C., Jr. and Askins, W., Advanced Wastewater Treatment for an Organic
Chemicals Manufacturing Complex, U.S./U.S.S.R. Symposium on Physical/Chemical
Treatment, Cincinnati, Ohio, November, 1975.
Marek, A. C., Jr., and Pikulin, M. A., Start-Up and' Operation of the 20 MGD
Granular Activated Carbon Treatment Facility, Proceedings 33rd Annual Purdue
Industrial Waste Conference, West Lafayette, Indiana, May, 1978.
Powell, S. T. , Lamb, J. C., and Ritter, R. H., Industrial and Municipal Coop-
eration for Joint Treatment of Wastes, SIW, 31:1044-1058, September, 1959.
122
-------
CASE HISTORY NO. 5
VALLEJO SANITATION AND FLOOD CONTROL DISTRICT
VALLEJO, CALIFORNIA
PHYSICAL-CHEMICAL WASTEWATER TREATMENT
HISTORY AND INTRODUCTION
In 1972/ work was started to convert Vallejo's existing conventional 7 mgd
primary wastewater treatment plant to a 13 mgd physical-chemical plant utiliz-
ing GAC. The plant was to produce an effluent with an average BOD of 30 mg/L
and an average suspended solids of 30 mg/L. The decision to use physicalchemi-
cal treatment rather than biological secondary treatment was reached after
engineering studies of the existing treatment facilities and after pilot plant
tests were completed. Plant design and construction were completed- in 1978. The
revised and enlarged plant has been in start-up and operation since that time.
Some mechanical problems have been experienced, but the physical-chemical
treatment plant appears to be capable of meeting the discharge standards for
Carquinez Strait. This is based on rather limited operating experience.
DESIGN DATA
The plant is designed for an average dry weather flow of 13 mgd and a max-
imum daily flow of 27.7 mgd (Figure A-10). The plant processes include:
1.. Prechlorination
2. Screening
3. Grit removal
4. Lime addition to pH = 10.2
5. Polymer addition
6. Rapid mixing
7. Flocculation
8. Sedimentation
9. Single stage recarbonation with standby hydrochloric acid feed to pH
= 8.0
10. Granular activated carbon adsorption
11. Chlorination
12. Polymer addition
13. Dual media filtration
14. Sludge :. torage
15. Vacuum ^iltration of sludge
16. GAC reactivation in a multiple hearth furnace
The removal factors of the plant unit processes at mean daily flow are
given in Table A-11.
123
-------
SETTL/VG t
KECA&S-
QUATIOKJ
DOAL
Mr'0!A
ro
FOR SLUDGE
OEWATEPIUG
QUEKICH
TAUK
TO
COLUMJUS
Figure A-10. Flow diagram of Vallejo wastewater physical-chemical treatment plant, Vallejo, CA.
-------
TABLE A-11. REMOVAL FACTORS OF UNIT PROCESSES AT MEAN DAILY FLOW
Data for 6 months
ending 8-31-79
Element
COD
BOD
SS
Raw Floe. &
sewage sed.
424 424 to 140
= 284
67%
223 228 to 80
= 143
64%
368 368 to 70
= 298
81%
GAC
adsorption
175 to 90
= 85
49%
104 to 45
= 59
57%
66 to 22
= 44
67%
Design estimates based on
pilot plant tests
Filtration
90 to 45
= 45
50%
45 to 25
= 20
45%
22 to 7
= 15
46%
Effluent
45
89%*
25
89.5%
7
97%*
Unit
mg/L
mg/L
percent
mg/L
mg/L
percent
mg/L
mg/L
percent
*0verall removal.
Basis of GAC System Design
At low flows, carbon columns are taken off the line as necessary, to limit
carbon contact time to not more than 30 minutes, limiting the time avoids
hydrogen sulfide production that might occur under longer detention times. The
content of carbon column influent is about 5 mg/L. Figure A-11 .is a crosssec-
tion through a carbon bed.
Carbon adsorption;
Number units (contactors
Size-length, width, depth
Area per unit
Flow rate (upflow mode)
Flow per unit
Carbon size
Carbon bed expansion
Carbon head loss
Air scrubbing
Flushing rate
Carbon bed expansi jn
when flushing
Carbon head loss
when flushing
Carbon depth
Carbon volume per unit
Carbon active capacity
in 6 beds
Min.
Max
Six
18 ft x 30 ft
324 ft2
6 gpro/ft2 @13 mgd 10 gpm/ft2 @28 mgd
1,540 gpm 3,300 gpm
12 x 40 mesh
10% 40%
4 ft 5 ft
5 scfm/ft2
15 gpm/ft2
50%
5.92 ft
16 ft
5,190 ft3
31,000 ft3
125
-------
Carbon adsorption;
Carbon bulk density,
drained
Carbon contact time
Carbon active capacity
in 6 beds
Carbon inactive storage
in 2 beds
Carbon inventory
Carbon adsorption
Carbon dose
Carlson exhaust rate
Exhaustion of active carbon
Carbon loss per regeneration
Carbon refill rate
Carbon reactivation;
Number units
Total hearth area
Number hearths
Diameter hearths (I.D.)
Furnace loading
Furnace loading
Furnace operation time
Operation
Residence time
Steam required
Afterburner
Scrubber
Carbon dewater unti
Carbon quenching unit
Carbon transport pumps
Min.
24 lb/ft3
Avg.
or
Fixed
27 lb/ft3
25.2 min.
840,000 Ib
Max
30 lb/ft3
11.7 min
280,000 (storage)Ib
1,120,000 Ib
0.5 Ib COD/lb carbon
1,410 Ib/mil gal
18,400 Ib/day
46 days
7.5%
505,000 Ib/yr
One
402 ft2
Six? BSP furnace
12 ft I.D.
3 Ib carbon/ft2/hr
1,2 06 Ib carbon/hr
64% . , '
19 '24-hr days/mo
1 hr
1 Ib steam/Ib carbon
One
One
One
One
Two
Spent carbon is moved directly from the contactors to a small bin above
the dewatering and conveying screw that feeds the reactivation furnace using
eductors permanently mounted at four fixed locations in each contactor.
Eductors also convey the reactivated carbon back to two extra contactors, which
are used as storage for makeup and reactivated carbon.
The construction specifications for the project included a requirement for
testing the performance of the carbon reactivation system. A copy of this
report concludes this plant description.
CAPITAL OPERATION AND MAINTENANCE COSTS
The available construction cost data are limited to those given in Table
A-12.
126
-------
Water
level—
I&'-O" Square
"*~2-*fey;
Reinforced concrete
Effluent troughs
(-
12*40 mesh GACj
Plenum
• Air for air scour
' Influent f backwash
Figure A-ll. Cross section through carbon column (expanded bed upflow),
Vallejo Sanitation and Flood Control District, CA.
127
-------
TABLE A-12. GAC SYSTEM CONSTRUCTION COST SUMMARY*
Item . . Bid price*
GAC $ 443,000
GAC contactors 2,486,000
Gac regeneration system 1,430,000
*GAC contactors bid price includes all intra-process piping and instrumentation
and GAC storage beds. GAC regeneration system bid price includes a gas cool-
ing, conveyance, and compressor system used in the recarbonation process. No
bid. price break out is available for this gas system.
tBids were received in October 1974.
The cost data for O&M is also rather limited because of the short tiiae the
plant has been in full operationo
Plant operators estimate that they spend 3 hr/day/shift, or a total of 9
hr/day, in backwashing and air scouring the carbon adsorbers. At $11 per hour,
this comes to a total of $99 per day for operation of the carbon adsorption
system.
Cost data for GAC reactivation are not available at this time.
128
-------
PERFORMANCE TEST REPORT*
CARBON REGENERATION FURNACE
VALLEJO, CALIFORNIA
JOB NUMBER 5014-21t
This performance test was successfully conducted on the 12'-9" O.D. x 7
Hearth Carbon Regeneration Furnace System to meet the objectives of specified
carbon loss and stable furnace operation. The test was conducted starting at
10:35 AM, November 15, 1978 and ending at noon on November 17, 1978.
The activated carbon regenerated was ICI America granular carbon, 12 x 40
mesh, which had been used to adsorb organics from line treated and recarbonated
sewage at the Vallejo Sewage Treatment Plant.
The individuals involved during the testing periods were:
BSP Envirotech Kaiser Engineers
Gene Augustine Robert C. Carr
Tom Hynes
Jerry Isaak
Scott Hume • C. Norman Peterson
Ivan Joya
Dave McKissic
' Larry Thompson
ICI America
Lou Billello
Ron Sheaffer
SUMMARY AND CONCLUSIONS
The test was conducted in general accordance with the Performance Test
Procedure of August 21, 1978 which also provided for a Pre-performance Test
operating period.
The following objectives were accomplished during the Pre-performance Test
period:
1. The method of measuring the carbon bed depth was established and
agreed upon by Kaiser Engineers, ICI and BSP.
2. The rotary feed valve was calibrated to establish a curve of furnace
feed rate vs. rotary valve set point.
*Decen:iber 1, 1978. I. Joya, Senior Process Engineer
tBSP division of Envirotech Corporation
129
-------
3. The carbon regeneration system was operated for approximately five
days during the week of November 6 and early in the week of November
13 to establish reliable operation prior to the Performance Test.
During, the 48-hour Performance Test of November 15 - 17, 1978, the follow-
ing objectives were accomplished:
1. The carbon furnace was continuously operated at a feed rate of
approximately 1310 Ib/hr which is in excess of the specified 1206
Ib/hr.
2. The carbon loss from adsorber through regeneration and returned to
adsorber was 3.5% which is well within the specified loss of 5.5%
maximum.
3. Regeneration of the carbon was successful based on apparent density
and iodine number analyses made on composite samples collected during
the test. The average analytical results are summarized below and
the complete report by the independent laboratory appears in Appendix
III.
Spent carbon Regenerated carbon Virgin carbon*
Apparent density,
gm/cc .441 .403 .4
Iodine number .451 634 600 min.
*Specification values. Abrasion number, 25 max.
4. Sieve analysis of the collected samples indicated no significant
change in particle size distribution. In all cases, the spent and
regenerated samples exhibited the virgin carbon specification values
of no more than 5% of the particles larger than 12 mesh and no more
than 5% of the particles smaller than 40 mesh.
5. Successful regeneration was accomplished at temperatures in the
neighborhood of 1500°F.
TEST PROCEDURES
Bed Height Measurement
Measurements were made to establish bed height in Test Adsorber No. 2
before and after the 48-hour regeneration. The method of measurement was first
to establish a backwashed and settled bed with a surface as level as possible,
this was done by:
a. air scouring for 5 minutes,
b. settling until most of the air was evolved,
c. backwashing at high flow rate for 10 minutes,
d. settling for 10 minutes.
130
-------
Measurements were then taken at 17 points within the adsorber from a ref-
erence point at the top edge down to the top of the bed. Each measurement was
made by suspending a weighted screen attached to a steel measuring tape. The
bed measurements recorded appear in Appendix I.'
This method of measurement was established as reliable by Kaiser and BSP
during Pre-preformance Test trials of August 12 and 14, 1978.
Feed Rate Measurement
Furnace feed rate was established by calibrating the rotary feed valve
which regulates the flow of carbon from the furnace feed tank to the dewatering
screw. This calibration was made by routing the valve discharge into a 20 gal-
lon drum; then for several valve set points, the time interval, revolutions and
volume of carbon delivered were measured. Using these values along with a car-
bon density of 25 dry Ib/cf, a graph of feed rate vs. valve set point was made.
This graph appears in Appendix II. During the test, a valve set point of 1.8
was maintained.
Carbon Loss Determination
Prior to the test, the 5.5% specified carbon loss was equated to a loss of
bed height of about 4 3/4 inches. This is indicated in a letter of September 1,
1978 from Kaiser Engineers which appears in Appendix I. The calculations to
arrive at the value of 4 3/4 inches are as follows:
Allowable Carbon Loss = 1208 ^ x 48 hr x 0.055 = 3189 Ib.
TTr
Carbon Bed Inventory = 18 ft. x 18 ft. x 25 lb/ft3 = 675 ib/in.
w in/rt.
Allowable Loss of Bed = ? , . = 4.72 say 4 3/4 in.
Height b/5 iryin
The actual loss of bed height was 3 in. which corresponds to a loss of
3.5%; less than the specified 5.5%. The record of bed measurements appears in
the Appendix.
Data Collection
Furnace operating data was taken at hourly intervals and recorded in log
sheets which appe? r in the Appendix.
Sampling and Analyses
Composite samples of furnace feed carbon and furnace product carbon were
made by placing approximately 6 oz. into a quart container every 2 hours. Every
8 hours a composite was completed. These samples were analyzed by Ultrachem
Corporation of Walnut Creek, California for apparent density, iodine number and
131
-------
sieve analysis. The procedures used were those appearing in the D.S. EPA
Process Design Manual for Carbon Adsorption.
A sample of the carbon transported from the adsorber to the regeneration
area was analyzed for grease content. The value determined equates to
approximately 0.2% by weight.
Regeneration efficiency was monitored during the test by frequent analyses
of apparent density and some iodine number determinations. These analyses were
mainly conducted by Mr. Lou Billello of ICI. Adjustments in furnace shaft speed
anci steam rate were made based on these analytical results.
132
-------
APPENDIX III.
ACTIVATED CARBON ANALYTICAL RESULTS
November 30, 1978
B.S.P. Envirotech
One Davis Drive
Belmont, CA 94002
Attention: Mr. Ivan. Joya
Report #17473-revised Vallejo Job #5041-21
Subject; Analysis of activated carbon.
Samples Submitted; On November 20, 1978, we received twelve activated carbon
samples for apparent density, sieve analysis and iodine number. We also
received one activated carbon water sample in a 300 mL glass jar for grease
analysis. It was requested that all samples be analyzed on a "rush" basis.
The activated carbon ' samples were hand labeled: "BSP Vallejo 8 hr Composite"
and were distinguished by.the following labeling:
UC# Date Time Sample type
1 11-15-78 1200-1800 MHF Feed
2 11-15-78 1200-1800 MHF Product
3 11-15-78 2200-0400 MHF Feed
4 11-15-78 2200-0400 MHF Product
5 11-16-78 0600-1200 MHF Feed
6 11-16-78 0600-1200 MHF Product
7 11-16-78 1600-2200 MHF Feed
8 11-16-78 1600-2200 MHF Product
9 11-17-78 0000-0600 MHF Feed
10 11-17-78 0000-0600 MHF Product
11 . 11-17-78 0800-End MHF Feed
12 11-17-78 0800-End MHF Product
13 BSP Vallejo 2030 11-16 Sample from Bed Eductor, Grease
Procedures; Procedures followed for activated carbon sample analyses were sup-
plied by Mr. Ivan Joya to Ultrachem.
Note: Per telephone conversation of November 21, 1978, with Mr. Vic Feudale the
procedure 'for sieve analysis was modified to extend shaker time to a total of
five minutes instead of the recommended three minutes.
133
-------
All activated carbon samples were dried at 140°C for one hour preceeding
analysis.
Sample from bed educator was analyzed for grease. The jar contained approxi-
mately 250 mLs; 90% water and 10% activated carbon. The water portion was
extreicted with freon in a separatory funnel and the charcoal portion was
extracted with hexane using a sohxlet extraction.
134
-------
uc #
1
2
3
4
5
6
7
8
9
10
11
12
Results
1.
Appareni
Density
gm/cc
0.453
0.404
0.439
0.389
0.442
0.414
0.436
0.402
0.442
0.412
0.433
0.399
.
L-
Iodine
Number
331
624
403
650
438
627
516
619
52.5
645
492
639
Sample from Eductor
Sieve Analysis Wt % Retained on Screen
#12
0.09
0.20
0.13
0.24
0.08
0.12
0'.24
0.10
0.17
0.12
0.22
0.16
(#13)
#14
0.70
1.46
0.66
1.19
0.60
0.72
1.21
0.69
1.01
1.04
1.00
0.93
Aqueous portion (freon extractable)
Charcoal portion (hexane extractable)
#20
45.1
62.3
44.2
51.7
41.3
45.8
47.9 •
39.7
43.6
45.6
41.2
43.2
#30
49.1
33.7
47.5
42.7
51.1
47.2
44.7
48.2
47.1
46.4
48.0
46. .6
#40
3.53
1.94
5.04
3.41
' 4.71
5. .06
4.20
6.92
5.83
5.35
6.61
6.42
Pan
1.51
0.47
2.47
0.77
''2.22
• 1.13
1.74
4.31
2.25
1.52
2.95
2.72
mg grease
1.7
19.2
Total mg
20.9
135
-------
CASE HISTORY NO. 6
ORANGE COUNTY WATER DISTRICT
FOUNTAIN VALLEY, CALIFORNIA
MUNICIPAL WASTEWATER RECLAMATION
AND GROUNDWATER RECHARGE
HISTORY AND INTRODUCTION
Water Factory 21 is the AWT plant operated by the Orange County Water Dis-
trict as part of their groundwater injection project. The plant has been
designed to improve the quality of secondary effluent by providing lime treat-
ment, air stripping, recarbonation, chlorination for algae control, filtration,
granular activated carbon adsorption (GAC), reverse osmosis, and final chlorin-
ation for disinfection and ammonia removal. A schematic of the basic process
train is given in Figure A-12.
DESIGN DATA
Carbon Columns
Flow 15 mgd (influent to plant is secondary
- . .• -effluent) • • • • '
GAC columns . . (see Figure A-13) ' . •
Number Seventeen
' ;:Size "... . . •'. 12-ffc diameter; 4'1-ft height; 24-ft
'••'•' •" ' " ' .., sidewali height; conical top and bottom
Empty bed contact time 34 min
Surface loading 5o9 gpm/ft
Flow Originally upflow countercurrent changed
to downflow to reduce escape of carbon
fines
Carbon Filtrasorb 300 8 x 30 mesh; 26 lb/ft3
bulk density; 38 tons/column
GAC Reactivation System (Figure A-14)
The carbon reactivation process follows several basic steps:
1. Exhausted carbon extracted from column as slurry is transferred to
dewatering bins.
2. Carbon is drained for 10 min.
3. Moist carbon is transferred by stainless steel screw conveyor to top
of furnace.
136
-------
STRIPPING
CHEMICAL
CLARIFICATION
Q1*
INFLUENT
CaO
i
LIME SLUDGE
CHLORINATION
Q8
FILTRATION
RECARBONATION
A
i,
t
1
Q9 EFFLUENT
22A
22B
REVERSE OSMOSIS
Q6
ACTIVATED
CARBON
ADSORPTION
Figure A-12. Schematic diagram of 15 mgd (3 mgd reverse osmosis) advanced water reclamation
plant, Orange County Water District, CA. (*Denotes sampling point.)
-------
INFLUENT TO
WASTE VALVE
DRAIN VALVE
MANIFOLD
XX
•MANIFOLD
INFLUENT LINE
FLOW TUBE
EFFLUENT FLOW
CONTROL VALVE
CROSS OVER
CONNECTION VALVE
WASTE DRAIN LINE
Lgure A-13. Schematic diagram of granular activated carbon contactor
vessels, Orange County Water District, CA.
138
-------
EXHAUSTED CARBON SLURRY
CARBON
DEWATERING
TANK
CARBON-^
COLUMN
CARBON
FILLING
CHAMBER
REGENERATED
AND MAKE UP
CARBON SLURRY
MAKEUP
CARBON
ft
MAKEUP
CARBON
WASH TANK
CARBON
REGEN-
ERATION
FURNACE
PRESSURE
WATER
MAKEUP CARBON
PRESSURE
WATER
PRESSURE
WATER
REGENERATED
CARBON SLURRY
REGENERATED
CARBON WASH TANK
SCREENED
OVERFLOW
QUENCH TANK
PRESSURE
WATER
CARBON
SLURRY
PUMP
Figure A-14.
Carbon regeneration system, Orange County Water District,
CA.
139
-------
4. Carbon moves through furnace with the aid of four stainless steel
rabble arms on each hearth.
5. Regenerated carbon is cooled in a quench tank and pumped to a
de-fining chamber.
6. De-fined carbon is returned to GAC adsorbers for reuse.
The reactivation furnace has the following characteristics:
Type Six-hearth, gas fired
Rated capacity Six tons dry carbon/day
Temperature 1700°F
Afterburner To combust volatilized organics
Venturi wet-scrubber To remove carbon fines from furnace
exhaust
CARBON SYSTEM PERFORMANCE
Between January 1976 and January 1979, there were three distinct operating
periods for Water Factory 21:
• 1/76 to 10/76 - Trickling filter influent, no breakpoint chlorina-
tion, no reverse osmosis, no injection
o 10/76 to 3/78 - Trickling filter influent, breakpoint chlorination,
no reverse osmosis, injection
e ' 3/78 to 1/79 - Activated sludge influent, no forced air circulation
in stripping, partial ammonia removal by chlorination, reverse
osmosis, injection
During the second and third periods, the GAC columns were effective in
removing COD and TOC from the flow. During period two, the COD was reduced by
60 percent and the TOC by 51 percent. About 14 percent of the GAC in each col-
umn was regenerated every 40 to 70 days during this time. In period three, the
COD removal was 49 percent and 50 percent of the column carbon was regenerated
every 6 months. Trace organics and heavy metals were removed to various degrees
during these two test periods,
Figure A-15 illustrates the effect of carbon regeneration on COD remov-
als. The change in influent to the plant from trickling filter effluent to
activated sludge effluent significantly reduced the COD loading to the carbon
columns and also reduced the frequency of regeneration.
Test results showed that trihalomethanes (THM) are removed much more
effectively with fresh GAC than with old GAC. Also, the more highly brominated
THM are removed more efficiently than chloroform. The biggest problem in asses-
sing the results of the THM testing was the lack of sufficient analytical
precision when concentrations are near the detection limit.
140
-------
60
50
40
o
3 30
o
u
20
10
°t
•ft \ inn nw
r* uj i uun
-• TRICKLING FILTE
EFFLUENT
-A
rv* '
CARBON REGENER/
x\/
1 1 t 1 1
1
n MI
i
\
\\
\ \
V
v
VTION
\s*
nnwHFi nw
\J\jnv\r\- .VJn ^
« A/ — nvATrn ^i unrr FPFI IIFKIT b.
^ AL 1 IVA 1 LU1 JLUUUL ul i LU.LN 1 *•
CARBON REGENERATION
\
\
ty\^r\^j ^
\^\^ — v^
y \^- *-
) 20 40 60 80 100 120 140 160 180 200 220 240 260 280
U/
V
300 320 340
360
THROUGHPUT VOLUME, MG
Fiqure A-1.5. Influent and effluent COD for a typical GAC column during the latter part of Period 2 and
into Period 3.
-------
CAPITAL AND OPERATION AND MAINTENANCE COSTS
Capital Costs
The capital costs for the GAC system at Water Factory 21 totalled $3.3
million in 1972. These costs include sitework, landscaping, electrical and yard
piping, but not land. The breakdown of the costs is as follows:
Adsorption $1,432,000
Furnace 489,000
Building, controls, etc. 1,386,000
$3,307,000
OPERATION AND MAINTENANCE COSTS
The unit costs of operating the GAC system at Water Factory 21 from July
1, 1978, through June 30, 1979, are presented in Table A-13. The total quantity
of effluent treated was 3,991.4 mil gal, the total quantity of carbon processed
was 694 tons, and the total make-up carbon was 40 tons.
TABLE A-13. ANNUAL OPERATION AND MAINTENANCE COSTS FOR THE GAC SYSTEM
AT WATER FACTORY 21 (1978/79)
' . . Cost, $/mil gal
Adsorption
Operations labor, administration, and engineering . / $'6.26
Maintenance labor 8.76
Electricity 10.78
Maintenance materials, tools, etc. 4.00
Subtotal 29.80
Amortized capital • 55.43
Total $ 85.23
Reactivation
Operations labor . $ 25.05
Maintenance labor 8.75
Electricity 3.59
Gas 4.65
GAC make-up 12.02
Maintenance materials 6.63
Subtotal 60.69
Amortized capital ' 20.55
Total $ 81.24
Total for GAC system $166.47
142
-------
BIBLIOGRAPHY
Argo, D. G., "The Cost of Water Reclamation by AWT", Presented at 51st WPCF
Conference, Anaheim, California, October, 1978.
David Argo, personal communication, August, 1979t
McCarty, P. L., Argo, D. G., and Reinhard, M., "Operational Experiences with
Activated Carbon Adsorbers at Water Factory 21", May, 1979.
McCarty, P. L.,.Argo, D. G., 'and Reinhard, M., "Reliability of AWT", Presented
at Water Reuse Symposium, Washington, D.C., March, 1979.
McCarty, P. L., Reinhard, M., Dolce, C., Nguyen, H., and Argo, D. G., "Water
Factory 21: Reclaimed Water, Volatile Organics, Virus, and Treatment Perform-
ance". EPA-600/2-78-076. MERL, EPA, Cincinnati, Ohio, 1978.
143
-------
CASE HISTORY NO. 7
CITY OF NIAGARA FALLS, NEW YORK
MUNICIPAL WASTEWATER TREATMENT
HISTORY AND INTRODUCTION
Niagara Falls' first municipal wastewater treatment plant was completed in
1939. In 194Q, the plant was treating an average dry weather flow of 62 mgd. A
series of intercepting sewers were built at the same time to collect domestic
and industrial wastes and convey them to the treatment plant. A diversion sewer
was built to collect unpolluted cooling water from industries in the Buffalo
Avenue industrial complex and discharge it through the Adams Generating Plant
Tailrace Tunnel to the river just upstream of the Rainbow Bridge.
This system served the needs of the City and industry until the early
1960's when concern about the adequacy of the sewerage system was expressed. In
1964, final plans were completed for upgrading the then existing plant at Ash-
land Avenue to provide primary treatment. These plans were abandoned when new
state and federal anti-pollution regulations required a minimum of secondary
treatment. Further progress on the improvement of the City's wastewater facili-
ties was delayed during preparation of comprehensive sewerage studies, which
were undertaken for all of Niagara County and for several municipalities in the
Niagara Falls area.
In early 1970, a large environmental engineering firm was retained to make
a preliminary study, prepare final designs, provide services during construc-
tion,, and assist during start-up of new municipal wastewater treatment facili-
ties., Analyses indicated that extensive pilot plant studies were required to
select a treatment methodology. Although the City had to provide the equivalent
of secondary treatment to comply with international, federal, and state efflu-
ent regulations, the large volume of highly toxic waste generated by local
industries precluded use of conventional biological treatment processes.
A 1971 report, based on several months of pilot testing, recommended con-
struction of a plant incorporating physical-chemical treatment of the combined
industrial and municipal wastes. The City and regulatory agencies accepted the
report and subsequently gave the go-ahead to prepare final plans for a 48 mgd
AWT plant and a number of collection system improvements, such as a 2-mile
long, 6-ft diameter, concrete-lined rock tunnel intercepting sewer; a 20 mgd
wastewater pumping station located on the old plant site; force main and
gravityflow conduit to the new treatment plant on Buffalo Avenue; and connec-
tions to convey 29 mgd of industrial waste and 19 mgd of municipal wastes and
estraneous flow to the interceptor system.
Construction on the system improvements and the treatment plant began in
1973. The site of the wastewater treatment plant was originally developed
between 1890 and 1904 as the Adams Generating Plant. Construction of the
improvements was essentially completed in 1977, and flow was pumped to the
144
-------
treatment processes in mid-April 1977. The treatment plant is located on a
22.6-acre site about a mile above the celebrated Niagara Falls in the south-
western section of the City.
The cooperative spirit between industry and the City of Niagara Falls made
the improvements possible; 18 of the larger industries provided the local share
of construction costs. These industries, through the Industrial Liaison Commit-
tee, provide valuable guidance and support in regulating the quality of the
sewage received at the wastewater treatment plant. In addition, the technical
expertise and financial help of the state and federal pollution control agen-
cies made the project a successful reality. Niagara Falls is proud of this
spirit and considers it an integral part of the areawide wastewater treatment
system mandated by the Water Pollution Control Act Amendments of 1972.
DESIGN DATA
The Niagara Falls Wastewater Treatment Plant is a physical-chemical plant
that incorporates trash removal, chemical addition, rapid mixing, flocculation,
sedimentation, acid addition and mixing, GAC content in downflow filters, and
chlorination (Figure A-16).
Basic design data for the plant are given in Table A-14.
TABLE A-14. BASIC DESIGN DATA, NIAGARA FALLS AWT PLANT
Constituent
Population served
Design wastewater flows
Average mgd
Peak mgd
1970
86,000
1990
114,000
48
85
2020
132,000
48
85
Key system units
Flocculation basins (60 ft x 60 ft x 12 ft)
Sedimentation basins (250 ft x 60 ft x 12 ft)
Activated carbon contact beds (17 ft 4 in. x 42 ft
with carbon depth 8 ft 9 in.)
Chlorine contact tanks (120 ft x 40 ft x 15 ft)
Outlet to lower Niagara River (18 ft x 21 ft x 7,000 ft)
Sludge thickening basins (70 ft in diameter)
Sludge vacuum filters (11 ft 6 in. in
diameter drum :: 14 ft wide)
Carbon regeneration furnace (14 ft 3 in. in diameter)
Number of units
5
5
28
2
1
2
4
1
Basis of GAC System Design
The GAC contactors are rectangular, open-top, downflow beds of reinforced
concrete construction. There are 28 beds each 17 ft, 4 in. by 42 ft with an 8
ft, 9 in. depth of GAC. The filter underdrains originally consisted of a false
145
-------
GORGE P.S
SOUTHSIDE
INTERCEPTOR
DEWATERED
SLUDGE TO
LANDFILL
CHEMICAL
ADDITION
ACID ADDITION
TRASH TO
•LANDFILL
SLUDGE
THICKENER
VACUUM
FILTRATION
r
t
CARBON
REACTIVATION
CI2
GRANULAR
ACTIVATED
CARBON
CONTACTORS
EFFLUENT DISCHARGE TO
•LOWER NIAGARA RIVER
CHLORINE
CONTACT
BASIN
Figure A-16. Process flow schematic, Niagara Falls, NY.
-------
bottom made up of concrete panel plates fitted with plastic nozzles located
above a plenum. At design flow, 40 min contact time is provided with the car-
bon. The surface flow rate is 1.67 gpm/ft . Air-hydraulic backwash is pro-
vided for cleaning the beds.
Carbon is transferred in water slurry using hydraulic eductors. Makeup,
spent, and regenerated carbon is stored, each in separate cone-bottomed ves-
sels. Within each carbon bed, just above the filter bottom, a pressure water
header with water jet nozzles assists in removing the bed of carbon from the
flat bottomed beds.
Carbon is thermally reactivated in a single 14-ft 3-in* diameter, 6-hearth
furnace equipped with an afterburner and a wet scrubber. Spent carbon is dewa-
tered and conveyed into the furnace by an inclined screw.
CARBON SYSTEM PERFORMANCE
. In early 1978, .during the few months that the GAC system operated, the
quality of the plant effluent was basically as anticipated and all discharge
requirements were met.
By mid-1978, however, operation of the carbon system was shutdown because
of mechanical failure rather than any process problem.
When the treatment plant was first started in April 1977, intolerable
amounts of chlorine and chlorine components were being discharged into the
sewer system from industrial wastes, resulting in equipment damage, hazard to
personnel, and possible loss of carbon by chlorine oxidation. The carbon beds,
which had to be bypassed during this period, were not started up on a limited
flow basis until January 1978. The flow to the beds had to be limited because
of the then extremely high pH in the plant influent, rapidly changing pH, and
requirements for large quantities of sulfuric acid to reduce the alkalinity.
The carbon operation had to be shut down for the entire month of March 1979,
but began operation again in April of that year.
In early summer, the plant experienced a rupture of several carbon bed
bottoms and, after the carbon was removed, extensive corrosion and damage to
the precast filter bottoms and gullet wall projections were discovered. As a
result, the operation of the carbon filter beds was suspended, and the plant
has been providing primary treatment to the wastewaters flowing to it since
that time.
The filter b.ttoms consist of a grid of 3-in. thick x 2-ft square precast
concrete panels, each holding nine plastic nozzles screwed into inserts; the
panels are supported on 9-in. diameter by 3-ft high concrete columns at each
corner. These panels were cast with a V-groove around all four edges. The hold-
down system consists of four No. 3 reinforcing bars projecting from the support
piers, field-bent, to lie along the V-groove panel edges, and the space between
panels is packed with no-slump cement mortar to lock the panels and reinforcing
together.
147
-------
Analysis of the filter bottom design indicates that the system (columns,
reinforcement, and precast panels) is adequate to sustain the gravity loads and
the uplift loads due to the normal expected variations in the backwash flow and
operating pressures, normal impact forces, or those forces expected when some
nozzles need cleaning.
Examination of the ruptured panels, carbon transport gullet wall, and
exposed hold-down reinforcing bars showed extensive corrosion buildup on the
reinforcing bars, somewhat aged cracked seams in the grout, and some aged
cracks along the failure plane of the panels. In addition, severe corrosion
buildup was observed on the air and spray piping brackets and supports.
It is apparent that extraordinarily strong wastewaters in the backwash
system penetrated cracks or voids in the cement mortar joint and severely
attacked the reinforcing rods creating a substantial corrosion buildup.
Observations of the damaged areas indicate two conditions caused the rup-
ture. The first was the expansion of the assembly, causing the plates to bundle
or the gullet wall projection to shear. The second condition causing rupture
was a vertical upward pressure causing precast panels to move upward at the
support columns. The probable cause for the first condition is the corrosion of
the reinforcing bars in the panel joints, causing an expansion of the diameter
of the reinforcing bar and imparting a horizontal thrust on the plates. A prob-
able cause of rupture for the second condition is pressure beneath the plates
caxising rupture at the panel joints. With a structurally weakened panel, as
described above, it required only a small amount of net upward overpressure1 to
cause rupture. '
Failure by uplift could have been the' final act of a condition created :in
another manner. The corrosion of the reinforcing may have imparted a thrust -in
the plates and caused a horizontal crack to develop. The weakened plate would
then rupture by uplift pressures within the normal backwash range. In summary,
where definite evidence was found that the horizontal thrust alone caused
individual ruptures and that uplift pressures during normal operation resulted
in ruptures, it was not possible to find positive evidence that uplift pres-
sures, by themselves, were the cause. This indicates that a major problem in
the filter bottom was corrosion of the reinforcing steel in the precast panel
joints.
Evaluation of corrective measures includes modifying the anchorage assem-
blies and modifying the concrete bottom inself including placing new all-fiber-
glass reinforcing .plastic filter bottoms in the filters, and placing overpres-
sure relief vent riser pipes in each bed. In addition, positively to prevent
future back-up of strong primary settled effluent from the chlorine contact
chamber, a flap gate is to be installed at the chlorine contact chamber end of
the filter effluent pipe.
The restoration program will include design modifications to allow back-
wash of the beds as two distinct units; this will eliminate the possibility of
all components failing simultaneously with the resultant loss of the whole
system of beds.
148
-------
The carbon beds need rehabilitation to provide increased reliability and
maintainability features after difficulty with the six beds in mid-1978 and the
consequent need • to remedy .all 28 beds similarly constructed. Piping changes
will provide additional operational flexibility.
The rehabilitation project (at an estimated $3.1 million) includes provid-
ing and installing temporary piping changes in the carbon-transfer piping sys-
tems, transferring GAC from and returning it to the carbon-filter beds; remov-
ing carbon filler bed nozzles and plates; furnishing and installing new filter-
bottom panels and nozzles; changing the filter-bottom panel supporting masonry;
cleaning all spaces into which GAC or other contaminating matter has penetrated
because of the rupture of filter beds; providing and installing new supports of
air wash piping and spray water piping in the beds; changing piping and repair-
ing and adjusting butterfly valves in the backwash water piping, and providing
and installing backwater flap gate.
CAPITAL AND OPERATION AND MAINTENANCE COSTS
The total construction cost of the plant was $61 million. local, state,
and federal cost participation was $8.91 million for the City of Niagara Falls;
$7.47 million for the New York State Department of Environmental Conservation
Grant, and $44.82 million for the EPA ($61 million total).
Costs for operation and maintenance under normal conditions are not avail-
able at this time.
149
-------
CASE HISTORY NO. 8
CITY OF FITCHBURG, MASSACHUSETTS
INDUSTRIAL & MUNICIPAL WASTEWATER TREATMENT
HISTORY AND INTRODUCTION
In 1968, a plan for domestic and industrial wastewater disposal was pre-
pared for the City of Fitchburg, Massachusetts. It outlined pollution abatement
facilities for the City and its industries. An 11-mgd, two-stage activated
sludge treatment plant was proposed for handling the municipal wastewaters from
East Fitchburg and several of its industrieso The industrial wastewaters from
the Weyerhaeuser Co«, Paper Division, the Fitchburg Paper Co., and the Town of
Westminster were to be treated in a second, 15-mgd conventional activated
sludge treatment plant.
In the spring of 1970, pilot plant studies were conducted to determine the
efficiency of using a physical-chemical means to treat the wastes in the second
plant. Treatment included coagulation, flocculation, and sedimentation of the
wastewaters followed by treatment in activated carbon columns. Waste loadings
from the two mills and town amount to approximately 46,000 Ib/day of suspended
solids and 8,000 Ib/day of BOD. The sedimentation basins removed about 95 per-
cent of the suspended solids and about 40 to 50 percent of the BOD.
The pilot studies indicated the effectiveness of the coagulation, sedimen-
tation, activated carbon system .to produce an effluent.of superior quality, that
at least one of the paper companies wanted to recycle for reuse in its opera-
tions. Full-scale design of the West Fitchburg treatment facility was completed
in July 1971; construction of the Fitchburg plant was completed in 1975. Since
that time, the plant has operated intermittently because of a variety of
mechanical problems, rather than because of any process problems -with GAC
adsorption or regeneration. It appears that these mechanical problems will be
solved and corrected so that the plant can be in full operation on a continuous
basis about October 1979.
In the 1968 report, the paper companies had requested treatment for more
than 20 mgd of their wactewaters; since that time they have reduced their
requirements by about 30 percent. The study recommended that a biological
treatment facility be constructed for West Fitchburg. The plant was to have
facilities for coagulation, flocculation, and sedimentation, which would pre-
cede the proposed activated sludge process.
Later studies conducted in 1968 and 1970 indicated that with proper pH
control and the use of sodium aluminate and certain polyelectrolytes, the
"primary" system would effectively remove 90 to 95 percent of the suspended
solids and approximately 50 percent of the BOD in the wastewater. Because load-
ings to the activated sludge system would be about 35 to 40 mg/L of BOD lower
than originally anticipated—alternative methods of wastewater treatment were
150
-------
studied. The activated carbon process was considered, and a decision was made
to run a pilot plant study.
GAC Pilot Plant Study
Isotherm studies of the wastewaters indicated that approximately 2.6 Ib
carbon/1000 gal would be required to treat the wastewaters and that the iso-
therm was approximately linear. Other investigations, however, showed poor
correlation between the isotherm results and exhaustion rates in column tests.
A pilot plant facility (Figure A-17) was, therefore, established at the
Weyerhaeuser Co., Paper Division, where waste was available from one of the
primary clarifiers. The Fitchburg Paper Co. daily trucked its wastewaters in
55-gal drums to the pilot plant facility. The laboratory work for the pilot
plant was done at the existing Fitchburg wastewater treatment plant's labora-
tory facilities.
Wastewaters from the Weyerhaeuser clarifier and Fitchburg Paper Co. were
settled in a tank truck and then pumped at a rate of approximately 0.35 gpm to
the adjacent columns. Four columns were used with 9.2, 11.3, 16.0, and 16.0 Ib
of GAC, respectively, and empty column contact times of 5.8, 7.2, 9.9, and 9.9
min, respectively.
The results of the pilot plant study were used to design the full-scale
carbon adsorption system. The system is based on an average daily flow of 15
mgd and requires 1600 Ib carbon/mil gal waste treated. This, represents approxi-
mately 2.1 times the carbon needed for an equivalent level of treatment for
settled domestic sewage.
A thorough economic analysis was made of the activated carbon system as an
alternative to the activated sludge process. Because the cost associated with
the primary treatment portion of the alternative facilities were assumed to be
similar, they were not considered. Although the activated carbon system proved
to be approximately 6 percent more expensive than the activated sludge system,
the flexibility afforded the City in the operation of the facility more than
outweighed the additional capital cost.
A request for a change in the process from activated sludge to activated
carbon (made to the Massachusetts Division of Water Pollution Control in 1970)
was granted. All parties concerned agreed that the activated carbon plant would
produce a better quality effluent than would the activated sludge facility;
primarily because a better and more consistent removal of BOD and COD could be
achieved and beca ise it would remove the color from the wastewater better than
would have been possible from the activated sludge plant. The plant would also
be less prone to upsets caused by the highly variable loads and would be cap-
able of being shut down or run at far less than full capacity—for example, to
treat the municipal wastes only.
In addition, one of the paper companies requested that, should the company
desire, water from the treatment facility would be made available to it for
recycling in its process. The unique location of the mills on the river means
that the Fitchburg Paper Co. uses effluent from the Weyerhaeuser Co. for its
151
-------
FITCHBURG PAPER CO.
PAPER MACHINES (MILL 4}
TRUCKED UP 330 GALYDAY
(?) PRESSURE GAGE
($\ SAMPLING PIONT
JABSCOPUMP
0.35 GAL:/MIN (1740 ml/m»n)
©
EFFLUENT
Figure A-17. Schematic of pilot plant flow in West Fitchburg, MA.
152
-------
process water. The Fitchburg Paper Co. also uses the effluent from the treat-
ment plant after it flows into the river, which could just as easily be trans-
ported directly to the company for reuse.
DESIGN CRITERIA AND SYSTEM OPERATION
The processes required at the West Fitchburg wastewater treatment plant
are shown in Figure A-18. Effluent from the coagulation-flocculation-sedimenta-
tion system is pumped to the carbon filters.
Maximum pumping capacity of each of the four units is 8,500 gpm at a
design head of approximately 145 ft. The pumps and wet well have been designed
to allow for a potential 10-ft headloss through prefiltraticn facilities should
they be deemed necessary due to rapid filter plugging.
Twelve carbon filters, each 20 ft in diameter with a 15.5-ft bed depth,
are provided for parallel operation. The filters and piping allow for two ves-
sels to be out of operation, one for backwashing and the other for regenera-
tion. Ten active vessels permit maximum utilization of the carbon. The surface
loading and empty bed contact time for the 10 active vessels is about 8
gpia/ft and 15 min, respectively, at peak hourly flow. The flow through the
vessels is divided uniformly and controlled by individual rate controllers. The
vessels have been designed to operate at a maximum pressure of 50 psi. The
overall vessel height from dished end to dished end is approximately 33 ft,
which allows for a 40 percent expansion of the carbon bed on backwashing. The
filter bottom consists of 1 ft of gravel over Leopold tiles.
The filters are backwashed with plant effluent using two variable speed,
vertical,, mixed-flow or vertical turbine pumps, one of which serves as a
standby. A maximum wash water rate of 9,100 gpm ensures 40 percent expansion of
the filter at 75°F, and the water is discharged to a backwash lagoon. Each fil-
ter is washed for 15 min daily, with one filter backwashed every 2 hr on a time
cycle; a surface wash is also used. A clear well with a volume equal to approx-
imately 10 times a single backwash is provided, but the effective volume is
greater since there is a continuous flow into the clear well at all times
except at shutdown. A backwash basin (two, 65-ft diameter tanks) is provided to
uniformly return the backwash water together with the carbon transfer motive
water to the head end of the plant. Under the most adverse conditions, the max-
imum rate of return to the head end of the plant is 2.2 mgd or about 12 percent
of the plant throughout. Because the oxygen is depleted from the filtrate, the
effluent is aerated to maintain a level of 5 mg/L of dissolved oxygen.
Approximately once a week it is necessary to replace the carbon in one of
the vessels. Spent carbon is transferred to a spent carbon storage vessel
through water eductors, which are sized to empty or refill one vessel in less
than 6 hr. The volume of the spent carbon and regenerated carbon storage ves-
sels is equal to two times the carbon in any one filter vessel. The carbon is
regenerated in a multiple-hearth furnace using steam that has been sized based
on continuous operation with a loading rate of 100 Ib carbon/day/ft2 of
hearth.
153
-------
1, TO»N OF WESTMINSTER I FUTURE)
2, MONTACHUSETT REGIONAL VOCATIONAL
TECHNICAL HIGH SCHOOL
O, CITY OF F.ITCH BURG
1
COMMINUTION
FLOW
MEASUREMENT
PRIMARY
SEDIMENTATION
PRIMARY SLUDGE
CHLORINATION
FLOW
MEASUREMENT
CHEMICAL ADDITION
EAST FITCHBURQ
STP
SLUDGE
LAGOONS
(OFF SITE)
CARBON
REGENERATION
T
I
i
k
I
I
L_
COAGULATION
FLOCCULATION
SEDIMENTATION
-Or
I
I
I
t
CARBON
FILTERS
FILTER BACKWASH
10 WEYERHAEUSER CO.
2. FITCHBURG PAPER CO.
m
-
m
z
-------
In transferring the carbon, two water streams are required: fluidizing
water for maintaining the carbon-to-water ratio (1 to 2 Ib carbon/gal slurry)
and motive water to operate the eductors. Fluidizing water is pumped against
essentially the static head and motive water at the most efficient combination
of pressure and volume. Fluidizing water and motive water to and from the fil-
ter vessels is provided by separate, constant speed pumps that operate at 100
psi with suction from the process water header.
Operational Problems
A number of mechanical problems with the plant have been corrected:
1. Removal of pressure discharges from a gravity drain line beneath the
plant floor.
2. Synchronization of filter influent plant flows.
• 3. Replacement of basin sludge pumps.
4. Modification of the filter flow control system.
5. Control of a corrosion problem in the plant water supply system.
There are three other problems well on the road to correction:
1. Odor control of the reaeration process.
2. Deterioration of carbon column linings.
3. Failure of carbon column surface wash systems.
The plant has been through one complete carbon reactivation cycle.
CARBON SYSTEM PERFORMANCE
Only a very limited amount of data have been generated by the carbon sys-
tem to date because plant operation has been interrupted for prolonged periods
of time to correct mechanical defects. The adsorbers are producing an effluent
with a BOD concentration in the range of 20 to 30 mg/L. Waste discharge
requirements are 8 mg/L of BOD. The reactivation process has performed much as
expected.
The data col lected in the pilot plant studies are summarized in Table
A-15.
The resulting BOD and COD removals show that in Column 1 the average
removal of BOD and COD was about 45 percent whereas in Column 3, after 23 min
of contact time, almost 90 percent of the BOD and COD was removed from the
wastewater. The concentrations of the BOD and COD in the effluent stream after
only 13 min of contact time were less than 2.6 and 11.9 mg/L, respectively.
This is well below the requirements needed to meet stream standards. These test
155
-------
TABLE A-15. OPERATING RESULTS OF BOD AND COD REMOVAL IN PILOT PLANT
STUDY OF WEST FITCHBURG WASTEWATER TREATMENT FACILITY
Constituent
Raw
No. 1
Columns
No. 2
No. 3
Average wastewater
characteristics, mg/L
BOD
COD
Average removal, %
BOD
COD
13.2 7.6
53.4 29.2
42.0
45.0
2.6
11.9
80.0
78.0
1.7
6.8
87.0
87.0
results and similar results for the removal of suspended solids and headloss
across the filters showed that GAC was a practical form of treatment. No
extrapolation of the pilot plant was necessary to arrive at design parameter
values for the treatment facility.
PRELIMINARY ESTIMATE OF CAPITAL AND OPERATING COSTS
The estimated costs of construction and operation for the proposed treat-
ment facility have not been refined because the design was not complete at the
time the .estimates were made. . ... .
...The total, estimated capital cost for .tihe complete West Fitchburg plant is
detailed in Table A-16 and amounts to $12,270,000.
The annual operating costs for the plant, including the items outlined in
Table A-17 will amount to about $515,000. Approximately 70 percent of the oper-
ating costs of the activated carbon system ($140,000 of the estimated operating
expense) is associated with makeup carbon, assuming 5 percent of the exhausted
carbon is replaced. A reduction in the raw organic discharge by the industries
or an increase in the removal efficiency in the primary portion of the treat-
ment plant will have a greater effect on the operating costs of the activated
carbon system than it might have had on the activated sludge system.
The industries and the City of Fitchburg have negotiated contracts that
will ensure that the capital and operating expenses of the facility are funded
solely by the industries until such time as Westminster enters the plant.
156
-------
TABLE A-16. ESTIMATED CONSTRUCTION COSTS FOR WEST FITCHBURG
WASTEWATER TREATMENT FACILITY
Item
Estimated cost
Site work
Filter building
Sludge and chemical pumping station
Municipal influent structure
Municipal clarifiers
Flocculation and rapid-mix tanks
Wastewater clarifiers
Backwash lagoons
Post-aeration basin
Yard piping and miscellaneous structures
Electrical
Miscellaneous
$ 1,315,000
6,835,000
503,000
32,000
95,000
387,000
1,035,000
85,000
90,000
628,000
865,000
400,000
$12,270,000
TABLE A-17. ESTIMATED FIRST YEAR OPERATING COSTS* FOR WEST
FITCHBURG WASTEWATER TREATMENT FACILITY
Item
Estimated cost
Salaries
Power and heat
Chemicals
Lime
Polymer
Hypochlorite
Alum
Makeup carbon
Sludge, lagoon cleaning
Lab supplies jj
Maintenance
Total
$ 110,000
115,000
17,000
48,000t
1,00011
63,000§
102,000**
28,000
1,000
30,000
$ 515,000
*Based on expected initial loadings to plant.
tBased on dose of 1 mg/L; expected dose may be as low as 0.3 to 0.5 mg/L
llBased on chl srination of municipal wastes only.
§Based on maximum dose required.
**Based on present loadings.
157
-------
BIBLIOGRAPHY
Rimer, A. E., Callahan, W. F., and Woodward, R. L., "Activated Carbon System
for Treatment of Combined Municipal and Paper Mill Waste Waters in Fitchburg,
Massachusetts". Journal of the Technical Association of the Pulp and Paper
Indus;try; September, 1971.
158
-------
CASE HISTORY NO. 9
ARCO PETROLUEM PRODUCTS COMPANY
WATSON REFINERY
CARSON, CALIFORNIA
INDUSTRIAL WASTEWATER TREATMENT
HISTORY AND INTRODUCTION
The Watson Refinery of the ARCO Petroleum Products Company is located at
Carson, California, in Los Angeles County. It is adjacent to the Dominguez
Channel, a nonnavigable lined tidal estuary that is used for stormwater runoff
and discharges from industrial plants including refineries.
A 1968 study by the Los Angeles Regional Water Quality Control Board
showed that the oxygen demand of industrial wastes entering the channel was
depleting the oxygen in the channel. Consequently, a resolution was issued
limiting all discharges to the Dominguez Channel (Table A-18).
TABLE A-18. DOMINGUEZ CHANNEL DISCHARGE LIMITS
Parameter Limit
Temperature 140°F max
Sulfides . 1.0 mg/L max
Floatable oil ' No visible
Dissolved oil 75 mg/L
pH, range . 6.5 to .10.5
Cyanides Below toxic concentrations
COD 1,330 Ib/day max
Settleable solids 1.0 cc/L max
Many of the Watson Refinery sewers are combined so that process wastes and
stormwater runoff will be mixed during rainy periods. Because of the complexity
of the sewer system, segregation of wastes was not considered feasible. There-
fore, a scheme had to be developed that would provide for disposal of process
wastes during dry weather and combined wastes during the rainy season.
To dispose of normal dry weather process wastes, the Refinery and the Los
Angeles County Sanitation District agreed that the District would accept pro-
cess wastes with i minimum amount of treatment. The specific sewer discharge
limits establishea by the Los Angeles County Sanitation District are shown on
Table A-19.
In addition (to the above limits) any compounds that give off strong
odors (mercaptans) toxic or flammable gases are prohibited.
159
-------
TABLE A-1.9. LOS ANGELES COUNTY SANITATION DISTRICT SEWER DISCHARGE LIMITS
Parameter Limit .
Temperature 120°F max
Sulfides 0.1 mg/L max
Floatable oil 10 mg/L max
Dissolved oil 25 mg/L max
pH 6.0 min
Cyanides 10 mg/L
Nitrates 5% max
Because of large volumes of stormwater runoff, the County would not accept
any process wastes diluted with stormwater. Whenever rainfall amounts would
exceed 0.1 in., all flow was required to be diverted from the Sanitation Dis-
trict sewers until 24-hr after the rain ceased. That requirement was later
changed to a 4-hr wait.
To meet the discharge conditions established for both disposal points, it
was decided to discharge the normal dry weather process flows to the Sanitation
District sewer and to store all of the rainy season mixed wastewater (process
and stormwater) in a large (50 mil gal) reservoir. The stored water would be
treated to remove organics, which exert a COD, and the treated water would be
discharged to the Dominguez Channel.
Several alternative methods for treating the impounded wastewater were
considered including biological methods. Since treatment would only be required
during wet weatherhowever, .these methods were -considered too difficult to
start-up and control on an intermittent basis. Also the desired 95% removal of
COD would not be achieved by these methods.
Using activated carbon to adsorb the soluble COD producing compounds was
the best treatment choice since it could be placed in operation quickly when-
ever it was needed, be readily controlled, and be adapted to changing feed con-
ditions and discharge requirements as necessary.
To meet the discharge limitations and compliance schedule, it was neces-
sary to have the treatment plant in operation by February, 1971, so that the
maximum discharge of 1,330 Ib/day total COD would not be exceeded.
DESIGN OF GAC SYSTEM
Bench and pilot scale tests were run to determine the feasibility of using
GAC to reduce soluble COD and phenol concentrations to levels that would meet
the discharge limits. Pilot plant, tests were run using 5-in diameter Plexiglas
columns with a total carbon depth of 13 ft. Runs were made using various feed
COD concentrations and contact times. As a result of these tests, the following
design criteria were established:
160
-------
Flow rate
Influent COD (total)
Effluent COD (total)
Holding basin
Carbon dosage
3,000 gpm
150 mg/L average
37 mg/L average
50 mil gal capacity
1 lb/1,000 gal treated
In addition to the above, the plant was to be designed to run 90 days without
replacing or regenerating the carbon if the maximum expected rainfall amounts
were received. The spent carbon would be regenerated during the dry season
thereby allowing for a regeneration furnace of minimum size.
Basis of 6AC Design and Operation
The purpose of the activated carbon adsorption process is to remove solu-
ble organic compounds, which exert an oxygen demand on the receiving body of
water (Dominguez Channel), from the mixed process and stormwater. The discharge
requirements limit the effluent total COD to 1,330 Ib/day. The following design
basis was established to meet that limit.
Activated Carbon Adsorption
Number of carbon columns
Type
Size of columns
Carbon bed depth
Carbon size
Flow per column
Surface loading rate
Contact time
Carbon dosage
Weight of makeup carbon
per bed
Backwash
Supply
Duration of backwash
Backwash rate
Bed expansion
Twelve
Downflow, gravity
12 ft x 12 ft x 12 ft
13 ft
8 x 30 mesh
250 gpm
1.74 gpm/ft2
56 min
1 lb/1,000 gal
2,400 Ib
Treated effluent
10 min
2,900 gpm
40% max with surface washers
A downflow gravity carbon column arrangement was selected because of the
suspended solids expected in the carbon column influent. Although the storm-
water was to bfe stored in a large reservoir where solids would settle and
floating oil would be skimmed, the impounded water would still contain suffi-
cient amounts of suspended solids and oil to interfere with upflow carbon
column operation. Backwashing capacity was provided to expand the carbon beds
by 40 percent, and with a unique rotating surface washer; backwashing would
breakup and remove filtered solids and accumulated oil.
A schematic flow diagram of the wastewater treatment facility is presented
in Figure A-19. Water is pumped from the reservoir to the influent water
161
-------
TO RESERVOIR
FROM RESERVOIR
01
NJ
CARBON
GRAVEL
^
LEOPOLD TILE
BACKWASH
TROUGH
CARBON COLUMN
12 COLUMNS
EACH 12* x 12' x 26' DEEP
PUMPl
CHLORINATOR
PUMP
BACKWASH
EFFLUENT
SUMP
EFFLUENT BACKWASH
RETENTION SUMP
SUMP !
1 TO DOMINGUEZ
CHANNEL
Figure A-19. Granular activated carbon wastewater treatment system, ARCO Petroleum Products Co.
-------
distributor at the treatment plant. The rate is controlled to each column by a
flow control flume; it can, however, be varied to stagger the exhaustion of
carbon as desired. The water passes down through the carbon and underdrain
system and is conveyed to an effluent retention sump. Each bed of carbon is
supported by a 1-ft thick layer of gravel, which is supported in turn, by a
Leopold tile underdrain.
From the effluent sump, the treated wastewater flows to the Dominguez
Channel. Chlorine can be added at the sump inlet if desired. This sump also
serves as the reservoir for backwash water.
The carbon columns were designed for parallel operation in a downflow
mode. Since the water is not filtered before being introduced into the carbon
columns, the granular carbon acts as a filter for the remaining solid parti-
cles. The filtered solids gradually impede the water flow resulting in
increased headloss. To prevent excessive headless, the accumulated solids must
be washed from the carbon beds. This is provided by a backwash pump located at
the effluent sump. When the water level in a column rises to the bottom of the
backwash trough, that column is shut down for backwashing. Treated water is
pumped into the underdrain system of the column being washed at a design rate
of 2,900 gpm (20 gpm/ft ). The flow of water upward through the column causes
it to expand and release the lighter solids; these light solids are carried
into the backwash troughs and to a sump from which they are pumped with the
water to the reservoir. The rotating surface wash at each column supplies
additional agitation to increase solids removal from the expanded carbon bed.
Since the carbon treatment plant is used only when stormwater has accumu-
lated in the large reservoir, it can be run at a very uniform rate independent
of normal process wastewater flow variations. The concentration of organic
materials in the influent water cannot, however, be controlled so that varia-
tion can occur.
Basis of Reactivation System
As the activated carbon becomes spent and no longer provides the desired
removal of organic materials, it must be reactivated. A schematic diagram of
the carbon reactivation system is shown in Figure A-20.
Carbon removal troughs are provided at the bottom of each column. The
upper edges of the troughs are at the same level as the gravel that supports
the carbon. To reactivate a carbon column, the spent carbon is moved by gravity
along the troughs by high pressure water spray nozzles. The carbon slurry is
transferred from he column via a larger concrete trough to a sump from which
it is pumped to the spent carbon holding tank.
The holding tank is sized to contain all of the carbon from one column.
The carbon is then transferred to a dewatering screw conveyor where the excess
water is drained as the carbon moves up the inclined conveyor to the top of the
reactivation furnace.
163
-------
CARBON
TRANSFER
TROUGH
CARBON
COLUMN
SPENT REACTIVATED
CARBON CARBON
TANK TANK
DEWATERING
SCREW
SCRUBBER
WATER
EDUCTOH
Figure A-20. Granular activated carbon transfer and reactivation system, ARCO Products Co.
-------
As the spent carbon enters the top of the furnace, it is deposited on the
first hearth and is moved across each succeeding hearth by rotating rabble
arms attached to a central shaft. Teeth mounted as plows on each rabble arm
move the carbon inward or outward across the hearth. The temperature increases
as the carbon moves through the furnace to provide the thermal reactivation.
Automatically controlled burners are mounted at the lower hearths, and steam is
added to aid in the reactivation process.
The reactivated carbon drops from the bottom hearth into a water quench
tank where it is cooled before being transferred to the reactivated carbon
holding tank. It is then transferred by gravity when needed to an empty
adsorber cell.
Off-gases from the processing of the spent granular carbon require addi-
tional treatment before being discharged to the atmosphere. Air pollution codes
in the Los Angeles basin are the strictest in the nation. To meet Air Pollution
Control District requirements, an afterburner at the top of the reactivation
furnace raises the off-gas temperature to about 1450°F and a quencher-scrubber
system cools the gases and removes the final solid particles from the exiting
gases.
Design Criteria - Carbon Reactivation Furnace
Furnace type Multiple hearth
Furnace diameter 56-in. I.D.
Number of hearths Six
Capacity 8,500 Ib/day
Estimated carbon loss 5%
Fuel Natural gas
Fuel requirements 3,000 SCFH
(including afterburner)
Temperature, range 1,600°-1,750°F
Steam requirements 354 Ib/hr
Afterburner type Integral (zero hearth)
Apparent density - spent 0.54-0.56
- reactivated 0.48-0.50
Time to regenerate one bed 6 days
GAC SYSTEM PERFORMANCE
The full-scale carbon adsorption system was started up first on a trial
basis in ".ay, 1971, to get operating experience and permit correction of any
operational defic.encies that might become evident. Process wastewater diluted
with service water was used for the test since the rainy season would not begin
until la^e in the year. The initial feed rate to the treatment plant was set at
the design rate of 250 gpm carbon adsorber. The influent COD was 650 mg/L and
the effluent COD ranged from 44 to 80 mg/L compared with the design effluent
concentration of 37 mg/L COD. During the initial test run of 208 hr, the efflu-
ent became cloudly and a septic odor became evident. Washing the beds with a 2
percent caustic solution corrected the problem of cloudiness and septicity but
did not improve the effluent COD.
165
-------
In later stages of the testing period, the influent COD dropped to within
the design concentration but the total COD in the effluent was nearly always
greater than 37 mg/L. Other measures to improve COD removal such.as more vigor-
ous backwashing and more frequent backwashing were not successful.
It also became apparent that the carbon exhaustion rate was greater than
had been expected from the pilot plant studies. When one carbon cell was run to
complete exhaustion, the exhaustion rate was 7 lb/1,000 gal compared with the
design exhaustion rate of 1 lb/1,000 gal at the design effluent concentration.
The first stormwater treatment period began in December, 1971, when suffi-
cient rainfall accumulated for full operation of the treatment plant at the
design rate of 3,000 gpm. Because of the intermittent occurrence of rains dur-
ing that season, the treatment plant was operated six times during that season.
The results of these runs are given in Table A-20.
TABLE A-20. TYPICAL PERFORMANCE DATA FIRST RAINY SEASON
Number
1
2
3
4
5
6
Length of
run , hr
44
48
38
18
95
22
Feed
Rate , gpm
3000
3000-2000
2000
2000
1000
2000
Average COD
Feed Effluent
326
360
374
310
237
147
43
48
86
67
100
93
Initially, the influent COD was higher than design so that feed rate was
reduced in subsequent runs. The effluent COD, however, remained high, even
after the influent COD concentration was lower in later runs because of more
dilution. It was not determined whether the effluent COD consisted primarily of
soluble organic materials that normally would be adsorbed. At one time during
the first rainy season, the presence of algae in the influent to the carbon
columns was responsible for high effluent COD. The algae was neither adsorbed
nor removed by filtration so that it passed through the columns and was dis-
charged. Although the algae contributed a nonsoluble COD, it constituted a por-
tion of the total COD discharged. The COD loadings on the carbon in each column
varied from 0.2 to 0.3 Ib COD/lb carbon.
Suspended solids removal data are limited, but for one period during the
first rainy season, the influent to the activated carbon columns averaged 106
mg/L and the effluent suspended solids averaged 22 mg/L.
Operational changes were made before the second rainfall season of 1972-
1973 to try to improve the quality of the effluent and decrease the carbon
dosage. Five of the carbon columns were placed in continuous service so that
they would be reactivated and placed back in service during the season. The
other seven columns were operated as before. In addition, treated water was
recycled to the holding reservoir to reduce the concentration of the influent.
166
-------
Figure A-21 shows the performance of one of the five continuously operated
columns, which is typical of all five. Operation of the columns was discontin-
ued and the carbon reactivated when the effluent COP concentration exceeded 50
mg/L. After reactivation and replacement, the column was operated until the
effluent COD reached 100 mg/L.
Operating results for the remaining seven columns, which were operated in
the design mode of once through flow and no regeneration, are given in Figure
A-22. These columns were shut down for later regeneration when the effluent COD
reached 50 mg/L.
Oil removal for the period shows that 57 percent of the influent oil con-
centration of 28 mg/L was removed; 12 mg mg/L remained in the carbon column
effluent.
Average of data obtained in the 1972-1973 rainy period show that 102 mil
-.gal''of. wastewater were treated; the influent and effluent COD concentrations of
233 mg/L and 48 mg/L, respectively, resulted in a carbon loading of 0.26 Ib
COD/lb carbon.
Reactivation System
The first adsorber was reactivated in November, 1971, and the second in
February, 1972. Five other adsorbers followed in the April-July, 1972 period.
The time required for reactivation of each varied from 4 to 14 days compared
with the design regeneration rate of 5.7 days. Several factors contributed
initially to the extended reactivation time including such problems as gravel
plugging the feed eductor, eductor wear, poor temperature control, and high
center shaft rates. After the original 1-in. eductor was replaced with a large
1-1/2-in. unit and the center shaft was slowed down, the reactivation time for,
each column approached the design time and the apparent density of the reacti-
vated carbon was within the design range. The following year no major problems
were apparent in the carbon reactivation system, and the quality of the reacti-
vated .carbon was more nearly that of virgin carbon.
Actual carbon losses during reactivation were found to be about 5 percent
which was attributed primarily to attrition during handling and reactivation.
Any fines generated during the handling of carbon were backwashed out of the
adsorbers before the adsorbers were placed in operation.
In April, 1973, the air pollution regulatory agency tested the reactiva-
tion furnace stack and determined that the limit for carbon monoxide (CO) was
being exceeded. B.' increasing the afterburner temperature to about 1550°F and
opening the combustion air damper, the plant was able to bring the CO level
well below the limit.
Recent Discharge Modifications
In 1975, the Los Angeles County Sanitation District changed the sewer
regulations affecting .the ARCO Watson refinery. The new regulation permits
stored untreated rainwater to be discharged to the sewer beginning 24 hr after
167
-------
a
O
o
u
400
300
200
100
Feed
0.4
,-S 0.3
f3 O
J3«
O
Effluent
±
Effluent
02468
024 6
MHIIens of Canons of Water Treated
8 10 12 14 16
0246
8 02 4 6 8 10 12 14 IB
Millions of Gallons of Watei Treated
Figure A-21. COD removal and carbon loading of a continuously operated column.
-------
10
u
B
O
o
a
u
-100
300
200
100
0.4
I °-3
o
0.2
0.1
\7
1 2 3
45678
Minions of Gallons of Water Treated
9 10 11 12
3 4 5 6 7 8 9 10 11 12
Minions of GaHons of Water Treated
Figure A-22. COD removal and carbon loading - normal (design) operation.
-------
the cessation of rainfall. Up to 10 mil gal can be discharged by the refinery
during off-peak hours. In addition, the refinery can begin discharging normal
process wastewater to the sewer beginning 4 hr after the rainfall ceases.
As a result of the newer regulation, the activated carbon treatment facil-
ity has seen very little operation in recent years. It was operated most
recently during the 1977-78 rainy season. No operating or cost data are avail-
able for that period.
The refinery intends to keep the treatment system in operating condition
so that in .the future it can be placed in service if necessary. Very little
corrosion has occurred because carbon steel parts of the system have been
flushed well after each use period. Much of the system in which carbon is
stored was constructed of concrete or stainless steel thereby eliminating
excessive corrosion. To start up the adsorption and reactivation systems, only
minor instrumentation repairs on the furnace would be required.
CAPITAL AND OPERATION AND MAINTENANCE
Capital Costs
The cost of the activated carbon plant including adsorbers, carbon trans-
fer system, reactivation system, appurtenances, and initial fill of granular
activated carbon, as installed, was approximately $1.0 million (1971 dollars).
This does not include the value of the plant site or other associated costs
such as the holding reservoir or pretreatment facilities.
Operation and Maintenance Costs .
The average cost to operate and maintain the activated carbon treatment
plant over the first 2 years of operation (1971-1973) was $0.49 per 1,000 gal
of water treated based on actual figures supplied (not adjusted for later
years).
A tabulation of operating and maintenance cost is presented in Table A-21.
170
-------
TABLE A-21. OPERATING AND MAINTENANCE COSTS 1971-73
4 per 1,000 gal
Item $ of water treated
Utilities 21,067 12
Maintenance labor 10,251 6
Operating labor 23,443 14
Carbon (makeup) 21,779 13
Miscellaneous
(maintenance overhead 7,249 4
and costs other than
labor and transporta-
tion overhead)
Total 83,789 49
BIBLIOGRAPHY
Loop, G.C., "Refinery Effluent Water Treatment Plant Using Activated Carbon",
EPA-660/1-75-020, Final Report, EPA Grant No. 12050 GTR, 1975.
Mehta, P.L., "The Application of Activated Carbon in Tertiary Treatment of
Refinery Effluent", presented at 36th midyear meeting of the American Petroleum
Institute's Division of Refining, San Francisco, California, May 13, 1971.
Prosche, K.A., Loop, G.C., and Strand, R.P., "Operation and Performance of a
Refinery Waste Water Carbon Adsorption Plant", proceedings 29th Annual Purdue
University, Industrial Waste Conference, West Lafayette, Indiana, May, 1974.
171
-------
CASE HISTORY NO. 10
RHONE-POULENC INC.
PORTLAND, OREGON
INDUSTRIAL WASTEWATER TREATMENT
HISTORY AND INTRODUCTION
Rhone-Poulenc Inc.'s manufacturing facility at Portland, Oregon, formerly
known as Rhodia Inc., Chipman Division, manufactures herbicides such as 2, 4-D
acid, MCPA acid, 2, 4-DB acid, and esters of these products. Wastewater from
these processes originally discharged to an industrial lagoon located adjacent
to the plant site. Regulation of Rhone-Poulenc's discharge began in 1966 when
the Oregon State Sanitary Authority limited the discharge of phenols to a maxi-
mum of 25 mg/L and required that the discharged wastewater bypass the lagoon
and be diverted directly to the river via a new pipeline. The plant installed a
process for chlorinating the effluent followed by neutralization with lime and
caustic soda to comply with the phenol limit. Within a year, the plant was
notified that the City of Portland would build a system to collect and treat
industrial wastes from the area industries and that the Rhone-Poulenc discharge
could no longer be discharged to the river. Subsequently, it was determined
tha.t the effluent would be acceptable for discharge to the sewer except for
phenol, which would be limited to a maximum of 1.0 mg/L. The characteristics of
the untreated wastewater are.given in Table A-22.. -.,..-
TABLE A-22. RAW WASTEWATER CHARACTERISTICS
Constituent
Concentration
Phenol and cresol
Chlorophenols & chlorocresols
2,4-D and MCPA
Alcohols
Chlorides (as NaCL)
Sulfates (as Na2SO4)
BOD
COD
Total solids
Suspended solids
pK
10 mg/L
100 mg/L
100 mg/L
1000 mg/L
50000 mg/L
8000 mg/L
2000 mg/L
3600 mg/L
62000 mg/L
10 mg/L (varies with pH)
0.5
To reduce the phenol concentration to below the stated maximum, the Rhone-
Poulenc laboratory investigated various processes including conventional bio-
logical methods, treatment with chlorine and bromine, ozone, ion exchange, and
activated carbon. Activated carbon treatment was the method selected because it
was found to be the most effective based on phenol removal and cost.
172
-------
Pilot plant tests were run at the plant to develop adsorption data, design
parameters, and preliminary cost information. It was determined that not only
were phenolics removed but 2, 4-D, acid, solvents, and alcohol were also
removed. Adsorption was enhanced and suspended solids were minimized by main-
taining a low pH through the adsorption process.
DESIGN OF GAC SYSTEM
From the pilot plant results and evaluation of the untreated plant dis-
charge, the design parameters were established (Table A-23).
TABLE A-23. SYSTEM DESIGN CONDITIONS
Flow rate
Hydraulic surface loading rate
Carbon reactivation frequency
Effluent quality
pH (for adsorption)
pH (range for discharge)
150,000 gpd
2.0 gpm/ft2
One adsorber every 2 weeks
1.0 mg/L phenol max
1.0 max
7.0 - 9.0
Operation of the adsorption system was to be continuous with the reactiva-
tion system being operated on an intermittent basis as required.
Basis of GAC Design and Operation
The purpose of the activated carbon adsorption process as used at Rhone-
Poulenc was to remove phenolic compounds that would interfere with the biologi-
cal treatment process used by the City to treat wastewaters. To meet the dis-
charge criteria of 1 mg/L phenol maximum, the following design basis was
established.
Activated Carbon Adsorption System
Number of carbon columns
Type
Size of columns
Carbon bed depth
Carbon size
Flow
Surface loading
Contact time empty vessel volume)
Carbon quanticy
Two
Downflow, gravity*
8 ft ID x 35 ft ID x 35 ft high
14 ft 6 in. each adsorber
12 x 40 mesh
104 gpm
2.0 gpm/ft2
105 min
18,000 Ib/adsorber
*Later changed to upflow
A schematic diagram of the GAC wastewater treatment facility is shown in
Figure A-23. Untreated wastewater is pumped to a 3,500 gal mixing tank from the
manufacturing areas of the plant. After mixing, it is pumped at a constant rate
173
-------
Ren
ADSORBER
FEED
TANK
NEUTRALIZATION
TANKS
^
Urn* Slurry
LIME STORAGE
TANK
P
•flkienl
LIME Mam
TANK
Gasto«lmospli«re
Process Recycle
"iater| |~~^l»
1
OEWATERINO I
SCREW H-|
[—»• 1 |AFTEHWm»IEB
J
Figure A-23. Granular activated carbon wastewater treatment facility, Rhone-Pou1enc, Inc.
-------
to the adsorbers. The rate is set to provide a continuous and steady flow to
the adsorbers regardless of the upstream fluctuations.
As shown in the schematic, the carbon columns were designed to operate in
series although provision was made for parallel operation if desired. Waste-
water enters the first column in the series from the bottom and flows upward
through the carbon. From the top of the first column, it enters the bottom of
the second and passes upward through the carbon in that column and exits just
above the surface of the activated carbon bed. It then flows -" o a neutraliza-
tion tank where lime is added to raise the pH to about 7.0 before being dis-
charged to the sewer.
Although the columns are operated in an upflow mode, the original design
and installation was for downflow operation. It was discovered that when spent
carbon was removed from the column for reactivation, a "heel" of spent material
remained in the column. Dpon refilling with fresh carbon, adsorbed organics
would leach from the heel and cause the effluent to .contain phenol. Upflow
operation prevents leaching of this material to the discharged effluent
stream.
Since the wastewater contains no suspended solids at the low pH used in
the adsorption process, no facilities were provided for backwashing the carbon
columns.
All of the wastewater processing tanks including the adsorbers are made of
wood (Douglas Fir) to withstand the corrosive acidic wastes being treated. The
carbon bed is supported by a 1-ft layer of quartz gravel, which is supported on
a perforated wooden distributor plate.
As the carbon in the first column of the series becomes spent (as evi-
denced by high phenol after the first adsorber, approaching the influent con-
centration, or by effluent phenol concentrations above 1.0 after the second
adsorber) , it must be reactivated. The first column is then taken out of serv-
ice, and the second column continues to operate in a single-stage mode. Spent
carbon is removed from the first adsorber to a spent carbon holding tank.
Fresh carbon is then placed in the empty column, which is then operated as the
second or polishing adsorber.
Spent carbon is removed from the bottom of an adsorber by opening a valve
in the spent carbon line near the bottom of the adsorber just above the gravel
layer. It is transferred in a water slurry via an eductor. To prevent leaving a
large heel of spent carbon in the adsorber, a small quantity of water flows
upward continuous.' y through the distributor plate and gravel to slightly expand
the carbon and cause it to flow toward the carbon drain so that only a small
heel remains.
Basis of Carbon Reactivation Design
To reactivate the spent carbon, it is transferred from the spent carbon
holding tank via an eductor to a dewatering screw conveyor. It is moved up an
inclined trough by the screw conveyor, and the excess water is drained off back
175
-------
to the holding tank. Dewatered carbon containing about 50 percent moisture by
weight drops from the end of the conveyor onto the top hearth of a multiple
hearth furnace with the following design characteristics.
Carbon Reactivation Furnace
Furnace type Multiple hearth
Number of hearths Six
Furnace size 54 in.-I.D
Furnace capacity 8500 Ib/day
Estimated carbon losses 5%
Afterburner type External, horizontal
Fuel Natural gas
Steam requirements ..... 354 Ib/hr
Regeneration time 2.1 days/bed
Rotating rabble arms attached to the central shaft move the carbon through
the furnace by means of teeth attached to the rabble arms. The teeth act as
plows to move the carbon inward or outward on succeeding hearths. Drop holes
are located at the center of the furnace for inward moving hearths and at the
outside for outward moving hearths; these holes permit the carbon to drop from
hearth to hearth. In the upper hearths, the carbon is dried and some of the
adsorbed organic materials are volatilized. Burners are located on Hearths No.
4 and 6 (Hearth No. 4 is at the top and Hearth No. 6 is at the bottom) so that
the furnace is progressively hotter toward' the .bottom. Steam is added in the
lower part of the furnace at a rate of 1 Ib of carbon to aid in reactivation.
The maximum temperature is in the range of 1600°-1800°F at the fired hearths.
Off-gases from the reactivation pass upward through the furnace to an after-
burner where they are burned at about "I800°F in the presence' of excess oxygen
to destroy any odor and burn the volatile gases that would otherwise be
released to the atmosphere.
When reactivation of the carbon is complete, the carbon drops from the
furnace into a water filled quench tank where it is cooled before handling and
storage. From the quench tank it is transported in slurry form to a fresh car-
bon storage tank where it is stored until needed in one of the adsorber
columns.
All carbon transfers are done in a slurry form by means of hydraulic
eductors. Motive water used for transfer is recirculated to minimize the amount
of water discharged to the sewer.
System Start-up and Operation
The GAC treatment facility at Rhone-Poulenc was first placed in service
November 25, 1969. From the beginning it was able to produce an effluent meet-
ing the established phenol limit of 1 mg/L. Originally it was planned to oper-
ate the plant at 150,000 gpd and to take the first adsorber off-stream for
reactivation when it evidenced a phenol breakthrough, which was expected to
result in an organic loading on the carbon 'of about 0.15 Ib/lb carbon. At
start-up, however, the flow rate was about half the design rate, and it was
176
-------
decided to operate the adsorption system until the phenol concentration of the
first stage effluent approached that of the influent. This would provide for a
much lovjr rate of exhaustion than originally planned. The resulting organic
loading on the carbon—0.47 Ib/lb carbon—made reactivation more difficult, and
improved furnace operation; the adsorption system was returned to the designed
operating procedure.
Because the carbon column discharge nozzel is located above the carbon
bed, the depth of carbon in each column was reduced to 12 ft, instead of the
original design depth of 14-ft 6-in., to prevent carbon losses by carryover.
This results in a shorter empty vessel contact time of 87 min at the design
flow rate compared with 105 min as designed.
During the first 4 months of operation, an average of 1.144 mil gal of
wastewater was treated before reactivation of the first stage adsorber as com-
pared with the design of 1.04 gal/adsorber.
In 1977, a new waste discharge permit issued by the City of Portland
placed additional restrictions on Rhone-Poulenc for discharges to the sewer
system. It limited the discharge of sulfates to a maximum of 500 mg/L and
specified a minimum DO concentration of 1.0 mg/L. In addition, it required all
wastewater to be held in one of three storage basins each of which holds the
total process flow for an 8-hr period. No wastewater could be discharged until
a completed analysis showed compliance with the regulations.
Also in 1977, the Oregon Department of Environmental Quality issued a
permit for discharge of plant runoff and boiler and cooling tower blowdown to
the Willamette River. • Before that time, these discharges had not been
restricted. The limits set fourth in the permit are shown in Table A-24.
TABLE A-24. DISCHARGE LIMITATIONS—PLANT RUNOFF
AND COOLING TOWER BLOWDOWN
Constituent Average concentration
Suspended solids 50 mg/L
Total phenolics 0.5 mg/L
Chlorinated hydrocarbons 1.35 mg/L
TOC 154 mg/L
Requirements were also set to require holding all of the nonprocess waste-
water until analy.is of the water could be completed. Wastes not in compliance
must be returned to the plant for treatment.
To meet the terms of these permits, a new process wastewater treatment
facility was installed that included new activated carbon columns and other
treatment steps to ensure all other limits were met. The original carbon
adsorption system was retained for treatment of contaminated stormwater and
boiler and cooling tower blowdown on an intermittant basis as necessary. The
177
-------
new carbpn treatment system was installed under a contract whereby the con-
tractor would remove the spent carbon, reactivate it, and return fresh carbon
to Rhone-Poulenc for both treatment systems.
Although the multiple hearth furnace used for reactivation of GAG started
up in late 1969, and has operated intermittently as planned, it has been'the
primary area of major mechanical problems and maintenance expense. The plant
shuts down 1 month each year for repair and maintenance. Originally, any
reipairs to the granular carbon adsorption and reactivation systems were to be
done during the plant shutdown period. Major repairs to the furnace have been
resquired,. however, on a much more frequent basis.
Corrosion resistant materials were used throughout the furnace because of
the corrosive conditions expected. Rabble teeth and arms on the upper two
hearths were made of titanium, which lasted 8 to 9 months. The replacement
teeth of ceramic-coated stainless steel were no improvement over titanium.
Because of the cost of the coated teeth, the plant returned to using titanium
teeth. On the other hearths, stainless steel teeth were replaced every 6
months.
The upper two hearths were replaced annually because of deterioration.
After 2 years of operation, these upper hearths were coated with a moisture
resistant, high temperature refractory that . with minor operating changes,
improved hearth life.
The original dewatering screw conveyor was made of Haveg because corrosive
conditions were expected. The Haveg conveyor trough performed as expected, but
the screw required, replacement within 2 weeks because of.excessive abrasion.
Eventually,' a 316 stainless steel screw conveyor was installed that lasted
longer and cost less although it still was replaced at 3-month intervals
because of corrosion.
SYSTEM PERFORMANCE
A summary of operating results of the reactivation system for 1971 and
1972 are given in Table A-25.
TABLE A-25. OPERATING RESULTS OF REACTIVATION SYSTEM
~ Constituent 1971 1972
Total wastewater treated, mil gal 25.08 31.08
Organics adsorbed, 1,000 Ib 397.2 529.2
Carbon reactivated, 1,000 Ib • 1008 1344
Carbon exhaustion rate, Ib carbon/
1000 gal treated 40.19 46.49
Organic loading, Ib organics
adsorbed/lb carbon 0.39 0.37
Carbon losses, average percent 7.8 9.8
178
-------
The organic loadings shown are higher than would be expected for a phenol
removal system and reflect the adsorption of other organics. Carbon losses
included some activated carbon lost from the columns into the effluent dis-
charge line as well as reactivation losses. Actual losses ranged from 5 to 10
percent overall.
When initially started, low hearth temperatures prevented the spent gran-
ular carbon from being reactivated to virgin carbon quality. The original burn-
ers were replaced with larger burners to maintain an adequate temperature pro-
file for reactivation. Typical fuel usage for reactivation was approximately
6500 Btu/lb spent carbon at a nominal reactivation rate of 300 Ib/hr. That fuel
use rate does not include fuel burned in the afterburner and is not correlated
with furnace burner settings.
Since the contract adsorption and reactivation service was started in
1977, on-site reactivation by Khone-Poulenc has been discontinued. The reacti-
vation furnace has experienced severe corrosion of nearly all mechanical parts,
exterior surface, instrumentation, and control panel. It would require complete
rebuilding or replacement if on-site reactivation again became desirable.
CAPITAL AND OPERATION AND MAINTENANCE COSTS
Capital Costs
The total project cost including carbon columns, pumps, carbon handling
system, reactivation system, appurtenances, and initial carbon fill was
$300,000 (1969 dollars). Table A-26 shows the capital expenditure for the
individual parts of the GAC treatment facility.
Operation and Maintenance Costs
Operating costs including maintenance for the year 1973 are shown in Table
A-27 and are reported in 1973 dollars.
The operating costs shown are for the entire treatment facility and were
not separated as to systems. Based on an actual wastewater flow of 70 gpm
(100,800 gpd), the cost of treatment was $3.19/1,000 gal. Maintenance, the
largest single item accounted for 36 percent of the total.
The costs for reactivating spent carbon for the year 1975 (1975 dollars)
are shown in Table A-28.
Approximate!; 1,500,000 Ib of spent carbon were reactivated in 1975 at an
average cost of $0.125/lb. These figures do not include operating or mainte-
nance of the adsorption system or other treatment units.
Costs presented here were developed by Rhone-Poulenc for comparison with
the cost of a contract adsorption and reactivation service. Based only on these
numbers, the contract service offered no cost savings. When the cost of lost
production due to lack of reactivation system reliability was considered, how-
ever, the contract service was selected.
179
-------
TABLE A-26. CAPITAL COSTS SUMMARY FOR RHONE-POULENC SYSTEM
Component •"" 'Cost, 1969 dollars
Carbon colunms , .13,300
Carbon storage vessels . 10,400
Recycle water vessel 3,500
Quench tank . 500 .
Pumps and eductors 5,900
Reactivation furnace (excluding
installation) 35 ,.400
Afterburner and stack 6,300
Off-gas scrubber system 10,000
Carbon dewatering screw 6,900
Piping and miscellaneous
material 26,500
Electrical 11,000
Instruments 5,800
Structures and foundations 22,800
Carbon inventory 18,500
Construction labor . 31,400
Engineering . 21,800
Subtotal • . 230,000
Neutralization system ' 40,000
Water collection systems 30 ,000
Total . 300,000
TABLE A-27. ANNUAL OPERATING COST SUMMARY FOR
RHONE-POULENC SYSTEM
Component • Annual cost - 1973
Makeup carbon ($0.35/lb) 24,240
Fuel (natural gas) 7,680
Electricity 1,800
Steam 3,600
Operating labor 15,120
Maintenance 42,000
Depreciation. (10%/year) 23 ,000
Total 117,440
180
-------
TABLE A-28. COST OF CARBON REACTIVATION FOR RHONE-POULENC SYSTEM
Component Cost, 1975 dollars
Makeup carbon ($0.54/lb) 40,500
Maintenance 87,200
Operating labor 15,000
Utilities - water 1,000
- electricity 2,500
- steam 16,300
- fuel (natural gas) 24,900
Total 187,400
BIBLIOGRAPHY
Henshaw, T. B., "Adsorption/Filtration Plant Cuts Phenols From Effluent", Chem-
ical Engineering, May 31, 1971.
Henshaw, T. B., "Activated Carbon System Reduces Toxic Phenols from Herbicide
Plant Effluent", Filtration Engineering, p. 8, July/August, 1972.
Tom Henshaw, personal communication, September, 1979.
181
-------
CASE HISTORY NO. 11
REICHHOLD CHEMICALS, INC.
TUSCALOOSA, ALABAMA
INDUSTRIAL WASTEWATER TREATMENT
HISTORY AND INTRODUCTION
Reichhold Chemical Inc.'s (RCI) manufacturing facility at Tuscaloosa,
Alabama, is located adjacent to the Warrior Pool between the Oliver and Holt
locks and dams on the Black Warrior River. This navigable stream is part of the
Warrior-Tombigbee River system. The plant is relatively old; it began opera-
tions in the early 1940's with the manufacture of synthetic phenol. Through
the years, its manufacture had expanded to include formaldehyde, orthophenyl-
pheriol, pentaerythritol, sulfuric acid, sulfate and sulfite salts of sodium,
and various synthetic resins and plastics. Since the production units do not
all operate continuously or simultaneously, the plant produces a complex waste-
water stream of widely varying characteristics.
Until the mid 1960's, process wastewater mixed with 'large quantities of
once-through cooling water and stormwater runoff was discharged to the Black
Warrior River with little or no treatment and no discharge limits.
A combined study of RCI and the Alabama Water Improvement Commission in
1966 revealed the following approximate characteristics of the RCI discharge:
Flow (range) ' • , • .... VIQ to 15 mgd .
Biochemical oxygen demand 16,444 Ib/day
Chemical oxygen demand 26,718 Ib/day
Phenolics 1,540 Ib/day
pH (range) 5.4 to 12.3
Target reductions for waste discharges were established, and a requirement
for a 5-day holding capacity (emergency storage) for all wastewater flows
prompted RCI to begin searching for methods to reduce the total flow and to
treat the process wastes. In-plant changes, which included process modifica-
tions and sewer segregation, ultimately reduced the discharge stream by 95
percent to about 500,000 gpd.
Discharge of • the v;astewater to a municipal sewer and treatment facility
was not an available alternative since there was no secondary treatment plant
in the area. The treatment and disposal methods RCI considered and evaluated as
to their potential for meeting a desired discharge limit included deep well
injection, evaporation/incineration, biological oxidation, and adsorption. A
test well was drilled ind tested, but was not considered a feasible method for
disposing of the plant waste, although it remains a possible future discharge
method for treated wastes. Because of several disadvantages including applica-
bility limited to high strength wastes, evaporation/incineration was considered
infeasible.
182
-------
Based on the results of bench scale biological studies, an activated
sludge pilot plant was installed and operated for several months to develop
design criteria and establish a basis for estimating capital and operating
costs. During the pilot plant operation, several waste streams were discovered
which were not amenable to or interfered with biological treatment.
Adsorption methods using GAC were evaluated as a practical approach to
treat the combined chemical wastes. The use of activated carbon provided the
desired reduction of organics based on laboratory isotherm tests. A pilot plant
using plexiglass columns was operated to establish design parameters, pretreat-
ment requirements, expected effluent quality and cost estimates.
As a result of all the evaluations, RCI selected adsorption using GAC to
remove soluble organics as the treatment method to be installed.
DESIGN OF GAC SYSTEM
As a result of the pilot plant operation and laboratory data, the unit
operations selected for treating the wastewater included flow equalization,
neutralization, clarification, and adsorption. A schematic flow diagram of this
system is presented in Figure A-24. The combined process flow streams enter an
equalization basin where the variability of concentration and flow are mini-
mized. The wastewater is then pumped to a neutralization basin where acid is
added to reduce the pH. Solids are removed by gravity settling in a clarifier,
and the clarified wastewater is pumped to GAC columns. The treated wastewater
then flows to a treated water tank and finally to the Black Warrior River. In
an emergency or if the treated water is not of adequate quality for discharge
to the river, it is diverted to a large holding basin.
Maximum permitted limits for discharge to the Black Warrior River are
shown in Table A-29.
TABLE A-29. BLACK WARRIOR RIVER DISCHARGE LIMITS
Parameters
Limit
Biochemical oxygen demand
Chemical oxygen demand
Phenol
Total suspended solids
pK
1644 Ib/day
2672 Ib/day
27 Ib/day
200 Ib/day
6.0-9.0
Basics of GAC System Design and Operation
The purpose of GAC adsorption at RCI is to remove dissolved organics
including phenol in the wastewater to meet discharge requirements.
An earthen equalization basin designed to hold approximately 3 day1 s flow
was installed to level out flow and concentration variations. Untreated waste-
water flows by gravity to the equalization basin, is pumped to an acid mixing
183
-------
CO
\_1/
EQUALIZATION
BASIN
ACID
FEED
TANK
POLYMER
FEED
SYSTEM
I
I
ir
i i
1
1
1
CLARIFIER
I 1
ACID
MIXING
CHAMBER
hi^
FLOCCULATION
CHAMBER
.
V*
1
•
1
1
_L
CARBON CHARGE
TANKS
TODEWATERING
SYSTEM
COLUMN
FEED
TANK
COLUMNS
II
TREATED
WATER
STORAGE
TANK
THICKENER
{^J
TR
SLUG
REASURING
TANK
TO
FURNACE FEED
TANK
Figure A-24. Flow diagram of wastewater treatment facility, Reinhold Chemical, Inc.
-------
chamber where sulfuric acid is added to adjust pH to about 7.0; flows to a
flocculation chamber where a polymer is added to improve solids flocculation;
and enters a clarifier where the solids are settled and removed for disposal.
The clarified water containing less than 42 Ib/day of suspended solids (10
mg/L) is pumped to two GAC columns in parallel. The following was the design
basis for the GAC treatment system.
Granular Activated Carbon Treatment System
Flow rate (average)
(maximum)
Number of carbon columns
Type
Size of columns
Carbon size
Carbon quantity (each column)
Column flow (each column)
Surface loading rate
Contact time (empty bed)
Carbon exhaustion rate
500,000 gpd
1,000,000 gpd
Two in parallel
Dpflow, moving bed
12-ft diameter x 40-ft high
8 x 30 mesh
125,000 Ibs
175 gpm
1.55 gpm/ft2
100 min
65 lb/1000 gal
(35,500 Ib/day)
Each carbon column, made of carbon steel is lined with an epoxy material
to prevent corrosion. The vessels are totally enclosed with conical sections at
the top and bottom. Water enters each column via eight stainless steel conical
screens connected to an external circular header. After passing upward through
the column, treated water exits through an identical arrangement of screens and
header system. Treated water then flows to a treated water holding tank from
which it flows by gravity to the Black Warrior River. If it cannot be dis-
charged, it flows to a treated water holding basin with a capacity for 5 days
flow (2.5 mil gal).
As water passes up through the carbon column, the activated carbon in the
lower portion of the column, which is in contact with the highest concentration
of organics, loses its ability to adsorb additional organics (becomes spent).
Periodically, a portion of the carbon is removed from the bottom and an equal
amount of fresh carbon is admitted to the top of the column. In this way, the
column is operated in a countercurrent mode and the freshest carbon is in con-
tact with the highest quality water to achieve the maximum amount of organic
removal.
Basis of Reactivated System Design and Operation
Figure A-25 presents a flow schematic of the carbon handling and reactiva-
tion system. The spent carbon slug, taken from the bottom of the column, flows
to the slug measuring tank to ensure that the same amount of carbon is trans-
ferred each time. From the slug measuring tank, carbon flows by gravity to the
furnace feed tank and is then hydraulically moved via an eductor to a dewater-
ing screw conveyor located at the top of the multiple-hearth furnace. As the
carbon slurry enters the dewatering screw conveyor, excess water drains off
back to the furnace feed tank. The granular carbon is moved slowly up the
185
-------
SPENT CARBON
FROM SLUG
MEASURING
TANK
oo
DEWATERING
SCREW
FURNACE FEED
TANK
EXHAUST TO
ATMOSPHERE
SCRUBBER
SYSTEM
REACTIVATION
FURNACE
QUENCH
TANK
BLOW
CASE
Figure A-25. Carbon handling and reactivation system, Reinhold Chemicals, Inc.
-------
CD
1-4-
r |
CARBON 1
COLUMN 1
u
I
r*—
i
V
SPENT CARBON
TANK
Figure A-26. I
M
CARBON
' ' COLUMN
r
^ _t_
r"r>i
t
•
QUENCH!
TANK V
'low diagr
4- OVERFLOW FROM -» 1 j.
*—\l REACT. CATIBON TANK 1
PI,-— r^,}
CARBON . I '
• > COLUMN 4- Ij
I LLLLn I
* i^ "-* ' " • intATEO WATER mm? >
i TANKS LU?
rnnu — — - — n • 1
MISC. FLOWS 1 "1
_J * 1
^__ ^ ___ — > * ' j I ^ EFFLUENT
p 1 HOLDING TANKS j ' TO RIVER
xl i t
fll _. TO SPENT CARBON TANK |
II T- •
" 1 ^
' REACTIVATED
CARBON TANK
1 CARROM
1 . REACTIVATION CARBON TRANSFER LINE
"*" 1 TUI1NACE •
am of granular activated carbon treatment plant, Stephan Chemical Co.
-------
inclined conveyor trough and is discharged onto the top hearth (Hearth No. 1)
of the reactivation furnace. Spent carbon. is moved across the hearth by angled
teeth attached to rabble arms, which '. rotate. . by means of a central rotating
shaft. After passing across the hearth, the carbon drops to the next lower
hearth where the process is repeated, inward on one hearth and outward on the
next, alternating each time.
On the upper hearths, moisture and some easily volatilized organics are
driven off. Furnace temperatures increase from top to bottom; burners are
located at "Hearths 4 and 6 (Hearth 6 is the lowest hearth). Tempertures range
from about 700°F at Hearth No. 1 to about 1700°F at Hearth No. 6. Steam is
added on the- fired hearths .to improve reactivation.
As the hot reactivated carbon leaves the furnace, it drops into a water-
filled quench tank where it is cooled and returned to a slurry form for trans-
fer back to the top of the carbon column. A pneumatically operated blowcase
transfers the reactivated carbon to the columns. Virgin makeup carbon is stored
in a storage tank.
The following is the basis of design for the carbon reactivation system:
Granular Carbon Reactivation System
Furnace, type
Furnace size
Furnace capacity
Fuel
Afterburner, type
Carbon loss per cycle ''
Makeup: carbon .
Multiple hearth
.13-ft 6-in. O.D. x 6 hearth
132,500 Ib/day
Natural gas/L.P. gas
Internal (zero hearth)
5 percent
1,625 Ib/day
Furnace off-gases (combustion products, moisture, and volatilized organ-
ics) are heated in a gas-fired afterburner at the top of the reactivation fur-
nace. Excess air is added to ensure combustion of volatile gases and odor. The
gases then pass through a cooler and scrubber before being discharged to the
atmosphere.
Operational Problems
Operating problems in the wastewater treatment system have included high
turbidity in the final effluent and excessive pressure drop in the carbon col-
umns. These problems were believed to have been caused by carryover of sus-
pended solids from the clarifier and by biological activity in the columns.
From time-to-time, the polymer addition/flocculation processes are not
effective in agglomerating the solids so they will settle readily in the clari-
fier. The result is a highly turbid feedwater to the carbon columns; this
feeclwater may be filtered by the carbon, causing high headloss and reduced
flow, or the solids may pass through the columns causing effluent turbidity.
Occasionally, changing the type of polymer will regain control of solids.
188
-------
When biological activity increases in the columns, as sometimes occurs,
the pressure increases and flow is reduced. This occurrence is reduced when
adequate velocities are maintained in the columns.
Although the carbon columns were lined to prevent corrosion, leaks in the
bottom cone required repair three to four times a year. In 1975, the bottom
cones were replaced with 316 stainless steel cones. Rubber hose segments of
carbon transfer lines, which required periodic replacement because of leaks due
to abrasion, were replaced with pipe sections. Although leaks still occur, the
repair frequency has been reduced.
The carbon transfer and reactivation system has performed basically as
designed. Approximately 32,000 Ib/day of spent carbon have been reactivated at
about 5 percent loss.
Typical fuel usage is about 6,200 BTU/lb carbon reactivated, including
fuel used in the afterburner.
Carbon slurry transfer lines and eductors tend to plug with debris or
lumps of carbon. When this happens in the furnace feed line, no spent carbon
reaches the furnace. When a stoppage is noticed and feed is restored, furnace
control becomes difficult for a time and a poorly reactivated carbon may be
produced.
Reactivated carbon quality as measured by iodine number has not approached
that of virgin carbon because inoganics partially coat the reactivated carbon.
Reactivated carbon iodine numbers are usually about 650 compared with 950 for
virgin carbon. The furnace feed system was modified to use plant service water
instead of treated effluent for transferring spent carbon to the dewatering
screw. The intent was to remove some of the soluble inorganic salts from the
carbon before reactivation to help restore the iodine number. This was only
partially successful, and the iodine number remained low.
The reactivation furnace has always complied with established air pollu-
tion codes. No noticeable smoke or odor is discharged to the atmosphere
although the maximum afterburner temperature is 1100°F compare with the 1400°F
design temperature.
The most frequent and costly repairs have been to the multiple hearth
reactivation furnace. Major repair-replacement items include rabble arms and
teeth, hearths, and refractory. Shortly after startup, the material adsorbed on
or remaining witl the carbon in the furnace created an extremely corrosive
condition causiny the rabble teeth to wear down quickly. This condition still
persists so that a large number of teeth must be replaced approximately every 4
months.
Major hearth repair or replacement of hearths has been required approxi-
mately every 2 years. Repair of other refractories, although less extensive
than hearth repairs, are required when the furnace is shutdown for other
maintenance.
189
-------
The RGI wastewater treatment facility was started up in 1973. Quality of
the treated water discharged to the Black Warrior River has always been within
the established limits. Results of wastewater treatment by GAC for COD, BOD and
phenol are shown in Table A-30 for the years 1973-1977. The original design
conditions are shown for comparison.
TABLE A-30. ACTIVATED CARBON TREATMENT PERFORMANCE
Constituent,
COD
Year
1973
(April-Dec)
1974
1975
1976
1977
(Jan- July)
In
4668
9025
9174
9700
11099
Out
2140
1504
1830
1968
2170.
BOD
In
4687
3330
3229
3971
4031
Out
1208
871
1008
1138
1294
Ib/day
Phenol
In
897
712
761
1022
973
Out
4.78
1.24
0.95
2.05
1.54
Flow,
mgd
0.452
0.461
0.354
0.359
0.329
Design 28,000 2,672 16,000 '1,644 2,700 27 0.50
Initially, some difficulty was experienced in maintaining the effluent COD
within permitted limits. One stream containing organics not amenable to adsorp-
tion was the major cause of high effluent COD. When that stream was removed
from the treatment system, COD was controlled within the limits. Control of
COD and BOD can still be difficult at times when high concentrations of nonad-
sorbable organics or some types of inorganic materials are present.
Phenol removal has been much better than that required for discharge.
Removal has consistently exceeded the design efficiency of 99.0 percent and
typically exceeds 99.8 percent.
Although no data are shown for the period following July 1977, the quality
of the discharge has not varied appreciably. The monthly average discharge
quantities (in pounds per day) for July 1979 were 2,152 for COD, 1,364 for BOD,
0.14 for phenol. Flow for the same month averaged 277,000 gpd.
Ammonia nitrogen in the wastewater was not thought to be of concern, and
no treatment was provided for ammonia removal. Ammonia quantities in the
untreated wastewater are always less than any discharge limit. Ammonia concen-
tration, however, increases through the carbon columns. In July 1979, ammonia
averaged 107 Ib/day at the carbon column inlet and 170 Ib/day in the effluent;
an increase of 58.9 percent. This increase may be because biological activity
converts urea to ammonia within the carbon columns.
The carbon exhaustion rate has averaged about 79 lb/1000 gal wastewater
treated compared with the design rate of 65 lb/1000 gal. Exhaustion of the
carbon, which controls the column slugging frequency, is determined by COD
190
-------
measured at the outlet. Although routine tests also included phenol and BOD,
COD (which is easily analyzed) breaks through more quickly. Carbon loadings
have averaged 0.26 Ib COD, 0.09 Ib BOD, and 0.03 Ib phenol/lb carbon. If carbon
were regenerated on the basis of phenol breakthrough, the loadings would be
much higher.
CAPITAL AND OPERATION AND MAINTENANCE COSTS
Capital Cost
Estimated capital cost for the complete wastewater treatment facility was
$1.3 million. This estimate included the cost of engineering, equipment, and
construction for the complete wastewater facility, i.e., earthen holding
basins, pretreatment units, adsorption system, reactivation system, and initial
charge of activated carbon. No breakdown of the capital cost for adsorption or
reactivation systems is currently available.
Operation and Maintenance Cost
The annual cost of operating this wastewater treatment facility for the
years 1973-1976 is presented in Table A-31.
TABLE A-31. ANNUAL OPERATING COST*
Item 1973 1974 1975 1976
Labor 88,080 82,607 95,539 102,845
Maintenance 57,324 111,034 172,828 161,871
Utilities • 56,908 62,721 . 91,944 111,945
Carbon 182,892 258,119 348,462 320,671
Fixed expenses 109,.234 122,967 129,602 144,373
Other 74,549 68,034 87,336 104,939
Total $568,987 $705,482 $925,711 $946,644
•Actual dollars each year.
The cost figures are not separated as to treatment or process systems
since the wastewater facility is operated. and maintained as a unit. For this
plant, the largest single expense item, about 35 percent of the total, is for
makeup carbon. Furnace maintenance is reflected in the increase in maintenance
cost from 1973 to 1976.
Although a ^eakdown of operating costs for later years is not available,
the total cost of treatment plant operations for 1978 was $1,060,000.
Operating costs normalized to the actual quantity of wastewater treated
each year are:
191
-------
. Year ..• $/100,0 gal ,
. V '.*.". • 1.973 ... .... .. $3.79
'.'.-• 1.974 ' ".' ." ' "'"•• • ' " . '.". .-'4V6Q,' .... ,....-. ... : .-,'. ' • •••', . .... . .
' • ' -'- •.--•'' •'• ' 1975 ••'• . '• . 7..8.7 "-'/ '' • ' .
1976 '7.93 .
The 1979 monthly budget for wastewater treatment is approximately $90,000.
BIBLIOGRAPHY
Norton, D. and Bonner, H. personal communication, September, 1979.
Shumaker, T. P., "Carbon Treatment of Complex Organic Wastewaters", presented
at Manufacturing Chemists Association Carbon Adsorption Workshop, Washington,
D.C., November 16, 1977. • • .
192
-------
CASE HISTORY NOo 12
STEFAN CHEMICAL COMPANY
BORDENTOWN, NEW JERSEY
INDUSTRIAL WASTEWATER TREATMENT
HISTORY AND INTRODUCTION
The Stepan Chemical Company's Fieldsboro plant at Bordentown, New Jersey,
is located adjacent to the Delaware River south of Trenton, NJ. The plant is
part of the surfactant division which makes surfactants for use in soaps and
cleaners and for other uses. Sources of process wastewater include leakage from
pumps and lines/ tank washings, and minor spills. The present manufacturing
facility has been in operation less than 15 years. About 10 years ago, the
Delaware River Basin Commission placed discharge limits on the Stepan Chemical
outfall that included a BOD^ limit of 100 mg/L. To meet that limit, the company
evaluated various treatment methods. A treatment, method using GAC was found
to provide the best results although it would not achieve the desired BOD5
reduction. The State of New Jersey agreed with the company that a granular
carbon treatment facility should be built and additional testing conducted in
an attempt to meet the BOD5 discharge limit.
Construction of the GAC wastewater treatment facility was completed for
start-up in 1972. Since that time the plant has normally met the discharge
requirements, including those established later by a NPDES permit, which are
shown in Table A-32.
TABLE A-32. STEPAN CHEMICAL CO. - NPDES DISCHARGE REQUIREMENTS
Parameter Limits, avg* (kg/day)
pK, range units 6-9
Suspended solids 0.6 kg/day
BOD5 3.2 kg/day
COD 61 kg/day
TOG 12.3 kg/day
Oil and grease 0.55 kg/day
M3ASt 0.2 kg/day
*Maxiraum value for single sample is twice the average
tMethylene blue active substances
DESIGN DATA OF GRANULAR ACTIVATED CARBON COLUMNS AND REACTIVATION SYSTEM
The granular activated carbon treatment facility was designed to treat a
waste having the following general characteristics: flow, 15,000 gpd; BOD,
2,000 mg/L; and COD, 9,000 mg/L.
193
-------
Basis of 'GAG Design
The _activated carbon adsorption system (Figure A-26) was designed accord-
ing to the following conditions, which were established as the result of small
column tests.
Carbon Adsorption System
Number of carbon columns
Type of columns
Column size
Bed depth
Quantity of carbon
Type of carbon
Required contact time
Carbon exhaustion time
Three in series
Downflow, pressure
6 ft dia x 10 ft high
7.5 ft
6,000'; lb/ column
8 x 30 mesh
50 min ,
430 lb/1000 gal
No preliminary treatment unit was included in the design so that untreated
process wastes are pumped from a collection sump to the adsorption system.
Wastewater enters the top of the first column/ flows downward through the car-
bon bed, and passes out through the underdrain system. From there it flows to
the top of the second column, through the second bed, and through the third
column in the same manner. Each column in the series contains the .same quantity
of carbon and intermediate pumping is not necessary. After carbon treatment,
•the water is discharged to the Delaware River.
Whenever the carbon in the first column in the series cannot adsorb
additional organic material (becomes spent), the ,; column is taken off stream for
reactivation of its carbon. This occurs about once each day. The column which
was second in the series becomes the lead or first column, and column number
three becomes number two. Spent carbon is removed from the column taken off-
stream and transferred in a water slurry by hydraulic, pressure to the. spent
carbon storage tank. This tank is sufficiently large to hold all of the spent
carbon from one adsorber. Fresh carbon is then transferred to the empty carbon
column, and it is placed in service as the new third column.
A small fourth carbon unit, not a part of the original design, was added
to the treatment facility as a polishing filter for control of suspended
solids. This filter, containing about one third of the standard column volume,
is not considered one of the treatment columns and because of its low loading
is reactivated approximately monthly.
Basis of Reactivation System Design
A reactivation furnace and carbon handling system is provided to reacti-
vate the spent carbon to near its original characteristics. The following are
the design criteria for this system.
194
-------
Reactivation System Design Criteria
Furnace type
Number of hearths
Furnace size
Furnace capacity
Afterburner type
Fuel
Multiple hearth
Six
54 in. I.D.
6480 Ib/day
Internal, zero hearth
L.P. gas
Spent granular activated carbon is transferred in a water slurry from the
spent carbon tank via an eductor to a dewatering screw conveyor. In the
dewatering screw conveyor, which is sloped upward, the spent granular carbon is
gradually moved up the slope above the water level. Excess water is drained
back to the spent carbon tank. Partially dried granular carbon containing
approximately 40 to 50 percent moisture by weight falls from the end of the
conveyor onto the top hearth (Hearth No. 1) of the reactivation furnace. It is
moved, across the. hearth by teeth acting as plows, which are attached to rabble
arms. The arms are fastened to a hollow central rotating shaft, and cooling air
is circulated within the shaft and arms. As the carbon moves across the first
hearth, it is dried and, upon reaching the center of the furnace, drops to the
next lower hearth. The lateral movement of the carbon is repeated outward and
continues in this fashion until it drops from the lowest furnace hearth (Hearth
No. 6). As the carbon passes through the furnace, it becomes progressively
hotter since burners are located near the bottom of the furnace (two burners
each on Hearths Nos. 4 and 6). Maximum temperatures are normally in the range
of 1600° to 1800°F.
Furnace off-gases including unburned volatile materials and combustion
products, pass upward through the furnace to an afterburner. This unit is
included as a special zero hearth at the top of the furnace where additional
burners are mounted. Here the temperature is raised to the combustion point of
the gases to burn the volatiles and odors that may exist. The furnace gases
then are discharged to the atmosphere via the stack.
As the now-reactivated carbon exits the furnace, it drops into a water-
filled quench tank where it is cooled. From there, it is transferred via an
eductor to a reactivated carbon holding tank until it is needed in one of the
carbon columns. At that time it is transferred in a water slurry via another
eductor to the empty column.
Untreated river water is used to transfer carbon and to maintain minimum
tank levels in the carbon handling system. This water becomes, part of the plant
discharge, account, ng for approximately half of the outfall flow. It is not
considered to need treatment as contaminated waste but must be filtered before
discharge.
Recent changes in regulations regarding disposal of contaminated storm-
water runoff have necessitated constructing a large 350,000 gal holding basin.
This basin can be used to hold wastewater in the event of a treatment system
breakdown as well as excess runoff. It is anticipated that a new treatment unit
will be needed to treat this wastewater to augment the existing wastewater
treatment facility,
•-.- -.- •.....- 195
-------
SYSTEM PERFORMANCE
Although the adsorption system design was not considered' capable of meet-
ing the 100 mg/L BOD limit as originally required at Stepan Chemical , it has
met or exceeded all requirements most of the time. This is because waste mate-
rials such as, methanol, which are not amenable to adsorption on activated car-
bon, are removed from the adsorption system. These materials are held for dis-
posal by incineration in, a boiler. Reduced concentration of contaminants at the
outfall is also partly .due to dilution by the carbon transfer water.
..a-- ' ." '
The final carbon filter previously mentioned was installed to enable the
discharge to comply with the outfall suspended solids limits. Carbon fines have
continually plagued the wastewater treatment system by building up in the
transfer and holding tanks and discharging to the river. The filter serves to
prevent discharge of these fines and to control the ' discharge of suspended
solids. . ' •
Table A-33 shows the average effluent quality at the outfall for 1977 and
1978. • - •• : , ....
__ TABLE A-33. TREATED EFFLUENT WATER QUALITY •-'- "
. Parameter . . 1977 avg _ . 1.978 -avg'-^ : _
Suspended solids " "'-^.24 mg/L 5 mg/Lf
BODj '" . ' 26 mg/L • ;.„• 16. mg/L,.;
COD '•""•": 126 mg/L • -117 mg/L
TOC ; 32 mg/L - \ ; ,. : ' 19 ,mg/L ;
Oil and grease ... 1.0 mg/L . • 4 mg/L'
Surfactants (MBAS) ' 0.26 mg/L '0.12 mg/L1-;,.-
Flow rate 28,800 gal/day 33,540 gal/day
.
One maintenance problem is erosion of elbows and piping in the carbon
transfer system. In spite of the long radius elbows and flexible hoses cur-
rently used in the transport system, these items require replacement .at regular
intervals. • . : ' • .< '• '• i|
1
The carbon adsorption system was designed to treat 15,000 gpd of process
wastewater. The actual amount treated averages about 11,000 to 12,000 gpd. The
carbon exhaustion rate has averaged about 500 lb/1,000 gal compared with the
design rate of 430 lb/1,000 gal. :
The reactivation furnace has been operated at a nominal rate of 6,000
Ib/day. This rate is determined by the amount of carbon in one vessel as one
carbon column is reactivated each day. Early in the operation of the treatment
facility, the effluent from each column was analyzed for COD on a frequent
basis. It was soon discovered that one column should be reactivated each day to
maintain the optimum maximum organic removal consistent with good furnace
196
-------
operation. Consequently, the routine daily reactivation of the lead column was
established, and frequent analysis of the waste stream was discontinued.
No routine evaluation of reactivation process parameters such as the
apparent density test is currently practiced. Instead, the furnace is operated
on the basis of furnace draft and hearth temperatures, which are closely moni-
tored. Current losses average about 6 percent, based on a reactivation rate of
6,000 Ib/day and annual carbon purchases of 130,000 Ib (350 Ib/day).
Shortly after start-up, the furnace experienced a high rate of corrosion
of the rabble teeth and rabble arms. Replacing the original teeth with teeth of
different materials was done with no positive results. Some of the teeth
required replacement as often as monthly regardless of materials used. The
problem was eventually solved when the steam lines became plugged. Without
steam, corrosion of the teeth was much less severe so that replacement is now
required about every 6 months.
Operation of the afterburner has been discontinued with no noticeable
effect on stack emissions. No odor is noticeable, and air quality limits have
not been exceeded according to tests. The decision to discontinue its operation
was based on fuel costs (L.P. gas) when it was found that afterburner fuel
demand equaled the reactivation fuel demand.
CAPITAL AND OPERATION AND MAINTENANCE COSTS
Capital Cost
The initial capital cost of the GAC wastewater treatment and reactivation
facility, including the initial charge of activated carbon, was about $225,000
in 1973 dollars.
Operation and Maintenance
Recent cost calculations show that the treated water costs are $0.025/gal
(?25/1,000 gal). This is the total cost including makeup carbon, fuel, elec-
tricity, labor (both operating and maintenance), maintenance materials and
depreciation. Fuel (L.P. gas) use for reactivation is 110,000 gal/yr, which
results in a heat demand of 6,000 Btu/lb carbon reactivated. Fuel cost for
reactivation is $0.02/lb carbon.
BIBLIOGRAPHY
Byung Lee, personal communication, October, 1979.
197
-------
CASE HISTORY NO. 13
REPUBLIC STEEL CORPORATION
CLEVELAND, OHIO
INDUSTRIAL WASTEWATER TREATMENT
INTRODUCTION AND HISTORY
The Cleveland district steel mill of Republic Steel Corporation at Cleve-
land, Ohio, is a completely integrated mill producing a variety of finished and
semi-finished steel products. At the beginning of the process, coke for use in
blast furnaces is produced from coal in the coke plant. In the process of coke
production, wastewaters containing organic and inorganic materials are gener-
ated or released and, because of their nature, cannot be discharged without
treatment. Among these materials are oils, ammonia, phenols, cyanides, and
suspended solids. They are most commonly contained in three major process
streams: crude ammonia liquor, barometric condenser water, and intercepting
sump water. Typical concentrations of contaminants in 'the coking waste for a
6,000-tpd coke plant are shown in Table A-34.
TABLE A-34. CONTAMINANTS IN COKE PLANT* WASTE STREAMS
Concentration, mg/L CTR
Stream
Flow
(gpm) Ammonia
Phenol Oils Cyanide
Crude ammonia
liquor
Barometric con-
denser water
Intercepting sump
water
150
1,000
300
5,000
20
150
2,000
40
300
1,000
20
10,000
20
40
100
*6,000 tpa code production
In. the early 1970's, Republic Steel Corporation began to look for an
economically feasible treatment process for coke wastewater.
Some of the conventional processes considered by Republic Steel for treat-
ment of the coke waste streams included phenol removal by solvent extraction,
ammonia removal by limed ammonia still, and biological treatment of the com-
bined wastes. Based upon a review of available treatment technologies, adsorp-
tion using GAC for phenol and organic removal was chosen for testing and devel-
opment. Among the reasons for its selection was its ability to perform
resgardless of normal fluctuations of pH, temperature, flow,.and organic concen-
trations. Since GAC can also be reactivated and .reused, the process was con-
sidered to be potentially economically feasible.
198
-------
GAC Testing
Laboratory testing was conducted by treating samples of the wastewater
streams with activated carbon to develop preliminary adsorption isotherm data.
Based on favorable result's of the laboratory tests, pilot scale carbon columns
were operated on the barometric condenser water stream. Pilot plant data veri-
fied the laboratory data and yielded a carbon exhaustion rate of 5.3 lb/1,000
gal of barometric condenser water.
All three streams were then blended according to their respective flow
rates for testing in the pilot plant. These streams were found to be incompat-
ible for combined testing; because the crude ammonia liquor contained 30 to 50
times more phenol than the other streams, extremely high carbon exhaustion
rates resulted. A separate carbon adsorption system would be required for the
crude ammonia liquor in a full-scale treatment plant.
Pilot plant results also showed that the intercepting sump water from the
light oil refining plant contained large amounts of free and emulsified oils
that were not suitable for adsorption. The emulsified oils passed through the
column .without .being adsorbed, and the free oil coated the carbon particles
thereby reducing their ability to adsorb phenol. Pretreatment of that stream to
remove oils would be required before combining it with the barometric condenser
water for phenol removal in the activated carbon adsorption process.
An oil removal method was developed during the pilot plant studies that
included chemical treatment with spent pickle liquor and caustic soda to form
an insoluble iron precipitate to which oil would adhere. The chemically treated
water was then clarified in a dissolved gas flotation unit. Some of the cyanide
was also removed by'precipitation in this clarification process. The clarified
liquid waste stream was then suitable for adsorption on activated carbon.
During the pilot plant studies, the barometric condenser was replaced with
a surface condenser, significantly reducing the amount of water to be treated.
Also light oil refining was discontinued before completing the activated carbon
system design; this resulted in a significant reduction of emulsified oil.
DESIGN AND OPERATION
On the basis of laboratory and pilot plant test results and of the need to
limit discharges from the Cleveland district coke plant outfall, a treatment
system was designed to treat crude ammonia liquor to meet the following
limits:
Ammonia 250 Ib/day*
Phenol 2.5 Ib/day
Cyanide 40 Ib/day
SS 100 Ib/day
Oil and grease 25 Ib/day
pH 6-11
199
-------
The system as designed provides for separate treatment of crude ammonia
liquor and ammonium sulfate surface condenser condensate (Stream A) and of
intercepting sump water, surface condenser ejector, and secondary condenser
waters (Stream B). Stream B .is used for quenching coke after treatment.
Basis of GAC System Design and Operation
The GAC adsorption systems were designed to reduce the phenol concentra-
tion in Stream A from 4,216 Ib/day to 2.5 Ib/day and in Stream B from 1,363
Ib/day to 2.5 Ib/day. Since Stream B is not discharged, the effluent phenol in
that stream is not included in the outfall loading. Figure A-27 shows the flow
diagrams for both activated carbon treatment systems.
Before entering the treatment system, Stream A first passes through a
dephenolizer where the phenol concentration is reduced from 1,600 mg/L to about
30 mg/L. If the dephenolizer is not in operation, the wastewater can be treated
with, spent pickle liquor and caustic soda and sent to a dissolved gas flotation
system. The stream is then passed through a multi-media filter where the sus-
pended solids are reduced to about 15 mg/L.
After filtration, the wastewater enters the top of the first . carbon
column, flows downward through the carbon, and -exits from the bottom. It then
enters the top of the second column, flows downward through the carbon bed and
out to the ammonia still from which it is discharged to the receiving stream.
Since the carbon columns are pressurized steel vessels, no intermediate pumping
is required. '
Periodically one column is taken off stream for reactivation. This occurs
whenever the first column .effluent contains phenol greater than 5 mg/L. After
the spent carbon is removed, reactivated carbon is placed in the column and the
column is returned to service as the second or polishing column.
. In a similar fashion, Stream B is treated with waste pickle liquor and
caustic soda, passed through a dissolved gas flotation unit and mixed media
filter, and it then flows through two series-connected downflow activated car-
bon columns as previously discussed. Effluent from this treatment system
receives no further treatment since it is suitable for coke quenching.
Because of the low solids concentration in the filter effluent and the
rate of carbon exhaustion, no provision was made for backwashing of the GAC
columns.
When the carbon in one column in either system loses its ability to adsorb
additional phenol (becomes spent), as determined by effluent phenol concentra-
tion, it is transferred in a water slurry from the column to the spent carbon
storage tank near the reactivation furnace. It then flows in a slurry by grav-
ity into a dewatering screw conveyor that moves the carbon slowly up an
inclined trough and allows excess water to drain from the carbon. The drained
carbon is then discharged into the reactivation furnace.
'Ammonia limit was later changed to 500 Ib/day
200
-------
STREAM A
FROM DITHENOIIZER *
I.TERNATE STREAM A
EPMENOLIZEM DOWN) ^
Gk 1 1
r
1 i
CHEMICAL DISS. GAS
MUINR FiniAiiriN
>.
1
k
i
f
>
..*-,
l»
1
MIXED MEDIA
FILTER
ACTIVATED
CARBON COLUMN
TO AMMONIA STILL
AND OUT I ALL
ACTIVATED
CARBON COLUMN
K>
O
STREAM B
IT
CHEMICAL
MIXING
DISS GAS
FL01ATION
k
^
k.
Ml
KEO
>
k.
" ' r
MEDIA
k,
1
1
J
k...
r_
b.
1
1
TO COKE PLANT
FOR QUENCHING
, k
ACTIVATED
CARUON COLUMN
ACTIVATED
CARBON COLUMN
Figure A-27.
Flow diagram of granular activated carbon wastewater treatment system.
Corporation.
Republic Steel
-------
The empty carbon column is refilled with reactivated carbon from the
reactivated carbon.storage tank. A small quantity of virgin carbon is-added to
the column to makeup for that lost (usually about 5 to 8 percent) during trans-
fer and reactivation.
The adsorption system design criteria is as follows:
Granular Carbon Adsorption System
;Stream A ,:. . • Flow 220 gpm
Number of carbon columns Two
Type Downflow, series
Column size 11.0 ft diameter x 24.5 ft high
Carbon bed depth 18 ft
Surface loading rate 2.3 gpm/ft^
Contact time (empty bed) 116 min total
Carbon quantity 48,000 Ib each
Stream B Flow 440 gpm
Number of carbon columns Two
Type Downflow, series
Column size. 11.0 ft diameter x 24.5 ft high
Carbon bed depth 18 ft•
Surface loading rate 4.6 gpm/ft^
Contact time (empty bed) 58 min total
Carbon quantity 48,000'Ib each
Basis of Reactivation System Design and Operation
The purpose of the carbon reactivation system is to restore the activity
of the spent or used carbon to near its original characteristics for reuse at a
more economical cost than if virgin carbon were used. This reactivation takes
place in a furnace at maximum temperatures of 1,600° to 1800°F in the absence
of excess oxygen. Steam is used to improve the quality of the reactivated gran-
ular carbon. The following is the design criteria for the reactivation
furnace:
Carbon Reactivation Furnace
Furnace type Multiple hearth
No. of hearths Eight
Furnace size 16-ft O.D.
Furnace capacity 68,000 Ib/day carbon, dry basis
Afterburner type Internal (zero hearth)
Fuel, primary . Natural gas
alternate Coke oven gas
Dewatered spent carbon containing 45 to 50 percent moisture by weight
enters the furnace as it falls from the dewatering screw onto the top hearth
(Hearth No. 1). It is moved across the hearth by means of rotating rabble arms
attached to a central shaft. Stainless steel teeth mounted on the rabble arms
202
-------
act as plows to move the carbon and expose it to the hot furnace gases. As
the carbon moves across the hearth, it drops through a hole to the next lower
hearth where the rabbling process is repeated; the carbon is moved inward and
outward on succeeding hearths. As the carbon moves downward through the furnace
it becomes hotter. Drying of the carbon and volatizing of the organics takes
place in the upper hearths, and reactivation is accomplished on the lower
hearths. Burners are located at Hearths No. 4, 6, 7, and 8 (the bottom hearth),
and steam is added at a rate of 1 Ib steam/lb carbon. Maximum furnace tempera-
tures occur at Hearth No. 8.
Off-gases from reactivation are drawn upward through the furnace and
afterburner where they are incinerated at 1600°F to prevent the discharge of
objectionable odors to the atmosphere. The heated gases then pass through a
quencher and scrubber that cools the gas and removes much of the fine particu-
late material including carbon fines and ash.
After reactivation, the carbon falls into a quench tank .at the. furnace
outlet. Water is maintained in the tank to cool the hot incandescent carbon and
to maintain an air-tight seal on the furnace outlet. From the quench .tank, the
reactivated carbon drops into a blow-case that is pressurized with air to move
the carbon to the reactivated carbon storage tank. The carbon handling system
and reactivation flow diagram are shown on Figure A-28.
Operational Problems
The carbon columns are made, of carbon steel with a vinylester lining to
prevent corrosion. These linings have failed on occasion and have been repaired
using the same lining material. A more suitable lining material is currently
being sought.
Operating problems with the multiple hearth reactivation furnace have
included insufficient temperatures in the afterburner and instances of corro-
sion. The afterburner was designed to operate at about 1600°F. During actual
operation under steady-state conditions, the maximum afterburner temperature is
typically 1,350°F. No serious odor problems have been detected as a result of
the lower temperatures nor is there any record of violation of air quality
limits.
The water recirculation line to the off-gas precooler and scrubber, orig-
inally of carbon steel, was replaced with stainless steel because of corrosion
caused by sulfur dioxide (S02) in the gas. Portions of the stack also corroded
because of the S02- In some areas of the stack 304 SS has been used in place of
316 SS. When these were replaced by the proper material, corrosion was no
longer a problem.
SYSTEM PERFORMANCE
The GAC treatment system began continuous operation in July, 1977. From
the beginning, the treated effluent quality has met or exceeded the design
phenol performance. The discharged effluent from treatment Stream A typically
contains 0.025 mg/'L, phenol.
203
-------
K)
o
t*.
RFACTIVAiri)
CARBON
TO COLUMNS
RIACTIVATT.D
CARBON STORAGE
VIRGIN CARBON
FROM TRUCK
VIRGIN CARnON
STORAGE
Figure A-28. Flow diagram of granular activated carbon handling and reactivation system,
Republic Steel Corporation.
-------
Carbon exhaustion rates have averaged about 35 lb/1,000 gal and 25
lb/1,000 gal for Streams A and B, respectively. The reactivation furnace,
designed to reactivate 68,000 Ib/day of spent granular carbon, has been oper-
ated at an actual rate of 20,000 Ib/day. Quality of the reactivated carbon as
measured by iodine number is in the range of 860 to 870 compared with 900 mini-
mum for virgin carbon.
Initial carbon losses were high with some reactivation cycles showing 20
percent loss. More recently they have been averaging about 8 percent.
During the initial start-up period in late 1976 and early 1977, effluent
phenol would occasionally exceed the discharge limits even though the carbon
columns were fresh. This difficulty was traced to a defective valve that
allowed untreated wastewater to bypass the columns. The valve was repaired and
the problem was alleviated.
Phenol would leach from a heel of spent carbo.n that remained in the
columns after spent carbon was removed. Even though reactivated carbon was
placed in the column, the small amount of spent carbon would cause the effluent
phenol concentration to exceed discharge specifications. A method was found
whereby all of the spent carbon could be removed from the column during
transfer.
CAPITAL AND OPERATION AND MAINTENANCE COSTS
Capital Costs
The original capital cost for the treatment facility which included chem-
ical treatment, dissolved gas flotation, mixed-media filtration, solids
dewatering and disposal, GAC adsorption and reactivation, was about $10 mil-
lion. An exact figure is not available since Republic Steel did the construc-
tion. The cost for the adsorption or reactivation systems apart from the other
treatment units also could not be identified.
Operation and Maintenance Costs
Operating and maintenance costs for the wastewater treatment systems
including chemical treatment, dissolved gas flotation, filtration, sludge
dewatering, adsorption, and granular carbon reactivation have averaged about
$8,000/day. This amount does not include depreciation nor the cost of operating
the dephenolizer and ammonia still.
Natural gas i^ used as fuel in the reactivation furnace, but fuel usage
rates or cost of reactivation are not available.
BIBLIOGRAPHY
Naso, A.C. and John, E.T., "Physical-Chemical Treatment of Cleveland District
Coke Plant Waste Waters", presented at the American Iron & Steel Institute 85th
General Meeting, New York, N.Y., May 25, 1977.
205
-------
Naso, A.C., and Schroeder, J.W., "A New Method of Treating Coke Plant Waste
Waters", Iron And Steel Making, March, 1975.
Naso, A.C., and Schroeder, J.W., "A New Method of Treating Coke Plant Waste
Water", Iron and Steel Engineer, December, 1976.
J.W. 'Schroeder', personal communication, September, 1979.
206
-------
CASE HISTORY NO. 14
VILLAGE OF LEROY, NEW YORK
PHYSICAL-CHEMICAL WASTEWATER TREATMENT
HISTORY AND INTRODUCTION
In 1969, New York State's Division of Pure Waters advised the Village of
LeRoy that it would be required to install tertiary treatment and disinfection
if discharge was to continue in the vicinity of the existing primary effluent
outfall.
Further, it was expected that such discharge would meet groundwater clas-
sifications and standards established by the Water Resources Commission. After
several years of planning, .and then delays while awaiting Federal funding,
construction began in December, 1975.
The receiving water body for the wastewater from the Village of LeRoy in
Oatka Creek, one of the finest trout fisheries in western New York. To protect
this fishery, an extremely high degree of treatment was required. Detailed
analysis determined that the point of discharge was critical. The treatment
required for adequate stream protection was complete wastewater renovation. An
extensive pilot plant demonstration program indicated the most economical sol-
ution was the use of a physical/chemical treatment process.
The selected major processes were:
1) Clariflocculation to remove gross solids, fine and colloidal solids,
and phosphorus;
2) Rapid sand filtration to remove residual floe and solids to -protect
subsequent floe and solids to protect subsequent processes;
3) Activated charcoal absorption/adsorption to remove residual organics;
4) Breakpoint clorination, followed by activated carbon de-chlorination
to remove ammonia.
These processes are supported by recarbonation of lime-softened water,
final aeration, an-', on-site activated carbon regeneration. Figure A-29 is a
flow schematic for Jie plant.
The control panel is designed so that major systems in the plant operate
automatically and provides the operator with the necessary details to monitor
the entire plant. The displays monitor the positions of critical valves and the
condition of the primary elements of the process system. Automatic systems
controlled by this panel are chemical feeds, backwashing of the sand filters
and carbon columns, as well as the carbon transfer processes for regeneration.
207
-------
ENTRANCE MANHOLE ••
RAW SEWAGE
ENTRANCE STRUCTURE
2 COMMINUTORS BAR SCREEN
BYPASS,
PLANT DRAINAGE,
^ •»
GRIT
___!
CHAMBER & F
- .^J
CLARI-
FLOCCULA-
TOR
RECARBONATION TANK
PRE CHLORINATION
FEED
PLANT
SANITARY
SEWER
BYPASS ANY
OR ALL UNITS
BACKWASH
BYPASS ANY
OR ALL UNITS
FILTER EFFLUENT PUMPS
2 UNITS
/CAR
(ADSOR
I VESS
V 4 Ut
CHLORINE
f
BON\
BTION\*
ELS j
>JITS J
BLENDER ^
SODIUM
HYDROXIDE FEED
J BREAK POINT
ICHLORINATIONFEEC
BYPASS ANY
OR ALL UNITS
CHLORINE
ADSORBTION
VESSELS
2 UNITS
• BACKWASH
CHLORINE CONTACT CHAMBERS
2 UNITS
AERATION CHAMBER
POST CHLORINATION
FEED
OUTFALL SEWER
Figure A-29. Wastewater Flow Schematic, Village of LeRoy, NY.
208
-------
The control panel has various alarms that notify the operator of system mal-
functions and provide for manual override should the automatic mode fail.
GRANULAR ACTIVATED CARBON DESIGN DATA
Design capacity of GAC system
Average flow through GAC system
System Configuration
Adsorption
Number of contactors
Vol. per contactor
Diameter of contactors
Empty bed contact time
Reactivation
Type of furnace
Furnace size—physical description
Number of hearths
Diameter
Length
Frequency of reactivation
Activated carbon loss
Loss in adsorptive capacity per
reactivation cycle
Furnace capacity
Average furnace throughput
One mgd
0.450 mgd
Four
1285 ft3
11 ft
12 min
Fully automatic multi-hearth
Five
9-ft 3-in., O.D.
24-ft, 3 1/4-in., overall
13-ft, 6 1/2-in., shell
Quarterly
8.5 percent
Unknown
11,376 Ib/day
12,000 Ib/day
Figure A-30 is a cross-section through a carbon contactor.
PLANT PERFORMANCE
There is very little experience, particularly with regard to costs, in
operation of the plant. Plant operation has been intermittent and plagued with
mechanical problems and problems of waste septicity. By November, 1979, most of
the mechanical problems of the carbon adsorption system and carbon reactivation
system had been corrected. Some mechanical difficulties with the lime slaker
remained, and breakpoint chlorination treatment had not yet been practiced.
The principal remaining roadblock to full plant operation and anticipated
plant performance is septicity. The raw sewage is septic when it enters the
plant and remains septic throughout all of the treatment processes. Since the
plant is a physical-chemical treatment operation, the facilities installed to
correct septicity are rather limited. Until aerobic conditions can be estab-
lished in the carbon columns, the results of adsorption treatment are likely to
remain unsatisfactory and excessive reactivation of carbon will be needed.
Meaningful data on carbon system operations and costs cannot be obtained
until the pretreatment process and mechanical problems are corrected.
209
-------
•LEVEL .INDICATOR
\Yi 'LOW PRESSURE
VACUUM RELIEF VALVE
2H PRESSURE
RELIEF VALVE
DRAIN TO
BACKWASH LINE
AIR OPERATED
BUTTERFLY VALVE
2. SURFACE WASH
PNEUMATICALLY
OPERATED
BUTTERFLY VALVE -*'
13.8'DEPTH OF
ACTIVATED CARBON
PNEUMATICALLY OPERATED
3 - WAY VALVES
SEWAGE FROM MIXED
MEDIA FILTERS
BACKWASH TO DEBRIS TANK
GRAVEL
FILTER BLOCK
CONCRETE FILL
PNEUMATICALLY OPERATED
3-WAY VALVES
V \_/AIR OPERATED
J BALL VALVE
^CARBON SLURRY FROM
REGENERATED CARBON
HOLDING TANK
PIPING SUPPORT
.SURFACE WASH
' 'ASSEMBLY
LEVEL SENSING
PLUMB BOB
ACTIVATED CARBON
AIR OPERATED
BALL VALVE
CARBON SLURRY TO SPENT
CARBON HOLDING TANK
12" BACKWASH
TO DECHLORINATION VESSELS THROUGH
CHLORINE IN-LINE BLENDERS
Figure A-30. Carbon absorption vessel single coluinn schematic,
Village of LeRoy, NY.
210
-------
CAPITAL AND OPERATION AND MAINTENANCE COSTS
Capital Cost (1975)
Total system cost for piping, pumping,
surge tanks, adsorbers, and reactiva-
tion furnace
Amortization period
Interest rate
$2,500,000
40 years
5 percent
Adsorption Operation and Maintenance Costs (1979)
Operations and maintenance labor
Number of persons employed
Number of hours worked per year
Average wage rate
Electricity
'unit cost
Maintenance materials
Six
14,350 hr
$6.60/hr plus benefits
82,080 kwh/mo
$0.0394/kwh
Unknown—system still belongs
to Contractor
Regeneration Capital and Operating Costs (1979)
Capital cost
Furnace cost
Operating cost
Operations Labor
Number of persons employed
Total hours
Average wage rate
Maintenance labor
Number of persons employed
Number of hours worked per year
Average wage rate
Electricity
Total annual cost
Unit cost
Fuel
Type of fuel used
Total annual cost
Quantity
Unit cost
Maintenance mat* rials
Carbon
Total initial volume
Makeup
Unit cost
Included in figures above
Four
6,000 hr/yr
$6.74/hr plus benefits
One
100 hr
$6.35/hr plus benefits
103,000 kwh/yr (estimated)
$4,058
$0.0394/kwh
#2 Fuel oil
$18,900
30,000 gal/yr
$0.63/gal
None yet—have not accepted
furnace from the Contractor
7,500 ft3
68,000 Ib/yr (estimated)
$68.36/cwt
211
-------
CASE HISTORY NO. 15
CITY OF MANCHESTER, NEW HAMPSHIRE
PUBLIC WATER SUPPLY
HISTORY AND INTRODUCTION
Manchester, New Hampshire, is a city greatly concerned about its future
growth and prosperity. While promoting economic expansion and residential
growth, Manchester is actively preparing to accommodate the needs posed by this
development. One step in this direction has already been taken—that of provid-
ing a safe, sufficient water supply for public and commercial consumption.
For more than a century, Lake Massabesic has served as the water supply
for Manchester. When pollution for a growing population and industry jeopar-
dized other domestic water supplies, Lake Massabesic—because of its natural
characteristics and the constant efforts of the Manchester Water Works to safe-
guard its watershed--continued to supply the City with good quality water.
In recent years, however, Lake Massabesic water has been plagued by exces-
sive color and severe periodic taste and odor caused by aquatic organisms. Oil
and chemical spills caused by tanker trucks, which .must pass over a narrow neck
between the front and back ponds of Massabesic while traveling on the Route 28
Bypass, .could contaminate 'the entire water supply. Projections for future
water consumption indicate that, by the mid 1980's, Lake Massabesic will no
longer be able to satisfy Manchester's average daily consumer demand. The poor
quality water from the Merrimack River, now planned to supplement the Messa-
besic supply, will definitely require treatment.
These factors prompted the Manchester Board of Mayor and Aldermen to auth-
orize the Board of Water Commissioners in May, 1970, to file an application for
a federal grant to help finance the construction of a new water treatment
plant. A few months later, the Department of Housing and Urban Development
approved a $1,500,000 grant for this project. The Manchester Water Works then
engaged consulting engineers bo work with department engineers on the design of
a water filtration plant capable of treating an average capacity of 30 mgd
(maximum 40 mgd) to be built adjacent to Lake Massabesic.
Facility Planning and Testing
During the following months, the Water Works staff, assisting the consult-
ants, undertook a comprehensive series of engineernig evaluations in such areas
as filtration, standby power, treatment systems, and chlorine generation before
formulating a facility that would incorporate the most innovative concepts and
sophisticated techniques in the field of water treatment.
An extensive predesign study included comprehensive raw water analysis and
testing; evaluation of alternate treatment techniques in view of their
212
-------
efficiency, operating costs, and initial construction costs, and model testing.
This study revealed that complete conventional pretreatment including flash
mixing, flocculation, and sedimentation must be undertaken before filtration.
Treatment for the first 10 years of operation would concentrate on remov-
ing raw water color and controlling periodic taste and odor. Removing volatile
organics, turbidity, and other contaminants from Merrimack water would be done
in the future. Initially, powdered activated carbon was considered to remove
taste, odor, and other volatile organics, but because of the unpredictability
of the treatment process, continuous filtration through GAC was chosen to guar-
antee reliable.performance.
To ensure maximum use and life of the activated carbon, the carbon filters
were designed to operate in series with the sand filters. This arrangement
permits sand filters to remove all pinpoint floe and other particulate matter
remaining in the settled water? the carbon filters serve only to absorb organic
compounds. A basic flow schematic of the processes used at Manchester is given
on'Figure A-31.
DESIGN.AND OPERATION
Basis of GAC System Design
The Manchester facility is the first water purification plant in the coun-
try ; to .be , designed for .series, sand and carbon filtration. Both the sand and
carbon filters are automatic backwash filters (ABW). ABW filters need not be
removed;:from service for backwashing. ..Each: filter consists of. several indepen-
dent cells that can, be isolated individually fand backwashed while the remaining
icelis continue to process the'water.
Backwashing is accomplished in the same way for both sand and carbon fil-
ters. A carriage mechanism that spans the width of the filter travels about 2
ft/min across the bed. Water is pumped beneath a single cell, up through the
sand to the backwash hood, and into the washwater pump for disposal in the
washwater launder located along the filter wall. After filtration, water flows
into a clearwell located beneath the finish water pump room from which water
will be pumped into the distribution system. Backwash water from the sand fil-
ters flows by gravity into three sludge settling lagoons outside the plant.
Filter backwash water and carbon transport water are recycled to the headworks
of the water treatment plant.
The following are the design data for the system
Filtration and GAC adsorption
Number of sand filters Four
Number of carbon filters Four
Filter surface loading at 30 mgd 2.96 gpm/ft2
Sand media size 0.60-0.65 mm
Sand media depth 11.in.
Carbon media size 8 x 30 mesh
213
-------
LAKE MASSABESIC
CARBON 1
(REG) (VIR)
CARBON 2
CARBON 3
CARBON 4
FLUIDIZED BED
REGENERATOR "
STORAGE AND
DISTRIBUTION
Figure A-31. Regeneration and.treatment schematic, Manchester, NH.
-------
Carbon media depth 48 in.
Clearwell volume 400,000 gal
GAC System Operation
The Manchester Water Treatment Plant features conventional chemical pre-
treatment, coagulation, flocculation, and sedimentation, followed by series
filtration through four, 11-in. deep rapid sand filters and four, 48-in. deep
GAC filters. The sand filters serve to remove floe carryover and other particu-
lates that might tend to interfere with adsorption and thereby reduce GAC life
expectancy. All eight of the filters have a maximum design surface loading rate
of 4.0 gpm ft2 and are presently operated in series at a rate of about 1.3
gpm/ft2 and empty bed contact time (GAC filters., only) ranging from 14 to 22
min. The carbon itself is of coal base origin having been used successfully for
the continuous removal of tastes and odors due to algae since startup of the
treatment plant in November, 1974.
Briefly, the cycle that the carbon will pass through in going from the
filters to the regeneration building and then back to the filters is described
as follows. First, the GAC will be hydraulically educted from each filter and
directed to a spent carbon storage tank located in the regeneration building.
At this point, it is then passed through a dewatering and metering screw and
then fed to the upper drying stage of the regenerator described earlier. Once
the carbon passes through the regenerator and exists at the bottom of the fur-
nace, it will be placed into a quench tank for cooling and then into a regener-
ated carbon storage tank for subsequent return to the filters via water jet
eductors. . .
Reactivation of GAC
In light of the recent EPA regulations for trihalomethanes (THM) and for
carbon treatment, water utilities will more commonly use GAC. In the past, the
use of carbon in water treatment has been minimal; the cost for using it on a
throw-away basis has been moderate. Recently, faced with using large quantities
of carbon at ever-increasing costs, certain water utilities have begun using
on-site reactivation systems.
Manchester has installed a WESTVACO fluidized bed furnace to reactivate
the spent carbon from its GAC system (Figure A-33). EPA sponsored the initial
operation and testing of this system.
The WESTVACO fluidized bed design operates on a two-stage approach (illus-
trated on Figure A-T3); the wet spent carbon is dried in the upper stage of the
furnace using the off-gas from the lower regeneration stage as the primary heat
source. In the drying zone, a small amount of the organics may be volatized
from the carbon, but the majority remain adsorbed on the carbon and pass to the
lower regeneration zone. In this area, part of the adsorbates are volatized
from the carbon and the remaining pyrolyzed portion is removed by selective
reaction with steam until complete regeneration is achieved. Volatile adsor-
bates are combusted in the incineration zone above the regeneration bed while
the off-gas is passed through an air scrubber where carbon dust is removed.
215
-------
SPENT
CARBON
STORAGE
TANK
DEWATERING SCREW
REGENERATED
CARBON
COMBUSTION
AIR BLOWER
EXHAUST GAS
JL
OFF GAS
SCRUBBER
WESTVACO
FLUID BED
REGENERATOR
EXHAUST GAS
BLOWER
TEMPERING GAS
EXHAUST
STACK
RECYCLE
GAS BLOWER
EDUCTOR
Figure A-32. Process flow for activated carbon regeneration system, Manchester, NH.
-------
( PATENT APPLIED FOR)
STEAM
FUEL
Figure A-33. WESTVACO fluidized bed furnace for reactivating spent carbon.
217
-------
Reported operating advantages of this type of system include better fuel
economy and no internal moving parts, which may result in less maintenance and
lower costs.
The fluid bed regeneration system now installed at Manchester has a design
operating', capacity of 12,000 Ib/day (500 Ib/hr) of regenerated carbon and a
projected estimated total operating cost of 8.9
-------
TABLE A-35. SUMMARY OF ESTIMATED* DIRECT AND INDIRECT COSTS FOR
OPERATION OF FLUID BED REGENERATION SYSTEM
MANCHESTER, NH
Item
Direct costs
Makeup carbon ( 1 ) t
Labor (2)
Maintenance and repairs (3)
Fuel (4)
Power ( 5 )
Water (6)
Year
S210,000
• "20,800
13,608
58,800
8,820
8,142
Month
$17,500
1,733
1,134
4,900
735
679
Day
$600.00
59.43
38.88
168.00
25.20
23.26
Ib GAC
$0.0500
0.0049
0.0032
0.0140
0.0021
0.0019
Total
Indirect costs
Depreciation (7)
Insurance (8)
Administration and overhead (8)
Total
Total direct and indirect costs
Total
$320,170 $26,681
$ 43,750
695
8,950
$ 3,646
58
729
$ 53,195 $ 4,433
$915.77 $0.0761
$125.00
1.98
25.00
$0.0104
0.0002
0.0021
$152.00 $0.0127
$373,365 $ 31,114§ $7,068.00 $0.0890
*Estimate is based on steady state operation, 350 days/yr. Actual costs of
fuel, power, and labor are expected to rise above figures shown.
tcost components
(1) 10 percent loss, $0.50/lb GAC
(2) 24 operator hr/day, $208/wk/oper
(3) 3 percent of equipment cost/yr
(4) 10 gal/hr fuel consumption, $0.70/gal fuel cost (for regenerator and
steam boiler)
(5) 0.06 kw/lb GAC 3 1/2/2
-------
TABLE A-36. SUMMARY OF FINAL CONSTRUCTION COSTS FOR PHASE ONE OF
CARBON REGENERATION PROJECT - MANCHESTER, NH PROJECT,
. . EPA GRANT R8057371 ..'••'"'
Item Number Construction Amount
1. Carbon regeneration building $226,852
(34 ft x 39 ft x 30 ft high)
2. Fluid bed carbon regeneration system 528,630
(capacity - 6 ton/day)
3. Semi-automatic carbon transport 120,517
system (4 filters)
4. Miscellaneous costs 6,407
5. Total construction cost $882,406*
*Total cost of Item 7 of Financial Report to EPA, September, 1979. The above
costs have been reported to and are currently being reviewed by the EPA and
are not to be published without the prior written approval of the EPA Project
Officer. •. . • ...
BIBLIOGRAPHY
Manchester Water .Department. Brochure on dedication of new water treatment
plant, 1974.
Manchester Water Works. 107th Annual Report, 1978.
Manchester Water Works. Summary of Final Capital Costs for Phase One of Carbon
Regeneration Project. EPA Grant No. R805371.
.Manchester Water Works. Summary of Final Construction for Phase One of Carbon
Regeneration Project. EPA Grant No. R805371.
Kittredge, D., "The Economics of Carbon Regeneration - State of the Art".
Presented at Conference of the NEWWA, Lake Placid, New York, September 20,
1978.
WestVACO Chemical Division. The Manchester, New Hampshire, Water Works Fluid
Bee. Carbon Regeneration Facility. Nuclear Newsletter No. 3, WestVACO.
WestVACO Chemical Division. The WESTVACO Fluid Bed Carbon Regeneration System.
WestVACO.
A visit to the Manchester Water Treatment Plant in October, 1979, and a discus-
sion with David Kittredge and other Department staff members.
220
-------
CASE HISTORY NO. 16
PASSAIC VAT.T.KY WATER COMMISSION
LITTLE FALLS WATER TREATMENT PLANT
CLIFTON, NEW JERSEY
MUNICIPAL WATER TREATMENT
HISTORY AND INTRODUCTION
The Little Falls Filtration Plant of the Passaic Valley Water Commission
is located on the Passaic River near Towata, New Jersey. The Passaic River,
which is the source of supply, is a highly polluted stream that receives indus-
trial wastes from hundreds of upstream discharges. Organics have long been a
problem and GAC has been used experimentally for adsorption of organics.
Because of the organics problems and the local interest in their control,
the Little Falls Plant was one selected by EPA to study various methods for
reactivation of GAC that has been saturated with organics adsorbed from drink-
ing water.
The infrared electric furnace is the thermal reactivation equipment to be
evaluated at Little Falls. The potential advantages of this method are: (1) the
general availability of electric power as compared with the fossil fuels
required by other types of furnaces, (2) shorter warm-up and cool-down periods
as compared with alternative equipment, and (3) lower attrition of carbon on
the traveling belt used in the electric furnace as compared with attrition
losses in furnaces involving movement of the carbon by rabbling or fluidizing.
GAC REACTIVATION SYSTEM
The Little Falls Filtration Plant employs prechlorination, alum coagula-
tion, settling, dual-media (anthracite coal over sand) filtration, and, at
times, dechlorination. The average flow is 55 mgd. A side stream of filtered
water (2.2 mgd) is diverted to three pressure vessels each containing a differ-
ent type granular activated carbon with an empty bed contact time (EBCT) of
approximately 8 min.
GAC Reactivation Furnace Design
The furnace being evaluated is a 100 kw, 100 Ib/hr infrared tunnel unit
manufactured by Shirco, Inc. of Dallas, Texas, with the following characteris-
tics (see Figure A-^4).
Dimensions
Activated carbon residency time
Temperature
Drying zone
Activation zone
Warm-up time
Weight
4-ft wide x 20-ft long
20 to 25 min
1150°F
1650°F
30 min
17,000 Ib
221
-------
REMOTE CONTROL PANEL
to
IO
to
GAC IN
4V
ret#
i 1 -t— 1— *••»!>'
*-' ^v
FEED MODULE
f-
•h
u
•
1
L
— T-I
DRYING, PRYROLYSIS AND
I
^
^
\J
DISCHARGE
I-*- TO EXHAUST
--REACTIVATED
GRANULAR
ACTIVATED CARBON
TO QUENCH TANK
ACTIVATION MODULES
MODULE
Fiqurp A-34. Cross section of infrared tunnel furnace. Little Falls Water Treatment Plant, Clifton,
N J.
-------
OPERATION AND PERFORMANCE
The GAC adsorbers became operational in March, 1978 and carbon was exposed
continuously (without backwash) for 3 months. The adsorbents were then sequen-
tially educted into .a storage bin, fed into the furnace, then returned to the
adsorbers for quenching. By May, 1979, the granular carbon had been reactivated
twice and plans were to have at least two more exposure/reactivation cycles
before the project was completed.
For the initial reactivation, the contact time for the granular carbon
within the furnace was 20 min at a temperature of 1650°F. These conditions were
selected by the furnace vendor on the basis of thermographic analysis and
iodine numbers.
The first carbon to be reactivated in this system was the lignite base HD-
1030 (a product of ICI Americas, Inc.). Unfortunately, because of inexperience,
problems in handling the adsorbent and in the operation of the furnace occur-
red, and excessive atypical losses resulted. Further, performance of the lig-
nite activated -carbon could not be equitably judged. The characteristics of the
two coal-based granular carbons (WESTVACO WV-W and Calgon F-400) before and
after reactivation are shown in Table A-37. Iodine numbers were restored to
within 12 percent of the virgin granular carbon, and apparent densities
increased about 3 percent through the cycle.
TABLE A-37. COMPARISON OF VIRGIN TO REACTIVATED GRANULAR CARBON, CYCLE 1
Granular carbon
Virgin
Carbon
Exhausted
Reactivated
Iodine number
HD-1030
WV-W
F-400
601
850
1023
416
711
790
(See text)
761
890
Apparent density
HD-1030
WV-W
F-400
0.44
0.57
0.47
0.44
0.61
0.51
(See text)
0.57
0.49
The real test, of course, was how the GAC performed on the actual treated
Passaic River water. The total organic carbon (TOO concentration in the influ-
ent to the virgin activated carbon ranged from 1.5 mg/L to 3.9 mg/L, with a
mean, of 2.7 mg/L. tfher. the adsorbents were reactivated and put back into ser-
vice, the TOC had a mean concentration of 4.4 mg/L. Because of the different
loadings, comparison of virgin to reactivated carbon is difficult, but if these
differences are neglected, the decline in TOC removal is approximately 12 per-
cent for the 11 weeks in service after reactivation (see Figure A-35). Subse-
quent cycles should allow for more direct comparisons.
223
-------
to
*»
INF. TOC: 1.5-5.3 mg/l
EBCT=8mln
4 6 8
TIME IN SERVICE. WEEKS
10
12
Figure A-35. Effectiveness of reactivation measured by total organic carbon.
-------
Organic halogen (TOX) concentrations on samples of activated carbon col-
lected from the top and bottom of each adsorber were averaged and the results
shown in Figure A-36. Again, the reactivated carbon was about 10 percent less
effective than the virgin material. Because of a modification in the TOX analy-
sis, the influent and effluent TOX concentrations before and after reactivation
cannot be compared directly. For perspective, however, the nonpurgeable organic
halogen concentrations applied to the adsorbents after reactivation ranged from
106 g/L to 260 g/L, and the breakthrough pattern was very similar to the TOC
curve.
Although both UV absorbance and fluorescence data were collected, the
latter parameter is more consistent and is potentially a good operational mon-
itor for this system. Figure A-37, a plot of fractional removal of fluores-
cence, shows about 4 percent difference between virgin and reactivated granular
carbon.
In the adsorption of organics from highly treated municipal wastewater at
South Tahoe and Orange County (Water Factory 21), the apparent loss in GAC
adsorptive capacity during the first regeneration as measured by the iodine
number was 10 to 15 percent, but there was no observed loss in capacity to
remove COD. After the initial regeneration at these two locations, there was no
further decline in iodine numbers. At South Tahoe, where GAC has been in ser-
vice for 11 years, the carbon has been through a full 20 cycles of
reactivation.
Discussion of Results
The study by the Passaic Valley Water Commission at the Little Falls Fil-
tration Plant is providing very useful information for understanding the hand-
ling and reactivating of granular carbon. The infrared furnace in the first
reactivation restored the adsorptive properties (depending on which parameter
performance is judged on) to within 85 to 96 percent of virgin.
The second reactivation was completed, and in May, 1979, twice reactivated
carbon was being exposed to filtered water. Furnace conditions were changed
slightly for the second reactivation. The temperature in the activation zone
was raised to 1700°F and the retention time lengthened to 25 minutes. The
iodine numbers and apparent densities improved and, when compared with the
first reactivation, these preliminary indications are encouraging that other
removal parameters such as TOC and TOX will also improve.
Improvement in iodine numbers and apparent densities of the GAC obtained
by use of higher re activation temperatures and more time in the furnace may be
accompanied by higher carbon losses due to burning. Experience with reactiva-
tion of GAC in food processing plants and in wastewater treatment has shown
that it is possible to strike a satisfactory economic balance between burning
losses and restoration of carbon activity by proper manipulation of furnace
temperatures and retention time. In these applications, the optimum carbon loss
per cycle (total attrition and burning loss) is usually in the 8 to 10 percent
range.
225
-------
M
to
(n
x
o
NOTE: THESE VALUES ARE AVERAGES
OF TOP AND BOTTOM SAMPLES
6 8
TIME IN SERVICE, WEEKS
Fi.qi.ire A-36. F.ffechiveness of reactivation measured by total organic halogen.
-------
4 • •
TIME IN SERVICE. WEEKS
10
Figure A-37. Effectiveness of reactivation measured by fluorescence.
-------
Operational Problems
The methods used to transfer the exhausted carbon from the adsorbers to
•the storage hopper are unsatisfactory. The greatest deficiency, however, lies
in the inability to feed spent carbon to the furnace at a uniform metered rate.
The present system for doing this is totally unsatisfactory. Carbon is fed to
the furnace at an unknown, highly variable rate. Further, the method of
dewatering the carbon results in a highly variable moisture content in carbon
fed to the furnace. It is surprising that the furnace has worked as well as it
has considering that the layer of carbon on the furnace belt is of nonuniform
thickness and with a variable moisture content. The furnace cannot possibly
operate at reasonable efficiency under these conditions. A fair assessment of
furnace capabilities can only be made if the present facilities for dewatering
and feeding the carbon to the furnace are completely replaced with a satisfac-
tory system.
Severe corrosion problems have been experienced in exhaust gas scrubbers.
There are two potential corrective • measures: the pH of the scrubber water may
be elevated by adding sodium hydroxide, or the scrubber may be fabricated
entirely of 316 SS.
SUMMARY
Despite the problems experienced, the performance of the electric furnace
at Little Falls is very encouraging. The problems encountered are mechanical in
nature and are not GAG process problems.
Information available on the operation of this unit .in- May, 1979 is sum-
marized in Table A-38. Preliminary indications are that the second reactivation
is improved from the first. Separating burning losses from attrition losses is
not possible, and the losses shown are aggregates. These losses can perhaps be
reduced as more experience is gathered in operating the system. The power
requirements are reassuring. Assuming a power cost of $0.03/kwh, the cost of
running the furnace would be slightly over 2
-------
TABLE A-38. COMPARISON SUMMARY FOR VIRGIN AND REACTIVATED CARBON
Operation parameter
Iodine number
Apparent density
Total organic carbon
Fluorscence
Total organic halogen
Carbon losses*
Power requiredt
Cvcle 1
Percent
change
-12
+ 3
-11
- 4
-10
7%
0.7 kw/lb
Cycle 2
Percent
change
-6
+ 1
Not yet available
Not yet available
Not yet available
4%
0.7 kw/lb (1.6 kw/kg)
*This is for the entire system—sum of losses from initial backwash through
transporting, reactivation, and quenching.
tThis is the electrical power for the furnace only.
BIBLIOGRAPHY
Love, 0. Jr., and Inhoffer, W. R., "Experience with Infrared Furnace for Reac-
tivating Granular Activated Carbon - A Progress Report", MERL, EPA, Cincinnati,
Ohio, May, 1979.
Love, 0., Jr., "Experience with Reactivation of Granular Activated Carbon",
AWWA Seminar Proceedings, June, 1979.
A visit to the Little Falls Water Treatment Plant in October, 1979.
229
-------
CASE HISTORY NO. 17
COLORADO SPRINGS, COLORADO
MUNICIPAL WASTEWATER TREATMENT
HISTORY AND INTRODUCTION
This wastewater treatment facility is located in Colorado Springs,
Colorado, and became operational in December, 1970. Two unique circumstances
make this facility somewhat different from other plants. The tertiary treatment
is preceded by bio-filtration, and the bio-filtration process was heavily over-
loaded during operation of the tertiary plant. The effluent from the heavily
loaded secondary plant presented some difficult mechanical and operational
problems in subsequent treatment.
The City of Colorado Springs constructed the Advanced Wastewater Treatment
Plant with an EPA Research and Demonstration Grant. The total cost of the pro-
gram was $1.0 million, with the EPA and the City of Colorado Springs each pay-
ing one half.
The tertiary plant consists of two "circuits", each involving different
processes. These circuits are called the Irrigation and the Industrial. The
completed plant was first put on-line in December, 1970. It operated continu-
ously for 4 years and has been operated only intermittently since 1975.
• Irrigation Circuit consists of four dual-media pressure filters oper-
ated in parallel. The media is made up of 3 ft .of 1.5 mm sand on the
bottom and 5 ft on 2.8 mm Anthrafilt on the top. Each filter has a
surface area of 113/ft^. At a design loading of 15 gpm/ft2, they
produce a total of 9 mgd.
• Industrial Circuit (see Figure A-38) water receives a much higher
degree of treatment than the Irrigation Circuit water. It has a
design capacity of 2.0 mgd. The Industrial Circuit utilizes lime
addition, recarbonation, filtration,, and carbon adsorption. The plant
also has lime recalcining and carbon reactivation facilities.
DESIGN DATA
Carbon Adsorption System Design and Operation
After filtration, the wastewater is passed through two down-flow carbon
adsorption towers in series. These towers are 20-ft in diameter with a total
side wall depth of 14 ft. There is a 10-ft deep bed of 8 x 30 mesh granular
activated carbon in each tower. This amounts to 94,000 Ib carbon/tower. At
design flow of 2 mgd, the loading rate is 4.25 gpm/ft^, giving a total resi-
dence time in the carbon beds of 34 min. Carbon loading is calculated on pounds
of COD removed per pound of carbon (usually between 0.50 and 0.60).
230
-------
The GAC used at Colorado Springs is WESTVACO WV-L 8 x 30. The Colorado
Springs plant determines need for carbon reactivation by the carbon loading in
terms of pounds of total COD removed per pound of carbon. Generally, the
adsorption service cycle is stopped when a value of 0.45 to 0.60 is attained.
The carbon tower is not removed from service when a certain COD, BOD, TOC, or
MBAS breakthrough occurs, but rather when the desired carbon loading is
reached. COD loadings in excess of 0.60 makes adequate regeneration more diffi-
cult, i.e., higher temperatures are required, slower carbon feed rates to the
furnace are necessary, and longer time for reactivation results.
The average COD loading for 12 regeneration cycles has been 0.05 Ib
removed/lb carbon. The two-stage system involves changing columns from the
"lead" to the "polish" position. It allows the carbon to adsorb 41 percent more
COD loading. Fifty-eight percent of the total COD removal is accomplished when
the tower is in the lead position, and the average lead position loading is
0.292 Ib/lb carbon. The average polish position loading is 0.207 Ib/lb carbon.
When the Colorado Springs tertiary plant operated under severe loading
conditions because of the overloaded secondary trickling filters, frequent
backwashing was necessary. Another reason for daily backwashing is because of
the production of hydrogen sulfide. If the towers are allowed to become anaero-
bic, sulfates in the water are reduced to sulfides—creating severe odor prob-
lems. Both carbon towers were backwashed daily, as were the dual-media fil-
ters. To backwash, the tower is taken off-line, then backwashed with an air
scour at 3.5 cfm/ft2 for 15 min. This procedure is repeated two times per back-
wash and results in an average backwash water use of 10 percent, which is
rather high. Ordinary operation should require backwashing only every other
day.
Carbon losses were excessive when the carbon towers were backwashed at 20
gpm/ft2 with a resulting bed expansion of 50 percent. The carbon towers have
14-ft side walls and 4-ft conical tops and were designed for a 10-ft depth of
carbon. Because 2 ft of the total sidewall depth is occupied by the under-
drains, a bed expansion of only 2 ft, or 20 percent, is permissible without
loss of carbon. The space provided for carbon expansion is adequate to retain
the carbon in the towers during backwashing.
The result was that much carbon was lost in the backwash system. Some-
times as much as 1 ft of carbon (9,400 Ib) was lost during a few backwashing
cycles. To avoid this, the designed 10-ft bed depth as reduced to 8 ft. This
allows a 50 percent bed expansion within the tower, and practically eliminates
carbon losses due to backwashing. Future plans include screening of tower out-
lets and allowance for 65 percent bed expansion. The capital cost of this
increase will very quickly be paid for by reduced carbon losses.
Complete monitoring of backwash functions should not be overlooked.
Accurate flow meters monitoring the rate of backwash water are of utmost impor-
tance. Colorado Springs uses air to scrub the carbon. Simultaneous air and
water backwashing can be done only under close control and at very low rates.
Using air and water together for backwashing is not recommended because carbon
231
-------
FROM BIO-
FILTRATION
SAMPLING
SOLIDS
CONTACT
CLARIFIE
SPENT LIME
HOLDING
TANK
TO PRIVATE
TO
BIO-FILTRATION
CENTRIFUGE
SPENT
CARBON
TANK
TURBIDITY
METER
SAMPLING
SAMPLING
MAKE-UP,,/QUENCH
CARBON \ TANK
INDUSTRIAL WATER
Figure A-38. Flow diagram of the industrial circuit, Colorado Springs, CO.
232
-------
loss will be severe and there is very little improvement over using the
alternating air-water cycle procedure.
Carbon Reactivation System Design and Operation
When the carbon in a tower has become exhausted, the carbon is conveyed by
a water eductor, to a spent carbon holding tank. The carbon is removed from the
holding tank through a rotary proportioning valve to eductors and then fed by a
dewatering screw to a 36-in. diameter, 6-hearth, multi-hearth furnace, which is
fired at about 1650°F. After passing through the carbon furnace, the carbon is
quenched and moved by water eductors back to the carbon tower. The furnace has
a throughput capacity of 75 Ib/hr, thus requiring between 50 and 60 days to
reactivate the contents of one carbon column.
Prom 1970 to 1974, there were 12 reactivation cycles at Colorado Springs.
Number One tower was reactivated three times, Number Two tower five times, and
Number Three tower four times. One of the carbon towers experienced consider-
ably higher reactivation losses than the other two. No explanation is available
as to why high losses would occur on carbon from the same batch, being split
into three adsorbers, and receiving exactly the same on-line and reactivation
conditions. If this adsorber is excluded, the average loss on reactivation
losses do not include losses due to backvashing while on-line, but do include
carbon losses from eduction and backwashing immediately after reactivation is
completed. To determine carbon losses, the tower is first backwashed and
drained before reactivation (the backwashing levels the carbon completely). A
measurement is then taken of the carbon bed surface. After reactivation is
completed and all reactivated carbon has been educted back to the adsorber, the
tower is again backwashed to remove fines, drained, and measured. The differ-
ence in this bed depth is the "loss on reactivation."
When a carbon tower is exhausted, it is taken off-line and the carbon is
moved by water eduction to a spent carbon holding tank. There are five, 4-in.
pipe draw-off points in the floor of the tower, and all of the draw-off lines
are connected to a main 4-in. header. This header connects to a 3-in. eduction
line that transports the carbon to the spent carbon tank. Each draw-off line
enters the tower and reaches different areas in the floor of the vessel. These
lines may be opened individually or all at once depending upon the water pres-
sure available to operate the eductor. It takes about 2 days to educt 94,000
pounds of carbon to the spent carbon tank.
When all of the carbon has been withdrawn from a tower, the reactivation
begins. The carbon slurry is withdrawn from the spent carbon tank through a
variable speed rot. ry displacement valve that measures the amount of carbon
being fed to the rurnace. This measured amount of carbon is moved by water
eduction to an inclined dewatering screw. The screw includes a water seal
baffle to prevent noxious odors from escaping the furnace and also to ensure
positive pressure on the furnance. Spent carbon or carbon feed is sampled at
this point in the process.
The dewatered carbon then enters the furnace at about 40 percent moisture.
Temperatures of the six hearths may vary from one regeneration cycle to the
233
-------
next, depending upon the COD load contained on the carbon, the desired iodine
number of apparent density, and the rate of carbon feed to the furnace. A typi-
cal temperature profile of the furnace hearths is: No. 1. (or top) - 700°F; No.
2 - 95G F; No. 3 -.1,250°F; No. 4 - 1,600°F; No; 5 -'"1,550°?;. and No.. 6 -
1,650°F. The furnace has two burners each on No. 4 and No. 6 hearth only.
Because the furnace has a throughput capacity of 75 Ib/hr, it takes between 50
and 60 days to reactivate the contents of one carbon column. After reactiva-
tion, the carbon is quenched and moved by water eduction back to a tower. Vir-
tually no plugging or erosion problems with this transportation system have
been experienced.
Colorado Springs controls the carbon reactivation operation by the use of
four simple laboratory analysis: iodine number, apparent density, sieve analy-
sis, and the ash content. Results of a series of these tests are summarized in
Table A-39. More complex analysis can be performed on activated carbon to
determine various properties, but they are of little value in actually control-
ling the reactivation. Also, as the reactivation progresses, continual'monitor-
ing of all furnace functions is important. Significant controls are the temper-
atures of the various' hearths, the carbon feed rate, and a positive pressure in
the furnace. These controls are important- in preventing "overburn" and high
losses on reactivation.
TABLE A-39. ANALYSIS OF REGENERATED. CARBON FROM ADSORBER NO. 2
Analysis/unit
Apparent dencity, g/cc
Iodine- number • .
Molasses decoloring index
Total ash , %
Effective size, mm
Mean particle diameter, mm
Uniformity coefficient
No. 4 hearth temperature, °F
No. 6 hearth temberature, °F
Loss on regeneration, %
Virgin
carbon
0
. 1040
7
6
1
1
.488
.6
.6
.97
.58
.71
-
-
"
Regenerated
Cycle 1
• 0
. 886
8
8
1
1586
1633
5
.469
.3
.2
-
.80
-
.3
Cvcle 2
0.
.824
8.
9.
1.
1.
1.
1595
1575
6.
503
5
0
02
63
62
7
carbon average
Cycle 3
0.
802
8.
8.
1.
1.
1.
1593
1658
6.
504
1
4
06
66
62
2
Cycle 4
0
785
7
1
1
1
1624
1634
3
.508
-
.9
.07
.68
.59
.2
Carbon make-up at Colorado Springs has been 83,400 Ib during 12 regener-
ation cycles, or 36.4 Ib/mil gal treated. The total carbon dose (original fill
plus make-up) is 160 Ib/mil gal treated. These carbon losses include both reac-
tivation and backwash losses. The reactivated carbon dosage only is 423 Ib/mil
gal of treated water. A total of 967,000 Ib of carbon has been reactivated in
the plant.
Operational Problems
Adsorption System—When backwashing a carbon adsorber there is always a
chance of debris entering the backwash supply and 'plugging the backwash
234
-------
underdrain system. Tremendous pressures are developed and applied over very
large areas during backwash. If the underdrainage system becomes partially
plugged, backwashing pressure could be great enough to physically damage the
vessel or piping. Precautions should be taken to prevent this damage from
occurring by: (1) selecting the proper underdrain system; (2) insuring a clean,
protected backwash supply; (3) installing pressure shut-off switches on back-
wash pumps to prevent over-pressuring underdrain system and piping; (4) instal-
ling pressure gauges above and below the underdrain system; (5) excluding air
from backwash water. These gauges will measure the pressure differential being
developed across the underdrain during backwashing. An automatic differential
shut-down switch may be used or an alarm may be sounded.
Reactivation Furnace—Some findings and suggestions resulting from the
experience at Colorado Springs regarding furnace operation are:
1. The installation of a fine screen basket in the overflow of the car-
bon dewatering screw proved to be a major step in preventing carbon
losses. This screen is checked hourly by the operator to ensure that
the screw speed is coordinated with the carbon feed rate from the
rotary valve.
2. Carbon feed rate is critical in maintaining low loss in regeneration.
If the furnace is overloaded, temperatures are difficult to maintain
and the burners are called upon for more heat. The burner gas con-
sumption and combustion air supply are increased and this condition
results in critical stack gas velocities. These velocities can be
great enough to entrain in the gas stream smaller carbon particles
that are subsequently burned in the afterburner.
3. Visual inspection of the furnace scrubber water can be employed to
predict high furnace efficiency or upsets in its operation; and
4. Positive pressure on the furnace should be maintained at all times.
The tertiary plant at Colorado Springs has been operating under severe
loading conditions. For this reason, the results of plant operation are not too
impressive from the standpoint of final effluent concentrations. When the high
concentrations of pollutants in the influent to the system are considered, the
tertiary process has been very satisfactory in removing gross amounts of con-
taminants. The tertiary process efficiency is strongly dependent upon the qual-
ity of secondary effluent being applied to it.
The carbon ads jrption system does smooth out and remove some high concen-
tration peaks, but to a limited extent; then effluent concentrations begin to
be proportionally higher as the feed concentration increases. Figures
A-39, A-^40, and A-41 illustrate seasonal fluctuations in feed concentrations
and the resulting effluent quality concentrations. The solids contact clarifier
absorbs most of the fluctuations, but trends can definitely be seen in the
carbon tower effluents. Table A-40 compares the tertiary water quality under
the heavily loaded conditions of 1972 and after the start of the new activated
sludge plant, which reduced the loading on the bio-filtration plant. Note that
235
-------
even though the carbon tower feed COD was reduced by 57 ppm or 50 percent, the
final carbon effluent still showed only slightly improved reductions of 20 ppm
or 39 percent. This information plus the curves illustrated in Figures 2, 3,
and 4 indicates that the carbon system has limited capabilities in removal
efficiencies under fluctuating feed concentrations. These data reveal that the
carbon adsorption towers will only remove a given percentage of what is
applied.
CAPITAL AND O&M COSTS
The unit cost of processed water at the Colorado Springs Tertiary Treat-
ment Plant was $343/mil gal for industrial quality water during 1972. A compar-
ison of these figures with the Advanced Water Treatment Plant at South Tahoe
for the same year is made below.
Colorado Springs
South Tahoe
Capital Costs
$ /mil/gal
$76.01
74.50
Operation & Maintenance Total
$/mil gal $
$267.19
142.40
$343.20
216.90
The capital costs compare quite closely, but there is an obvious differ-
ence in the respective operation and maintenance costs. The higher costs in the
Colorado Springs plant probably are due primarily to undersized equipment in
the calcining process, which results in constant overloading of the equipment
and more frequent maintenance and repair. Another major factor in the unit cost
discrepancy between the two plants is the cost of labor. The same number of
operators are required to operate the 2 mgd facility at Colorado Springs as are
needed to operate .the 7.5 mgd Lake Tahoe plant. Also, the effluent quality of
the bio-filtration plant may be a contributing factor in increased unit costs.
Improvement in secondary treatment could conceivably result in a savings of
approximately 5 to 15 percent in overall treatment costs.
As shown in Table A-40, costs were broken down into three categories:
capital, operation, and maintenance. For simplification, the expenses for oper-
ation and maintenance were combined after being calculated. A further breakdown
was made to determine the cost of individual systems in the tertiary plant.
These individual systems were the tertiary headworks, chemical precipitation,
lime calcining and feed, pH neutralization, filtration, carbon adsorption and
regeneration, and processed water storage. Operation and maintenance costs are
for 1972. The capital costs are at 1969 prices.
236
-------
NJ
U)
•-J
a
o
o
_j
Q
LU
340
320
300
280
260
240
220
200
160
140
120
100
80
60
40
20
0
fr
s!
5
^
>r
:\^=
NY
x*/
JAN-72 FEB MAR APR MAY JUN JUL AUO SEP OCT NOV DEC-72
Figure A-39. COD concentrations in various unit process effluents.
-------
M
LO
03
120
110
100
90
80
"i> 70
8 60
to
< 50
o
uj 40
a:
30
20
10
JAN-72 FEB MAR APR MAY JUN JUL AUG SEP OCT ,NOV DEC-72
Figure A-40. BOD concentrations in various unit process effluents.
-------
N)
Ul
VO
7.0
6.0
5.0
2
S 3.0
-I
D
o
£3 2.0
-------
TABLE A-40. GAC SYSTEM PERFORMANCE
Heavy loading period - 1972 Light loading period*-1973-74
Dual Lead Polish Dual Lead Polish
media carbon carbon media carbon carbon
Constituentt effluent effluent effluent effluent effluent effluent
BOD x
COD
TOC
TSS
Turbidity
P04
MBAS
Color
46
114
37
3.7
4.2
1.7
3.2
38
29
60
26
3.0
4.4
1.5
1.1
22
21
43
21
3.1
3.3
1.4
0.4
11
27
57
24
2.7
2.4
1.3
1.7
17
21
51
19
2.7
2.0
1.5
0.8
15
13
31
10
3.1
1.3
1.4
0.4
7.4
*0peration after new activated sludge system went on-line.
tmg/L except color.
BIBLIOGRAPHY
Grenwald, D., "Carbon Adsorption", presented at 3rd Annual International
Pollution Engineering Congress, September, 1974.
WESTVACO Chemical Division brochure, "Wastewater Treatment at Colorado Springs
with Activated Carbon", 1975.
Design, laboratory, and cost data supplied by the City of Colorado Springs.
A visit to the plant in August, 1979.
240
-------
TABLE A-41. 1972 UNIT COSTS AT THE COLORADO SPRINGS TERTIARY TREATMENT PLANT
System
Headworks
Reactor clarifier
Lime calcining
pH neutralization
Carbon towers
Carbon regeneration
Dual media filtration
Reservoir
Plant
flow,
mg
602.30
602.30
602.30
602.30
506.80
506.80
506.80
506.80
Capital
System
cost,
$
1,793.00
2,358.00
14,050.00
607 . 00
10,159.00
9,494.00
1,173.00
1,874.00
costs
Unit
cost,
$/mg
2.98
3.91
. 23.32
1.01
20.05
18.73
2.31
3.70
Operation
and maintenance
System
cost,
$
1,936.00
50,594.00
15,324.00
37,410.00
16,766.00
28,642.00
1,442.00
Unit
cost,
$/mg
3.21
84.00
25.44
62.11
33.08
56.51
2.84
Totals
Total
cost,
$
3,729.00
52,952.00
29,374.00
38,017.00
26,925.00
38,136.00
2,615.00
1,874.00
Unit
cost,
$/mg
6.19
87.91
48.76
63.12
53.13
75.24
5.15
3.70
TOTALS
41,508.00
76.01 152,114.00 267.19 193,622.00 343.20
-------
CASE HISTORY NO. 18
HERCULES, INC.
HATTIESBURG, MISSISSIPPI
INDUSTRIAL WASTEWATER TREATMENT
HISTORY AND INTRODT":TION
The Hattiesburg, Mississippi, plant of Hercules Incorporated produces a
wide variety of industrial chemical typical of the naval .stores industry
including organic chemicals made from raw materials extracted from pine trees
and stumps.
The plant contains 20 operating areas, each of which produces contaminated
process wastewater. Process wastewaters are separated from sanitary wastes and
noncontact cooling water and are discharged after treatment to a drainage ditch
leading to the Bowie River. The Bowie River is a tributary of the Pascagoula
River system. This river system has been the subject of several studies indi-
cating that water .quality problems, particularly low dissolved oxygen, exist at
least part of the time. Although poor • river water quality has not been
directly attributed to the Hercules, Inc. discharge, Hercules along with the
other major dischargers has initiated wastewater treatment practices to improve
river water quality.
Although preliminary at-source treatment .has been provided in most of .the
operating areas, treatment of the combined process wastes was required before
the wastewater could be discharged. An -impounding basin was 'Constructed in 1951
to remove floating material (oils) and settleable solids from the combined
process waste stream. Additional treatment was provided in 1972 when a dis-
solved air flotation (DAF) clarifier was installed. The impounding basin and
DAF clarifier make up the primary wastewater treatment system. Table A-42
shows the TOC concentration and pH through the primary treatment process.
TABLE A-42. PRIMARY TREATMENT TOC AND pH
Stream
Untreated process waste
Impounding basin effluent
DAF effluent
TOC, mg/L
1200
400-500
200-250
pH
3-4
3-4
6-7
When it became necessary to provide additional treatment for removal of
phenol and improved removal of TOC, a program to develop a secondary treatment
system was initiated.
Two secondary treatment methods were considered to have potential applica-
tion for use at Hercules, Inc.: biological treatment via the conventional
242
-------
activated sludge process and adsorption using GAC. Laboratory and pilot plant
studies were conducted to determine the most efficient and cost effective
method for treatment.
Results of activated carbon isotherm tests showed that 1 Ib of COD could
be adsorbed by each Ib of carbon when treating the Hercules wastes. Pilot plant
tests using GAC confirmed the isotherm results and showed that 45 min contact
time, based on empty bed volume, would be required for 75 to 85 percent COD
removal.
An activated sludge pilot plant was also operated on the Hercules waste to
provide design data. This tretment method would provide treatment results com-
parable to the GAC adsorption process. Consruction costs for both processes
were estimated in the same range. The adsorption process was chosen because of
its minimal land requirements and its ability to continue to provide treatment
in case of manufacturing process upsets.
The secondary treatment system, installed and started up in July, 1973,
included mixed media filtration, GAC adsorption, and carbon reactivation.
DESIGN OF GAC SYSTEM
An upflow (countercurrent) adsorption system using packed columns of GAC
was selected because of low estimated capital cost and anticipated efficiency
of carbon use. Because of suspended solids in the primary treatment system
effluent, filtration was required to prevent excessive headloss in the carbon
adsorbers. A high-rate downflow filter was designed to filter the wastewater so
that wastewater feed to the adsorption system would contain 5 mg/L or less
.suspended solids.
Table A-43 shows the performance criteria for the Hercules activated car-
bon adsorption system.
TABLE A-43. ADSORPTION SYSTEM PERFORMANCE CRITERIA
Parameter
Flow
COD
TOC
BOD
Suspended solids
Oil and grease
pH
Influent
700 mg/L
200 mg/L
250 mg/L
Effluent
3.25 mgd
125 mg/L
30 mg/L
50 mg/L
5 mg/L
5 mg/L
6-8
Basis of GAC Design and Operation
The purpose of GAC treatment at Hercules, Inc., is to remove organic mate-
rials including phenol from the process wastes. A schematic flow diagram for
243
-------
the GAC adsorption system is shown in Figure A-42. Partiallly treated waste
from the primary treatment unit flows to a sump from which it is pumped to the
mixed media filters and activated carbon columns. The activated carbon treat-
ment system design basis is as follows:
Activated Carbon Treatment System
Flow rate •
Number of carbon columns
Type
Size of columns
Granular carbon size
Carbon quantity (each column)
Column flow rate (each)
Surface loading rate
Contact time (empty bed)
Carbon exhaustion rate
3.35 mgd
Three in parallel
Upflow, moving bed
12-ft diameter x 40-ft high
12 x 40 mesh
135,000 Ib
780 gpm
6.6 gpm/ft
48 min
6.3 lb/1000 gal
(21,000 Ib/day)
The totally closed carbon columns are made of 316 stainless steel and were
designed with conical sections at the top and bottom. The vessels were designed
for countercurrent operation with filtered wastewater entering the bottom and
flowing upward to the top where the treated water exits the vessel via stain-
less steel conical screens. .The treated effluent then flows to a treated water
storage tank and discharges to the Bowie River. ' '
As wastewater passes upward through the column, the activated carbon in
the lower part of the column loses its ability to remove organics (becomes
spent). Periodically, a• small amount (or slug) of spent carbon is removed from
the bottom of the column and an equal amount of fresh carbon is added at the
top. In this way, the freshest carbon is in contact with the higher quality
water to achieve the maximum removal of organic materials.
The spent carbon sludge removed from each carbon column flows to the slug
measuring tank and then to the furnace feed tank. The slug measuring tank
volume (5 percent of the volume of a carbon column) ensures that equal and
exact volumes of spent carbon are transferred each time.
Spent carbon is moved from the furnance feed tank by a hydraulic eductor
to a dewatering screw conveyor where the carbon is drained (dewatered) in pre-
paration for thermal reactivation in a furnace. All carbon transfers are made
in a water slurry.
In the reactivation system as originally designed and constucted, the
dewatered carbon containing 50 percent moisture is discharged from the screw
conveyor to the top hearth (Hearth No. 1) of a multiple hearth furnace. Rabble
arms attached to a central rotating shaft move the carbon across the top hearth
by means of rabble teeth (plows) toward the center of the furnace where it
drops through a hole to Hearth No. 2. The rabbling process is repeated on each
hearth novinc the carbon outward and inward across successive hearths.
244
-------
CARBON CHARGE TANKS
N)
ib.
tn
INRUENT
BACKWASH TO
IMPOUNDING
BASIN
SUMP
V V V
AfTsFVsT
CARBON COLUMNS
ru
TO RIVER
TREATED WATER
STORAGE TANK
TO SLUG
MEASURING
TANK
Figure A-42. Flow diagram of granular activated carbon adsorption system, Hercules, Inc.
-------
The carbon reactivation system design basis is:
Granular Carbon Reactivation System
Furnace type ... •.;.,.. Multiple hearth
Furnace size 12'ft 0.0. x five hearth
Furnace capacity 33,600 Ib/day
Fuel Natural gas/fuel oil
Afterburner, type External
Carbon loss per cycle 5%
Makeup carbon 1050 Ib/day
Moisture and some easily volatilized organics are driven off in the top
hearths. As the carbon moves down through the furnace, the temperature
increases and reaches a maximum at the lowest hearth (Hearth No. 5). Burners
are located at Hearths 3 and 5 and temperatures range from 500° to 700°F at
Hearth No. 1 to 1600° to 1800°F at Hearth No. 5. .Steam is added on the fired
hearths to improve reactivation.
As the hot reactivated carbon exits the furnace, it. drops into a water
filled quench tank where it is cooled and returned to a slurry form for trans-
fer to the reactivated carbon storage tank. Reactivated carbon is transfered
from storage to each of the three carbon columns by a pneumatically operated
blowcase. A .storage tank is also provided for virgin makeup carbon. Furnace
off-gases (combustion products, moisture; and volatilized organics) are heated
in a gas-fired afterburner located adjacent to the multiple-hearth furnace.
Excess air is added to ensure complete combustion of volatile gases and odor.
The hot gases then pass through a cooler and scrubber before being discharged
to the atmosphere.' '
The reactivation and carbon handling system schematic diagram is shown in
Figure A-43.
Operational Problems
Operation of the secondary treatment units was interrupted many times,
particularly in the early operating years, by mechanical failure of many parts
of the system. Failure of the mixed media filter discharge laterals caused the
media to migrate to the carbon columns. This permitted high concentrations of
suspended solids and oil and grease to enter the upflow packed activated carbon
beds. High adsorber headloss resulted and carbon had to be removed and recti-
vated frequently in an effort to maintain effluent quality. When filter media
mixed with spent activated carbon entered the reactivation furnace, the media
melted and plugged portions of the furnace. The filter media also caused
severe erosion of the carbon transfer system. On many occasions, the adsorbers
experienced high headloss that could not be traced to unusually high suspended
solids in the influent. The high headloss reduced the flow rate through the
carbon, and effluent quality deteriorated. Odors became evident indicating
septicity due to anaerobic biological activity within the carbon bed. As septic
gases were generated, the headloss continued to increase. Headloss 'could be
reduced-and flow rate restored only by more frequent carbon column slugging and
increased reactivation rate.
246
-------
NJ
*>
•O
FURNACE
FEED
TANK
REACTIVATED
CARBON
TO COLUMNS
SPENT CARBON FROM
CARBON COLUMNS
SLUG MEASURING TANK
MAKE-UP
CARBON
STORAGE
REACTIVATED
CARBON
STORAGE
DEWATERING
SCREW
QUENCH
TANK
EXHAUST TO
ATMOSPHERE
AFTER
BURNER
OFF-GAS
SCRUBBER
GRANULAR
CARBON
REACTIVATION
FURNACE
Figure A-43. Carbon handling and reactivation system, Hercules, Inc.
-------
To improve operation of the GAC adsorption system and to control headloss,
the packed ''bed adsorbers were modified in 1978 to permit operation in an
expanded mode. The stainless steel vessels were extended at the top to permit
expansion of the carbon bed and the outlet septa were removed to prevent bind-
ing by solids. With proper attention to flow rate through each adsorber to
maintain expansion of the carbon bed, instances of high headloss have occurred
less frequently.
The multiple hearth furnace has also been a factor in the operation of the
adsorption system. The furnace was designed for 60 percent excess capacity to
permit periodic high reactivation rates. Although the additional capacity was
provided, the multiple hearth furnace experienced periodic mechanical failure
requiring shutdown for repairs. Any furnace outage exceeding 48 hr resulted in
poor performance of the GAC system because fresh carbon .was not available for
transfer to the adsorbers.
Major mechanical failures in the multiple hearth furnace have included
failure of hearths, rabble arms, and center shaft. Each of these failures
required repairs resulting in a minimum of 6 to 7 days downtime. During the
first 4 years of :operation, the furnace was inoperable 15 percent of the time
and continued increasing each year.
Reactivation System Modification
In 1977 . the multiple hearth furnace was replaced with a fluidized bed
carbon reactivation furnace. .This WESTVACO fluid bed reactivation furnace,
which contains no unsupported brick work or high temperature moving parts, was
.installed and started up in 1978. As in a multiple hearth furnace, the wet
spent carbon is 'dried at the-top of the furnace at 300° to 500°F and is reacti-
vated in the lower section at 1600° to 1800°F. Off-gases containing combustion
products and water vapor pass through the afterburner, gas cooler, and
scrubber, which are part of the original multiple hearth furnace system, before
being discharged to the atmosphere. Although the afterburner can be fired if
necessary for air pollution control, it has not been necessary to do so since
the fluidized bed furnace began operation.
In Table A-44, the fuel and steam use in the Hercules multiple hearth
furnace can be compared with that in a fluid bed activated carbon reactivation
furnace at a typical carbon processing rate of 21,000 Ib/day.
TABLE A-44. CARBON REACTIVATION FUEL AND STEAM CONSUMPTION
Reactivation Fuel Steam Total
rate, consumption, consumption, energy,
Furnace Ib/day Btu/lb Ib/lb Btu/lb Btu/lb
Fluid bed 21,000 2474 1.14 1323 3797
Multiple hearth 21,000 6000-7000 1.0 1161 7161-8161
248
-------
Carbon losses in the fluidized bed furnace have ranged from 1.0 to 2.5
percent compared to the design estimate of 2.1 percent.
1 . •'•">
SYSTEM PERFORMANCE i -, '
* i
The secondary wastewater treatment facility consisting of mixed media
filtration and GAC adsorption at Hercules, Inc. plant at Hattiesburg, Missis-
sippi, was started up in July, 1973. Due to in-plant modifications, the flow
rate to the secondary treatment unit upon start-up was 2.6 mgd as compared to
the design rate of 3.35 mgd. Dissolved organic quantities were, however, 35
percent greater than the design amount.
Results of wastewater treatment by GAC for removal of TOC are shown in
Table A-45.
TABLE A-45. ACTIVATED CARBON PERFORMANCE, AVERAGE TOC REMOVAL*
TOC
Year
1973-75
1977
1978
1979 (Jan - June)
Influent,
mg/L
300
285
281
302
Effluent,
mg/L
116
177
189
188
Removed
61
38
33
38
Flow
mgd
3.2
1.79
1.66
1.39
*The average effluent phenolics concentrations (calculated as phenol) for the
years 1977, 1978, and 1979 (first six months) were 0.13, 0.13, and 0.14 mg/L
respectively.
Exhaustion rates for the GAC averaged 9.5 lb/1000 gal wastewater treated.
Carbon loadings have ranged from 0.1 to 0.44 Ib TOC/lb carbon, and losses have
average 8 percent during periods of typical operation. Typical carbon losses
attributed to the reactivation furnace have been 5 percent.
CAPITAL AND OPERATION AND MAINTENANCE COSTS
Capital Cost
The capital cost for the GAC adsorption and reactivation systems was
$1,422,000 in 1973. That included the cost of vessels, piping, reactivation
furnace, controls and instrumentation, and initial fill of GAC. In 1979, the
three stainless ste^l carbon columns were extended at a cost of $200,000.
Operating and Maintenance Cost
The annual costs of operating the Hercules secondary wastewater treatment
facility are presented in Table A-46.
249
-------
TABLE A-46. AVERAGE MONTHLY OPERATING COST
Item
Operation
Maintenance
Utilities '
Carbon
Total
1973-75 Avg.
$ 4,831
7,619
4,828
12,908
$30,186
1977
$12,258
8,167
8,000
23,825
$52,250
1978
$19,208
13,617
9,258
22,375
$64,458
1979
$18,046
16,737
10,117
26,679
$71,579
During the first 2 years of operation, the average monthly use of
activated carbon was 33,273 Ib. Total cost of carbon treatment was $0.314/1000
gal.
As the cost of operation and maintenance has risen, the amount of waste-
water requiring treatment has been reduced. Operating costs based on the quan-
tity of wastewater treated each year are:
Year $/1000 gal
1973-75 0.314
1977 0.97
'1978 V.29
1979 . 1.47
None of the operating cost figures reflect or include amortization of
capital.
BIBLIOGRAPHY ..,:••,
Gardner, F.H., Jr., and Williamson, A.R. Naval Stores Wastewater Purification
and Reuse by Activated Carbon Treatment, EPA-600/2-76-227, USEPA, Cincinnati,
Ohio, 1976.
Naval Stores Wastewater Purification and Reuse by Activated Treatment - Update
presented at: AICHE 69th Annual Meeting, Chicago, Illinois, December 12, 1976.
Gardner, F.H., Jr. Activated Carbon Treatment of Wastewater from a Wood Naval
Stores Plant presented at Symposium on Advances in the Treatment of Industrial
Waste Discharges, Second Joint Conference, Chemical Institute of Canada and
American Chemial Society, Montreal, Canada, May 30, 1977.
Gardner, F.H., Jr. Carbon Adsorption Secondary Treatment of Chemical Plant
Wastewater, presented at: Carbon Adsorption Workshop, Manufacturing Chemists
Association for the USEPA, March 21, 1978.
Johnson, H.R., and Massey, M.L. Energy Savings by Fluid Bed Regeneration of
Granular Activated Carbon presented at: Energy Optimization of Water and
Wastewater Management for Municipal and Industrial Applications Conference, New
Orleans, Louisana, December 10-13, 1979.
R.A. Van 3eek, personal communication, September, 1979.
250
-------
CASE HISTORY NO. 19
INDUSTRIAL SUGAR, INC.
ST. LOUIS, MISSOURI
SUGAR REFINERY
HISTORY AND INTRODUCTION
The Industrial Sugar, Inc., sugar refinery at St. Louis, Missouri,
manufactures high purity liquid cane sugar for use in the soft drink industry
and in other industries that use liquid sugar. Raw granular cane sugar pur-
chased from cane sugar mills is used as the raw material. Before refining, the
liquid sugar normally contains 800 color units and is a dark amber color. The
sugar may, however, contain as high as 3200 color units depending on the source
of the raw granular sugar. The final product is a high purity water clean solu-
tion containing 77 percent solids (77 Brix).
Sugar refining operations at Industrial Sugar, Inc., began in 1958 with
the use of GAC for decoloring the liquid sugar. In 1965, a multiple hearth
furnace was installed to reactivate the granular carbon for reuse.
DESIGN OF GAC SYSTEM
Raw granular sugar is first placed in solution by addition of water and
steam. Refining steps include filtration for removal of particulate impurities,
adsorption by GAC for color removal, and ion exchange for removal of inorganic
impurities.
The original activated carbon adsorption system at Industrial Sugar, Inc.,
included six downflow columns each 9 ft in diameter and 9 ft high, containing
nearly 600 ft^ of 12 x 40 mesh coal based carbon. The columns were operated in
two parallel "trains" of three each in series. Piping was arranged so that any
of the columns in each train could be operated in any position from lead to
final polishing.
The refinery was later expanded by adding four 10-ft diameter downflow
'columns each containing 1200 ft^ of carbon. The larger vessels are operated in
series and also in series with the smaller columns. Figure A-44 presents a
schematic flow diagram of the adsorption system.
All of the columns are made of carbon steel and are lined with an abrasion
resistant food grade epoxy lining to prevent corrosion. Stainless steel screens
(Nevaclog) are mounted in the columns to support the carbon beds. The original
columns were built with standard dished heads and the screens were installed
horizontally. Carbon is removed through a pipe nozzle located in the vertical
column side immediately above the screen. The more recently installed columns
have slightly conical bottoms, and carbon is discharged through an opening in
the center of the cone.
251
-------
10
in
CAnnorj AosonoEns
Iff OIAMETEH VESSELS
CAHOON ADSORBERS
9- OlAMElEn VESSELS
SUGAR
LIQUID
SPENT CARBON
TO REACTIVATION
CLEAR
SUGAR
LIQUID
Figure A-44. Schematic flow diagram of granular activated carbon adsorption system,
Industrial Sugar, Inc.
-------
All carbon transfers from the columns are made by pressurizing the vessel
and blowing the carbon out in a water slurry. A small heel of carbon is left in
the column each time carbon is removed, but its effect on the operation has not
been considered significant.
The carbon columns are designed for downflow operation with the sugar
solution entering the top of the column and exiting from the bottom. It then
flows to the top of the next column in series and continues in that manner
through the adsorption system. .:
As color molecules are adsorbed on the GAC/ the ability of the carbon to
continue to remove color becomes limited and the carbon becomes exhausted or
spent in relation to the material being adsorbed. When the carbon in the lead
column in the series becomes spent or approaches a spent condition, it is
removed from service for reactivation. At that time, the second column in the
series becomes the lead column-.
When the column is taken out of service, the spent carbon is "sweetened
off" by rinsing with clear water to recover as much sugar as possible before
reactivation.
When GAC becomes spent, the activity can be restored thermally in a reac-
tivation furnace for reuse. Industrial Sugar, Inc., uses a multiple hearth fur-
nace for reactivation. After a column is taken out of service and "sweetening
off" is complete, the carbon is transferred in a water slurry to a spent carbon
storage tank where it is dewatered by allowing the water to drain down through
a screen. The dewatered carbon containing 39 to 40 percent moisture is trans-
ferred by a screw conveyor to the top of the reactivation furnace. The design
bases for the adsorption and reactivated systems are:
Adsorption System
Nuraber of adsorbers
Mode of operation
Depth of carbon (total)
Contact time
Type and size of carbon
Reactivation System
Furnace type
Furnace size
Furnace capacity (design)
Fuel
Six, 9-ft diam
Four, 10-ft diam
Downflow, series
90 ft
18 hr
Granular, 12 x 40 mesh
Multiple hearth
6-ft O.D. x six hearth
10,000 Ib/day
Natural gas
The dewatered spent carbon enters at the top of the multiple hearth fur-
nace where it is dried first at about 400CF. As it passes downward through the
furnace, the temperature increases to a maximum of 1600°F at the bottom hearth
(Hearth No. 6). Two burners, one located on Hearth No. 4 and one on Hearth
253
-------
No. 6, provide the necessary heat for reactivation. Steam may also be added at
the fired hearths if desired. .
Reactivation off-gases, drawn off the top of the furnace and through a
cyclone separator by an induced draft fan, are then discharged to the atmo-
sphere. Any particulates removed by the cyclone are returned to the furnace.
After the carbon is reactivated, it drops from the furnace into a water-
filled tank for cooling and a rubber lined centrifugal pump pumps it to a reac-
tivated carbon storage tank. When sufficient reactivated carbon is available,
it is transferred by an eductor in a water slurry to the empty column. That
column then becomes the final or polishing column in its series.
A schematic flow diagram of the carbon handling and reactivation system is
shown in Figure A-45.
SYSTEM OPERATION AND PERFORMANCE
Since startup of the multiple hearth reactivation furnace in 1965, the
adsorption and reactivation systems have operated, with a minimum of operator
attention. The adsorption system operates at a constant flow, which provides 18
hr of contact time. Whenever the flow of filtered liquid sugar to the adsorp-
tion system is interrupted or the desired flow rate cannot be maintained, the
adsorbers are automatically placed in a recycle mode so that all of the flow is
recycled within the adsorption system. Similarly> if downstream processes
demand less decolorized liquid sugar, the refinery operator can initiate
recycle. • .
The reactivation rate of spent carbon—normally 2 Ib carbon/hundred weight
of raw sugar—is controlled by the reactivation furnace capacity. Activated
carbon in an adsorber that is the most heavily loaded with organic color is
transferred to the reactivation system as soon as a dewatering tank is
available. Because of this method of operation, none of the carbon becomes
completely exhausted with respect to the influent color concentration.
The reactivation furnace, designed for a nominal reactivation capacity of
10,000 Ib/day is operated at a rate of 12,000 Ib/day. The spent carbon is
dewatered as completely as possible in a dewatering tank and is conveyed to the
furnace at a minimum moisture concentration (usually 39 percent). Furnace
operating temperatures at the bottom reactivation hearths are closely main-
tained at 1600°F. Although steam can be added at the fired hearths, no improve-
ment in reactivation efficiency or reactivated carbon quality has been noted at
Industrial Sugar, Inc. with its use, so it is not currently used in the reacti-
vation process. By operating the furnace at a constant rate and temperature,
all of the reactivated carbon in the system is of a consistent quality.
Apparent density of the GAC is the parameter used for controlling and
evaluating the reactivation process. Samples of reactivated carbon are
collected every 30 min and a composite sample of spent carbon is collected over
254
-------
Ul
U1
SP£NT CARBON
FROM ADSORPTION
SYSTEM
SPENT CARBON
DEWATERING BIN
6* DIA. X 6 HEARTH
REACTIVATION
FURNACE
"V
OUENCH^
REACTIVATED
CARBON STORAGE
REACTIVATED CARBON
TO ADSORPTION
SYSTEM
TANK
Figure A-45. Schematic flow diagram of carbon handling and reactivation system, Industrial Sugar,
Inc.
-------
a 24-hr period for the apparent density test. Reactivated carbon apparent
density is always 5 g/100 cc less than that of the spent carbon as shown by the
following typical values.
Spent carbon Reactivated carbon Makeup carbon
59-60 g/100 cc . 54-55 g/cc 49-50 g/cc
The furnace off-gas system consisting of an induced draft fan and cyclone
separator has never experienced any air quality violations.
The major repair item has been relining the reactivation furnace; this has
been done twice. The last relining occurred in 1978 for an average lining life
of 6.5 yr.
Although the furnace is shut down for 48 hr every weekend, there have been
no serious refractory or hearth failures. Because of the size of the furnace
and its construction, the hearths must be removed and replaced with each relin-
ing.
CAPITAL AND' OPERATION AND MAINTENANCE
Capital Cost
The reactivation system, as originally installed, included dewatering
bins, furnace feed screw conveyors, multiple hearth furnace, quench tank, and
carbon transfer pump. The total capital cost of the reactivation system,
installed in 1965, was $125,000.
Operation Cost
No record of costs associated with operating the adsorption or reactiva-
tion system has been maintained. Although the furnace uses natural gas for
fuel, the fuel use for reactivation is not known because it is not metered
separately. Industrial Sugar, Inc., plans to install a gas meter for the fur-
nace because of rising fuel prices.
Carbon losses are not monitored on a short-term basis but have been calcu-
lated to be 2.5 percent based on annual purchases of GAC for make-up.
The annual cost of maintenance and repair of the activated carbon adsorp-
tion and reactivation systems have not been documented but are considered
minimal.
256
-------
BIBLIOGRAPHY
Jeep, Robert D., Industrial Sugar, Inc., personal communication, January, 1980.
257
-------
CASE HISTORY NO. .20
AMERICAN WATER WORKS SERVICE COMPANY
HOPEWELL, VIRGINIA
MUNICIPAL WATER PURIFICATION
HISTORY AND INTRODUCTION
The American Water Works Service (AWWS) Company is the largest investor-
owned water utility in the United States, currently supplying about 5 million
people through 100 systems in 20 states from coast to coast. Their experience
with GAC began in 1960 at Hopewell, Virginia, where they installed GAC to
remove taste and odor. This was the first such installation of this kind in
•this country at' a drinking water plant. By 1975, 46 U.S. plants had partially
or completely converted to GAC. Only six used GAC after conventional filtra-
tion. About .-.one-third of these -46 plants are part of the AWWS system. Since
1975, many more plants have come to use GAC. AWWS Company recently completed
its largest installation at Chattanooga, Tennessee, (72 mgd) which makes 19
locations within the Company at the present time. Table A-47 shows the current
list. Hopewell, interestingly, is the only 'plant that uses GAC as a post con-
tactor; the GAC can be expected to remove the earthy odor caused by geosmin for
a period of only 30 to 60 days.
TABLE A-47. AMERICAN WATER WORKS SYSTEM LOCATIONS PRESENTLY USING GAC
Bed depth, in
Gac, Ib
Capacity, mgd
Illinois'
E. St. Louis
Granite City
peoria
Indiana
Kokomo
Muncie
Richmond
Terre Haute
18
18
30
20
24
24
30
412,200
174,000
165,000
162,000
90,500
100,900
127,000
35
13
30
11
8
8
6
Iowa
Davenport
Kentucky
Lexington
Ohio
Ashtabula
Marion
24
14
24
18
331,700
177,600
75,000
80,000
30
20
4
7
258
-------
TABIE A-47. Continued
Bed depth, in Gac, Ib Capacity, mod
Pennsylvania
N€:w Castle 30 190,800 8
Pittsburgh 30 1,947,400 86
Tennessee
Chattanooga 30 927,400 72
Virginia
Hopewell . 60 81,000 3
West: Virginia
Hvmtington 30 337,500 24
Madison 18 7,300 1
Princeton 18 • 24,300 3
Wtsston 18 14,600 1
Total 19 — 5,426,200 370
GAC SYSTEM
Figure A-46 shows the Hopewell plant flow diagram. The Appomattox River,
sourrce of supply for Hopewell, receives back water from the nutrient-rich James
Rivisr during high tide, thereby providing ideal conditions for growth of acti-
nomycetes. Because the James River water quality has improved over the .years, a
1-ytsar service life is now obtained from the GAC.
Regeneration capability was not installed at Hopewell. After the first few
exchanges of GAC, some filter sand was replaced with partially spent GAC, to
get a longer contact time. The carbon could still dechlorinate after it was
exhausted for taste and odor control. Since the plant had been dechlorinating
with sodium bisulfite, enough sand filters were converted to spent carbon fil-
ters to accomplish the same degree of dechlorination. When carbon was tested in
parallel with sand for suspended solids removal, the carbon did slightly better
than sand without affecting its dechlorinating properties (See Table A-48).
After 10 years, some carbon was still dechlorinating. Since then, many others
hava verified GAC's ability to remove suspended matter, and it is now fully
accepted as a filter medium.
Over the years, it was found that the iodine number (an indication of
available surface area) of GAC decreases with use, which means that the carbon
pores are becoming filled with organic compounds. When the iodine number gets
down to around 300 (virgin material ranges from 650 to 1100), the carbon is no
longer effective enough to ensure taste and odor control. While the iodine
number drops, the apparent density (the weight of carbon per unit volume)
increases, also indicating pickup of organic chemicals. An increase of about 20
259
-------
APPOMATTOX RIVER
3MGD
TRANSFER PUMP
THREE VIRGIN
CARBON FILTERS;
60 INCHES DEEP
FILTERED
WATER STORAGE
DISTRIVUTIVE
PUMPS
LOW SERVICE PUMP
2-STAGE HIGH ENERGY MIX
FIVE SETTLING BASINS
49 FILTERS .AT 0.67 MGD EACH
WITH. SPENT CARBON MEDIUM
FILTERED
WATER STORAGE
DISTRIBUTIVE
PUMPS
TO INDUSTRIAL SYSTEM
TO DOMESTIC SYSTEM
Figure A-46. Schematic of American Water Works Service System, Hopewell,
260
-------
TABLE A-48. IMPURITY REMOVAL BY ADSORPTION - FILTRATION CARBON BEDS*
AND BY SAND MEDIUM
Imouritv
Threshold odor number
Color
Manganese, ppm
Iron, ppm
Turbidity
Chlorine , ppm
Influent
35-140
4-14
0.066-0.15
0.2-0.37
0.45-1.4
1.4-2.8
Carbon bed
Effluent
0-4
0-2
0.008-0.017
0.006-0.025
0.07-0.15
0-0.25
Sand medium
Effluent
35-70
1-2
0.008-0.017
0.012-0.087
0.10-0.25
1.4-2.8
*24-in. deep at 2 gpm/ft .
tExtracted from: "Removal of Organic Contaminants by Granular Carbon Filtra-
tion" by Hager and Flentje, JAWWA, November, 1965.
§Odor samples over 60 day period; other data 150 day period.
percent (0.52 to 0.60 gm/cm3) seems to accompany the drop in the iodine number.
Figures £-47 and A-48 show these effects. A threshold odor number of 3 seems to
cor:rolete with the Iodine number and the apparent density as described above;
knowing this, they can be used as operating parameters.
SYSTEM OPERATION AND PERFORMANCE
Although the AWWS Co. installed GAC only in plants where severe taste or
odo:: problems had been experienced, it was recognized that other organics were
being removed, and the extent of such removals was investigated.
Work done by EPA and others concerning the adsorptive capacity of GAC for
trihalomethanes and other organics has indicated the life of the GAC is short.
At aopewell, data indicated a 78 percent THM removal after 9 months of opera-
tion ana the processing of 0.23 mil gal of filtered water/ft3 GAC (see Table A-
49). Therefore, it appears that the bed life for trihalomethane removal can
possibly be greater than has been previously reported, but cannot be accurately
predicted without specific evaluation on a case-by-case basis.
.-.1 though the GAC has been installed in AWWS Company plants basically for
tas'ie snd odor removal, there are indications that specific organic removal
doe:= . occur over a long period of time. This was -indicated at the Hopewell,
Virignie., plant in the removal of kepone: the average kepone in ppb, in the
riv'sr water was 0.0'7; before GAC, 0.011; and after GAC, 0.001. These data were
developed from an .r. erage of 60 samples over a 2-yr period.
Han 51in? GAC
During the past 20 years, AWWS Company has handled millions of pounds of
carbon. They have used hand shovels, pails, drums, eductors, eductors on hoists
with trolleys, cranes with buckets, and vacuum trucks—all of which work.
261
-------
000
900 .
800 .
700
600
z
8 50°
400
300 -
200
3
YEARS
LCU
re A-47. Typical iodine number vs. time relationship.
262
-------
•,0.60 -
YEARS
Figure A-4S. Typical apparent density vs. time relationship.
263
-------
TABLE A-49. TRIHALOMETHANES AT HOPEWELL, VA PLANT EFFLUENT
Chloroform
Dibromochloromethane
Dichlorobromomethane
Bromoform
TTHM (ppb)
% Removal TTHM
7 -mo
Pre GAC
77
22
17
20
136
old GAC
Post GAC
55
5
<1
2
62
54%
9-mo
Pre-Gac
261
1.9
191
<1
454
old GAC
Post-GAC
68
3.3
27
<1
98
78%
Undoubtedly, the Hopewell plant personnel have the most experience. To remove
GAC from the filters, the water eductor is used in combination with a periodic
backwash at low rate. By maneuvering the suction hose around the filter, vir-
tually all the carbon can be removed, even where gravel is supporting the car-
bon. When sand is present, it is more difficult to remove all the carbon with-
out getting sand mixed with it. But the system works very well, and with a
little equipment and a little experience, 800 cf can be removed from a 1 mgd
filter in less than 4 manhours.
Problems With GAC Use
Most of the AWWS Company experience has been with GAC from bituminous
material, but lately they have had good success with lignite. There were .some
problems with its use. Since GAC is lighter than other media, backwash rates
had to be lowered to prevent loss of material. In addition, the rate as a func-
tion of water temperature and particle size suddenly became- a crucial operating
parameter. Figure A-49 shows this relationship. Many times, more precise wash
flow control had to be installed to prevent loss of the relatively expensive
material. Recently, it was found that some carbons require higher wash rates,
which may demonstrate a change in carbon raw materials or in processing.
Nevertheless, with variation like this, filter washing of GAC requires closer
operational control than does any other medium. Even with good control, some
carbon gets washed away—perhaps 2-in./yr.
Another problem is the possible accumulation of mud in the media. This
does not happen in all cases, but when floe particles are particularly cohe-
sive, as in a highly polymer-dosed water, the floe does not wash away readily
at these low wash rates, even when wash times have been extended. Wash water
consumption does not always decrease with use of GAC (because of lower rates);
in actuality, wash water consumption may increase due to the extended times
required to clean the bed.
Because of the dechlorination capacity of GAC, bacteria may multiply in
carbon beds; the filtered water may contain bacteria many-fold higher than the
inlet water. Total plate count bacteria have been controlled to some extent by
backwashing with chlorine up to 50 ppm of combined residual. Fortunately, post
chlorination usually lowers the total plate count to less than 10 colonies/mL.
264
-------
z
o
Q_
X
LU
O
LU
CD
H
Z
LU
u
tt
in
0.
100
9.0
80
70
60
50
FILTRASORB 300-MEAN PARTICLE DIAMETER (MPD)=1.6
•AND
- FILTRASORB 400-MEAN PARTICLE DIAMETER (MPD)=1.0
30
20
10.-
8 12 16 20 24
SUPERFICIAL VELOCITY (GPM/FT2)
28
30
Figure A-49. Relationship between water temperature and particle
size of GAC during backwashing.
265
-------
CAPITAL AND OPERATION AND MAINTENANCE COSTS
Despite some problems, GAC is an excellent operating tool at a cost which
is not exorbitant, for taste and odor control. Table A-50 gives some recent
cost data, exclusive of capital cost of installation, for GAC operation after
several years operation. These costs do not reflect the decrease in other taste
and odor chemicals, any change in wash water, any change in chlorine consump-
tion, or any savings in sludge disposal. The overall cost reudction for just
taste and odor control at these locations has been 34 percent.
The AWWS Company plans and budget for calendar year 1980 at Hopewell are
to treat the anticipated flow of 1,460 mil gal with 85,000 Ib of virgin carbon
at a cost of $41,140. The GAC cost is 48.4
-------
CASE HISTORY NO. 21
AMERICAN WATER WORKS SERVICE COMPANY
DAVENPORT, IOWA
MUNICIPAL WATER PURIFICATION
HISTORY AND INTRODUCTION
The 106-yr-old Davenport filter plant treats an average of 18 to 20 mgd
(a iruiximum of 30 mgd) and uses a treatment sequence that includes prechlorina-
tion, coagulation with aluminum sulfate and cationic polymer, sedimentation,
filtration and GAC, postchlorination, fluoridation, and pH adjustment with lime
(Figure A-50).
Davenport's taste and odor, problems were chronic before 20 gravity sand
filters were converted to granular carbon in 1973. Algae blooms, upstream
industrial pollution, agricultural runoff, and accidental spills are the major
sources or organic contaminants that contribute to taste, odor, and turbidity
problems at Davenport. The plant must also contend with trihalomethanes.
The initial 1973 GAC installation, which eliminated the need for continued
PAC (except as backup in emergency situations), had a filter-bed depth of 6 in.
sand and 18 in. GAC. In 1975, GAC in the 20 plant filters was replaced for the
first: time and new carbon installed to the total depth of 24 in. Then, in 1978,
Davenport Water signed a 3-year contract with AWWS Company that specified peri-
odic changeout of used carbon and various analytical and monitoring services.
Filtration rate at Davenport is 2 gpm ft . Carbon contact time is 7.5 min.
Individual filters are backwashed every 24 hr.
SYSTEM OPERATION AND PERFORMANCE
The GAC installation at Davenport is aimed primarily at taste and odor
control. A secondary benefit is THM (trihalomethane) reduction.
It appears that the time span in which GAC will remove THM can be measured
in months or years at the Davenport water treatment facility. Research by EPA,
the American Water Works Association, and other groups on the adsorptive capa-
city of granular carbon for THM had indicated that removal of THM by GAC is
shortlived, measured in weeks. Therefore, it was of great interest to find the
resul.ts at Davenport illustrated in Table A-51. The THM reduction of 23 and 31
percent was achieved by an activated carbon that had been on-line for more than
2 years and had processed more than 1 mil gal settled water/ft^ GAC. To check
this result further, a pilot column was loaded with the spent carbon; tests
(Tabl.e «-52) again indicated a 25 percent reduction in THM.
267
-------
NJ
en
oo
-COAGULANT AID
,—CHLORINE DIOXIDE
POWDERED CARBON
•ALUM
LIME
FILTER
AID
*- DISTRIBUTION
Figure A-50. Flow diagram, Davenport Water Co., East River Station (AWWS Co.), Davenport, IA.
-------
The granular carbon in the 20 plant filters was changed in July and August
of 1978, and the percentage of THM reductions then ranged from a high of 60
percent to a low of 19 (Tables A-51 and A-53).
Use of granular carbon at Davenport also has helped reduce need for chlo-
rines. Specific types of organics that exhibit chlorine demand are effectively
removed by GAC. This reduces chlorine demand and allows for easier maintenance
of chlorine residuals. At Davenport, chlorine feed has been reduced by 18 per-
cent over a 5-year period.
TABLE A-51. TRIHALOMETHANE REMOVAL AT DAVENPORT WATER COMPANY
Dat<5
19713
5/8
5/8
5/8
7/20
7/20
7/20
7/20
*Calgon
tPotable
Location
Settled water
effluent basin 1
S'ettled water
effluent basin 2
Plant effluent
Settled water
basins 1 and 2
Filter 2 GAC*
Filter 9 PWSt
Plant effluent
FS 400
Water Service
CHC13
' 93
.
97
-71
104
73
42
42
TABLE A-52. PILOT COLUMN TEST
Date
1978
10/10
10/10
10/10
Location
Mississippi
River water
Settled water
effluent basins
1 and 2
Pilot column
effluent
CHC13
12
107
81
CHBrCl
4
4
3
5
2
1
2
Trihalomethanes , yg/L
3 CHBr2Cl CHBr3 TTHM
<1 <1 97
<1 <1 101
<1 2 76
<1 <1 109
<1 <1 75
<1 <1 43
<1 <1 44
EXHAUSTED GAC, DAVENPORT WATER
CHBrCl
<1
2
1
Trihalomethanes, vg/L
2 CHBr2Cl CHBr3 TTHM
<1 <1 12
<1 <1 109
<1 <1 82
Reduction
%
23
31
61
60
COMPANY
Reduction.
%
25
*Calgon FS 400
269
-------
TABLE A-53. TRIHALOMETHANE REDUCTION AFTER GAC CHANGE,
DAVENPORT WATER COMPANY
Trihalomethanes, g/L
Date
Filter
Location.
Reduction
Quenched
Potential
End
Nov.
1978
Dec.
1978
Jan.
1979
Influent
Effluent
Influent
Effluent
Influent
Effluent
152
120
93
97
71
62
87 .
15
56
8
26
7
239
135
149
105
97
69
44
30
29
Davenport also realized a significant savings in backwash water. With sand
used as a filter medium, 15 to 17 gpm/ft2 of finished water is needed for back-
washina. With GAC, however, the maximum backwashing requirement has been 14
gpm/ft". This translates into a 15 percent backwash-water reduction over a 5-
year period.
Another benefit of granular carbon as used at Davenport is that removal of
turbidity is improved. Raw water turbidities fluctuate from 0 NTU to highs of
150 to 200 NTU at other times. Normal levels are in the range of 18 to 20 NTU.
In the treatment process, most solids are settled out by sedimentation using
cationic polymers to help coagulate the solids. By the time the water reaches
the GAC filters, turbidity levels are generally less than 3 NTU; after passing
through the GAC, the polished water is at a.level of 0.32 NTU.
Because of a strong soil conservation program on the farmlands north of
Davenport, the amount of naturally occurring clays and other suspended solids
in the Mississippi River has decreased. This reduces turbidity in the plant's
raw-water intake—which is good. Lowered turbidity in the river, however,
allows more sunlight and more heat to penetrate the river and cause more heat-
and light-sensitive algal bloom. Thus, reduced turbidity can result in greater
taste and odor problems at the Iowa plant.
Removal of TOC at Davenport confirms the experience elsewhere—TOC remov-
als do not correlate with specific organic removals. This is illustrated by
Tables A-54 and A-55. Table A-55 also shows the decrease in TOC removal with
time, of GAC service.
The cost of using powdered and granular carbon were compared; over a 3-
year 'period. Davenport's management estimates it saved 24 percent by using GAC
instead of PAC ($5.18 mil gal of finished water produced). Also, GAC is thought
to do a better job of controlling taste and odor than does PAC. Unlike PAC
treatment, which must be constantly adjusted to compensate for changing water
conditions, granular carbon treatment requires no changes once it is in place.
Due to GAC's reserve adsorption capacity, emergency treatment responses to
sudden shifts in water quality are unnecessary.
270
-------
TABLE A-54. ANALYSIS OF ORGANICS AT THE DAVENPORT WATER COMPANY,
OCTOBER 1, 1976*
Compound
Alkane
Alkyl benzene
Benzothiazole
1 - Butyne
1 - (Chloroethyl) toluene
Chloroform
1 ,3-Dichloropropanone
4 , 4-Dimethyl- 1-pentene
Isoprene
Naphthalene
TOC
Concentration
Influent
0.3
0.2
4.2
0.1
0.1
120
0.2
0.2
1.7
2.0
g/L
Effluent
0.2
NDt
ND
ND
ND
12
ND
ND
0.6
ND
16000
•Carbon in service: 10 months; tWater processed: 0.391 mg/ft.
TABLE A-55. GRANULAR ACTIVATED CARBON REMOVAL OF TOC AT THE DAVENPORT
WATER COMPANY, NOVEMBER 1978—MAY 1979
Month of ooeration
4
5
6
7
8
9
1
Concentration
Influent
6560
7000
5475
3960
4250
4000
5900
g/L*
Effluent
5190
5100
4775
3750
4000
3550
2750
Removal , %
20.9
22.9
12.8
5.3
5.9
11.3
53.4
*Average of 2 to 4 samples per month.
Just before its 1973 GAC installation, Davenport experienced threshold
odor numbers (TON) of 8 to 10 in influent water. At the time, plant personnel
relied on powdered activated carbon (PAC) and sand filtration for control. A
significant taste and odor reduction followed the GAC installation, despite
increased organic loadings in 1975 when new chemical plants went on-stream
north of Davenport Table A-56).
Based on previous PAC operation at Davenport, management has concluded
that GAC gives more uniform results and is easier to handle.
271
-------
.TABLE A-56. TASTE AND ODOR REDUCTION FOLLOWING GAC USE
Influent, ' Effluent,
Year TON TON
1973-74 4.0 "Musty" <1
1975 (First GAC 3.5 "Musty" 1.5-2
changeout)
1;976 4.5 "Musty" <1
1977 5.0 "Musty"/"Grassy" 2
1:978 .(Second GAC
changeout) 5.5 "Musty" <1
Recently, Davenport plant operators developed a system to complete the
unloading of carbon from the truck (with hoppers and eductors) to the filter in
just 2 1/2 hr, and to load spent carbon .back into the truck in just 2 hr per
filter. The system makes use of a 1,400 gpm pump and a fire hydrant pressured
to 120 psig to pipe the carbon slurry.
Davenport modified a 3-in. sand eductor and found that it could take the
carbon'out of the filter without putting a man in. During carbon removal, the
eductor is kept flooded at all times by introducing backwash water. The back-
wash water keeps the carbon moving in a slurry. It flows downward and inward
toward the mouth of the eductor, which creates a vacuum effect. The appearance
is like that of a miniature volcano. This new operation requires . just two men
(a 50 percent reduction). •
Use of GAC at the Davenport Water Company has produced effluent TON con-
sistently less than one; a THM in the effluent of less than 100 ppb; and effec-
tive filtration of turbidities (to less than 1 NTU).
In taste and odor control, the primary function of the GAC installation,
the following results have been achieved;
1. A saving in backwash water and an increase in saleable water.
2. A 16.2 percent reduced overall cost for taste-and-odor control.
3. Clean, simple operation that requires no adjustments as the level of
raw water taste and odor changes.
The 1980 contract between the carbon supplier and the Davenport Water
Company calls for replacing the GAC in all 20 filter units—10 in the spring
and the other 10 in the fall. Based on the average daily water use of 18 mil
gal and a contract cost of $11,000 per month for 1980, the operating cost of
GAC treatment for taste-and-odor control plus side benefits is 2
-------
BIBLIOGRAPHY
Moser, R. H., and Blanck, C. A., "Experience with Granular Activated Carbon for
Taste and Odor Removal". Proceedings of AWWA Seminar on Controlling Organics in
Drinking Water, June, 1979.
Blanck, C. A., "GAC Value Exceeds Expectation". Water & Sewage Works, January,
1980.
Plant inspection; information provided by Clarence A. Blanck and Don Wulf of
AWWS CO.
275
-------
CASE HISTORY NO. 22 .
SPRECKLES SUGAR DIVISION
AMSTAR CORPORATION '
WOODLAND, CALIFORNIA
SUGAR REFINERY
HISTORY AND INTRODUCTION
The Woodland, California, sugar factory of the Spreckles Sugar'Division,
Amstar Corporation, one of five Spreckles sugar factories, began manufacturing
granulated sugar from sugar beets in 1937.
Harvesting of sugar beets near Woodland begins in July or August and con-
tinues until the rains begin in November or December. The harvest of overwin-
tered beets resumes in March after the rains and continues for 80 to 90 days.
The Woodland factory fall and spring campaigns correspond to the fall and
spring beet harvests. The factory, which has a processing capacity of 3,450
tons of beets/day, maintains beet storage capacity for only 5 to 7 days of
normal operation since beet quality starts to deteriorate soon after
harvesting.
When beets arrive at the factory, they are prepared for processing by
trash separation and washing. Sugar beet processing includes slicing, diffusion
(leaching of sugar from the pulp), purification,"evaporation, and crystalliz-
ation. Valuable by-products of, sugar, beet processing include molasses and .dried
pulp, 'which ..are used :in .animal .feed. . •. : .': • J .v ;' . ; -..'.-• /•'•.„ ' '..
The GAC adsorption process is used in beet sugar refining to remove both
color and other impurities that would otherwise cause the, formation of floe in
the sugar solution.
The Woodland factory was the first of the Spreckles Sugar plants to
install a GAC facility for removing color and' floe from beet sugar thick
juice. This facility uses the patented* Continuous Adsorption Process (CAP),
which was installed in 1958.
DESIGN OF GAC SYSTEMS
Adsorption System Design
The following is the design basis for the Woodland Factory adsorption
system.
Number of adsorbers Three
Size (each) 9-ft 6-in. diameter x 47-ft 6-in. high
Mode of operation Upflow expanded, parallel
Carbon bed depth 31 ft
Type and size of carbon Granular, 12 x 40 mesh
Quantity of carbon (each) 62,000.1b
'*U.S. Parents No. 2,954,305;2,769,297; and 2,969,298 assigned to the American
Sugar Refining Company. 274
-------
Each of the three adsorption columns is .cylindrical with an 8-ft high
inverted cone bottom section, which has a 60° slope. The cylindrical section of
each is epoxy lined and the cone bottoms are stainless steel coated. Tempera-
ture control in the column is provided by steam tracing, insulation, and
aluminum jacketing.
Factory thick juice (beet sugar juice after diffusion, purification, and
partial evaporation) containing color and floe is injected into each adsorber
via nozzles located near the bottom of the vertical cylindrical section and
flows upward to the top where the clear sugar liquor discharges over a weir.
Fresh activated carbon is continuously added at the open top of each adsorber
and exhausted carbon is continuously removed at the bottom; this results in a
continuous counter-current process. The process is carefully controlled so that
the carbon beds are maintained in a slightly expanded (not fluidized) condi-
tion. To maintain expansion, the flow rate, the sugar solids concentration, and
the temperatures are carefully controlled. The thick juice is kept at 90°C and
60-65 Brix.
Carbon fines and other particulate material that may be contained in the
decolorized sugar liquor are removed by filtration in plate-and-frame filters
after adsorption. A schematic flow diagram of the adsorption system is shown in
Figure A-51.
In the continuous adsorption process, the activated carbon in the lower
part of the adsorber is exposed to the highest concentration of color and floe
in the thick juice. As it adsorbs more of the impurities it moves downward by
gravity into the cone section and finally exits the adsorber at the apex of the
cone. The spent carbon then flows in a sugar liquor slurry to the top of a
sweetening-off column adjacent to the adsorbers. There steam and water are
added to remove and recover a sugar that accompanies the carbon during trans-
fer. One sweetening-off column, 24-ft high and 18-in. in diameter is used for
processing the spent carbon from the three adsorbers. Flow rate of the carbon
is controlled by an adjustable weir at the top of the spent carbon discharge
lines before entering the sweeting-off column.
Reactivetion System Design
The carbon was originally reactivated in a rotary kiln. In 1962, the kiln
was replaced with a multiple hearth furnace in an effort to reduce carbon
losses and improve reliability of the reactivation system.
The design basis for the present reactivation furnace is:
Furnace type Multiple hearth
Furnace size 8-ft 4-in. x six hearth
Furnace capacity (design) 15,000 Ib/day
Fuel Natural gas
275
-------
KJ
~J
CTi
GRANULAR
ACTIVATED CARBON
ADSORBERS
UNTREATED
SUGAR LIQUOR h
^
A
/ \
/ \
i «-
+
A
A
REACTIVATED CARDQ"
MULTI-HEARTH FURNACE
DEWATERING SCREW
CONVEYOR
TREATED SUGAR LIQUOR
—t TO FILTRATION
j'
iii
IV Iv. (V ill
i y 1 i '— i ii i 111
lf:
ll
"ii L-
SWEETENING-OFF
COLUMN
SPENT CARBON
TO REACTIVATION
Figure A-51. Schematic flow diagram of granular activated carbon adsorption system, Amstar Corp. .
-------
The desweetened spent carbon flows by gravity in a water slurry to the
reactivation system shown in Figure A-52.
Spent carbon slurry is discharged into the low end of an inclined
dewatering screw conveyor and is moved up the inclined conveyor toward the
reactivation furnace. As the carbon moves gradually upward, excess slurry water
is drained off. A sufficiently high water level is maintained at the lower end
of the dewatering screw to keep a water seal on the furnace to prevent excess
air from entering or combustion gases from escaping from the furnace.
The dewatered spent carbon, containing about 50 perent moisture by weight,
falls from the discharge end of the dewatering screw onto the upper furnace
hearth for drying. From there the spent carbon is rabbled through the furnace
where it increases in temperature as it falls to the lower hearths. Three gas
burners located on Hearth No. 4 and three on Hearth No. 6 provide • heat for
drying, pyrolizing, and reactivating the carbon. To prevent combustion of the
hot carbon, excess air is not allowed to enter the furnace. Steam is added at
the fired hearths to assist in the reactivation process.
Hot reactivated carbon drops from the bottom hearth (No. 6) into a water-
filled quench tank from which it is educted in a water slurry to a dewatering
screw conveyor (similar to the spent carbon dewatering screw), located in a
headhouse above the adsorbers. Excess slurry water is drained from the carbon
before it is returned to the adsorbers via belt conveyors.
Off gases from the reactivation furnace (800° to 900°F) pass through a
water scrubber before being discharged to the atmosphere.
OPERATION AND MAINTENANCE
The adsorption system is controlled to provide 20 to 30 min conntact time
in the adsorbers and to maintain the carbon bed in an expanded condition.
Although the adsorbers were designed to contain 31-ft deep carbon beds, it
was found by experience tht they could be operated as efficiently at shallower
depths thereby reducing the activated carbon inventory. The adsorbers are now
operated with 13 to 14 ft of carbon above the inlet nozzles. Residence time for
carbon in the adsorbers is 4 to 5 days and is varied as necessary to prevent
carryover of floe or color in the treated sugar liquor.
Control of the reactivation process is based on the apparent density in
g/cc and iodine number of the reactivated granular carbon. These tests are run
daily during the c; upaign. Average values for apparent density and iodine num-
ber for the 1979 fo.il campaign (July-December) are shown in Table A-57.
277
-------
to
^1
o>
SPENT CARBON
FROM SWEETENING-OFF
COLUMN
DEWATERING SCREW
REACTIVATED
CARBON TO '
ADSORBERS
TO ATMOSPHERE
OFF GAS
SCRUBBER
8'-4" x 6 HEARTH
CARBON REACTIVATION
FURNACE
QUENCH
TANK
Figure A-52. Schematic flow diagram of granular carbon reactivation system, Amstar Corp.
-------
TABLE A-57. ACTIVATED CARBON QUALITY
Granular carbon Apparent density Iodine no.
g/cc
Virgin 0.453 1091
Spent 0.521 250
Reactivated 0.43 516
The reactivation furnace is operated to produce reactivated carbon with an
apparent density essentially equal to that of virgin activated carbon. Operat-
ing temperatures on the lower three hearths are typically 1660°, 1600°, and
1720°F at hearths 4, 5, and 6, respectively. Steam is added at hearths No. 4
and 6 (the fired hearths) at a rate of 0.25 to 0.5 Ib steam/lb carbon.
Make-up carbon (virgin GAC) is added to the adsorbers at an average rate
of 720 Ib/day. Based on a furnace operating rate of 15,000 Ib/day, carbon
losses are 4.8 percent/cycle.
Scrubbing of the reactivation furnace off-gases before discharge to the
atmosphere has resulted in continuous compliance with air quality requirements.
Major maintenance of the GAC adsorption system has been limited to
replacement of the epoxy (Plasite) linings every 3-years.
Reactivation system maintenance has included replacement of the furnace
shell in 1975 and other periodic, minor repairs. Although no major internal
furnace damage has occurred such as collapse of a hearth, occasional repair and
replacement of refractories has been necessary.
Since the plant operates 24 hours/day, 7 days/week during the beet pro-
cessing campaigns and shuts down between them, most of the required maintenance
is performed during the shutdown periods.
CAPITAL AND OPERATION AND MAINTENANCE COSTS
Capital Cost
The GAC adsorption and reactivation system as originally designed and
installed, including a .used rotary kiln reactivation furnace, was built in 1958
at a cost of $238,000. Included in that cost was the initial carbon fill at
$53,000.
No capital cort figures are available for the existing multiple hearth
furnace built in 1?'_2 to replace the rotary kiln.
Operation Cost
Operation of the adosrption and reactivation systems at the Woodland plant
requires one operator on a continuous basis. Average use of fuel, steam, and
make-uo carbon are:
279
-------
Fuel /natural gas) 1785 Btu/lb carbon
Steam 0.25-0.5 Ib/lb carbon
Make-up carbon 720 Ib/day
(4.8% loss)
Actual costs associated with the operation and maintenance of these
systems are considered proprietary by Spreckles and are not available for pub-
lication.
BIBLIOGRAPHY
Baker, T.W., Haskell, D.R., and Resch, I.A. "Factory Treatment of Process
Liquors Using Activated Carbon in a Continuous Adsorption Process to Improve
Sugar Quality", presented at California Sugar Beet Technologies, Monterey,
California, February 3, 1961.
Cooley, J.E., "Further Experience with the Continuous Process Using Pittsburgh
CAL Carbon". Presented at 12th General Meeting, American Society of Sugar Beet
Technologies, Denver, Colorado, February 5-8, '1962.
"Beet Sugar Technology", 2nd Ed., Beet Sugar Development Foundation, Ft.
Collins, Colorado, 1971.
Lee Lofton, Spreckles Sugar Division, Amstar Corporation, personal communica-
tion, February, 1980.
Michael Angelbeck, Spreckles Sugar Division, Amstar Corporation, personal com-
munication, February, 1980.
280
------- |