EPA-660/2-75-004
APRIL 1975
Environmental Protection Technology Series
Activated Carbon Treatment of
Unbleached Kraft Effluent for Reuse
I
I
S32Z
LU
CD
National Environmental Research Center
Office of Research and Development
U.S. Environmental Protection Agency
Corvaliis, Oregon 97330
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development,
U.S. Environmental Protection Agency, have been grouped into
five series. These five broad categories were established to
facilitate further development and application of environmental
technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in
related fields. The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY STUDIES series. This series describes research
performed to develop and demonstrate instrumentation, equipment
and methodology to repair or prevent environmental degradation from
point and non-point sources of pollution. This work provides the
new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
EPA REVIEW NOTICE
This report has been reviewed by the Office of Research and
Development, EPA, and approved for publication. Approval does
not signify that the contents necessarily reflect the views and
policies of the Environmental Protection Agency, nor does mention
of trade names or commercial products constitute endorsement or
recommendation for use.
-------
EPA-660/2-75-004
APRIL 1975
ACTIVATED CARBON TREATMENT OF
UNBLEACHED KRAFT EFFLUENT FOR REUSE
by
E. W. Lang
W. G. Timpe
R. L. Miller
Grant No. 12040 EJU
Program Element 1BB037
ROAP/TASK 21 AZX/027
Project Officers
John S. Ruppersberger
George R. Webster
Pacific Northwest Environmental Research Laboratory
National Environmental Research Center
Corvallis, Oregon 97330
NATIONAL ENVIRONMENTAL RESEARCH CENTER
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CORVALLIS. OREGON 97330
For sale by the Superintendent of Documents, U.S. Government
Printing Office, Washington, D.C. 20402
-------
ABSTRACT
A four-year pilot plant program was carried out to investigate the tech-
nical and economic feasibility of treating unbleached kraft pulp and paper
mill effluent for reuse. Preliminary laboratory studies and cost estimates
indicated that the following treatment sequences should be investigated
in the pilot plant: 1) primary clarification, carbon adsorption; 2) lime
treatment, carbon adsorption; 3) primary clarification, bio-oxidation,
carbon adsorption.
Water of reusable quality can be provided from unbleached kraft effluent
by several combinations of treatment utilizing activated carbon. Un-
bleached pulping effluents typically contain about 1000 color units,
250 mg/1 TOC, and 250 mg/1 BOD. Reusable water quality as defined in
this study is 100 color units and 100 mg/1 TOC. The most economical
treatment is the microlime-carbon process that utilizes low dosages of
lime and clarification followed by carbon adsorption in down-flow granular
carbon beds. Capital cost for treatment by this process of 9.6 mgd of
unbleached kraft effluent from an 800-ton-per-day mill was estimated to
be approximately $6.7 millions. Operating costs, inclusive of capital
depreciation, were estimated to be $0.30 per 1000 gal and $3.58 per pulp-
ton, including credit for the reused water. Carbon adsorption in con-
tinuous counter-current stirred contactors was found to have promise of
lower operating cost and substantially lower capital costs as compared
to adsorption in fixed beds.
This report was submitted in fulfillment of Grant #12040 EJU by St.
Regis Paper Company under the partial sponsorship of the Environmental
Protection Agency. Work was completed in December, 1973.
ii
-------
CONTENTS
Abstract it
List of Figures viii
List of Tables x
Acknowledgments xiii
SECTIONS
I CONCLUSIONS ' 1
II RECOMMENDATIONS 3
III INTRODUCTION 4
IV LABORATORY STUDIES 6
Introduction 6
Selection of Carbons for Evaluation 9
Results of Carbon and Combined Treatments 9
Final Selection of Commercial Carbon for Pilot Plant 19
Additional Studies 21
V SELECTION OF EFFLUENT TREATMENT SYSTEMS TO BE EVALUATED
IN THE PILOT PLANT 28
Introduction 28
Selection of Effluents to be Treated 29
Preliminary Selection of Treatment Sequences and Carbon
Adsorption Systems 30
Order of Magnitude Cost Estimates 38
Final Selection of Treatment Sequences and Carbon
Adsorption Systems 46
VI PILOT PLANT DESCRIPTION AND OPERATION 48
Objectives 48
Description of Pilot Plant 48
Operating Procedures 57
iii
-------
CONTENTS (Continued)
VI PILOT PLANT DESCRIPTION AND OPERATION (Continued) Page
Sampling and Analytical Procedures 61
Manpower Requirements 63
VII OPERATING RESULTS:
PRIMARY CLARIFICATION AND BIO-OXIDATION 64
Bio-Oxidation 64
Primary Clarification 66
VIII OPERATING RESULTS:
LIME TREATMENT AND CARBONATION 67
Objectives and Description of Operations .67
Removal of Color and TOC in Lime Treater 68
Sludge from Lime Treater 74
Clarification in the Lime Treater 77
Operation of Carbonator 78
IX OPERATING RESULTS:
BIO-CARBON COLUMNS 80
Objectives and Description of Operation 80
Removal of Color and TOC by Carbon Adsorption 81
Dosage of Carbon Required 86
Loadings on Carbon 88
Rates of Removal of Impurities by Carbon 89
Effect of Flow Velocity on Rates of Adsorption 94
Biological Activity in Carbon Columns 94
Backwashing of Carbon Columns 94
Correlation of Variables nc
X OPERATING RESULTS:
PRIMARY-CARBON COLUMNS gg
Objectives and Description of Operation nr
Primary Clarification and Filtration ofi
Removals of Color and TOC by Carbon Adsorption 97
-------
CONTENTS (Continued)
X OPERATING RESULTS: Page
PRIMARY-CARBON COLUMNS (Continued)
Dosage of Carbon Required IQQ
Rates of Adsorption 102
Correlation of Variables 105
Pressure Drops and Backwashing of Columns 106
Displacement of TOC and Color from Columns During
Backwashing 107
Non-Adsorptive Removal Mechanisms in Carbon Columns 108
XI OPERATING RESULTS:
LIME-CARBON COLUMNS 111
Objectives and Description of Operation 111
Removals of Color and TOC; Effect of Soluble Calcium
Concentration 111
Dosage of Carbon Required 118
Rates of Removal of Color and TOC 122
Effect of Operating Variables on Removal of Color and TOC 122
Cumulative Removals by Carbon 124
Removal of Low Molecular Weight Compounds 127
Removal of BOD and Turbidity 127
Removal of Metal Ions 128
Pressure Drop and Backwashing of Columns 129
XII OPERATING RESULTS:
LIME-FACET ADSORPTION 130
Objectives and Description of Operation 130
Results from FACET Operation with Lime-Treated Water 130
Removals of Color and TOC 132
Adsorption Performance by Lime-FACET Compared to That by
Lime-Carbon Columns 133
Attrition of Carbon in FACET Operation 135
v
-------
CONTENTS (Continued)
Page
XIII SUPPORTING LABORATORY STUDIES 137
Color Increase in Presence of Activated Carbon 137
Adsorption Isotherms 141
Dynamic vs. Equilibrium Concentrations in Pilot Plant
Columns 145
Foaming Tendency of Treated Effluent 145
Effect of Operating Conditions on Adsorption of Color and
TOC on Carbon 146
XIV TREATMENT PLANT DESIGN STUDIES AND COST ESTIMATES 147
Overall Design Criteria 147
Design Methods and Assumptions 150
Cost Estimates 156
Results of Cost Estimates 157
Conclusions From Cost Estimates 163
XV REFERENCES 166
XVI PUBLICATIONS AND PATENTS 170
XVII GLOSSARY 172
XVIII APPENDICES 173
A. Water Quality Standards for Specific Uses 175
B. Procedures for Isotherms, Laboratory Lime Treatment,
Laboratory Bio-Oxidation and Analyses 179
C. Detailed Information on Pilot Plant Equipment 183
D. Printout of Daily Summaries of Conditions and Results
for Bio-Carbon Sequence 185
E. Printout of Daily Summaries of Conditions and Results
for Primary-Carbon Sequence 186
F. Printout of Daily Summaries of Conditions and Results
for Lime Treatment 187
G. Printout of Daily Summaries of Conditions and Results
for Lime-Carbon Sequence 188
vi
-------
CONTENTS (Continued)
XVIII APPENDICES (Continued) Page
H. Generalized Procedure for Designing Plant for
Treating Kraft Pulp Mill Effluent by Adsorption
in Granular Activated Carbon Columns 189
I. Relationships Useful in Design of Downflow Carbon
Adsorption Columns 193
J. Estimate No. 1 - Microlime - Carbon Treatment of
9.6 MGD of Effluent from a New 800 Ton/Day
Unbleached Kraft Pulp and Paper Mill 194
vii
-------
FIGURES
No.
f"
1 Carbon Adsorption Isotherms for Color Removal from Total
Mill Effluent 11
2 Carbon Adsorption Isotherms for Total Organic Carbon Removal
from Total Mill Effluent ' 12
3 Comparison of Cumulative Loadings on Darco and WV-M Carbons
in Simultaneous 2-Inch Column Tests 20
4 Carbon Adsorption Isotherms for Specific Compounds in
Turpentine Underflow 26
5 Effluent Treatment Pilot Plant 50
6 FACET System 53
7 Exterior of Pilot Plant Showing Lime Treater, Carbonation
Tank, Storage Tank, and Basin 55
8 Interior of Pilot Plant with Carbon Columns, FACET Tank
(foreground) and Duomedia Filter (right foreground) 56
»
9 Influence of Calcium in Lime-Treated Water on Percent of
Color Removed and on Color Remaining 72
10 Soluble Calcium in Lime-Treated Water Versus Increased
Conductivity due to Lime Treatment 73
11 Concentrations of Color and TOC During Bio-Oxidation-
Carbon Adsorption, Feb. 1 through Mar. 11, 1972 82
12 Color and TOC Remaining Versus Length of Carbon Bed During
Bio-Carbon Sequence 87
13 Cumulative Removals of TOC on Carbon, Feb. 1 to Mar. 11, 1972 90
14 Rates of Removal of Color and TOC by Carbon Adsorption During
Bio-Carbon Sequence, Feb. 4 to Mar. 11, 1972 92
15 Rates of Removal of Color and TOC by Carbon Columns as a
Function of Average Concentration During Bio-Carbon Operation,
Feb. 4 to Mar. 11, 1972 93
viii
-------
FIGURES (Continued)
No. Page
16 Color and TOG Remaining Versus Length of Bed During Primary-
Carbon Sequence 101
17 Rate of Removal of Color and TOC by Carbon Adsorption
During Primary-Carbon Operation at 10 gpm 103
18 Effect of Calcium Concentration on Removal of Color and TOC
by Carbon Adsorption (Averages from Operating Periods) 112
19 Concentrations of Color and TOC Remaining Versus Length of
Carbon Bed 117
20 Remaining Color in Water from Columns During Lime-Carbon
Sequence 119
21 Remaining TOC in Water from Columns During Lime-Carbon
Sequence 120
22 Color and TOC Versus Stage of Lime-Carbon Treatment for
Medium Calcium Run 121
23 Cumulative Removal of Color Versus Time for Lime-Carbon
Operation 125
24 Cumulative Removal of TOC Versus Time for Lime-Carbon
Operation 126
25 Changes in Color of Dilute Black Liquor with Time of Stirring
with Activated Carbon 139
26 Average Color Isotherms After Various Treatments During Pilot
Plant Operations 142
27 Average TOC Isotherms After Various Treatments During Pilot
Plant Operations 143
28 Flow Diagrams and Water Quality for Cost Estimates 1-11 149
29 Process Flow Diagram for FWPCA-Kellogg Study (17) 153
30 Flow Diagram for Microlime-Carbon Treatment of Unbleached
Kraft Effluent for Reuse
ix
-------
TABLES
No. ISSS
1 Laboratory Treatments of Effluents 8
2 Carbons Selected for Adsorption Studies 9
3 Comparisons of Carbons for Removal of Color and TOC from Two
Mill Effluents 14
4 Comparison of Carbons for Removal of Color and TOC from Paper
Machine White Water from Mill B, at pH 7.7 15
5 Effect of Treatment Sequences on Removal of Color from Mill
Effluents 16
6 Effects of Treatment Sequences on Removal of TOC from Mill
Effluents 17
7 Lime Treatment of Unbleached Kraft Effluent 22
8 Lime-Carbon Treatment of Unbleached Kraft Effluent 22
9 Color Change of Kraft Effluent 24
10 Ranges of Desired Properties of Process Water for Unbleached
Kraft Manufacture 32
11 Effect of Treatment Sequences on Removal of Color from Total
Mill Effluents 34
12 Effects of Treatment Sequences on Removal of TOC from Total
Mill Effluents 35
13 Order-of-Magnitude Cost Estimates for Various Treatments of
Total Mill Effluent from Mill A Effluent 40
14 Order-of-Magnitude Cost Estimates for Various Treatments of
Total Mill Effluents from Mills A and B 41
15 Summary of Conditions of Treatment and Estimated Capital and
Operating Costs for Various Treatments of Total Mill Effluent
from Mills A and B 42
16 Properties of Water Routinely Measured 62
-------
TABLES (Continued)
No. Page
17 Bio-Oxidation and Clarification for Period of 2/4/72 to
3/11/72 at Flow Rate of 20 gpm 65
18 Average Results from Lime Treatment 70
19 Relation of Specific Gravity and Solids Concentration of Lime
Treater Sludge 74
20 Distribution of Calcium in Lime Treatment 76
21 Operating Summary of Biological Oxidation - Carbon Adsorption
Sequence at 15 gpm, Low Carbon Dosage 84
22 Average Results for Bio-Carbon Sequence for Period of
February 4, 1972 to March 11, 1972, Low Carbon Dosage 85
23 Results from Bio-Carbon Adsorption with Differing Carbon
Dosage Rates 88
24 Summary of Primary Clarification-Carbon Adsorption Sequence 98
25 Average Results for Primary-Carbon Sequence for Period of
April 4 through May 20, 1972 at Flow Rate of 10 gpm 99
26 Displacement of Color and TOG during Backwashing of Columns 108
27 Comparison of Removal Rates of Exhausted Carbon with Normal
Operation 109
28 Lime-Carbon Sequence of Treatment: Removals of Color and TOC 114
29 Summary of Carbon Adsorption Following Lime Treatment:
Removals of BOD and Turbidity 115
30 Carbon Adsorption Treatment of Lime-Treated Effluent Flow=10 gpm 116
31 Organic Compounds in Lime-Carbon Operations 127
32 Metal Ion Concentrations in Lime-Carbon Sequence 129
33 Results from FACET Operation with Lime Treated Water at 10 gpm 131
34 Comparison of Lime-FACET and Lime-Carbon Column Operation at
a Water Flow Rate of 10 gpm 134
xi
-------
TABLES (Continued)
No.
35 Average Equilibrium Loadings on Carbon from Isotherms
Prepared During Pilot Plant Operation 141
36 Bases Used for Design and Cost Studies 148
37 Average Composition of Water To and From Treatment 148
38 Comparison of Costs for Treating Unbleached Kraft Mill
Effluent in New and Older Mills by Various Treatment
Sequences with Carbon at $0.27/lb 158
39 Comparison of Costs for Microlime and Minilime Processes 162
40 Effect of Carbon Price on Treatment Costs 164
41 Potential Water Quality Standards for Specific Water Uses 176
xii
-------
ACKNOWLEDGEMENTS
The authors gratefully acknowledge the invaluable contributions to this
program made by the many people in operations, engineering and research
and development who made their time freely available during this program.
The list of these contributors is too long to mention separately. It
includes people from operating companies including St. Regis Paper Co.,
engineering firms, manufacturers of activated carbon, academic and feder-
ally sponsored research, development and demonstration facilities.
The National Council of the Paper Industry for Air and Stream Improvement
(NCASI), particularly through H. F. Berger, Southern Regional Engineer,
lent its assistance throughout this program.
Dr. C. W. Dence and Dr. P. Luner of Syracuse University and Dr. H. Dugal
of the Institute of Paper Chemistry assisted in developing an understanding
of problems connected with effluent color.
Mr. E. L. Spruill, Mr. C. L. Davis, and Mr. M. Gould lent valuable assis-
tance in the area of lime treatment.
The support of the program by the Environmental Protection Agency, and
the help provided by Mr. W. J. Lacy, Mr. G- R. Webster and Mr. J. Gallup,
is gratefully acknowledged.
Major St. Regis contributors to this study include Mr. W. W. Spangler,
Mr. D. Barnes, Mr. M. R. Garrett (laboratory studies and pilot plant
operations), Mr. J. N. Rockwell (analytical), Mr. J. L. Gillespie (pilot
plant construction), Mr. J. Moore (computer data analysis), and Mrs. B.
Halfacre (secretarial).
xiii
-------
SECTION I
CONCLUSIONS
1. Unbleached kraft mill effluent can be treated by sequences of lime
treatment or primary clarification plus bio-oxidation or primary
clarification followed by carbon adsorption in granular carbon
columns to provide water suitable for general reuse in the mill.
2. Cost estimates indicate that a microlime-carbon sequence would be
the least costly for treating effluent containing 1000 APHA-NCASI
color units (CU) and 250 mg/1 total organic carbon (TOG) to provide
reusable water containing 100 CU and 100 mg/1 TOC. For treating
9.6 mgd of effluent from a 800 ton/day mill, the capital cost was
estimated to be $6-75 million and the total operating cost inclusive
of capital depreciation was estimated to be $0.30 per 1000 gal, or
$3.58 per pulp ton, with a credit of $0.06 per 1000 gal for the re-
used water.
3. Use of lower-cost carbon produced on-site at 10$/lb vs. a purchased
cost of 27
-------
of 3.8 hrs. These requirements caused high capital costs and the
resulting total operating costs were $0.58/kgal or $10.50/pulp ton,
which were about double those for microlime-carbon.
9. The primary-carbon sequence provides reusable water quality at a
dosage of 23.4 Ib activated carbon per 1000 gal effluent treated and
a contact time of 5.3 hrs. These requirements caused high capital
and operating costs and made this sequence much more costly than the
lime-carbon sequence.
10. Granular carbon rather than powdered activated carbon is the economical
choice because high dosage requirements (including the microlime-carbon
sequence) necessitate regenerative use of activated carbon even if
available at very low cost from mill site production. Regeneration
systems for powdered carbon, though under development, have not as yet
been applied on a large scale.
11. Fixed-bed carbon adsorption of raw clarified effluent or of water
treated by bio-oxidation or lime gave no operating problems. Back-
washing of the beds every one or two days prevents buildup of pressure
drop.
12. A novel continuous countercurrent process (FACET) for carbon adsorption
using stirred tanks of carbon slurries was shown to have considerable
promise for reducing capital and operating costs, and its development
is being continued.
»,
13. Non-adsorptive mechanisms are responsible for as much as 30% of color
removal and 20% of TOG removal, particularly in the primary-carbon
sequence.
14. Anaerobic digestion in the equalization basin resulted in substantial
increases in non-adsorbable TOG in the form of low molecular weight
organic compounds.
15. Interactions between carbon and effluent lead to not-fully-understood
color and suspended solids phenomena, particularly in the FACET system.
16. The color of lime treated effluent is subject to increases of approx-
imately 50% upon aging or bio-oxidation.
17. Carbon adsorption removed almost all of the foaming tendency of
effluents.
-------
SECTION II
RECOMMENDATIONS
It is recommended that the concept of treatment and reuse rather than
discharge be given serious consideration by operating mills and by those
planning to install new production and effluent treatment facilities.
This concept appears to be a sound one from an economic standpoint partic-
ularly when, as it appears, less severe quality criteria can be applied
to treatment for reuse as compared to treatment for discharge, particular-
ly with respect to BOD and TOG. Although a working definition of reusable
water quality has been established in this study, i.e. 100 CU and 100 mg/1
TOG, and although many internal streams are used today in the kraft mill
of even poorer quality, no generally accepted criteria exist and it is
recommended that these be established by an industry committee. These
criteria necessarily must include additional parameters which were not
primary performance criteria in this study. It is also recommended that
grants be established to demonstrate reusability of treated effluents on
a mill scale, first in selected areas and subsequently on a maximum feasi-
ble basis while appropriate operating and product quality parameters are
monitored.
It is recommended that Part II of this grant be continued to conclusion
to assure availability of low cost carbon from mill sources for application
in effluent treatment and reuse.
It is recommended that the development of the FACET carbon adsorption
system be continued as a means of reducing the cost associated with acti-
vated carbon treatment. In view of the cost per 1000 gal of water treat-
ed, it is recommended that application of advanced effluent treatment for
color removal be preceded, in order to reduce overall water costs, by a
study and/or implementation of suitable and economically advantageous
methods of reducing; a) the total effluent load to the treatment system;
b) the color that is allowed to enter it; and c) the fluctuations in
concentration in the effluent.
Specific questions recommended for further investigation include permissi-
ble levels in reused water of dissolved solids such as sodium and calcium
salts, heavy metal salts and others; effect of treatment and reuse on
temperature levels in critical process areas, effect of dissolved salts
on sheet formation, drainage, corrosion, pulp washing, lime mud washing,
and scale formation.
-------
SECTION III
INTRODUCTION
St. Regis Paper Company with partial support by EPA has been engaged
since July, 1969, in a program for the development of an economical
system for maximum water reuse in the kraft pulp and paper industry as
a means of water pollution control and water conservation. This program
is based on two key concepts: (1) effluent treatment using activated
carbon and, (2) on-site production of activated carbon from readily
available raw materials, particularly black liquor, with full integration
into the kraft mill recovery and power systems to achieve the lowest net
cost of activated carbon.
This report covers the work accomplished under Part 1 of the program:
effluent treatment using activated carbon. Preliminary results from
the laboratory phase (15) and pilot plant (45,46,47) have been reported
previously.
Maximum water reuse, though not strictly definable, is assumed to be in
the order of 90% reuse. Ten percent of the usual fresh water input to
the kraft mill, it is assumed, will be discharged primarily to purge
inorganic salts at nonpolluting levels. This effluent would discharge
a total organic contaminant load equivalent to 10% or less of present
discharges.
Cost estimates made prior to starting this program indicated that treat-
ment of unbleached kraft pulp and paper mill effluent by process combina-
tions involving activated carbon, produced at the mill, and reuse of the
treated water would be less costly than treatment by other processes to
meet anticipated discharge standards, including standards for color.
The first task of this program was a survey of the literature and a
review of information from industry on treatment of effluents and reuse
of treated water in pulp and paper mills. The findings of this survey
were published in the first report of this program (40). All other
activities of Part I are covered by the present report.
The basic possibility of using activated carbon to obtain high-quality
effluents, particularly in terms of color removal, was indicated in
studies by McGlasson (10,11) and Berger (4).
Treatment of kraft effluents with lime, as reported a number of years ago
by Berger (3) and commercially applied by Davis (5), achieves substantial
color removal and thus appeared to merit consideration as a treatment
-------
step prior to activated carbon. Further reports on commercial installa-
tions for lime decolorization of pulp effluents have appeared more recent-
ly (7,8,14,18).
The success of biological oxidation systems, particularly aerated lagoons
in the southern states, in achieving as much as 90% reduction in BOD-5,
and the trend towards more installations of this type, provided sufficient
argument to include in this study an investigation of the effectiveness
of activated carbon treatment of bio-oxidized kraft effluent.
Thus it appeared that the role of activated carbon treatment might be as
the sole treatment step for color and dissolved organics, or as a major
second step after lime or biological treatment, or as a polishing step
following a combination of lime treatment and biological oxidation.
On the basis of laboratory studies and preliminary cost estimates, given
in this report, a decision was made to build and operate a pilot plant
at the Pensacola, Florida, St. Regis pulp and paper mill with a capacity
of about 30 gpm of unbleached kraft effluent to investigate three sequen-
ces of treatment:
Primary clarification, activated carbon adsorption
Lime treatment, carbon adsorption
Primary clarification, biological oxidation, carbon adsorption.
The results from operating this pilot plant from November, 1971, through
December, 1972, and cost estimates for treating pulp mill effluents for
reuse constitute a major part of this report.
-------
SECTION IV
LABORATORY STUDIES
INTRODUCTION
The purpose of the laboratory study was to help define where and how
activated carbon might be used most economically in order to achieve
treatment and up to 90% reuse of effluent from kraft pulp and paper
mills. Based on laboratory results, and with the aid of preliminary
designs and economic estimates as presented in Section V, the available
choices were to be narrowed down to three systmes and a pilot plant
designed and a pilot plant operation program developed. In this study,
two main approaches were pursued in terms of effluents to be treated.
These were the treatment of specific effluents for specific reuse, and
the treatment of the total effluent for general reuse. It was recognized
that the best approach might be a compromise of the two, such as separate
treatment of evaporator condensate and turpentine underflow for reuse in
lime mud and pulp washing, while combining all other effluents for treat-
ment and general reuse.
For the total effluent, the major treatments studied were primary clarifi-
cation followed by activated carbon; primary clarification, lime treatment,
activated carbon; primary clarification, lime treatment, biological oxida-
tion, and activated carbon; primary clarification, biological oxidation,
and activated carbon; and primary clarification, biological oxidation,
lime treatment, and activated carbon. In these laboratory studies,
"primary clarification" (simulated by filtration) was always carried out
separately as a first treatment. Mill scale applications (14,18) and
hence design and cost studies (Section XIV) use no clarification ahead
of lime treatment. Laboratory treatment methods are described in Appen-
dix B.
These same treatment combinations were studied with the caustic and acid
bleach effluents, although general reuse of bleach effluents is considered
unlikely until chloride removal can be achieved economically. Only minor
reductions can be expected in the chloride content as a result of any of
the treatment sequences investigated. Hence, reuse of the bleach effluents
in the pulp mill would eventually lead to an undesirable build-up of
chloride in the black liquor.
For other specific effluents, the use of activated carbon treatment alone
was investigated in this study, although other treatments may also be
required. For example, fiber removal from paper machine white water will
be necessary before applying activated carbon treatment for the removal
of dissolved organics.
-------
By using effluents from the three southern mills of St. Regis, and one
other mill with lime treatment, representing a spectrum of modern and
older plants, the chances were increased that this program would be of
value to the entire pulp and paper industry. The study included labora-
tory treatment of effluent samples from various mills before any treatment
and after full-scale primary, biological, and/or lime treatments. Table 1
summarizes the treatments used in the laboratory study with each of the
effluents.
In terms of activated carbon treatment per se, the primary function of
the laboratory study was to determine the dosages of carbon required to
reach desired concentrations of color and TOC in the various effluents
to be tested. These dosage data usually are obtained in the form of
adsorption isotherms as described in Appendix B. Carbon dosage data are
part of the information required to prepare preliminary designs and make
preliminary cost estimates for potential treatment systems. More specif-
ically, for instance, dosage data together with other cost data for the
carbon itself as well as for equipment, are major determining factors in
deciding between single use of carbon (usually as powdered carbon) and
regenerative use of carbon (usually as granular carbon).
Commercially available carbons vary widely in their characteristics and
more particularly in their effectiveness in removing dissolved TOC or
color from different effluent streams. Although it is an objective of
the two-part program to develop an activated carbon for effluent treatment
from pulp mill sources (particularly black liquor), this carbon was not
available in sufficient quantity for the laboratory program or the subse-
quent pilot plant operation. By using a wide variety of commercially
available activated carbons in the laboratory program, it was possible to
select suitable commercial carbons for the pilot plant operation, as well
as for potential full-scale applications, and it was also hoped that the
information gained would help in the definition of characteristics to be
obtained in the carbon made from black liquor. The major part of the
laboratory study was devoted to establishing adsorption equilibrium data
for carbon as discussed above.
The final selection of the carbon to be used in the pilot plant was made
on the basis of additional isotherm work on Pensacola No. 2 Mill effluent
(unbleached kraft) and on the basis of comparative adsorption column work.
This work was done after reaching the decision (see Section V) to treat
unbleached kraft total effluent using granular carbon columns, and it
also helped to determine the size of the pilot plant carbon columns.
After selecting the lime-carbon treatment sequence as one of the three
pilot plant treatment sequences, (see Section?), additional studies
were performed in the area of lime and lime-carbon treatment.
-------
Table 1. LABORATORY TREATMENTS OF EFFLUENTS
Specific effluents Type of treatment
Turpentine underflow Carbon
Evaporator condenstate Carbon
Paper machine white water Carbon
Condensate cooling tower Carbon
Woodyard Carbon fe
Caustic bleach Carbon and combined.
Acid bleach Carbon and combined
Total mill effluents
Total mill Carbon and combined
High solids sewer Carbon
Low solids sewer Carbon
Total mill without biological oxidation Carbon
Total mill with biological oxidation Carbon
Total mill lime treated Carbon and bio-oxidation
Simulated pulp mill effluents
0.05% black liquor solids Carbon
0.5% black liquor solids Carbon
All treatments preceded by filtration
Including lime and bio-oxidation in various combinations
-------
Special studies were carried out in two other areas. One of these areas
was studies leading to a continuous, countercurrent, fine carbon contact-
ing system called FACET, for which a patent has been awarded (41), and
which underwent initial development in the pilot plant. The other area
is that of carbon adsorption in an aerated system with or without biolog-
ical activity.
Significant observations which have some bearing on the selection of a
treatment system and its results are discussed in a separate part of this
section, including observed color reversion after lime treatment, and
adsorption or non-adsorption of compounds in effluents.
SELECTION OF CARBONS FOR EVALUATION
Each of the major manufacturers or suppliers of activated carbons for
water treatment was contacted to determine which of their carbons would
likely be best for removal of color and TOG from each of the effluents
to be studied. Help in selecting carbons for the demonstration was also
obtained from a study of the literature and discussions with other investi-
gators working on water treatment. In general, the cost of the carbon
and whether it is available in granular or powder form were disregarded.
Samples for adsorption studies in powder form were supplied by Westvaco,
Calgon, Atlas, Norit, and Barnebey-Cheney.
A list of the carbons used in this study is given in Table 2.
Table 2. CARBONS SELECTED FOR ADSORPTION STUDIES
Barnebey-
Westvaco Calgon Atlas Norit Cheney
AN F-100 S-51 A SC
C-190 F-300 KB F XZ
WV-L F-400 DARCO FQA
WV-G XPT NC
WV-W XPS SG
WV-M
RESULTS OF CARBON AND COMBINED TREATMENTS
Evaluation of Activated Carbons
The major means of determining the effectiveness of each type of acti-
vated carbon was carbon adsorption isotherms, the procedure for which
-------
is described in Appendix B. Six samples of the effluent were treated
with six different dosages of powdered carbon, and the results were
plotted on log-log graphs of loading of impurity on the carbon versus
concentration of impurity remaining in the water sample.
Isotherms for four carbons for removal of color from the total mill
effluent from mill A are shown in Figure 1. Note that the isotherms
all had definite breaks at a color concentration of about 300 CU.
Typical isotherms for removal of TOG are shown in Figure 2. These
isotherms employ the same four carbons and the same effluent as the
color isotherms above. The TOC isotherms are almost identical for all
four carbons and can be represented by a single line.
Such isotherms for color and TOC removal were prepared using two to
four commercial carbons for each of the various effluent streams and
after pre-treatment by lime or bio-oxidation or both. Over 150 isotherms,
not including duplicates and repeats, were prepared.
The reasons for using several carbons for each stream were (1) to indi-
cate which of the commercially available carbons recommended by the
manufacturers are best for treating each stream, and (2) to indicate
which physical properties, such as pore size, of carbons are best for
each stream.
Some of the results of these carbon treatments are less definitive
than had been hoped owing to several factors:
The feed concentration and composition of a given stream often
varied considerably over the period of time that samples were taken
for the pre-treatment and carbon adsorption experiments.
Some of the isotherms had very steep slopes owing to high concentra-
tions of poorly adsorbed compounds, such as methanol.
Some isotherms, particularly for color, were made up of lines of
several slopes caused by the several types of components in the
water.
Comparisons of Carbons
The most reliable method of comparing carbons for removal of impurities
from water samples was to compare the loading of color or TOC per unit
weight of carbon at selected concentrations of color and TOC. These
loadings were read from the isotherms at color concentrations of 1000,
500, 250 and 100 CU and TOC concentrations of 150, 100 and 50 mg/1,
representing intermediate and fairly complete treatment levels.
10
-------
5000
I
N
60
1=
U
O
53
H
I I I I I I I I
LEGEND OF CARBONS TESTED
50 100
500 1000
COLOR, APHA color units
Figure 1. Carbon adsorption isotherms for color
removal from total mill effluent
11
-------
SOOrTTTT
100
50
g
M
§
10
LEGEND OF
CARBONS TESTED 1
A
D
a
©
AN
F-400
KB
NC
5 10
50 100
500 1000
TOTAL ORGANIC CARBON, mg/1
Figure 2. Carbon adsorption isotherms for total organic carbon
removal from total mill effluent
12
-------
Another commonly used method is to compare the concentration of the
impurity remaining in the water after equal dosages of the several carbons.
Disadvantages of this method are that it requires that all carbons be
tested with the same sample of water and that it provides no information
on the effects of concentration on loading which isotherms would provide.
In addition to loading capacity, other properties of carbons must be
considered in choosing a carbon for a given application. These properties
include whether powdered or granular carbon is best, the rates of adsorp-
tion, the friability of the (granular) carbon, handling problems, and
cost. If a carbon is to be used in a column contactor, it should be
evaluated in continuous column tests with the particle size most likely
to be used in the future plant scale of operation.
In fully counter-current adsorption, the carbon dosage required to reduce
the color or TOG from any feed concentration to some desired level can be
calculated by using the following formula:
Dosage, g/1 = Co - D (1)
Xo
where:
Co= Initial concentration in CU or mgTOC/1,
D = Desired concentration of color or TOG to be reached
Xo= Loading at Co, CU/g or mgTOC/g.
If a batchwise single - stage treatment is used, the dosage required is:
Dosage, g/1 = Co - D (2)
Xd
Where Xd is the loading at the desired final concentration
In practice, the maximum loading reached in a packed column system using
granular carbon is about 65% of Xo from the isotherm, resulting in a.
corresponding increase in the required dosage amounting to about 50% in
excess of the dosage calculated using equation (1) above. This difference
in loading is due to two major factors. First, a packed column is not a
true counter-current contacting system, and only a fraction of the carbon
actually is in contact with effluent of the feed concentration, Co, while
most of the carbon is in contact with effluent of decreasing concentration
in the direction of effluent flow. Second, the equilibrium that is estab-
lished in a flow system is between the adsorptive surface and the effluent
in immediate contact with it, while the bulk effluent remains at a higher
concentration. This concentration difference provides the driving force
for the transfer of the dissolved substances from the bulk phase through
the stagnant water film surrounding each carbon particle and the water in
13
-------
the pores of the carbon. Results taken from a total of 29 isotherms are
given in Table 3. These results were obtained using the total mill
effluent from two southern kraft mills. The carbons are ranked according
to the relative effectiveness of each for removal of color at 500 CU
remaining in solution. Note that the ranking would have been different
at a lower remaining color concentration, except that the top carbon
performed best at either concentration. Note further that top performers
in terms of color removal are not the top performers in terms of TOC
removal. For Mill B effluent, S-51 and AN carbons were best, while for
Mill A the KB and AN carbons were best. For pulp and paper mill effluents,
color removal tends to be the more important characteristic of a good
carbon. Mill A effluent generally resulted in higher loadings than
Mill B effluent, probably because of the presence of sea water in Mill
A effluent.
Another comparison of carbons is shown in Table 4. This compares four
carbons used to treat paper machine white water from Mill B. Results
indicate that the AN and S-51 carbons were superior in color and TOC
removal. The C-190 carbon, which showed good color loading, performed
poorly on removal of TOCS
Table 3. COMPARISONS OF CARBONS FOR REMOVAL OF
COLOR AND TOC FROM TWO MILL EFFLUENTS
Color removal TOC removal
No. samples avg. loading,CU/g avg. loading, mg/1
Carbon averaged at 500 CU at 100 CU at 100 mg/1 at 50 tag/1
Mill A effluent
KB 1 1200 200 200 30
AN 4 840 170 110 24
F-400 1 700 125 350 22
NC 1 500 180 • 250 23
Mill B effluent
S-51 6 260 140 40 11
AN 8 250 90 30 7
NC 1 250 60 21 8
WV-G 1 145 120 30
WV-L 3 140 40 21 8
P-300 1 130 75 100 9
KB 1 110 100 110 30
WV-M 1 60 23 20 1
14
-------
Similar comparisons of carbons were prepared from isotherms of carbons
made with other in-plant and total mill effluents. It was concluded
that the most promising carbons for removal of color and TOG from the
major mill effluents were AN, S-51, KB, F-300, and NC. Of these five
carbons AN, KB, and NC are powdered carbons, and S-51 (as Granular Darco)
and F-300 are granular carbons.
Table 4. COMPARISON OF CARBONS FOR REMOVAL OF COLOR AND TOC FROM
PAPER MACHINE WHITE WATER FROM MILL B, AT PH 7.7
Color removal
No. samples avg. loading, CU/g TOC removal avg. loading,
Carbon averaged at 100 CU mg/g at 50 mg/1
AN 3 577 137
S-51 2 279 285
C-190 3 247 65
F-400 2 115 77
Comparison of Treatment Sequences Involving Activated Carbon
The purpose of this program was to determine the best means of treating
in-plant and final effluents of pulp and paper mills for water reuse.
The most economic method of treating a given stream might be with primary
clarification and carbon treatment, or with carbon following lime treat-
ment, primary clarification and bio-oxidation, or combinations of lime
treatment and bio-oxidation. The effects of such combinations of pre-
treatment and adsorption of color and TOC by carbon are illustrated by
the results in Tables 5 and 6 for the treatment of a total mill effluent
with AN carbon. The treatment methods are described in Appendix B.
(All treatment sequences carried out in the lab start with paper filtra-
tion, simulating primary clarification.) The carbon treatments are all
with 1 g of AN powdered carbon per liter of effluent.
15
-------
Table 5. EFFECT OF TREATMENT SEQUENCES ON
REMOVAL OF COLOR FROM MILL EFFLUENTS
Treatment3
Initial colorb
CU
Final color
CU
Using total mill effluent from Mill A
Carbon
Bio
Carbon
Bio
Lime
Carbon
Lime
Carbon
Lime
Bio
Carbon
948
833
570
833
570
45
1380
70
785
64
185
370
570
180
570
45
10
70
8
64
185
10
Using total mill effluents from Mills C and D
Mill C
No aeration
Carbon
Mill C
With aeration
Carbon
Mill D
Lime
Carbon
Mill D
Lime
Bio (in lab)
Carbon
4590
4590
2025
525C
525 c
479
2500
2025
700
525 c
90
525 c
479
190
Initial
PH
8.6
7.5
7.4
7.2
7.5
10.1
7.5
7.2
7.5
a Basis of treatments: filtration of the water samples before treatment
to simulate primary clarification, single-stage powdered carbon (AN)
with dosage of 1 g/1, biological oxidation for 40 hr with nutrients
and seed added, lime treatment with dosage of 5 g/1 of CaO.
b Four different samples of effluent used for five treatment sequences.
c The high color levels after lime treatment are not representative of
normal mill performance.
16
-------
Table 6. EFFECTS OF TREATMENT SEQUENCES ON
REMOVAL OF TOG FROM MILL EFFLUENTS
Treatments
Using total mill
Carbon
Bio
Carbon
Bio
Lime
Carbon
Lime
Carbon
Lime
Bio
Carbon
Initial TOG,
mg/1
effluent from Mill A
138
111
66
111
66
-
106
67
106
67
46
Using Mill C effluent: before and after
No aeration
Carbon
With aeration
Carbon
713
713
183
Final TOG
mg/1
75
66
30
66
-
-
67
38
67
46
23
aeration basin
570
183
90
Initial
PH
8.6
7.5
7.4
7.2
7.2
10.1
7.5
a Basis of treatments: filtration of the water samples before treatment
to simulate primary clarification, single-stage powdered carbon (AN)
with dosage of 1 g/1, biological oxidation for 40 hr with nutrients
and seed added, lime treatment with dosage of 5 g/1 CaO.
17
-------
When only carbon was used, the color was reduced from 948 to 370 CU
with a dosage of 1 g/1 (see Table 5). When this effluent sample was
first treated by bio-oxidation, the color was reduced only slightly
(to 570 CU). The subsequent carbon treatment reduced the color to
180 CU with a dosage of 1 g/1. When lime treatment was used following
bio-oxidation the color was reduced sharply to 45 CU, and a dosage of
1 g/1 of carbon reduced it further to 10 CU. The use of bio-oxidation
following lime treatment was not as effective as the reverse treatment,
since the color increased during the bio-oxidation. (This reversion
of color after lime treatment is discussed later in this section.)
This increased color was readily removed to a level of 100 CU with a
carbon dosage of 0.1 g/1 and to 10 CU with a dosage of 1.0 g/1.
The effects of the same treatment sequences on TOC removal from the
same total mill effluent are given in Table 6. With carbon alone, the
TOC was reduced from 138 mg/1 to 75 mg/1 with a single dosage of 1 g/1.
Bio-oxidation reduced the TOC from 111 to 66 mg/1, and carbon treatment
further reduced it to 30 mg/1 with a dosage of 1 g/1.
Tables 5 and 6 also compare treatment methods on the basis of color and
TOC removals using effluents from Mill C before and after bio-oxidation
at the mill. Carbon treatment of the effluent before bio-oxidation with
a dosage of 1 g/1 reduced the color from 4590 to 2500 CU and reduced
the color after bio-oxidation from 2025 to 700. After lime treatment
at Mill D, carbon treatment with a dosage of 1 g/1 reduced the color of
the mill effluent to 90 CU. Bio-oxidation of this lime-treated effluent
reduced the color only slightly, and carbon was less effective for color
removal than before the effluent was bio-oxidized. The data in Table 6
for treatment of samples from Mill C show that the aeration basin reduced
the TOC from 713 to 183 mg/1 and that treatment at 1 g/1 further reduced
the TOC to 90 mg/1.
The tentative conclusions drawn from this comparison of treatment se-
quences on total mill effluent, when the goal is reduction to less than
100 CU and less than 50 mg/1 TOC, are that primary emphasis in the
subsequent studies should be placed on carbon treatment, on bio-oxidation
followed by carbon treatment, on lime treatment, and on lime treatment
followed by carbon treatment. It appears that treatment sequences in-
cluding both bio-oxidation and lime treatment before carbon treatment
offer comparatively less promise for economical treatments of mill
effluents for reuse. These conclusions are based on activated carbon
being produced at low cost from black liquor at the mill site, which is
the original basis of this program. Goals of different color or TOC
levels for the treated water will affect the comparisons of treatment
sequences, as will changes in the relative feed levels of color and TOC.
It should be noted that in the comparison presented, the controlling
impurity is color.
18
-------
FINAL SELECTION OF COMMERCIAL CARBON FOR PILOT PLANT
As will be discussed in Section V, the selection from the many alternatives
was narrowed to further work on three treatment sequences using total mill
effluent (unbleached) and using a granular rather than powdered carbon
adsorption system.
A review of available information, such as is presented in Table 3, on
treating Mill B effluent (Pensacola No. 2 Mill, chosen as the pilot plant
site), indicated that Atlas S-51, Westvaco AN, Atlas KB, and Calgon F-300
gave the best overall results. Of the carbons available in granular form,
the ranking of the carbons is (1) Atlas Granular Darco, (2) Calgon F-300,
and (3) Westvaco WV-L. Similar comparisons were made of the candidate
carbons using additional samples of Pensacola No. 2 Mill effluent, in-plant
streams, and synthetic effluent prepared by diluting spent pulping liquor
to the concentration equivalent to that of mill effluents (containing 0.05%
black liquor solids).
The conclusion from these isotherm evaluations was that the most promising
carbons were Atlas Granular Darco and Westvaco granular carbons. The rates
of adsorption of these carbons were compared in a series of dynamic column
adsorption runs using Pensacola No. 2 Mill effluent at a superficial flow
velocity of 1 to 2 gpm/ft^in 2-in. columns containing equal weights of
carbon (about 4-ft beds). The carbons included Atlas Darco 20 x 40 and
8 x 30 and Westvaco WV-L 8 x 30, WV-W 20 x 50, WV-M 20 x 40. In these
tests, the two Darco carbons were superior to the Westvaco carbons of
corresponding size in the rate and degree of removal of color and TOG.
These column adsorption runs showed that the removal of color from the
mill effluent would be more difficult than the removal of TOG in reaching
the selected goals of 100 CU and 50 mg/1 TOG. Therefore the properties of
loadings and rates of removal of color were of prime importance in judging
which carbon should be used in the pilot plant.
Results of one of the side by side comparisons are shown in Figure 3,
comparing cumulative loadings of color and TOG on granular Darco and WV-M
carbons.
The low rates of removal of color in both column tests indicated that the
contact time of water in the pilot plant or commercial columns would be
much longer than is commonly used in carbon treatment of municipal waste-
water. To provide the longer time, it was decided to use a low flow rate
of 1 to 2 gpm/ft of cross sectional area, of the columns. At this low
flow rate, pressure drop would be low and pumping cost would not be an
important factor in operating costs in a commercial unit. Therefore, it
was decided that to gain as much external area of carbon as possible while
still maintaining a reasonable pressure drop, 20 x 40 mesh carbon should
be used. The 20 x 40 mesh carbon has an external surface area about four
times that of the more generally used 8 x 30 mesh carbon and thus gives a
19
-------
Column A
Column B
Darco 20 x 40
lOOOg, 2.381 liters
1.01 gpm/ft2
2.1 bed vols./hr
Westvaco WV-M 20 x 40
lOOOg, 1.981 liters
1.01 gpm/ft2
2.6 bed vols./hr
250
o
o
H
00
C3
U
W
O
i-J
O
u
PN
o
o
a
M
5
200 -
150 -
A, DARCO, COLOR
100 -
50
100 150
CUMULATIVE BED VOLUMES
200
250
Figure 3. Comparison of cumulative loadings on Darco and
WV-M carbons in simultaneous 2-inch column tests
20
-------
much higher rate of adsorption. A finer size of Granular Darco (Darco XPT)
was available which had a particle size range (40 x 140 mesh) that was
close to the 60 x 140 size considered desirable for the FACET system as
described in Sections V and VI. Therefore, Darco XPT was also evaluated
in isotherm and column tests and found to be very similar to Darco 20 x 40
in equilibrium adsorption properties and the rates of removal were about
twice those with Darco 20 x 40.
Both the Darco 20 x 40 and the 40 x 140 carbons were evaluated for resis-
tance to attrition in stirring tests and by pumping the Darco XPT as a
slurry. The resistance of the Darco carbons to attrition was not as high
as some carbons but was considered satisfactory.
Atlas Granular Darco was finally selected for use in the pilot plant on
the basis of the following points:
It had generally superior adsorption capacities and rates of
adsorption for color and TOG.
The Atlas carbon was available in the two size ranges needed for
the pilot plant.
The raw material for the Atlas carbon (lignite) is similar to the
lignin-derived raw material to be used in producing activated carbon
from kraft pulping liquor under Part II of the EPA - St. Regis program.
- The Atlas carbons were lower in price - $0.25/lb. for the 20 x 40 mesh
carbon and $0.17/lb for 40 x 140 mesh carbon than the other candidate
granular carbons, which sold at about $0.29/lb.
Attrition loss of the Atlas 40 x 140, which is an important considera-
tion for the FACET carbon, was only slightly greater than that of
other granular carbons.
ADDITIONAL STUDIES
Lime and Lime-Carbon Treatment
After the selection of the lime-carbon treatment sequence of total mill
effluent as one of the sequences to be piloted (see Section V), further
experimental work was done to determine more closely the effect of
different combinations of lime and carbon dosage on the resultant color
and TOG levels. As described in Appendix B, a dosage of 5000 mg/1 CaO
was used in earlier laboratory studies, as compared to the dosage of
1000-2000 mg/1 CaO that was later found to be satisfactory in plant
tests at Interstate Paper Co. at Riceboro, Ga. and Continental Can Corp.
at Hodge, La., (14,18).
21
-------
Table 7 shows results obtained at different lime dosages in one particu-
lar effluent sample (Pensacola No. 2 Mill effluent). Table 8 shows the
effect of different combinations of dosages of lime and carbon, where
the sample was only filtered after lime treatment and before carbon
treatment, i.e. no carbonation was used to remove soluble lime.
Table 7. LIME TREATMENT OF UNBLEACHED KRAFT EFFLUENT
CaO dosage,
mg/1
Raw
200
400
800
1200
Color,
CU
1585
1375
584
277
257
TOC,
mg/1
245
204
181
146
136
Table 8. LIME-CARBON TREATMENT OF UNBLEACHED KRAFT EFFLUENT
Dosage, mg/1
CaO
0
0
200
200
400
400
800
800
1200
1200
Carbon3
0
5000
1000
5000
1000
5000
1000
5000
1000
5000
Darco S-51
Color,
CU
1675
413
438
177
61
6
56
8
21
1
CaO
mg/lb
14
14
154
98
539
504
832
714
PH
,7
.1
10.
10.
11.8
11.8
12.3
12.2
12.4
12.4
12.7
12.7
Concentration found in filtered treated effluent
22
-------
Results in Table 7 indicate that 63% of the color was removed at a
dosage of 400 mg/1 CaO, and that removal in this case leveled off at
about 82%. Table 8 shows, at very similar initial concentrations,
that 1 g/1 activated carbon is sufficient to reduce the color below
100 CU if 400 mg/1 CaO had been used. Higher lime dosages were un-
necessary in conjunction with 1 g/1 carbon, while higher carbon dosages
were unnecessary after 400 mg/1 CaO or higher, and insufficient at lower
lime dosages. Table 7 indicates that higher lime dosages alone were
also unable to reduce the color below 100 CU in this effluent. These
results indicated generally that an optimum combination lime and carbon
treatment might be developed in the pilot plant requiring substantially
lower lime dosages than are employed if only lime is used for color
removal.
Color Changes of Raw and Treated Effluents
In the course of laboratory work on the lime-bio treatment sequence,
color increases were observed when lime treated effluent was bio-oxidized.
The first two lines in Table 9 summarize these results. It was believed
that these increases were rather significant, particularly since color
increases were also observed at that time in the natural aeration bio-
oxidation ponds of the Riceboro, Ga., mill of Interstate Paper Co. (18),
and later at Continental Can Corp. at Hodge, La., (14) where lime treat-
ments before bio-oxidation are being operated on a commercial scale.
Subsequent laboratory work shown in lines 3-7 of Table 9 indicated that
color of lime treated effluent usually increases upon aging by about 50%,
with the exception of one out of 6 samples which showed a 28% decrease.
At the same time it was observed that carbon treated effluent tends to
decrease in color, while raw effluent exhibits some instability, showing
a decrease first and then an increase.
Table 9 (last line) also shows results observed during pilot plant bio-
oxidation of unbleached effluent for comparison, indicating a 15% color
reduction during bio-oxidation.
These results indicate a lower degree of stability of color than expected,
particularly since the color of kraft mill effluent is generally held to
be non-degradable in the usual aerated lagoon treatment system. Partic-
ularly in the case of lime treatment, treatment results should be assess-
ed giving full consideration to potential color increases during subsequent
bio-oxidation or storage.
23
-------
Table 9. COLOR CHANGE OF KRAFT EFFLUENT
to
Effluent
No. of
Samples
Caustic bleach 2
Unbl. total
Unbl
Unbl
. total
. total
Unbl. total
Unbl
Unbl
Unbl
Unbl
Unbl
a
b
. total
. total
. total
. total
. total
mill 3
mill 2
mill 1
mill 1
mill 1
mill 1
mill 1
mill 1
mill0
Initial color, CU Final color, CU
lime tr. carbon tr. raw bio-oxid. aged
67 130
46 104
166 - 242
68 100
184 ... 133
184 ... 313
184 - 268
119 - - 60
1180 - 1136
1305
950 810
Aging
time,
days
-
0.21
6
5
4
4
20
4
11
-
Sample
size, Sample
gal. initial
7.0
7.0
40 7.0
40 7.0
40 7.0
0.26 7.0
0.26a 7.0
0.26
400 10.1
400 b
10.2
PH
final
-
-
-
-
-
9.0
8.8
-
8.2
8.6
kept dark
ft (\C\ rt. «1 1 fl *-IWW« 1 rt
c pilot plant bio-oxidation
-------
Adsorption of Specific Compounds
Each of the effluents studied contains a mixture of organic chemcial
compounds. Usually, a few or none of these compounds are identified in
effluent treatment work beyond being reported as BOD, color, COD, or, in
this study, primarily as color and TOC. The adsorption of organic sub-
stances from mixed solution is a complex phenomenon, possibly owing to
the effect of one substance on the adsorption of another. This can
manifest itself in preferential adsorption of one substance over others,
non-adsorption if a substance is only weakly adsorbed at best, or the
displacement of a weakly adsorbed by a strongly adsorbed substance.
This selective or preferential adsorption must be kept in mind in the
evaluation of equilibrium adsorption data.
Concentrations of some specific compounds have been determined in a
number of tests in this study for two reasons. One is that, from a
commercial applications standpoint, this information may indicate
opportunities for recovery of by-products by adsorption - desorption
cycles with activated carbon. The return from the scale of such by-
products could be credited against effluent treatment costs. The
turpentine decanter underflow, which was studied in the laboratory for
removal of specific compounds, may be promising in this respect due to
its content of turpentine fractions and other organics. Figure 4 pre-
sents the adsorption isotherm of loading vs. concentration for carbon
treatment of turpentine underflow. Also given is an adjusted isotherm
with methanol subtracted from the TOC content, an isotherm for acetone,
and an isotherm for dimethyl sulfide plus dimethyl disulfide. These
compounds were identified by using gas-liquid chromatography. In this
experiment, methanol was found not to be adsorbed and about 207o of TOC
was not identified. The initial conclusion based on overall TOC ad-
sorption in this example would have been that this carbon is ineffective.
However, isotherms for specific compounds and the methanol-free TOC
would suggest that this carbon is reasonably efficient for removing
components other than methanol, thus potentially yielding a deodorized
water containing some methanol that could be reused in the mill. This
phenomenon has meanwhile been further explored in a separate study and
a patent application has been filed.
Adsorption Rates
During the course of the laboratory studies, a limited amount of time
was devoted to an examination of the factors affecting adsorption rates,
i.e. the rate at which impurities defined as TOC or color are removed
from effluent by activated carbon.
25
-------
5000
fcO
I
o
55
H
1000
500
TOTAL,
w/o MeOH
50 100
500 1000 5000 10000
TOTAL ORGANIC CARBON, mg/1
Figure 4. Carbon adsorption isotherms for specific
compounds in turpentine underflow
26
-------
A major factor affecting adsorption rate is particle size. If one
assumes that the equilibrium loading is not affected by particle size,
then the smaller particles should adsorb at a higher rate because the
distance a solute must travel within the porous structure of the carbon
is decreased. Weber concluded (15) that rate is proportional to the
inverse of the diameter squared, or l/d2. However, in another study,
Weber reports for 50% ABS removal from solution by two different sizes
(350 and 670 micron) that rates increased proportional to 1/d .
Melter (11) reports rate proportional to 1/d2, while Berger (10) found
on pulping effluents that rates were proportional to
In this laboratory study, rates were evaluated in stirred batch tests
where carbons of four different particle size ranges were added to
unbleached kraft total mill effluent from Mill A at dosages of 1, 5
and 10 g/1. Results indicate a proportionality of TOG adsorption rates
to the inverse of diameter to some power, with the power varying between
1.6 and 3.0 depending on the degree of loading at which rates were
evaluated, i.e. power of 1.6 at 20%, 2.0 at 30% and 3.0 at 50% of load-
ing. In other words, the more the carbon is loaded the more of a dis-
advantage the larger particle size presents compared to the smaller
size. However, this information was not later borne out fully by the
comparison of rates in the adsorption columns and the FACET systems in
the pilot plant. In these pilot plant runs the rate vs particle size
relationship or TOG removal appeared to be closer to proportionality
to 1/d2 at loadings normally above 50% of isotherm equilibrium loadings,
and perhaps even closer to 1/d ' if rates had been compared at strictly
equal loading on the carbon.
The effect of dosage on rate would be expected to cancel out if rate is
expressed as mgTOC/g carbon x hr. However, rates at higher dosages were
actually found to be lower.
27
-------
SECTION V
SELECTION OF EFFLUENT TREATMENT SYSTEMS TO BE
EVALUATED IN THE PILOT PLANT
INTRODUCTION
A laboratory program (see Section IV) was carried out to determine the
properties of various streams in three St. Regis kraft pulp and paper
mills and to provide a basis for selecting which streams have the greatest
promise for reuse when treated. Treatments of the various in-plant and
final effluent streams were evaluated in the laboratory. These treatments
included bio-oxidation, lime treatment, and activated carbon adsorption.
Combinations of these treatments were also evaluated to permit selection
of treatment sequences to be used in the pilot plant.
As a further input to selection of treatment sequences for the pilot plant,
order-of-magnitude cost estimates were made for treatment of effluents
from two St. Regis pulp and paper mills with the candidate treatment se-
quences. The selection of effluents to be treated and of effluent treat-
ment sequences, as well as the result of the cost estimates, are presented
in this section.
These estimates were based on total mill effluents from kraft pulp mills
not having bleaching operations because the chloride of the bleach water
would not be removed in the treatments and would make it unsuitable for
reuse as general mill feed water. The cost estimates included the follow-
ing combinations of primary clarification (primary), lime treatment (lime),
primary clarification combined with lime treatment (primary/lime), (as
carried out on a mill scale (14,18)), biological oxidation (bio), and
carbon adsorption (carbon):
primary, bio
primary, bio, carbon
primary, carbon
primary, lime
primary, lime, carbon
primary/lime, carbon
The carbon treatments included use of powdered carbon without regeneration
of the carbon, granular carbon with regeneration, and a new counter-current
continuous process called FACET in which an intermediate size carbon is
used.
28
-------
These estimates indicated that the use of powdered carbon was not practical
nor economical and that both column adsorption using granular carbon and
slurry contacting using the FACET process should be investigated in the
pilot plant. The treatment sequences selected for study in the pilot
plant to give reusable quality water were:
primary/lime, carbon
primary, bio, carbon
primary, carbon
(The actual execution of the primary/lime, carbon sequence in the pilot
plant used a holding pond to smooth out concentration variations and
inadvertently acted as a primary clarifier before lime treatment.)
In each sequence both granular carbon column and FACET slurry carbon pro-
cesses would be used. The cost estimates indicated that the total fixed
and direct operating costs for these sequences would range from $0.16 to
$0.29 per kgal in commercial installations.
In view of the demonstrated success and reliability of carbon column
adsorption process for treatment of municipal wastewater, compared with
the unproven status of the FACET process, it was decided that the major
emphasis in the pilot plant would be on the carbon column process.
SELECTION OF EFFLUENTS TO BE TREATED
Selection Procedure
The selection procedure consisted of eliminating the less desirable
effluent systems on the basis of quantitative comparison and judgement
against different criteria which were derived from the project objectives.
The effluent systems considered in this study are listed below, including
a brief argument for the acceptance or rejection of each.
Total Mill Effluent Without Bleach Effluent
One of the effluent systems considered was the treatment of total mill
effluent exclusive of bleach plant effluent, by carbon or by carbon
following bio-oxidation and/or lime treatment. This was the approach
selected for the pilot plant.
Production of a general mill feed water from an unbleached kraft total
mill effluent appeared to be technically feasible. This was indicated
by laboratory results of this program and by results reported by others.
Activated carbon plays an important role in various treatment sequences
capable of producing reusable water.
29
-------
Economically, the production of reusable water appeared to be compet-
itive when compared with treatment and disposal at anticipated more-
stringent near-future disposal standards.
Total Mill Effluent Including Bleach Effluent
Treatment of total mill effluent, including bleach plant effluents, by
carbon or by carbon sequentially combined with bio-oxidation and/or
lime treatment had also been considered and had been rejected. The
inclusion of bleach plant effluents (from chlorine-based bleach systems)
in the total mill effluent was considered undesirable because the chlo-
ride ion would build up in the recycled water. Laboratory investigations
showed that the chlorine content of bleach effluents is essentially not
reduced by activated carbon, or by any of the treatment sequences consid-
ered in this program. If a chloride removal process such as ion exchange
or reverse osmosis were to be considered at a later time, separate treat-
ment of bleach plant effluent would probably be the approach to take.
Chloride removal could then follow lime or lime plus activated carbon
treatment of bleach plant effluent.
In-Plant Effluents
Treatment of in-plant effluents by activated carbon in a separate unit
for each effluent had been considered as the major alternative to total
mill effluent treatment. It had been found that a substantial part of
the reuse of in-plant effluents that can ultimately be achieved does
not, or not yet, require activated carbon treatment. Moreover, the
favorable cost estimated for the total mill effluent treatment approach,
and the lesser complexity of that approach in terms of actual operation
and of its development, argued against pursuing treatment of individual
effluents under this program.
PRELIMINARY SELECTION OF TREATMENT SEQUENCES AND CARBON ADSORPTION SYSTEMS
Selection Procedure
The preliminary selection of treatment sequences to be studied in the
pilot plant proceeded along the following steps:
(1) Identification of potential sequences.
(2) Determination of reusable water quality requirements.
(3) Laboratory evaluation of the potential sequences with effluents
from the three southern mills of St. Regis and from one other
mill which uses lime treatment of the total mill effluent.
30
-------
(4) Preliminary selection of sequences based on laboratory
results and judgement of comparative economics.
(5) Consideration of variants within these sequences, particularly
with regard to the carbon adsorption step.
Identification of Potential Treatment Sequences
The following treatment sequences were considered to have sufficient
merit to warrant study in the laboratory:
(1) Primary clarification followed by activated carbon adsorption;
(2) Primary clarification, lime precipitation, activated carbon
adsorption;
(3) Primary clarification, biological oxidation, activated carbon
adsorption;
(4) Primary clarification, lime precipitation, biological oxidation,
activated carbon adsorption;
(5) Primary clarification, biological oxidation, lime precipitation,
activated carbon adsorption.
Determination of Reusable Water Quality Requirements
Water quality requirements are usually expressed in terms of a numerical
maximum or range of a number of properties such as color, dissolved solids,
etc. Table 10 presents such numerical values based on several references.
Although the TAPPI values are used as guidelines for new mills, their basis
is a survey of industry practice dating back to 1948, rather than a specific
determination of requirements. A TAPPI committee was set up at the 1970
TAPPI Air and Water Conference for the determination of actual water quality
requirements at various points of use in the kraft process. The absence
of BOD specifications in the TAPPI standards in Table 10 is noteworthy.
31
-------
Table 10. RANGES OF DESIRED PROPERTIES OF PROCESS
WATER FOR UNBLEACHED KRAFT MANUFACTURE
St. Regis TAPPI TAPPI
Property proposal 5/1/69* E601 S-53D £603 S-49C
Turbidity, mg/1 5-25 100 25
Color, CU 10-80 100 5
pH 6.5-8.0
Total alkalinity, mg/1 20-150 150 75
Hardness, mg/1 CaCOo 5-200 200 100
Dissolved solids, mg/1 50-500 500 250
Chloride, mg/1 10-150 . 200 75
Fe, mg/1 0.5 1 0.1
Mn, mg/1 0.3 0.5 0.05
COD, mg/1 0-12
BOD, mg/1 0-5
a Based on NCASI Bulletin 203 (48)
b Specification for chemical composition of process water for kraft paper
manufacture; tentative standard, corrected August, 1953.
c Specification for chemical composition of process water for manufacture
of soda and sulfate pulps; tentative standard, December, 1949.
Early in this program an effort was made within St. Regis to collect water
quality standards for the various points in the process of making pulp and
paper. The results of this effort, compiled in table form with accompany-
ing comments, are included in this report as Appendix A.
As a result of this effort, it was realized that the tentative TAPPI stan-
dards in Table 10 may well be over-specifying some of the water quality
requirements of an unbleached kraft mill, except boiler feed water which
requires further treatment. Water of 100 CU and 50 mg/1 TOG (total organic
carbon), equivalent to about 25-100 mg/1 BOD, was tentatively considered to
be adequate as a general mill feed water (later changed to 100 CU and 100
mg/1 TOC). It was considered possible that problems such as slime growth
might arise at some of the water use points due to specific organic compounds
or due to the general BOD. However, such problems as might arise will most
probably have an economical remedy, such as the addition of slimicides or
32
-------
a downward adjustment of the TOG specification for part of the mill
feed water. Thus, the tentative targets of 100 CU and 50 mg/1 TOG
were considered a reasonable basis to proceed on.
Other water properties were not considered in the selection of treatment
sequences, although their ultimate importance is recognized. Consider-
ation of color and TOG was believed to be sufficient, since removal of
TOG and color is the main purpose of the treatment systems under devel-
opment. Differences in efficiency and cost between the various systems
are expected in this area and are of prime importance. Consideration
of possible differences in removal of dissolved inorganics or in recycle
water temperature was deferred.
Laboratory Evaluation and Preliminary Selection of Treatment Sequences
The laboratory evaluation of the above identified treatment sequences
has been discussed in Section IV and has also been reported (15).
Table 11 and 12 repeat the data on color and TOG removal presented in
Tables 5 and 6 and additionally present data on dosages for two- and
three-stage treatment with powdered carbon estimated from isotherm
data by the procedure described in Appendix B.
Carbon treatment alone is seen to be capable of achieving the tentative
standards of 100 CU and 50 mg/1 TOG. The carbon dosages for both color
and TOG removal are about equal when treating Mill A effluent. For
Mill C effluent, substantially higher carbon dosages are required with
TOG being the dosage determining factor.
The use of lime or bio pretreatments results in substantial reductions
in the carbon dosage in all cases. The bio treatment is, as expected,
more effective for removing TOG than for removing color; therefore,
color removal becomes the dosage determining factor after bio treatment
(unless a partial bio treatment is used). Lime pretreatment, in contrast,
is more effective for color removal. Thus, TOG removal after lime treat-
ment becomes the dosage determining factor. (Caution is indicated, how-
ever, by color increases observed in laboratory biological oxidations
following lime treatment, as discussed in Section IV.)
When comparing the carbon dosages required to reach 100 CU and 50 mg/1
TOG in a three-step treatment (lime-bio-carbon and bio-lime-carbon) with
dosages required in the two-step lime-carbon treatment, it appears that
their is little or no carbon dosage advantage in the three-step treat-
ments. Since volumetric flow through a treatment system is a major cost
determinant, a third treatment step cannot be justified on the basis of
33
-------
Table 11. EFFECT OF TREATMENT SEQUENCES ON REMOVAL OF COLOR FROM TOTAL MILL EFFLUENTS
C
Treatment with carbon
with lg/1 carbon
•a
Treatment
Mill A
Carbon
Bio
Carbon
Bio
Lime
Carbon
Lime
Carbon
Lime
Bio
Carbon
Mill C
Carbon
Bio (at mill)
Carbon
Mill C
Lime (at mill)
Carbon
Lime (at mill)
Bio
Carbon
Initial
color, CU
948
833
570
833
570
45
1,380
70
785
64
185
4,590
4,590
2,025
525e
-
525e
479
Final
color.CU
-
570
-
570
45
-
70
-
64
185
-
-
2,025
-
525e
e
525
479
-
Initial
pH
8.6
7.5
7.4
7.2
7.5
10.1
7.5
7.2
7.5
color, CU
370
180
10
8
10
2,500
700
90
190
7
10
Removal
61
68
78
87
95
46
66
83
60
Loading
CU/g
580
390
35
62
175
2,100
1,300
430
290
dosage to 100 CU,g/l
123
stage stages stages
3.4 0.91 0.58
1.6 0.61 0.48
d - -
,
d
0.10
17.3 2.3 l.io
8.4 1.77 1.02
0.80 0.30 0.24
5.30f 1.30 0.74
All samples of water were filtered before treatment to remove suspended solids (simulating primary
, clarification).
Four different samples of effluent were used for five treatment sequences.
Westvaco Aqua Nuchar.
d No carbon needed to reach 100 CU.
e High color not considered representative of normal mill performance.
f Poor adsorption indicated by steep isotherm.
-------
Ui
Table 12. EFFECTS OF TREATMENT SEQUENCES ON REMOVAL OF TOG FROM TOTAL MILL EFFLUENTS
Treatment with carbon
Treatment3
Mill A
Carbon
Bio
Carbon
Bio
Lime
Carbon
Lime
Carbon
Lime
Bio
Carbon
Mill C
Carbon
Bio (at mill)
Carbon
Initial
TOG,
mg/1
138
111
66
111
66
-
106
67
106
67
46
713
713
183
Final
TOG,
mg/1
66
-
66
-
-
67
-
67
46
-
183
-
Initial
PH
8.6
7.5
7.4
7.2
7.2
10.1
7.5
with
TOG,
mg/1
75
30
23
38
23
570
90
lg/1 carbon
7c
Removal
44
55
-
43
50
20
51
Loading,
mg/TOC/g
53
36
_
30
23
140
90
dosage
1
stage
2.70
0.20
_
0.19
-
29.0
4.40
to 50/mg/l TOC,g/l
2 3
stages stages
0.86 0.57
-
_ -
_ -
-
5.20 3.20
1.20 0.73
All samples of water were filtered before treatment to remove suspended solids (simulating primary
clarification).
Westvaco Aqua Nuchar.
-------
marginal contaminant load reductions for the remaining two steps. This
comparison of treatment sequences for total mill effluent, when the
goal is reduction to less than 100 CU and less than 50 mg/1 TOC, serves
to eliminate the three-step sequences from further consideration.
Goals of different color or TOC levels in the treated water will affect
the comparison of treatment sequences. Changes in the concentration
of color and TOC in the water being treated will also affect the eco-
nomic comparison. Because of the variability of effluents from mill
to mill, the effect of color and TOC of the feed water on capital and
operating costs must be included in the economic comparison of treat-
ment sequences.
Variants in Carbon Treatment Systems
The conventional choice to be made for carbon treatment is between
(a) a granular carbon system including adsorption and carbon regenera-
tion; or (b) a powdered carbon system including adsorption and carbon
disposal after single use.
A granular carbon adsorption system conventionally uses the carbon in
packed columns, the liquid being purified while flowing through the
bed of carbon. In a powdered carbon system, the carbon is contacted
with the liquid in agitated tanks, followed by clarifiers to remove
the carbon from the water. Clarification must usually be assisted by
the use of flocculants.
Thermal regeneration in a multiple-hearth furnace is the most common
method for regenerating spent granular carbon. The development of
thermal regeneration of spent powdered carbon has only recently been
vigorously pursued (20,21,49,50,51). Disadvantages in regeneration
of powdered carbon, compared to granular carbon, lie in the difficulty
of dewatering the carbon before regeneration and in high carbon losses
during thermal regeneration. No commercial system was in existence
for regenerating powdered carbon, hence powdered carbon was considered
as a single-use material in this evaluation.
The purchased cost of powdered carbon is substantially lower than that
of granular carbon $0.085/lb. vs. $0.25-0.30(at the time of this study).
The conventional choice between the granular and powdered carbon
systems for a specific application is based on an evaluation of several
technical and economic factors.
36
-------
Usually, the use of powdered carbon is favored when the carbon
consumption is low due to low liquid volume and/or low dosage require-
ments (e.g. due to low contaminant concentration). Granular carbon
would be favored where countercurrent adsorption is advantageous, and
where the separation of a powdered carbon from the liquid being purified
would present a problem. An economic evaluation must take into consid-
eration the capital and operating costs associated with single-use
powdered or granular carbon and with regenerative use of granular car-
bon. The latter would require the higher capital investment, but
operating cost for make-up carbon would be lower.
As a third choice, St. Regis suggested a counter-current contacting
system that uses a fine activated carbon of a size intermediate between
powdered and standard granular carbon. A preliminary engineering
assessment of this system (dubbed FACET) indicated that it combines
some of the advantages of both conventional systems. The FACET system
is shown schematically in Figure 6 (Section VI) as a three-stage counter-
current contacting system. A patent (41) has since been issued on this
process.
Carbon size and concentration are chosen such that contact time require-
ments are short, while the size is large enough to provide fact settling
without the assistance of flocculating agents. Therefore, FACET permits
use of simple stirred tanks as contactor-settlers and avoids the need
for preclarification of the water commonly required in granular adsorp-
tion systems to prevent column pressure build-up and avoids the need
for frequent back-washing. Present indications are that a suitable
particle-size range for FACET is about 100 to 250 microns (140 to 60
mesh). The FACET particle size is believed to be large enough to
permit use of a standard multiple-hearth furnace for regeneration. The
FACET process is discussed further in Section VI.
The choice among the three carbon systems as applied in this project
is influenced by the proposed production of activated carbon at the
mill for the treatment of mill effluent. Such mill-site production
of carbon makes all three types of carbon available at a substantially
lower cost and also at a smaller cost differential between granular,
FACET, and powdered carbon.
Moreover, another dimension is added to the economic evaluation since
carbon cost depends not only on type of carbon but on production
volume. The latter is very materially dependent on whether the carbon
is discarded after use or is regenerated.
As a result of this preliminary evaluation of alternative carbon systems,
it appeared that all three systems should be further evaluated in the
economic estimate.
37
-------
ORDER OF MAGNITUDE COST ESTIMATES
Bases of Cost Estimates
Costs of equipment and overall plant costs for primary clarification,
bio-oxidation, lime treatment, and carbon adsorption were obtained
primarily from references 17 and 23-34. Additional sources of cost
data were Infilco, Eimco, and St. Regis' internal sources. Most of
the reference data on activated carbon systems are based on municipal
effluent treatment.
The estimates are of the order-of-magnitude type which might have
inaccuracies of 20% to 40%. Allowances were made in the capital costs
for installation, engineering, and contingencies. Most capital costs
were based on overall plant costs from curves or from similar plants
with the cost adjusted to the desired flow rate by suitable values of
capacity-cost exponents.
St. Regis Mill A total mill effluent, on which the laboratory tests
were based, had a flow rate of 38 mgd, which included about 20 mgd of
cooling water. Even though the flow without the cooling-water dilution
would have been only 18 mgd and would have contained about twice the
measured color and TOC concentrations, cost estimates were based on
the 38 mgd flow because the lab data were based on this diluted flow.
The average composition of this stream was 140 mg/1 TOC, 1000 CU,
130 mg/1 BOD, and pH 8.6.
The average composition of the St. Regis Mill C total mill effluent
during the lab program was 640 mg/1 TOC, 4450 CU, and pH 9.9. The
color was abnormally high because of mill upsets during this period.
The average flow rate was 15.3 mgd.
A limited dosage lime process (1000 mg/1 CaO) similar to that used at
Interstate Paper Company's Riceboro Mill (5,18) was used in these
estimates. The amounts of TOC and color reduction by lime treatment
that were used in these estimates are those obtained in the lab pro-
gram with a dosage of 5000 mg/1 CaO which gives about the same color
reduction as a dosage of 1000 ppm. The capital costs for lime treat-
ment include lime slaking, reactor-clarifier, carbonation, CaCOo
clarifier, and sludge filters, but do not include any portion of the
capital and operating costs of lime calcination.
The TOC reductions by bio-oxidation used in the estimates were 140 to
66 mg/1 TOC for Mill A and 640 to 140 for Mill C, which were the
reductions obtained in the laboratory experiments. It was assumed
that the bio-oxidation would be carried out in mechanically aerated
basins with a retention time of 10 days.
38
-------
Three carbon treatment processes were considered: powdered carbon
(about 10 micron average particle size), fine-carbon FACET system
(100 to 250 micron particle size to 140 x 60 mesh), and granular
carbon which is used in a fixed bed. Carbon dosage estimates were
based on laboratory results with Westvaco Aqua Nuchar powdered
carbon. The dosage used in these estimates was that required to re-
duce the color to 100 color units and TOG to 50 mg/1 plus 50% excess
carbon to allow for incomplete equilibration. Water with these values
of color and TOG is considered satisfactory for general plant reuse.
The influence of the price of carbon on treatment costs was determined
by assuming prices of $0.02, $0.085, and $0.27 per pound of carbon.
The price of commercially available powdered carbon was $0.085/lb and
that for granular carbon was $0.27/lb at the time of these estimates.
Charges related to fixed capital are included in operating costs on
the basis of 16-year, straight line depreciation (6.257., of TCI/year),
plus a total of 12% of TCI/year for repair, maintenance, taxes, and
interest charges. General administrative overhead (GAO) was assumed
to be 75% of labor costs plus 1.5% of TCI/year.
Pertinent operating conditions are given in Tables 13, 14, and 15 for
each estimate, along with results of the cost estimates.
Results of Estimates
Overview -
Tables 13 and 14 present a summary of the conditions and cost data
of the estimates prepared. The presentation in Table 15 is less
detailed and shows a wider range of combinations including those of
Tables 13 and 14.
Estimates A-2, A-ll, and A-3 in Table 13 represent the three treat-
ment sequences under consideration, using granular carbon adsorption.
The effect of capital and operating costs on treatment system selec-
tion will be discussed first. The influence of carbon cost on capital
and operating costs and indirectly on treatment system selection will
be discussed subsequently.
Capital Investment -
In terms of capital investment, the primary/lime-carbon system (A-ll)
represents the lowest cost, mainly because of the significantly lower
cost of the carbon adsorption and regeneration portion of the total
system. The primary-bio-carbon system (A-3) requires the highest
capital investment.
39
-------
Table 13.
Estimated No.
Treatments
ORDER- OF-MAGNITUDE COST ESTIMATES FOR VARIOUS TREATMENTS OF TOTAL MILL
EFFLUENT FROM MILL A; EFFLUENT: 140 mg/1 TOG, 1000 CU
Effluent
Flow, mgd.
Carbon treatment, type
Dosage, Ib/kgal ,
Carbon cone'n.,lb/ft
Retention, hours, carbon
water
Carbon regeneration
Final filter
Treated water comp'n: TOC, mg/1
Color,CU
Capital costs. $ million
Price of carbon, $/lb 0.02
Primary 2.5
Bio-oxidation
Primary/lime treatment
Filters
Carbon contactors
Carbon dewatering
Carbon drying & regeneration
Carbon in system
Total cost of investment(TCI)
Operating cost. $/kgal
Amortization @6.25% TCI/yr 0.040
R&M, tax, int.@127. TCI/yr 0.077
Labor + GAO 0.026
Elec.(0.5<:kwh)& gas(35<:MMBTU) 0.010
Coagulants 0.005
Make-up carbon (5%) 0.005
Lime (1000 mg/lCaO)@ $16/t -
A-2
Primary-
granular carbon
Mill A
38
gran.
5.0
25
950
1
yes
no
50
100
0.085
2.5
6.3
0.6
9.4
0.042
0.081
0.026
0.010
0.005
0.021
0.27
2.5
6.3
2.0
10.8
0.049
0.094
0.026
0.010
0.005
0.067
A-ll
Primary/ lime-
granular carbon
Mill A
38
gran.
1.42
25
950
1
yes
no
50
100
0.02 0.085 0.27
4.0
4.0
2.8
0.031 0.032
0.060 0.061
0.034 0.034
0.004 0.004
0.001 0.004
0.067 0.067
4.0
2.8
1.1
7.9
0.036
0.068
0.034
0.004
0.013
0.067
A-3
Primary-bio-
granular carbon
0.02
2.5
3.6
h
j
0.1
11.6
0.052
0.100
0.036
0.009
0.010
0.004
Mill A
38
gran.
4.3
25
950
1
yes
no
50
100
0.085
2.5
3.6
i5"4
)
0.5
12.0
0.054
0.104
0.036
0.009
0.010
0.019
0.27
2.5
3.6
h
)
1.7
13.2
0.059
0.114
0.036
0.009
0.010
0.059
Total operating costs
0.163 0.185 0.251 0.197 0.202 0.222
0.211 0.232 0.287
-------
Table 14. ORDER-OF-MAGNITUDE COST ESTIMATES FOR VARIOUS TREATMENTS OF TOTAL MILL EFFLUENTS FROM MILLS A AND B
Mill A: 140 mg/1 TOC, 1000 CU,; Mill B: 640 mg/1 TOC, 4450 CU
Estimate No.
Treatments
Effluent
Flow, mgd.
Carbon treatment, type
Dosage, Ib/kgal _
Carbon conc'n.,Ib/ft
Retention, hours, carbon
water
Carbon regeneration
Final filter
Treated water comp'n: TOC, mg/1
Color,CU
Capital costs, $ million
Price of carbon, $/lb
Primary
Bio-oxidation
Primary/lime treatment
Filters
Carbon contactors
Carbon dewatering
Carbon drying & regeneration
Carbon in system
Total cost of investment (TCI)
Operating cost, $/kgal
Amortization (36.25% TCI/yr
R&M, tax, int.(§12% TCI/yr
Labor + GAO
Elec.(0.5ckwh)& gas(35cMMBTU)
Coagulants
Make-up carbon*
Lime (1000 mg/1 CaO)@ $16/t
Total operating costs
A-6
Primary-
powd. carbon
Mill A
38
powd.
13
0.1
1.5
1
no
sand
50
100
0.02 0.085
2.5
A-10
Primary/lime-
powd. carbon
Mill A
38
powd.
6.9
0.05
2
2
no
sand
50
100
C-2
Primary-
gran. carbon
Mill B
15.3
gran.
12.5
25
950
1.5
yes
no
50
100
0.02
0.085
0.02
1.5
0.085
1.5
C-3
Primary-
powd. carbon
Mill B
15.3
powd.
36
0.3
1.5
1
no
sand
50
100
0.02 0.085
1.5
-
1.7
0.8
0.5
1.2
negl.
6.7
0.030
0.058
0.035
0.047
-
0.013
_
-
1.7
0.8
0.5
1.2
0.2
6.9
0.031
0.059
0.035
0.047
-
0.055
_
1
1
0
ne
6
0.029
0.056
0.036
0.020
0.010
0.260
_
-
.7
.5
.8
gl.
.5
0.029
0.056
0.036
0.020
0.010
1.100
-
4.
1.
0.
0.
neg
6.
0.030
0.058
0.025
0.001
0.005
0.138
0.067
0
7
5
5
1.
7
0.030
0.058
0.025
0.001
0.005
0.590
0.067
-
_
I"
0.1
5.0
0.056
0.104
0.031
0.010
0.010
0.012
-
-
_
I"
0.5
5.4
0.060
0.112
0.031
0.010
0.010
0.053
-
-
0.
1.
0.
neg
4.
0.046
0.089
0.046
0.020
0.030
'0.720
-
8
0
8
1.
2
0.046
0.089
0.046
0.020
0.030
3.05
_
0.183 0.227
0.411 1.25
0.324 0.776
0.223 0.276
0.95
3.28
(5% for A-8, C-2; 100% for A-6, A-10, C-3)
-------
Table 15. SUMMARY OF CONDITIONS OF TREATMENT AND ESTIMATED CAPITAL AND OPERATING COSTS FOR
VARIOUS TREATMENTS OF TOTAL MILL EFFLUENT FROM MILLS A AND B
NJ
Cost
Estimate
Primary
Clarif n
Bio-
oxid'n
Lime
Activated carbon
3wd.fine8 erar
Final
filtra-
tion
Mill A total mill effluent
1
2
3
4
5
6
7
7A
8
9
10
11
x
X
X
X
X
X
X
X
X
X
X
X
X
R
R
R
NR
75,R
75,NR
150,R
x
x
X
NR
R
Mill C total mill effluent before oxidation pond
1
2
3
x
X
X
R
NR
x
a
b
c
d
e
f
g
Average particle size in microns
Based on minimum price of carbon - $0.02/lb
Based on intermediate price of carbon - $0.085/lb
Same as c, but incl. $0.06/kgal credit for reusable water
Product water contains 570 CU
Product water contains 63 mg/1 TOC
Product water contains 140 mg/1 TOC, 2000 CU
Capital
cost
$mi11ion
Operating Cost
$/kgalb $/kgalc $/kgald
_
-
_
-
-
X
X
X
X
-
X
-
6.1
8.9
11.6
5.5
8.3
6.5
7.4
6.2
6.7
4.0
6.7
6.9
0.12e
0.16
0.21
0.17f
0.23
0.41
0.19
0.37
0.18,
•p
0.14r
0.32
0.20
0.12e
0.19
0.23
0.17t
0.23
1.25
0.24
1.20
0.23,
T
0.14
0.78
0.20
0.12e
0.13
0.17
0.17f
0.17
1.19
0.18
1.14
0.17,
T
0.14
0.72
0.14
3.5
5.0
4.2
0.16»
0.22
0.95
0.166
0.28
3.28
0.16
0.22
3.22
R - regenerated
NR - not regenerated
-------
Compared to the primary-carbon system, the addition of a biological
treatment system requires a substantial added investment while not
substantially reducing the investment requirement in the carbon part
of the treatment sequence. Thus, when aiming at a reuse treatment
system for new mill design, the primary/lime-carbon and the primary-
carbon sequences appeared to be the logical first and second choices
for further development, based on capital cost estimates.
However, when aiming at a reuse treatment sequence adaptable as a
modification for an existing mill with a biological oxidation system,
the choice would appear to become much more complex. It would in-
clude the following options: continued operation of existing facili-
ties, with addition of a carbon system; shut-down of biological oxida-
tion, with addition of a carbon system; shut-down of biological oxida-
tion, addition of a lime treatment system including modification of
the existing primary treatment, and addition of a carbon system.
Based on the limited cost detail on individual steps of the three
treatment sequences presented in Table 13 the mill system modifica-
tion most closely represented by the primary-bio-carbon sequence
could represent the economical choice and thus appeared to merit
further development.
An elimination of any one of these three treatment sequences thus
did not appear warranted on the basis of the capital investment
comparison considering the objective of assuring broadest utility
and applicability of successful development of new treatment systems
to both new and existing mills.
Operating Cost -
Comparing operating costs, including fixed costs based on capital
investment, at a carbon cost of $0.02 or $0.085/lb., the primary-
carbon system represents the lowest cost system. Total operating
costs for the primary/lime-carbon system are 15% above the primary-
carbon system, and the primary-bio-carbon system is yet 10% higher
in cost. These differences are within the accuracy of these estimates
and at most indicate a possible cost trend. In the case of an exist-
ing mill with a biological oxidation system, and where that system is
already fully or partially depreciated, the primary-bio-carbon system
would again tend to compare more favorably. Thus on the basis of
operating cost, using mill-site produced carbon at up to about
$0.085/lb.t no justification was seen to eliminate any of the sequences
from further development under the objectives of this program.
43
-------
Carbon Cost -
The fraction of total capital investment (TCI) represented by the
carbon inventory merits closer inspection. It is seen to be an
insignificant factor at a carbon cost of $0.02/lb., and a fairly
low fraction (about 5%) of TCI at a carbon cost of $0.085/lb.
However, at a carbon cost of $0.27/lb. (equivalent to purchased
carbon), carbon inventory amounts to 13%-19% of TCI. Assuming
successful development of mill-site carbon production at a cost of
$0.02-.085/lb. carbon, system selection is essentially insensitive
to the capital investment cost aspect of the carbon inventory. Carbon
costs above the stated range tend to favor the primary/lime-carbon
treatment sequence.
Total operating costs, including fixed costs based on capital invest-
ments, of all three sequences vary with carbon cost for two reasons:
(a) make-up carbon represents a (variable) direct operating cost item;
(b) the initial carbon charge (inventory) represents a (variable)
capital investment cost item. However, the operating cost of the
primary/lime-carbon system is relatively insensitive to carbon cost.
Over the range of carbon prices examined ($0.02 to $0.27/lb.), the
operating cost of the primary/lime-carbon system increases by only
15%, while the operating cost of the primary-carbon system increases
by 54%. This makes the primary/lime carbon system the less costly
system to operate with (purchased) carbon at $0.27/lb.
The estimated operating costs of the primary/lime-carbon system,
using purchased carbon at $0.27/lb., do not compare too unfavorably
with the primary-carbon system using mill site-produced carbon at
$0.085/lb. or less. The lower indicated investment cost of the
primary/lime-carbon system using purchased carbon may offer suffi-
cient reason in a specific case to give it serious consideration in
comparison with systems offering lower operating costs.
Carbon Adsorption Systems
Table 14 brings out the effect of different carbon systems on capital
and operating costs. It should be noted that the estimates in Table 14
were prepared only for two different carbon costs of $0.02 and $0.085/lb.
The FACET system is represented by estimate A-8. It shows a moderate
increase in direct operating costs over the granular carbon system A-2
(Table 13). This increase is due to higher estimated carbon dosage
and correspondingly higher make-up requirements and to higher regen-
erating utility costs. Comparing the capital cost estimates A-8 and
A-2, the potentially significantly lower TCI of the FACET system is
observed. This lower capital requirement would appear to make the
44
-------
FACET system an attractive alternative to the granular carbon system,
if the lower costs can be substantiated on the basis of further
development and subsequent cost estimates.
The effect of using powdered carbon in the primary-carbon sequence
with no regeneration of the carbon is indicated in estimate A-6, and
may be compared with the corresponding granular carbon system estimate
A-2. Operating costs for A-6 are higher, even at the lowest assumed
carbon cost of $0.02/lb, due to higher carbon requirements in the
absence of carbon regeneration. On the other hand, the TCI for A-6
is lower in the absence of investment costs for carbon regeneration.
However, non-competitive operating costs of the powdered carbon system
provided a basis for ruling it out from further development.
A-10 is an estimate of the powdered-carbon version of the primary/lime-
carbon sequence. A comparison with A-ll (primary/lime-granular carbon)
shows no capital cost advantage of A-10, while operating costs are
substantially higher than A-ll, even at the lowest carbon cost of
$0.02/lb, and in spite of the relatively low carbon dosage required.
Based on this and on the preceding comparison of A-6 with A-2, it
appeared that further development of a powdered, non-regenerative
carbon system was not warranted.
Effluent Contaminant Concentration -
Estimate C-2 in Table 14 is for primary-carbon treatment of Mill C
total mill effluent which is more concentrated than that of Mill A
but has a lower daily flow. In comparison with estimate A-2 in
Table 13, C-2 shows a higher operating cost primarily because of
higher fixed costs per unit volume of water and to a lesser extent
because of increased carbon dosage. Estimates for treating Mill C
effluent by the other two sequences were not prepared, since it was
unlikely that the pilot plant system selection would have been
influenced by these estimates.
Estimate C-3 emphasizes the unattractiveness of a non-regenerative
powdered carbon system as discussed in conjunction with A-6 and
A-10. The higher carbon dosage required in C-3 makes it comparative-
ly even less attractive.
Comparison of Reuse Systems with Other Systems -
The comparison of treatment systems in Table 15 indicates that the
use of carbon treatment (either granular or FACET) might be only
slightly more costly than biological oxidation, A-l, Table 15.
Carbon treatment, and in some situations lime treatment, would
provide a water suitable for reuse. In general, primary-bio treat-
ment will not reduce the color sufficiently and primary-lime will not
45
-------
reduce the TOG, and perhaps color, sufficiently for reuse in the
mill, and will not reduce BOD sufficiently for discharge to rivers
in the future. The cost estimates indicate that lime treatment
would be 15% higher than bio-oxidation. However, firm design and
cost data were not available for the lime treatment process and there
was some doubt as to the reliability of the values used in the esti-
mation of the costs for lime treatment.
FINAL SELECTION OF TREATMENT SEQUENCES AND CARBON ADSORPTION SYSTEMS
The final selection of treatment sequences to be piloted was as follows:
(1) Primary-carbon
(2) Primary/lime-carbon
(3) Primary-bio-carbon
(The actual execution of the primary/lime carbon sequence in the pilot
plant used a holding pond to smooth out concentration variations and
inadvertently acted as a primary clarifier before lime treatment.)
An examination of the economics and consideration of other factors did
not provide a basis for eliminating any of the three treatment sequences
considered. This program's orientation towards both new mills and exist-
ing mills precluded the elimination from development, on economic grounds,
of the primary-bio-carbon system, as discussed previously. Such elimina-
tion, moreover, would require strong justification considering the fact
that the primary-bio-carbon system has the advantage of minimum departure
from established technology.
In the final selection of carbon systems within the above treatment
sequences, the single-use powdered carbon system was eliminated from
further development because of its comparatively high cost. The remain-
ing two systems; i.e., granular carbon and FACET, both indicated certain
comparative economic and technical advantages. In terms of cost, the
granular system indicated somewhat lower total operating cost, while the
FACET system indicated lower capital investment. From a cost estimate
reliability standpoint, the granular carbon system estimate was probably
a more reliable one than that for the FACET system because of the avail-
ability of substantial reference material for the granular carbon systems
compared with the conceptual status of the FACET system.
From the technical standpoint, the FACET system would appear to offer a
better chance of avoiding certain problems that could develop in the
operation of large-scale adsorption systems. Granular carbon columns
tend to develop anaerobic biological activity. In a FACET system, it
would be a relatively simple matter to control anaerobic activity or
46
-------
take advantage of aerobic activity by the injection of air. Another
problem is often caused by suspended solids, leading to plugging and
the need for frequent backwashing of granular carbon columns. Suspended
solids are not expected to be a problem in the FACET system. Problems
of this type can only be properly assessed in a pilot plant. Design
data for a more reliable capital cost estimate for a commercial system
also can only be obtained on the basis of pilot plant operating data.
Hence, it was considered desirable to include both the granular carbon
and FACET system in the plans for the pilot plant.
47
-------
SECTION VI
PILOT PLANT DESCRIPTION AND OPERATION
OBJECTIVES
A pilot plant was designed and constructed that would meet the objectives
of this program: to determine the technical feasibility of the selected
treatment sequences involving carbon adsorption for removal of color and
TOC from unbleached kraft mill effluent, to develop suitable conditions
of operation, and to provide engineering design data for commerical systems,
The pilot plant was designed so that effluent treatment by any of the
following three sequences could be investigated:
Primary-carbon
Primary/lime-carbon (abbreviated lime-carbon)
Primary-bio-carbon (abbreviated bio-carbon)
Although the system selection detailed in Section V provided for inclusion
of the primary/lime-carbon system, i.e., lime addition followed by simul-
taneous removal of lime-color bodies and settleable solids in a primary
clarifier, pilot plant plans were for practical reasons developed on the
basis of lime treatment following primary clarification. (A holding
pond, required to smooth out concentration variations , inadvertently
acted as a primary clarifier before lime treatment.) The design incor-
porated two carbon adsorption systems, i.e., downflow columns and FACET,
and provided for simultaneous operation of both, resulting in a total of
six treatment options.
DESCRIPTION OF PILOT PLANT
Location
The pilot plant was located alongside the effluent ditch from the
St. Regis Pensacola No. 2 Mill that normally produces 620 tons of
unbleached pulp per day from a charge of 25% hardwood and 75% southern
pine. This mill, built in 1948, uses batch digesters and produces
mainly bag paper. The Pensacola mill effluent was selected for use in
the pilot plant because this mill is typical of the more prevalent
older pulp and paper mills. In addition, the Pensacola location affords
the opportunity of testing a wide range of individual and combined
effluents including bleach effluents from the No. 1 Mill.
48
-------
Basis for Design of Pilot Plant
The pilot plant was designed to treat 30 gpm of mill effluent in the
bio-oxidation and lime treatments, 15 gpm in the carbon columns, and
15 gpm in the FACET system. These flows accommodate the desire for
large flows to provide reliable commercial design data and yet are
small enough to keep equipment and operating costs within reasonable
limits. The 30 gpm capacity of the bio and lime units gave sufficient
flow for side-by-side evaluations of the two carbon systems with the
same feed water.
The pilot plant was designed for continuous operation seven days per
week because intermittent operation would not have provided useful
data for evaluation and designs. To provide operation round-the-clock,
it was decided to use operators one shift per day for five days per
week and to provide sufficient automatic controls and instrumentation
to permit continuous operation without operator attendance during the
remaining shifts, thus avoiding the cost and manpower scheduling prob-
lems connected with round-the-clock operator attendance, particularly
for pilot plants.
An aeration basin (187,000 gal) was used for bio-oxidation because
this method is the most commonly used method for bio-oxidation treat-
ment of pulp mill effluents. A retention time of 4.3 days was used in
designing the equilization-aeration basin. The depth of the earthen-
dam rectangular aeration basin (8 ft) and the size of the floating
aerators were selected to make the operation of the aeration basin
typical of full-scale aeration basins.
A large amount of flexibility was incorporated in the pilot plant to
permit any treatment step to be by-passed or to permit simultaneous
operation of as many as three of the six treatment options.
Pilot Plant Equipment
A flow diagram of the pilot plant is shown in Figure 5. Effluent is
pumped from the ditch into a basin having a volume of 187,000 gallons
and providing 4.3 days retention time when operating at 30 gpm. The
basin is of compacted earth construction with the interior sloping
sides coated with gunnite concrete to prevent water erosion of the
sides. The basin serves to dampen the wide fluctuations of impurities,
which may vary as much as 6-fold within a few days. One section of ^
the basin provides primary clarification at a rise rate of 100 gpd/ft
at 30 gpm. Water from the basin passes through slots in a wooden wall
placed across one corner of the basin. The wall consists of vertical
boards with % to % inch spacing between boards. Only the lower one^or
two feet of the slots are open, the remainder is covered with plastic
to minimize water circulation from the aeration section.
49
-------
Ln
O
LIME
1 .
r \\ \
'- + Y CO
,i ' • ,! " i .2
'l -' iiis^alll •- A
i •! !! '••
! ! Jl M "
1 llpiHr, .
ik / M M '
1 V U'1 i: !: Jr-^
j X \i~jy
J-> SLUDGE <-} f clj Q^
ATER CARBONATOR pH
ADJUST-
MENT
5~\ r-Q-, (— 0~^
^ 1 " 1 ' S^
\ UC^ ! // ^
\ \ /
f-
\
\_
73k
j
^T±
FILTER
-4
— 1
^J
^
^
?,
i
i
i
i
1 1-.
/
k
^V
~s
^CT
f /
t
I
i
i
i
i
rpk
ACTIVATED CARBON COLUMNS
STORAGE
TANK
CLARIFICATION
No.2 MILL
EFFLUENT
EQUILIBRATION OR
BIO-OXIDATION BASIN
SPENT
CARBON
ACTIVATED CARBON
CONTACTORS
FILTER SiORAGE
TANK
Figure 5. Effluent treatment pilot plant
-------
The basin has a 5 hp surface aerator which is operated when biological
oxidation is desired as part of the operating sequence. In that case,
nutrients at normally used dosages are also fed to the basin. Further
details on the basin and on each portion of the pilot plant are given
in Appendix C.
Effluent from the basin is pumped either directly to one or both of
the two activated carbon adsorption systems, or it is pumped to the
lime treatment system and then to the activated carbon systems.
The lime treatment equipment consists of a 10 ft diameter reactor-
clarifier with a 39 in. diameter center well, providing a retention
time of 20 minutes in the reactor section and a clarifier rise rate
of 550 gpd/ft at 30 gpm. Lime addition in the reactor raises the
pH of the effluent to about 12 and causes precipitation of a mixture
of CaCOo and a. Ca-organic floe. Lime addition is by pump from a mix
tank into which hydrated lime is fed continuously from a hopper where
it is slurried with about 5 gpm of recycled lime-treated effluent. A
second reactor-clarifier permits carbonation of the lime treated,
clarified effluent for removal of excess dissolved lime at pH 10.5.
In the pilot plant,bottled COo is used for this purpose. At 30 gpm,
this 7.5 ft diameter reactor-clarifier provides a retention time of
27 minutes in the reaction zone and a clarifier rise rate of 1000 gpd/
ft . A 200 gallon tank is provided for final pH adjustment with CC^.
The lime treatment unit, the carbonation unit, the pH adjustment tank,
and a 10,000 gal surge tank are located on an outdoor concrete pad.
The remaining equipment items are located inside a 30 x 52 ft pre-
fabricated steel building having a 20 ft wall height. The building
also provides storage for truck-load quantities of lime and carbon.
Two activated carbon adsorption systems are provided in the pilot plant,
intended for simultaneous comparative operation if desired, both at
15 gpm. One system consists of four standard downflow carbon columns,
3 ft diameter by 15 ft high, each with a carbon bed of 10 ft (1600 Ib
of granular carbon of 20 x 40 mesh size). At 15 gpm, each column
provides 0.6 hour retention time in the carbon bed at a superficial
velocity of 2 gpm/ft . All water connections to the columns are by
means of 1% in. hoses with quick disconnects so that any desired
arrangement of series or parallel flow could be used, either up-flow
or down-flow through the columns. This arrangement greatly simplified
the piping, especially for backwashing.
A 4-ft diameter duomedia filter with a filter bed of 1 ft of anthracite
and 1 ft of sand, supported on graded gravel, is provided for filtering
the feed to the carbon systems when desired.
51
-------
A second carbon system, a novel system called FACET (for Fine Activated
Carbon Effluent Treatment) and subject of a U.S. patent (41), was in-
cluded in the pilot plant for initial development (Figure 6). The
FACET system is a multi-stage counter-current agitated system with
continuous counter-current transfer of carbon and liquid from stage to
stage. The tanks used in the pilot plant are sized 3 ft diameter by
5 ft high, with a mixing zone of 170 gallons and a carbon charge of
212 Ib per tank at 15 g carbon per 100 ml of slurry, or 15 wt/vol %.
Water to be treated flows by gravity through the three stirred tanks
in series and then to a product storage tank or is pumped through a
3-ft diameter duomedia filter and then to the product water tank.
Activated carbon, which has a mesh size of 40 by 140 mesh, is fed from
a feed hopper into the third FACET tank. Carbon slurry, which typically
has a concentration of 15 wt/vol %, is removed from the third tank at
the same rate as the carbon is fed to that tank. The carbon is thus
moved counter-current to the flow of water through the second tank,
through the first tank, and to a spent-carbon container.
The system endeavors to avoid some of the problems associated with
the conventional granular and powdered carbon contacting systems,
while combining advantages of both of these systems by using a carbon
size intermediate to the standard powdered and granular classifications.
Adsorption rate being inversely proportional to the square of the carbon
particle diameter (12,13), the intermediate size carbon loads up to
its capacity much faster than the coarser granular carbon. Although
the ultimate capacities of both granular and intermediate size carbons
are essentially equal, the higher adsorption rate of the intermediate
size carbon results in a corresponding increase in carbon turn-over
rate and a corresponding decrease in carbon inventory. Equipment size
for the adsorber decreases correspondingly.
The FACET system allows close control of the quality of the treated
effluent by adjustments in carbon transfer rate or carbon slurry
density, or both. Continuous discharge of spent carbon permits a
closer tie between regenerator capacity and average carbon throughput,
eliminating costly offstream inventory of carbon. Suspended solids
present in the liquid to be treated do not present a plugging problem
as in the granular carbon column system, hence primary clarifiers may
be omitted.
Comparing FACET to the powdered carbon system, the intermediate size
carbon has a higher settling rate than the powdered carbon. This
obviates the need for flocculants and substantially reduces the clar-
ification area compared to that of a powdered carbon system (clarifi-
cation area being a function of liquid flow rate and particle sedimenta-
tion rate). Use of the intermediate size carbon allows a properly
52
-------
EFFLUENT
CO
r-ACTIVATED CARBON
SPENT
CARBON CONTACTORS
FILTER STORAGE
TANK
Figure 6. FACET system
-------
baffled agitated tank to serve as both adsorber and clarifier. As a
result, equipment size as well as cost is significantly lower for the
FACET system than for the powdered carbon adsorption system.
Figure 7 shows the pilot plant building, the lime-treatment equipment,
and equalization basin. Figure 8 shows the carbon columns, the duo-
media filter in the right foreground, and the FACET tanks in the front
foreground.
Instrumentation
Instrumentation of the pilot plant is sufficient to permit continuous
24-hour operation with only one operating shift for adjustments and
sample taking. Conductivity, pH, turbidity, temperature, and flow
rate are recorded continuously.
Three water flow rates are automatically controlled and recorded: feed
to lime treatment, feed to carbon columns, and feed to FACET carbon
system. Automatic controllers are provided on the CC>2 feed lines to
the carbonator reactor-clarifier and pH adjustment tank to control the
pH in these tanks at a selected level. Timer controllers are provided
for the sludge valves from the lime treater and the carbonator.
The monitoring of the qualities of 12 streams in the process is pro-
vided by a step programmer at the instrumentation-control center of the
pilot plant. Lines from 11 sample points and a tap-water line are
brought to the continuously operated sampling instruments through
programmer controlled solenoid valves. The programmer opens one sample
line at a time in sequence and the water from that sample point flows
in series through a conductivity cell, pH measuring cell, a Hach light-
scattering turbidimeter, and a Bausch and Lomb Spectronic-70 spectrophoto-
meter for measurement of color. The signals from these four instruments
and a thermocouple are conditioned and recorded on a twelve-point record-
er during a five-minute measuring and printing cycle for the sample
stream that is on. The stream number that is being measured is identified
on the recorder. After the five-minute period on one stream, the pro-
grammer steps to the next sample stream that is to be analyzed. The
programmer can be modified to permit analysis of only the sample streams
of real interest, which reduces the overall cycle time. This programmed
stream analysis system permits the above five important parameters to
be measured and recorded during the unattended evening shifts and on
weekends. When an operator is present, the stream analysis system is
valuable for showing trends and rate of change which permits better
control of the operation. The pilot plant was built and shake-down
runs made during 1971. The total cost of the pilot plant, excluding
the building, was $194,000.
54
-------
Ln
i ,i
Figure 7. Exterior of pilot plant showing lime treater,
carbonation tank, storage tank, and basin
-------
Figure 8. Interior of pilot plant with carbon columns,
FACET tank (foreground) and duomedia filter (right foreground)
56
-------
OPERATING PROCEDURES
The normal procedures used in operating each part of the pilot plant,
and the sampling and analytical procedures are discussed in this portion
of the report. Procedures that were unique to a given treatment sequence
are discussed in the sections on each treatment sequence. A simplified
flow diagram is given in Figure 5.
Bio-oxidation and Clarification
The operation of bio-oxidation in the aeration basin was quite simple
and required little attention. Water was pumped continuously from
the mill effluent ditch through a 4-mesh strainer that was placed
beneath the surface of the water in the ditch. At times when there
was quite a bit of fiber in the effluent, the screen would accumulate
solids and would require cleaning with a brush. The water was pumped
into a basin at a flow rate of about 4 gpm greater than was being
used in the pilot plant. The excess water flowed through a constant-
head overflow attached to the drain line from the bottom of the basin.
The water being pumped from the mill effluent ditch was discharged into
the basin at the end opposite the withdrawal point. The retention
time of water in the basin during bio-oxidation was 6.5 days, based
on a flow rate of 20 gpm. Nutrients (ammonium nitrate and phosphoric
acid) were added once per day in the ratio of 2.5 Ib N and 0.5 Ib P
for each 100 Ib of BOD in the incoming water, which is the usual rate
of addition in southern kraft mills. The 5 hp surface aerator was run
continuously which kept the dissolved oxygen at about 6 mg/1. (A spare
aerator was also placed in the basin for use if needed.)
Primary Clarification
When the operating sequence called for primary clarification without
bio-oxidation, the aerator was turned off, nutrient feed was discontinued,
and the aeration basin then served as an equalization basin to help
level out the hourly and daily fluctuations of impurities in the mill
effluent. The water from the equalization basin passed through the
clarifier section of the basin and was pumped to the next treatment step.
Lime Treatment
When the operating sequence called for lime treatment, the aeration
basin was again used as an equalization basin, with aeration and
nutrient feed turned off, to level out impurities concentrations in
the effluent. The water also passed through the clarification zone,
although clarification was not called for at that point of the intended
57
-------
treatment sequence. Water from the clarification zone of the basin
was pumped through a flow controller to the center mixing zone of
the lime treater where a dilute slurry of lime was also fed continuously.
Bagged hydrated lime was charged to a hopper on the feeder every one
to three days. An oscillating screw feeder dropped the lime at the
selected dose rate into a mixing tank to provide the slurry that was
pumped to the lime treater. Water for slurrying the lime was recycled
from the lime treater through a flow meter and a float valve which
maintained a constant level of water in the slurry tank.
The rate of feed or dosage rate of lime was manually set at a constant
value for periods of at least several days and was varied only when
attempting to relate color reduction to dosage. Later it became evident
that for microlime treatment, the lime dosage should be varied to main-
tain a constant concentration of soluble calcium. Without a suitable
continuous monitor for soluble calcium, the control of dosage to main-
tain a constant soluble calcium content was barely satisfactory.
The mixture of lime and water in the axial cylindrical mixing chamber
of the lime treater was agitated by a slow-moving multiple-arm stirrer.
The water moved from the bottom of the mixing chamber upward through
the clarification zone and overflowed to a pipe leading to the carbonator.
Sludge that settled to the bottom of the lime treater could be discharged
to the drain through a timer-controlled valve but was normally withdrawn
once per day to a 50-gallon tank for measurement and sampling. Later
in the program, it was found that better lime treatment was obtained
by circulating sludge at about 5 gpm to the top of the mixing chamber
to increase the solids content of the water in the mixing chamber.
Carbonation
The carbonator was of the same design as the lime treater except that
the clarification zone had half the volume arid cross-sectional area
of the lime treater, which provided twice the rise rate of the lime
treater. Water, which flowed by gravity from the lime treater, was
treated with cylinder C02 in the mixing zone to remove the excess
lime in the water as CaCO-j. The C02 was fed from two banks of cylinders
through pressure regulators, a flow control valve, and a flow indicator
to a turbine-type gas disperser placed about 4 ft below the surface of
the carbonator mixing chamber. An inverted-trough C0? bubbler located
at the bottom of the mixing chamber was used initially for admitting
the C02 to the water but the efficiency of absorption was very poor.
The turbine type of CC^ disperser gave very good absorption efficiency.
Sludge recycle to the top of the carbonator mixing chamber was later
added by use of an air lift. Sludge from the carbonator was discharged
once per day to the sewer.
58
-------
pH Adjustment
Water from the clarification zone of the carbonator flowed by gravity
to a 200-gallon tank in which C02 was admitted through a turbine
disperser to adjust the pH to the desired level. The rate of admission
of C02 to the carbonator and to the pH adjust tank was controlled
automatically to the desired pH as measured by pH sensors in these tanks.
Carbon Column Adsorption
Water from the lime treatment unit, from bio-oxidation, or from primary
clarification alone was pumped either directly to the carbon columns
or through a duomedia filter and then to the columns. The duomedia
filter operated quite satisfactorily if it was backwashed every one
or two days for about 10 minutes at about 6 gpm/ft . Less frequent
backwashing of the filter resulted in the formation of mud balls in
the filter which were not broken up during backwashing. Backwash
water was provided by well water stored in the 10,000 gal surge tank.
Each column was charged with 1600 Ib of Atlas Darco 20 x 40 mesh
activated carbon for all runs, which gave a bed depth of 10 ft and
a bed volume of 70.7 ft . The columns were operated in the down-flow
mode during all of the pilot plant runs. The water from the bottom
of the first column flowed through the hose inter-connections to the
top of the second column, and so forth, for as many columns as were
to be used. The water from the final column flowed through a vertical
riser pipe to an overflow equal to the height of the columns. This
overflow device insured that the columns always remained full of water.
The water pressure after each column was measured and the pressure
drop across each column was checked several times each day and was
recorded daily. At the relatively low flow rates used (0.7 to 2.1 gpm/
ft ), the pressure drop across all four columns remained at low values
of 2 to 8 psi. Sample taps were available on each column at 2-ft
intervals, but samples for analysis were normally obtained only after
each column.
The columns were backwashed every other day with well water stored in
the 10,000 gal surge tank at 35 gpm (5 gpm/ft ) for 20 min. Before
the backwash was started, compressed air was admitted to the bottom
of each column to thoroughly agitate the carbon bed and break up any
mud layer that might be forming on the top of the carbon bed. The
backwash water was discharged to the sewer but, in a commercial plant,
it would be clarified and reused for backwashing or added to the feed
to the columns.
59
-------
When the quality of the product water from the columns became poorer
than the selected limits, the first column was removed from the system
and a column with a fresh charge of carbon was added as the last column.
The second column then became the lead column. The carbon from the
column that was removed from operation was discharged as a slurry by
opening a 6-in hand-hole at the base of the carbon bed (top of the
sand-gravel support bed). The spent carbon was discarded. Fresh
carbon from bags was recharged to the column through a manhole at the
top after water was added to the column. Bagged carbon was hoisted
by an electric hoist to a charging platform at the top of the four
columns.
FACET Adsorption
Water from the basin or from lime treatment was pumped through a flow
control valve to the first of the three FACET contacting tanks. - The
water flowed by gravity through the three tanks to a product storage
tank or through a pump and duomedia filter. The speed of each of the
slow-speed agitators in the FACET tanks was adjusted to barely keep
all of the carbon suspended.
Carbon to FACET (Atlas Darco 40 x 140 mesh XPT activated carbon) was
charged to the feed hopper from bags and fed by a feeder and vibrating
conveyor to the third FACET tank. The usual rate of carbon addition
for treating 10 gpm of lime-treated water was only about 2-4 Ib/hr.
The carbon was moved at this same rate through the three tanks counter-
current to the water flow, while the concentration of carbon in each
tank was maintained at the desired level (10-15 wt/vol %) by adjusting
the flow of carbon from each tank. The carbon was transferred as a
slurry from tank to tank by means of air lifts which performed reliably.
Originally, we tried rotary-vane pumps having neoprene impellers for
transferring the carbon slurry but these were not satisfactory. The
solids concentration in each tank was measured several times during
the day shift, and the rate of transfer from each tank was adjusted
to give the desired solids concentration by adjusting air flow to the
air lift. Spent carbon from the first FACET tank was collected, drained
of water, weighed, and sampled for determination of moisture content.
As indicated in Figure 6, baffles were provided in each tank to provide
a clarification zone at the top of each tank so that no carbon would
be carried by the water to the succeeding water treating tank. The
baffles were inclined at an angle of 45° to allow the settled carbon
to slide back into the agitated zone. These baffles had sufficient
openings around the edges that some turbulence from the mixing zone
persisted into the clarification zone which allowed some of the carbon
to be carried by the water in the wrong direction. This lack of good
60
-------
clarification was particularly troublesome in the third FACET tank
because 'some of the fresh carbon was lost from the system in the
product water. The rise rate in this zone was 1.4 gpm/ft2 at 10 gpm,
which allowed the 140-mesh particles to remain in the tank since they
settled at a rate of about 18 ft/hr which is equivalent to 2.2 gpm/ft .
SAMPLING AND ANALYTICAL PROCEDURES
Samples were taken once per day during most of the pilot plant runs.
Even though most of the properties of the water at each stage of the
treatment varied only slightly during a 24-hour period, some streams
were sampled more frequently if the changes were more rapid or if a
special run was being made. The previously described automatic sensing
and recording of pH, conductivity, temperature, color, and turbidity
provided a constant readout of these properties for any number of sampling
points up to twelve.
Samples taken each day were analyzed in the research laboratory using
the procedures and analytical equipment described in Appendix B. The
properties determined in the analyses and the usual frequency of analysis
are given in Table 16. In normal operation, we sampled and analyzed
seven streams during operation of the bio-carbon sequence, six during
the primary-carbon sequence, and eight during the lime-carbon sequence.
When the FACET unit was operated, an additional 3 or 4 samples were
analyzed. Most of these were analyzed daily but some much less frequently,
as indicated in Table 16.
61
-------
Table 16. PROPERTIES OF WATER ROUTINELY MEASURED
Property Frequency of analysis per week
Color, FPCU (paper filter) 7
MPCU (0.8 micron Millipore filter) 7
CU (standard color) 7
TOG (paper filter) 7
TIC (paper filter) 7
pH 7
Temperature 7
Turbidity 1-7
Conductivity 2-7
Suspended solids 1
Total solids 1
Total dissolved solids 1
BOD-5 1
Sulfate 1
Calcium 7
Volatile neutral compounds 1
Volatile organic acids 1
Foam tendency 1
Adsorption isotherms after each step of treatment 1 per month
Metals(Mg, Fe, Na, K, Al, Cr, Ni) about 2 per month
Solids in lime sludge 1
Calcium in lime sludge 1
62
-------
Based on pilot plant experience, it was found desirable to determine the
parameter of "color" not only by the industry standard APHA-NCASI method
but also by modified procedures which measure apparent color. In the
APHA-NCASI method, color is measured at 465 nm on a spectrophotometer
after adjustment of pH to 7.6 and filtration through a 0.8 micron Millipore
filter. All values reported in this report unless otherwise designated
were obtained by this method, and are reported as "CU". One modified
procedure involved 0.8 micron Millipore filtration prior to adjustment
to pH 7.6. (Used after lime treatment, this procedure ensures more
complete removal of suspended solids.) Another modified procedure con-
sisted of measuring color after filtration through a Whatman No. 2 filter
paper and adjustment to pH 7.6, referred to as "7.6 FPCU". A third
modified procedure consisted of measuring color after filtration through
Whatman No. 2 filter paper without pH adjustment, noted as "FPCU". A
fourth modified procedure consisted of measuring color after filtration
through an 0.8 micron Millipore filter without any pH adjustment, noted
as "MPCU".
The 7.6 FPCU for the different treatment sequences were higher than the
standard colors by the following factors: bio-carbon, 2.0; primary-
carbon, 1.7; lime-carbon, 1.7. Apparent colors (FPCU) measured with
pH adjustment (in the pH range of approximately 8 to 11,5) were higher
than pH-adjusted colors by a factor of about 2.0.
The color of effluents from pulp and paper mills decreases as the pH is
reduced from pH 12 to pH 7.6, which is the desired pH for discharge to
streams. As the pH is further reduced the color decreases and then
drops rapidly at pH 4 and below, at which point the lignin compounds
start to precipitate from solution and are most completely precipitated
at pH of about 2.
The ratios of FPCU/MPCU and also FPCU/TOC were used to indicate the
concentration of large colloids and agglomerated color bodies in the
water that passed through a paper filter but were retained by the 0.8
micron Millipore filter.
MANPOWER REQUIREMENTS
The operation of the pilot plant normally required one operating
engineer, one operating technician, and one analytical technician for
40 hr per week. Analysis of data, planning, supervision of support-
ing laboratory studies required 40 hr per week of an experienced
development engineer. Maintenance, especially for the pH controllers
and continuous water analyzers, required a craftsman about 8 hr per
week.
63
-------
SECTION VII
OPERATING RESULTS:
PRIMARY CLARIFICATION AND BIO-OXIDATION
BIO-OXIDATION
The bio-carbon treatment sequence was the first to be evaluated in the
pilot plant after shake-down runs were made to check out each treatment
stage. The procedure used in operating the aeration basin for biological
oxidation was discussed in Section VI.
The aeration basin was operated continuously from December 1, 1971 to
March 18, 1972, with the primary purpose of providing bio-treated water
as feed to the carbon columns that would be typical of bio-treated water
from aeration basins used by a large number of pulp and paper mills.
The bio-carbon sequence of treatment would provide operating and cost
data that could be used by those mills to evaluate the treatment of the
water from their aeration basins by carbon adsorption to provide water
suitable for reuse in the mill.
The rates of removal by bio-oxidation of BOD and TOC in the basin were
only slightly less than those obtained by aeration basins in the pulp
and paper industry. At the usual flow rate of mill effluent through the
basin of 20 gpm, the retention time in the 187,000-gal basin was 6.5 days.
The 5-hp surface aerator maintained an average dissolved oxygen content
of 6 mg/1. The temperature of the incoming water was about 115°F, and
the average temperature in the basin was 74°F. Average results from bio-
oxidation are given in Table 17 for the period of February 4, 1972 to
March 11, 1972, which was the period used for evaluating the results
from the bio-carbon sequence.
64
-------
Table 17. BIO-OXIDATION AND CLARIFICATION FOR PERIOD OF
2/4/72 to 3/11/72 AT FLOW RATE OF 20 GPM
Percent
To basin From basin removal
BOD, mg/1 278 69 75
TOC, mg/1 278 139 50
Color, FPCU 1610 1329 17
MPCU 1330 871 35
CU 950 810 15
pH 10.2 8.6
Temperature, F 116 74
Conductivity, microtnhos 1580
Turbidity, JTU 34 23 32
Suspended solids, mg/1 422 202 52
Methanol mg/1 35 4.1 88
Acetaldehyde, mg/1 9 10 0
During this test period, the average removal of BOD was 75%, the removal
of TOC was 50%, and the removal of standard color was 15%. For compar-
ison, the removals in the aeration basins treating the total effluent
from the Pensacola mill for a seven-month period were 92.5% of the BOD
and 28% of the TOC, and the color increased 7%. Most mills find no
color reduction in bio-oxidation and less reduction of TOC than was
obtained in the pilot plant. No explanation can be given for these
differences in pilot plant and mill results, other than pointing out
that the pilot plant bio-oxidation basin received raw effluent, while
the mill bio-oxidation basin ponds receive clarified effluent.
Bio-oxidation and incidental air stripping removed 88% of the methanol
but none of the acetaldehyde, on the basis of a single set of samples
taken during this period of operation. In this set of samples, there
were no significant changes in the concentrations of metal ions.
The concentrations of color, TOC, and BOD in the mill effluent as it
entered the aeration basin varied widely - as much as four fold from
one 24-hr composite sample to another. Such surges were partially
dampened by the 6.5-day retention time in the basin, but these surges
still caused wide swings in color and TOC concentrations in the bio-
treated water fed to the carbon columns, as will be discussed later
under carbon column operation.
65
-------
PRIMARY CLARIFICATION
The basin was used during the period of April, 1972 to December 22, 1972
as an equalization basin and a primary clarifier to provide water to the
primary-carbon and the lime-carbon sequences of operation. Mill effluent
water was pumped into the basin at a flow rate about 4 gpm greater than
was used for the pilot plant, with the excess over-flowing to waste. The
aerator was not used. A low degree of mixing in the basin was provided
by jetting the incoming water into the basin. Most of the solids in the
incoming water settled in the basin, but some further clarification took
place in the clarifier section of the basin.
During the period in which the basin was used for equalization and
clarification, it was noted that there was anaerobic activity presumably
because of the 1-ft to 2-ft layer of sludge (primarily bark fines and
fiber) on the bottom of the basin that had accumulated during the bio-
oxidation operation. On the basis of 12 analyses, the passage of the
water through the equalization basin caused the average methanol concen-
tration to decrease from 45 to 33 mg/1, for a reduction of 277o, and the
average acetaldehyde concentration to decrease from 22 to 17 mg/1, for
a reduction of 23%, the formic acid concentration to decrease from 10 to
7 mg/1, and the lactic acid concentration to decrease from 1.5 to 0.7 mg/1.
The glycolic acid content remained at about 10 mg/1, and the acetic acid
content increased from 26 to 92 mg/1, for an increase of 254%. Presumably,
this large increase of acetic acid resulted from anaerobic biological
growth in the basin. The TOG contributed by these volatiles increased
from 48 to 79 mg/1, an increase of 65%. As discussed later, these volatile
neutral and acidic compounds contributed substantially to the TOG remain-
ing in the water after carbon adsorption. There was no significant
decrease of color, TOG, or BOD on the basis of eight sets of samples, but
accurate values of changes in these parameters were difficult to get be-
cause of the 5-day time lag of the water going through the basin.
The average amount of suspended solids in the water to the basin was 422
mg/1 during the bio-carbon operation but was only 103 mg/1 during the
lime-carbon operation from September through December, 1972. The amount
of suspended solids in the water from the settling basin ranged from 48
to 265 and averaged 145 mg/1. The average turbidity from the settling
basin during the primary-carbon and lime-carbon operating periods was
22 JTU.
66
-------
SECTION VIII
OPERATING RESULTS:
LIME TREATMENT AND CARBONATION
OBJECTIVES AND DESCRIPTION OF OPERATIONS
The lime treatment --carbon adsorption (lime-carbon) sequence was operated
from June 30, 1972 to December 22, 1972. A flow rate of 22 gpm was used
most of the time and 11 gpm the rest of the time. The lime-treated water
was used as feed to the carbon columns at 10 gpm and for FACET carbon
adsorption at 10 gpm during most of the period. The lime dosage ranged
from 318 to 980 mg/1 CaO. This dosage compares with 750-1800 mg/1 CaO
used or practiced by other pulp and paper companies (5,14,18) as the so-
called minimum lime treatment.
Results obtained in the pilot plant in the first few weeks of operation
on this treatment sequence, and related results obtained earlier in
laboratory inveatigations, had given indications that a color reduction
to the 100 CU range by a lime-carbon treatment sequence might be achieved
most economically by using a lime dosage substantially below that required
to achieve maximum color removal in the lime treatment step. Indications
were that the lime dosage might be reduced to the level where the lime-
carbon treated water could be reused in the mill without prior removal
of soluble calcium by carbonation, thus reducing lime treatment capital
and operating costs. This system was dubbed "micro" lime (microlime)
treatment, as compared to the "minimum" lime (minilime) treatment in
which carbonation for calcium removal is required. The full results
obtained in the pilot plant in fact support the microlime - carbon treat-
ment as the treatment of choice to achieve treated effluent quality of
100 CU.
Because of the promising results obtained with microlime treatment follow-
ed by carbon adsorption without carbonation of the lime-treated water,
most of the data obtained on lime treatment was with low lime dosages of
350 to 700 mg/1 of CaO. This microlime period of operation was from
July 16, 1972 through December 6, 1972. Carbonation of the microlime
treated water was not used because the residual calcium concentration
at about 80 mg/1 Ca was considered to be low enough for reuse. In
December 1972 the minilime treatment process was evaluated at a dosage
of about 970 mg/1 of CaO. Carbonation of the minilime treated water was
used since it would be required in mill operations to recover the soluble
calcium and to make the water suitable for reuse in the mill. Carbon-
ation with lime kiln gas removes soluble calcium in one lime treatment
installation (14), while natural carbonation accomplishes calcium removal
in another (5,18).
67
-------
In the course of pilot plant operations, it was found that color removal
in the carbon colums as well as in lime treatment is directly influenced
by the concentration of soluble calcium in the water from the lime treat-
er or to the carbon columns, while lime dosage itself exerts only an in-
direct influence. However, lack of automatic feed-back control based on
soluble calcium dictated continuation of operations based on a predeter-
mined lime dosage. This led to swings in the soluble calcium with chang-
ing calcium demand of the effluent based on fluctuations in carbonate
ion and color concentration. The control of residual soluble lime in
the water from microlime treatment was difficult, not only because of
variations in concentrations of carbonate ion and color in the feed water
to lime treatment, but also because of long time lags in receiving ana-
lytical results on the calcium content of the water. About 40% of the
lime added in the microlime treatment was used in converting carbonate
ion to CaC03, but since the carbonate concentration varied as much as
25% in a 24-hr period, the demand for lime for this reaction also varied
by 25%. In general, the color in the feed water increased at the same
time as the carbonate increased, which caused additional fluctuations
in the demand for lime.
Other factors, largely mechanical in nature, in both the lime treater
and carbonator caused occasional upsets in the operation. Most of these
operating problems were overcome quickly, and changes in equipment or
procedure were made to prevent reoccurrence of the same problems.
REMOVAL OF COLOR AND TOG IN LIME TREATER
Data Analysis
As a result of the variations in the composition of the feed water to
lime treatment and variations due to mechanical problems, there were
few periods for which the effect of independent operating variables
could be related to the dependent variables without careful analysis
of the data. The dependent variables primarily used in judging perfor-
mance were color and TOG concentrations of the treated water. The
data were analyzed by several means. For a typical day, 30 to 40 items
of primary data were logged from operating log sheets and laboratory
analysis sheets onto a summary table. These data were also inserted
in a computer program which gave a print-out of all primary as well as
many secondary calculated values (see Appendix F). Most of the primary
items of data were also plotted each day on elapsed-time graphs. These
graphs permitted visual correlations of many of the variables. One
method of analyzing the data consisted of selecting periods of operation
having similar conditions of operation and grouping these into "runs".
There were 15 such runs which lasted for periods of 3 to 20 days each.
68
-------
Another method of data correlation consisted of computer-plotting all
values of one property against another property suspected of showing
some relationship or of providing more insight to the operation. A
third method consisted of obtaining correlation coefficients by computer
programs for nine independent versus nine dependent variables and run-
ning multiple-regression analyses of these variables. Since many of
the dependent variables were largely influenced by the soluble calcium
content of the treated water, the data for both lime treatment and
carbonation steps were divided into three ranges of calcium concentra-
tions. By this means, this important variable (calcium concentration)
could be held essentially constant while the effects of other variables
were compared.
Results of Color and TOG Removals in Lime Treater Alone
The results from the treatment of mill effluent with lime in the lime
treater without carbonation are given in Table 18. All results, ex-
cept for days when there were equipment failures and upsets, are aver-
aged under low, medium, and high concentrations of calcium in the
product water from the lime treater. The average calcium concentrations
in the three ranges were 67, 138, and 228 mg/1. Also, results obtained
closer to the preferred microlime level of 80 mg/1 have been averaged
from the applicable runs in the low and medium calcium range.
As seen in Table 18, the calcium concentration had little relationship
to the dosage of lime. This lack of correspondence was caused primarily
by variations of Na2CO.j (or TIC) and also by variations in color content
of the mill effluent to the lime treater. As indicated in Table 18,
the total inorganic or carbonate carbon (TIC) of the feed water happened
to be high during the low-calcium runs and low during the high-calcium
runs. Lime treatment reduced the average color of the feed water 63%
when the product water contained 67 mg/1 of Ca, 68% with 71 mg/1 Ca,
78% with 138 mg/1 Ca, and 81% with 228 mg/1 Ca. Lime treatment also
reduced TOC by 34 to 40% on the average for these ranges of soluble
calcium. The BOD content was reduced only 9-13%, which shows that the
compounds making up much of the BOD are not very susceptible to removal
by lime treatment.
69
-------
Table 18. AVERAGE RESULTS FROM LIME TREATMENT
Number 'of days averaged
Lime dose, as mg/1 CaO
as mg/1 Ca
Calcium (soluble), mg/1
range, mg/1
Color, FPCU
cua
removed, CU %
TOC, mg/1
removed, 7°
TIC, mg/1
BOD, mg/1
removed, %
pH
Conductivity, micromhos
Conductivity increase
Turbidity, JTU
Volatile organic compounds
methanol, mg/1
acetaldehyde, ing/1
acetic acid, mg/1
Sludge density, g/ml
Sludge, % dry solids
Ca tied up by organics, mg/lc
Atoms of C/atom Ca in organics
Medium calcium
36
478
34
443
342
feed
9
1446
712
248
64
165
9.7
1225
19
32
30
115
prod.
67
17-83
1071
262
63
163
34
14
151
9
11.8
2245
1020
53
34
28
105
1.40
24
117
2.4
feed
9
1310
761
206
59
187
9.8
1471
18
37
29
70
317
prod.
71
38-103
977
243
68
139
33
12
_
-
11.8
2819
1348
45
.
_
-
-
-
98
2.3
feed
15
1670
956
275
52
261
10.1
1593
24
36
21
85
69
528
377
prod.
138
103-164
975
211
78
169
39
11
276
6e
11.9
3597
2004
49
41
19
59
1.14
25
117
3.0
16
576
412
feed
13
1386
842
246
34
416
10.1
1247
33
12*
3*
46b
prod.
228
172-422
770
156
81
146
40
11
348
16
12.0
3749
2502
72
"H
6b
42b
1.27
35
120
2.8
a
b
c
d
In the product water, filtration preceded adjustment to pH 7.6 in the standard color measurement.
Results from a single sample.
Ca added plus Ca in feed water, minus Ca in product water and Ca in CaCO-j formed with TIC.
Atoms of TOC removed per atom of Ca tied up by organics.
Increase
-------
Influence of Soluble Calcium on Removals
The influence of calcium concentration on percent removal of color in
lime treatment and on color remaining is shown in Figure 9. Each point
is for a run which consisted of 3 to 20 days during which conditions
were fairly constant. The results show that when the average soluble
calcium concentration was 80 mg/1, the color was reduced by about 70%
and that the calcium content had to be doubled to increase the removal
to 80%. The color was reduced to about 230 CU at 80 mg/1 Ca and to
about 100 CU at about 400 mg/1 Ca (which is the calcium content from
minilime treatment before carbonation). The TOG data in Table 18 show
that removal of TOG in lime treatment was influenced somewhat less than
was color removal by the soluble calcium concentration.
The influence of calcium concentration on removals of color and TOG
are also discussed in Section XI under results from carbon adsorption
following lime treatment. In the lime-carbon sequence it was found
that the soluble calcium content had to be at least 60 to 80 mg/1 to
achieve good removals of color and TOG in the carbon columns.
Spruill (14), in full-scale lime treatment of kraft pulp and paper
effluent by Continental Can Co., found a similar influence of calcium
content on color removal. The Continental Can Co. uses minilime treat-
ment and a soluble calcium concentration of about 400 mg/1 Ca to pro-
vide reliable operation and satisfactory removals of color. (In this
plant the excess calcium in the water is recovered by carbonation.)
The conductivity of the water was increased by the addition of lime,
primarily from the increased NaOH content from the reaction of Ca(OH)2
with Na2C03 but also from the increased concentration of Ca(OH)2« The
solubility of Ca(OH) (about 1500 mg/1) was not exceeded by the lime
dosages used in this program, and therefore essentially all of the lime
not used in reaction with organic compounds or ^2663 went into solution.
Any free Ca(OH)2 found in the sludge at these applied lime dosages is
a result of insufficient mixing or time for dissolution. The percent
increase of conductivity in lime treatment correlated fairly well with
the soluble calcium concentration of the treated water, as shown in
Figure 10. When lime was added to give a soluble calcium concentration
of 80 mg/1, the average conductivity increased from 1600 to 3200 micro-
mhos, for an increase of 1600 micromhos or 100%. The relationship of
soluble calcium to conductivity increase might be useful as a rough
control of the lime dosage to give a desired concentration of calcium
in the lime-treated water.
The turbidity of the feed water ranged from 15 to 35 JTU and that of
the lime-treated water averaged about 58 JTU (see Table 18). The
turbidity of the product water was largely due to slowly settling floe
that was not removed in the clarifier section of the lime treater.
71
-------
100
S3
IJ
>
Q
S
O
Pn
O
30 -
20 -
- 400
- 300
P4
3
o
CJ
- 200
CO
- 100
100 200
SOLUBLE CALCIUM, mg/1
300
400
Figure 9. Influence of calcium in lime-treated water on
percent of color removed and on color remaining
-------
450
400
MEAN
RANGE OF 2/3 OF VALUES
100 200
% INCREASE OF CONDUCTIVITY
300
Figure 10. Soluble calcium in lime-treated water versus
increased conductivity due to lime treatment
73
-------
Lime treatment removed very little of the volatile organic compounds,
as indicated by the results in Table 18. This result is not surprising
since these compounds would not be expected to form insoluble calcium
compounds. As discussed in Section VII, it is believed that most of
these volatiles were produced by anaerobic activity in the equalization
basin. These compounds contributed about 60 out of the 169 mg/1 TOC
in the product water from the medium-calcium runs.
SLUDGE FROM LIME THEATER
The solids from the lime treater were found to settle rapidly and to form
a dense sludge. The average sludge, as withdrawn from the lime treater
each morning, contained 28% solids. The solids content of the sludge
varied almost linearly with the specific gravity of the sludge as shown
in Table 19.
Table 19. RELATION OF SPECIFIC GRAVITY AND SOLIDS
CONCENTRATION OF LIME TREATER SLUDGE
Specific gravity, Solids in sludge,
g/ml weight %
1.03 5
1.05 9
1.10 17
1.15 24
1.30 42
The amount of sludge produced per unit of water treated as determined
from measurement of sludge volume and solids content was quite variable.
Therefore the amount of sludge and sludge solids were calculated from
the carbonate removed from the water, the organics removed (TOC), and
calcium removed in the sludge with the assumption that no solids were
carried out in the treated water. These calculations indicated that
during the 69 days of operation with the medium calcium concentration
range, the average amount of sludge, expressed as concentration in the
product water, was 629 mg/1 or 5.2 lb/1000 gal. The sludge solids con-
sisted of about 40% Ca of which 54% existed as CaCOo (from removal of
carbonates), and 46% as calcium-organics. This rather high content of
calcium carbonate is evidently the major factor causing the sludge to
settle and to filter readily. In contrast, the sludge from minilime
treatment of mill effluents is difficult to filter (3,7,8,14) and special
74
-------
precautions are used to add fiber or lime mud from mill operations to
make the color-removal sludge filterable. It is also possible that the
color compounds which are removed only when the higher lime dosages of
the minilime treatment are used are primarily responsible for the poorer
settling and filtering characteristics of the sludge from the minilime
treatment.
In a filtration test with a representative sample of the sludge from
lime treatment, the filtration rate with a vacuum of 24 in. Hg and a
final cake thickness of 0.7 in. was 380 Ib dry solids per hour per
square foot. The cake had a dry solids content of 74%. It appears that
the sludge would be suitable for recycle to the lime kiln by adding it
to the lime mud before the mill lime mud filters.
The average distribution of calcium in lime treatment is shown in Table 20
for operation at microlime, medium, and high calcium concentrations.
The calcium in the product water ranges from a low 71 mg/1 (or 22% of the
calcium added plus the calcium in the feed water) in the microlime treat-
ment to a high of 228 mg/1 (or 54%) in the high soluble calcium treatment.
Similarly, the Ca in Ca-organics ranges from 98 to 120 mg/1, increasing
with increasing color removal. The Ca in CaCOo is determined only by
the TIC concentration of the feed water since added calcium is consumed
in CaCOo formation before forming Ca-organics and raising the soluble
calcium concentration.
It is expected that with good control of lime dosage to give a constant
concentration of calcium in the treated water, good lime-carbon operation
will be obtained at about 80 mg/1 Ca in the treated water and that the
loss of calcium to the water will be about 25%.
The amount of calcium in the sludge as Ca-organics was 98-120 mg/1 for
the four groups of runs (Tables 18 and 20). The average ratio of atoms
of carbon in the sludge to atoms of calcium associated with the organics
(which is the calcium in the feed plus the calcium added less that in
the product water and less that tied up as CaC03 by the removal of
carbonates) was about the same for the four groups of data from lime
treatment and ranged from 2.3 to 3.0. The average ratio for all of
the lime treatment operation was 2.74 atoms of carbon per atom of
calcium. If it is assumed that the organic material removed is "typical"
kraft lignin having 9 carbon atoms and 3.3 oxygen atoms per monomeric
group, the above average Ca-organic sludge contained 3.3 calcium atoms
per monomeric group. The Ca-organics contained, on the average,
0.82 mg TOC/mg Ca, 5.3 CU/mg Ca, and 6.9 CU/mg TOC.
The above Ca-organic atomic composition suggest some difficulty in con-
sidering the removal of color from unbleached kraft effluent as a stoi-
chiometric reaction of calcium ions arid acidic organic groups such as
carboxylic and enolic groups, as Bennett and Dence (52) have shown for
kraft bleach effluent. At least a part of the functionality of calcium
ions may be explained as coagulation of colloidally suspended substances.
No doubt some of the color is originally in colloidal forms, while prod-
ucts of stoichiometric reaction with calcium may form additional colloids.
75
-------
Table 20. DISTRIBUTION OF CALCIUM IN LIME TREATMENT
Microlime concentr.
of soluble calcium,
average of 34 days
Medium concentr.
of soluble calcium,
average of 69 days
High concentr.
of soluble calcium,
average of 16 days
TIC
TOC
Calcium in:
Ca in feed water
Ca added as lime
Ca in, total
Calcium out:
Ca in product water
Ca in CaC03
(from TIC removed)
Ca in Ca-organics
(by difference)
Ca out, total
Feed water
mg/1
, 59
206
9
317
326
)
Prod. water
mg/1 % of tot.
12
139
71 22
157 48
98 30
326 100
Feed water
mg/1
52
275
15
377
392
Prod. water
mg/1 % of tot.
11
169
138 35
137 35
117 30
392 100
Feed water
mg/1
32
246
13
412
425
Prod. water
mg/1 % of tot.
11
146
228 54
77 18
120 28
425 100
-------
CLARIFICATION IN THE LIME TREATER
The rise rate in the clarification zone of the lime treater was 0.31
gpm/ft2 or 450 gpd/ft at the normal flow rate of 22 gpm. At this rise
rate some of the ca-organic floe failed to settle even though slude from
the reaction-mixing zone settled very rapidly. Several modifications
were made to provide increased contact between the feed water and the
lime to increase the degree of reaction and utilization of the lime.
The dosage of lime, in the range we used (including the minilime range)
appeared to have little influence on the settling of the small amount
of suspended solids in the water.
With the improved conditions of mixing, the average turbidity for the
three groups of runs ranged from 19 to 33 JTU in the feed water to the
lime treater and from 49-72 JTU in the water from the lime treater.
Corresponding values of suspended solids were 170 mg/1 in the feed water
and 550 mg/1 in the water from the lime treater.
Gould (7,8) also encountered floes that settled slowly in the Georgia
Pacific lime treatment of bleach effluent and used synthetic flocculants
to remove them.
Carry-over of suspended solids from the lime treater was no problem in
the operation of the carbon adsorption columns, although it would be
prudent to minimize such carry-over. On the other hand, for direct
discharge or reuse of micro-or minilime treated effluent, the level of
turbidity and suspended solids observed in this study suggests that
means for improved suspended solids removal must be further developed,
or provided- for in the form of synthetic flocculant, in a commercial
installation. One of the changes mentioned above to improve the degree
of clarification was to increase the solids concentration in the mixing
zone of the lime treater by externally pumping sludge from the bottom
of the lime treater to the top of the mixing zone where feed water and
lime slurry entered. This modification increased the solids content
in the mixing zone from about 4 to 15 g/1. The suspended solids in the
water from the lime treater was apparently reduced about 20% by this
change.
Another modification was made to improve the degree of mixing and to
increase the solution of the particles of Ca(OH)2 by increasing the flow
of lime-treated water to the lime slurry mixing tank at the lime feeder.
The flow was increased from 2 gpm to 5 gpm which was calculated to be
sufficient to dissolve about 40% of the lime from the feeder. Undissolved
particles of lime were further dispersed by the centrifugal slurry pump
from which a portion of the slurry was recycled to the slurry mixing tank.
This increased flow of water to the slurry mixing and the recycling of
the slurry through the feed pump caused a markedly visible reduction of
white lime particles in the sludge from the lime treater and evidently
increased the utilization of the lime.
77
-------
The turbidity and filterable solids apparently consisted of slow-settling
flocculated color bodies. These did not contribute to the measured color
and TOC because the water is filtered before these determinations are
made. We observed that, if the pH of the samples were reduced to 7.6
before the sample were filtered, the flocculated color dissolved and
increased the color of the water after the normal filtration.
OPERATION OF CARBONATOR
Most of the operation of the lime treatment section of the pilot plant
was with the microlime process which doesn't require carbonation to re-
move the excess calcium. When using microlime treatment, the water from
the lime treater was allowed to flow through the carbonator as a means
of getting the water to the pump that transferred it to FACET carbon
adsorption or through the duomedia filter to the carbon columns. The
carbonator was used in conjunction with microlime treatment for a period
of 11 days to test the effects of low concentrations of calcium (7-23 mg/1)
in microlime-treated water on the performance of the carbon columns.
Another run was made with the carbonator to provide minilime treatment
of water for an 11-day test of carbon adsorption following minilime
treatment.
In the minilime run, the lime dosage averaged 920 mg/1 CaO with a water
flow of 22 gpm. In the lime treater, the color was reduced from 730 to
81 CU for a reduction of 89%, and the TOC was reduced from 212 to 126 mg/1,
for a reduction of 40%. The higher lime dosage did not cause noticably
improved settling of floe nor reduce the turbidity of the water from the
lime treater, as compared to operation at lower lime dosages. Of the
calcium going into the system, 63% or 413 mg/1 Ca, ended up in the water
from the lime treater, compared to 22 to 54% (71 to 228 mg/1) in earlier
operations. This concentration of calcium in the water and the degrees
of removal agree closely with those obtained by Spruill (14) in full-
scale treatment of kraft mill effluent.
The carbonator was operated to reduce the pH to 10.5, the point of least
solubility of CaCOo, and to reduce the calcium content from about 400 to
10-20 mg/1. Further reduction of pH to about 8 was obtained with the
addition of C02 to the pH-adjust tank. Malfunctioning of the pH sensors
and C02 flow controllers caused occasional increases of calcium in the
water from the carbonator above the desired range. The concentrations
of color and TOC in the water from the carbonator and pH-adjust tank
were normally greater than in the water from the lime treater. This
increase was caused by the small amount of floe in the lime-treated
water that dissolved at the lower pH of the carbonator and pH-adjust
tanks.
78
-------
The sludge from the carbonator, which is essentially all CaCOo, settled
and filtered rapidly and was quite similar to the sludge from the lime
treater. In commercial plants, this sludge would be pumped back to the
lime treater to be combined with lime treater sludge.
The turbidity of the water from the carbonator was rather high (about
80 JTU). Only a slight reduction of this turbidity was obtained by
duomedia filtration. Three potential sources of this turbidity were
examined. The first, i.e., floe carry-over from the lime treater, was
ruled out because laboratory tests showed that the carried-over lime-
color floe redissolves upon carbonation, resulting in an increase in
color and in formation of CaCOo. The second potential source of turbidity
is carry-over of CaCO_ formed in the carbonator, particularly in the
colloidal form. Acidification of samples to pH 6 to dissolve CaCOo
resulted in a turbidity decrease of only about 50%, leaving a white
colloidal material in suspension. The nature of this third potential
source of turbidity was determined as most likely to be colloidal elemen-
tal sulfur formed in the carbonator due to pH reduction of polysulfide-
containing effluent, as discussed below.
At times the water from the equalization basin contained as much as 20
mg/1 of sodium polysulfide, which caused the lime-carbon treated water
to have a bright yellow hue in the absence of carbonation in the carbon-
ator. When effluent samples containing the sodium polysulfide were
adjusted to pH 6 or lower, the polysulfide liberated colloidal elemental
sulfur which had a white colloidal appearance. Several spot tests indica-
ted significant increases of polysulfide in the water as it passed through
the equalization basin. Polysulfide formation probably involves action
of anaerobic bacteria on thiosulfate and/or partial oxidation of sulfide
to sulfur.
79
-------
SECTION IX
OPERATING RESULTS:
BIO-CARBON COLUMNS
OBJECTIVES AND DESCRIPTION OF OPERATION
Objectives
The bio-carbon sequence was the first to be operated in the pilot
plant. As discussed in Section VII, mill effluent was bio-oxidized
and clarified to provide water for the carbon adsorption operation
that was typical of the effluent from pulp mills having secondary
treatment.
The objectives of the pilot plant operation with the bio-carbon se-
quence were to determine how successfully carbon adsorption in columns
could remove the color and TOC that remained after typical clarification
and bio-oxidation of pulp mill effluent and to obtain data needed to
design full-scale plants if this method had technical and economic
promise. The needed design data included the amount of fresh carbon
that is required per 1000 gal of water treated, the contact time
(column length and flow rate) of water in the column needed for remov-
als to given levels, pressure drops through the columns, backwash
procedures needed, rates of removal of color and TOC, and the effects
of anticipated biological activity in the carbon beds.
Clarification and.Filtration of Water to Columns
The bio-oxidized water was partially clarified in a section of the basin
and then filtered in a duomedia filter to insure that the operation of
the carbon columns would not be adversely affected by suspended solids
in the water.
A wooden slat fence used to section off a portion of the basin as a
damper was only partially effective in eliminating circulation through
this section, hence the water discharged was rather high in suspended
solids. Toward the end of the bio-carbon sequence, a plastic film
covering of the fenee resulted in achieving effective clarification
in this section.
The duomedia filter contained 12-in layers of anthracite and sand
supported on graded gravel. With the normal flow rate of 15 gpm used
in the bio-carbon sequence, the flow velocity in the filter was 1.2
gpm/ft. By trial and error, we found that mud balls would form in the
80
-------
00
N>
H
M
o
o
5|
I
O
CJ
o
o
H
2000
1500
1000
500
300
200
100
MILL EFFLUENT
TO BIO-OXIDATION
TO COLUMNS
FROM COLUMNS
III!
TO BIO-OXIDATION
TO COLUMNS
FROM COLUMNSV^
2/124 68 10 12 14 16 18 20 22 24 26 28 3/1 3 5 79 11
DATE
Figure 11. Concentrations of color and TOG during bio-oxidation-
carbon adsorption, Feb. 1 through March 11, 1972
-------
to March 11, 1972), the flow was 15 gpm (2.1 gpm/ft2) through four
columns in series. Four fresh columns were added resulting in an
average dosage rate of fresh carbon of 8 lb/1000 gal. Note that the
concentrations of color and TOG in the 24-hour composite samples of mill
effluent to bio-oxidation have a close correspondence, and that both
varied widely from day to day.
These wide swings of color and TOG in the water to bio-oxidation were
dampened quite a bit by the average retention time of 6.5 days in the
equalization basin. The changes of concentration in the feed to the
columns usually caused corresponding changes in the product water. The
amount of color and TOG removed remained" fairly constant in spite of
the changes of feed concentration to the columns. The average reduc-
tions of color and TOG by carbon adsorption were 578 CU and 91 mg/1
IOC.
Average results from the bio-carbon operation for the operating period
of February 4, 1972 to March 11, 1972 are given in Table 21. More
complete operating results are given in Table 22 for the bio-carbon
operation. Included in Table 22 are average results for each of the
four column positions and for other measured parameters not included in
Table 21. A printout of daily summaries of conditions and results is
given in Appendix D.
This mode of operation at a column feed pH of 8.6 and a product pH of
8.2 gave a water with an average concentration of 58 mg/1 of TOG, which
is expected to be acceptable as a general mill feed water. However,
the color level of 230 CU may be too high for some areas in the mill.
It is estimated that an additional 25 ft of carbon bed would be required
to reach a target of 100 CU in the treated effluent.
This increased length increases the amount of carbon in use (carbon
inventory) in the same proportion, i.e., by 63%, while it increases the
average dosage rate only by the 25% for the 25% greater amount of color
removed. Both factors are important in the economics of the total system.
A rather careful determination must be made by each mill, and by the
industry as a whole, of actual water quality requirements in terms of
color and TOG in reuse water obtained from effluent.
The amount of color and TOG removed in the lead column (Table 22) was
greater than in any of the other three columns even though it was fairly
well loaded on the average. The amount removed per column became less
for each succeeding column. On the other hand, each column removed a
nearly equal percentage of its feed.
83
-------
Table 21. OPERATING SUMMARY OF BIOLOGICAL OXIDATION
CARBON ADSORPTION SEQUENCE AT 15 GPM
LOW CARBON DOSAGE
Number of fresh charges of carbon 4
Average temperature of water to columns F 74
Cumulative removal when discharge - CU/g carbon . 646
- TOG, mg/g carbon 98
Average dosage rate, Ib carbon per 1000 gal water 8
Average retention time of water through 4 columns, hr 2.34
Volumetric flow rate, v/(v hr) 0.43
Color removal Range Average
Feed to bio-oxidation, CU 460-2700 1210
Feed to carbon columns, CU 500-1200 810
Product from carbon columns, CU 46- 440 232
Removal by bio-oxidation plus filter, % - 33
Removal by carbon, % of feed to carbon - 71
Total removal, % of feed to bio-oxidation - 81
Rate of removal by carbon, CU/(g hr) 0.40-0.87 0.67
TOG removal
Feed to bio-oxidation, TOC, mg/1 120- 966 277
Feed to carbon columns, TOC, mg/1 90- 350 149
Product from carbon columns, TOC, mg/1 10- 90 58
Removal by bio-oxidation plus filter, % - 46
Removal by carbon, % of feed to carbon - 61
Total removal, % of feed to bio-oxidation - 79
Rate of removal by carbon, TOC, mg/(g hr) 0.06-0.18 0.11
84
-------
Table 22. AVERAGE RESULTS FOR BIO-CARBON SEQUENCE FOR PERIOD
OF FEBRUARY 4, 1972 to MARCH 11, 1972,
LOW CARBON DOSAGE
Flow rate, gpm
Temperature water to basin, °F
to columns, °F
Color, to basin, FPCU
CU
to columns, FPCU
CU
from columns, FPCU
CU
TOG, to basin, mg/1
to columns, mg/1
from columns, mg/1
BOD, to basin, mg/1
to columns, mg/1
from columns, mg/1
D.O., from basin, mg/1
pH
Conductivity, micromhos,
Turbidityb, JTU
Filterable solidsb, mg/1
Pressure drop, columns0, psi
Color removal rate, CU/(g hr)
TOC removal rate, mg/(g hr)
Color removed, 7<> to each column
Color removed, % to lead column
TOC removed, % to each column
TOC removed, % to lead column
Color remaining, % to lead column
TOC remaining, 70 to lead column
Color removal, CU
TOC removal, mg/1
Cumulative color removed, CU/g
Cumulative TOC removed, mg/g
& Four fresh columns were added during this period, so four individual
columns served as the lead column, four as the second, etc.
Average of two analyses.
c Ap per column just before backwashing.
General or
overall data
14.8
112
74
1610
1210
1408
810
277
149
278
67
6.0
8.6
16
55
0.67
0.11
71
61
578
91
Lead
column
1170
621
109
8.4
1580
3.6
0.87
0.18
23
23
25
24
77
75
189
40
603
87
Second
column
•
934
445
83
8.3
2.4
0.81
0.12
28
22
23
18
55
57
176
26
394
44
Third
column
771
319
70
8.2
2.3
0.56
0.06
28
16
16
9
39
49
126
13
204
20
Fourth
column
666
232
58
16
8.2
1580
11
22
2.3
0.40
0.06
27
11
19
9
29
40
87
12
78
8
85
-------
The concentration of color and TOG remaining in the water as a function
of length of column is shown in Figure 12. This plot can be used to
estimate the length of bed required at these operating conditions to
reach any desired color or TOG concentration. For example, under these
conditions of operation, the length of bed to reduce the color to 100 CU
would be about 65 ft, or about 1.63 times that used in the pilot plant
(40 ft) which reduced the color to an average of 230 CU. Alternatively,
the color could have been reduced to a lower value if the contact time
had been increased by reducing the flow rate, but this would have re-
quired a much lower flow velocity in the carbon bed, which causes a
reduced rate of adsorption (as discussed later in this section) and
would also have required a greater amount of carbon per 1000 gal of
treated water.
In the case of TOG, Figure 12 shows that the original goal of 50 mg/1
could have been reached if the bed length had been increased to 48 ft.
These plots show that, with the bio-carbon sequence, the reduction
of color to a suitable level is more difficult than the reduction of
TOG. This is true only of bio-carbon adsorption because the bio-oxidation
in the aeration basin removed less of the color than TOG. In contrast,
a greater percentage of color than TOG was removed by the lime treatment
step. The concentration of BOD was reduced by 76% by bio-oxidation to
67 mg/1 and further reduced 76% to 16 mg/1 by the carbon columns, for
an overall reduction of 94% by bio-carbon (see Table 22).
The pH decreased from 8.6 to 8.2 as the water passed through the columns.
As discussed later, this was perhaps caused by bio-activity in the carbon
beds. The average turbidity decreased from 16 JTU in the feed water to
the columns to 11 JTU in the water from the columns. The average temper-
ature of the water in the columns was 74°F.
DOSAGE OF CARBON REQUIRED
One of the primary objectives of the pilot plant operation was to deter-
mine the amount of fresh carbon or regenerated carbon that is needed per
1000 gal of water to provide water suitable for reuse in the mill. This
information cannot be obtained from adsorption isotherms with sufficient
reliability. Rather it must be obtained in continuous bench-scale or
pilot plant columns and is determined from the average amount of water
that can be treated per new column of carbon.
The dosage rate for the test period of February 4, 1972 to .March 11, 1972,
during which time four fresh columns were added, was 8 Ib of carbon per
1000 gal of water, or 940 mg/1. For the period of March 11, 1972 to
March 18, 1972 the rate of addition of fresh columns was increased to
determine how much the product water quality would be improved by the
high dosage rate. On the basis of two column changes, the dosage rate
for this period was 26 Ib per 1000 gal, or 3.3 times that of the low-
dosage period. The results from these two periods are given in Table 23.
86
-------
10
20 30 40 50 60 70
LENGTH OF BED, ft
80
Figure 12. Color and TOC remaining versus length of
carbon bed during bio-carbon sequence
87
-------
Table 23. RESULTS FROM BIO-CARBON ADSORPTION WITH
DIFFERING CARBON DOSAGE RATES
Low carbon
dosage rate
2/4 to 3/11/72
Conc'n. to aeration basin, CU or mg/1
to carbon columns, CU or mg/1
from carbon columns, CU or mg/1
Removal by carbon adsorption, % of feed to
columns
Rate of removal, CU/(g hr) or mg TOC/(g hr)
No. of fresh charges of carbon (1600 Ib ea.)
Cum. removals when discharged, CU/g or mg TOC/g
Av. dosage rate, Ib carbon per 1000 gal water
Color
1210
810
232
71
0.67
4
; 646
8.0
TOG
277
149
58
61
0.11
-
98
-
High carbon
dosage rate
3/11 to 3/18/72
Color TOG
1210 532
716 161
226 49
68
0.46
2
327
26.0
70
0.10
42
Considering, of course, the short duration of the test of high carbon
dosage rate, a few points may be noted. Somewhat surprisingly, these
results show that the high dosage rate gave only slightly lower concentra-
tions of color and TOG, and a mixed response in terms of percentage remov-
als, rather than the expected strongly increased percentage removals of
color and TOG. Because of these unexpected results, it may be speculated
that perhaps a dosage rate even somewhat reduced below the 8 Ib per 1000
gal of water of the low-dosage period might have achieved nearly equal
removals. However, this would not be expected on the basis of equilibrium
adsorption isotherms. The average cumulative removal in the first column
during operation at 8 Ib per 1000 gal was 646 CU/g, which is about 65% of
isotherm equilibrium loading and is in line with percentage-of-equilibrium
loadings achieved in other carbon column operations.
LOADINGS ON CARBON
The apparent loadings of color or TOG on the carbon are actually cumula-
tive removals per unit weight of carbon. The cumulative removals are
greater than the actual loadings of adsorbed impurities because of remov-
als by bio-activity and because a substantial amount of the impurities
removed from the water is not actually adsorbed within the pores of the
carbon but is apparently held on the outer surfaces of the carbon. Much
of this adhering impurity is removed during backwashing, as reported
later under a discussion of backwashing.
88
-------
The cumulative weight of TOG removed by each column per unit of carbon,
expressed as mg TOC/g carbon, is shown in Figure 13 as a function of
operating time. The cumulative removals of TOG by four columns, by the
time they were removed from operation, ranged from 70 to 130 mg/g of
carbon during the period of February 4, 1972 to March 11, 1972, and the
average removal was 98 mg/g, or 9.8% based on the original weight of
carbon (see Tables 22 and 23). The cumulative removals of color for the
same period ranged from 495 to 667 CU/g, for an average of 646 CU/g.
Average cumulative removals during the period of March 11, 1972 to
March 18, 1972, when the carbon dosage was raised to 26 lb/1000 gal, were
expectedly lower, i.e., 42 mg/g TOG and 327 CU/g.
RATES OF REMOVAL OF IMPURITIES BY CARBON
The rates of adsorption of impurities on carbon have been studied by a
number of investigators. The proposed rate equations based on adsorption
of single compounds from solution tend not to fit operating data from
complex multi-component systems. The rate of adsorption in weight of a
given impurity removed per hour is usually expressed as a direct function
of weight of carbon, concentration gradient driving force, temperature
of the water, and as an inverse function of carbon particle size. Other
factors affecting rate are the pore size and pore size distribution of
the carbon, the molecular size and polarity of the material being adsorb-
ed, the degree of turbulence around the particle, and the diffusivity
of the impurity in the water. Development of an applicable rate equation
for pulp mill effluent is complicated by the countless compounds in the
water and the wide variability of the types of compounds. The concentra-
tion gradient driving force used in rate equations is the difference in
concentration of the water within the pores of the carbon, or its equilib-
rium adsorption concentrations, and the concentration of the bulk solution.
Again, the development of an applicable rate equation is complicated by
the fact that the equilibrium concentration within the carbon is not known.
The equilibrium concentration could be determined from an equilibrium
adsorption isotherm for the same water and carbon if one knew the loading
on the carbon. But the indicated loading or cumulative removal of the
impurity per unit weight of carbon that is measured is not the same as
the true loading, as discussed above.
Because of these complicating factors, we used a simplified rate expres-
sion which is the amount removed per weight of carbon per hour, or CU/(g hr)
and mg TOC/(g hr). The rates were calculated (by the computer program)
each day for each column from the amounts of impurity removed and the
weight of carbon in each column. These rates are of value in comparing
the effects of operating variables on rates and for designing full-scale
plants. The rates are directly applicable only to water of the same
composition and temperature, the same type and size of carbon, the same
velocity and retention time of the water in the columns, and the same
rate of addition of fresh carbon (dosage rate).
89
-------
td
o
oo
o
o
H
140
120
8 10 12
14 16 18 20 22
DATE
24 26 28 3/1
11
Figure 13. Cumulative removals of TOG on carbon,
Feb. 1 to March 11, 1972
-------
The rates of removal in the pilot plant columns during the bio-carbon
sequence were highest in the lead column, in which the concentrations
were highest, and were lowest in the last column in which the concentra-
tions were lowest (see Table 22). During the period from February 4, 1972
to March 11, 1972, the average rates of removal by column position ranged
from 0.87 CU/(g hr) for the first column to 0.40 CU/(g hr) for the fourth
column, and the overall average rate was 0.67 CU/(g hr) (see Table 22).
The corresponding rates of removal of TOG were 0.18 and 0.06 mg/(g hr),
with an average of 0.11 mg/(g hr). These results are shown graphically
in Figure 14 as a function of bed length. Note that the highest rates
were observed with the lead column that had the most fully loaded carbon
and contained the highest concentrations of impurities in the water.
The rates of adsorption in a given column as it was moved through the
four column positions are also indicated by the slopes of the cumulative
removal curves of Figure 13.
The rates of removal are plotted versus average concentration in each
column in Figure 15. These curves show that rates are very dependent
on bulk solution concentration. The effect of concentration would be
much greater if the depressing effect of loading on the carbon had not
been present.
The equations for rates of removal of color and TOG for the curves given
in Figure 15 are:
Color removal rate = 0.523 In C - 2.55, for 250 to 800 CU
TOG removal rate = 0.194 In C-- 0.77, for 70 to 150 mg/1 TOG
Where rate is in CU/(g hr) or mg TOC/(g hr), C is average concentra-
tion in CU or mg/1.
As stated above, backwashing of the columns removed some of the color
and TOG that had been loosely held by the carbon. There was no evidence
that the rate of adsorption was increased by a higher frequency of back-
washing, even though this would be expected.
The rates of removal were less when the high dosage rate of carbon was
used (March 11, 1972 to March 18, 1972), even though the cumulative
removals, or degree of loading of the carbon, were only 50% of those
during the low dosage period (February 4, 1972 to March 11, 1972) (see
Table 23). It is postulated as one possible explanation, that the
external coating of organic matter on the carbon, as discussed earlier,
was an overwhelming rate-controlling mechanism, which nullified the
effect of fresher carbon from the high-dosage operation. However, the
short duration of operations at high dosage rates makes this comparison
less than definitive.
91
-------
00
^x
O
*l
w
I
1.0
0.8
0.6
0.4
0.2
0
0.20
J» 0.15
3 0.10
O
O
o
o
H
0.05
603
715
87
125
i
533
394 204
CUM REMOVAL, CU/g
533 382
AVG CONC'N, CU
44 20
CUM REMOVAL, mg/g
95 78
9 AVG CONC'N, mg/1
I I i
10 20
LENGTH OF CARBON BED, ft
30
78
275
8
64
40
Figure 14. Rates of removal of color and TOG by carbon adsorption
during bio-carbon sequence, Feb. 4 to March 11, 1972
92
-------
1000
500
00
s
u
o
H
o
iJ
O
w
200
100
50
30
0
I 1 1 1 1-
0.5 1.0
COLOR RATE, CU/ (g hr)
1
0.1 0.2
TOG RATE, mg/ (g hr)
1.5
I I I 1 L.
0.3
Figure 15. Rates of removal of color and TOG by carbon columns
as a function of average concentration during bio-
carbon operations, Feb. 4 to March 11,
93
-------
EFFECT OF FLOW VELOCITY ON RATES OF ADSORPTION
The superficial (empty-column) velocity of water through the columns
at 15 gpm was 2.1 gpm/ft2. A lower velocity would give a longer contact
time in the four columns for a greater reduction of color, but it was
not known whether the lower velocity would cause lower rates of removal
per unit of carbon due to an increase in the thickness of the stagnant
layer of water around each carbon particle. Special runs were made at
0.7 and 1.75 gpm/ft which indicated that at the higher velocity of
1.75 gpm/ft the rate was only slightly higher than at 0.7 gpm/ft . The
results of these tests are discussed more fully later in conjunction
with tests made during the primary-carbon sequence of operation (see
Section X).
BIOLOGICAL ACTIVITY IN CARBON COLUMNS
There was evidence of some aerobic and anaerobic bio-activity in the
carbon columns, as is more fully discussed under the primary-carbon
sequence in Section X. The average dissolved oxygen content of the
feed to the columns during the bio-carbon sequence was 6 mg/1 but de-
creased rapidly to less than 1 mg/1 in the water from the first column.
However, very little removal of TOG (about 3 mg/1) can be attributed to
biological consumption of the 6 mg/1 of dissolved oxygen based on stoi-
chiometric considerations. Most of the aerobic activity appeared to be
concentrated in the top one or two feet of the first column, evidenced
by caking of carbon particles and column plugging unless the columns
were backwashed every one or two days.
Anaerobic activity was also present in the columns, as indicated by
bacterial analyses made near the end of the bio-carbon operation. The
total bacterial count was low in the mill effluent to the basin and very
high from the basin. The concentration was reduced from 3.6 x 10 to
2.1 x 10°/ml by the sand filter. The concentration was further reduced
to 1 x 106/ml in passage through the carbon columns. All but about 0.1%
of the bacteria were retained on a 0.8 micron filter. Hydrogen sulfide
producing bacteria were present in the feed to the columns, but the
water from the columns had very low concentrations of these bacteria.
Occasionally, there was evidence of H2S odor in the water from the columns.
Anaerobes were found in the water to and from the columns.
BACKWASHING OF CARBON COLUMNS
The columns were backwashed by a high rate of flow of compressed air for
about 20 sec followed by an up-flow of 30 gpm of well water for a period
of 20 minutes every other day. The backwash water was black and contain-
ed about 10% of the color and TOG that had been removed by the columns
94
-------
during the two days of adsorption. When the columns were backwashed at
the two-day frequency, the average pressure drop through the four columns
in series was about 10 psi before backwashing and about 3 psi after back-
washing.
CORRELATION OF VARIABLES
The interactions of 12 independent variables with 18 dependent variables
during the bio-carbon operation of February 4, 1972 to March 11, 1972
at 15 gpm were examined by a computer regression analysis. The results
of this analysis generally confirmed the conclusions from the direct
analysis of the data. The 12 independent variables included temperature,
pH, and concentrations of color and TOG to the columns, and cumulative
removals of color and TOG from each of the four columns. (The term
cumulative removal is used in preference to loading for reasons stated
earlier in this section.) The 18 dependent variables included color and
TOC concentrations from each column, removal rates of color and TOC for
each column, and percentage removals of color and TOC from the collective
columns.
This regression analysis of the data showed that the color of the water
from the fourth column was largely a direct function of the color to the
columns, the cumulative removals of color and TOC by the first two columns,
and the cumulative removal of TOC of the fourth column, i.e., the product
color increased as the feed color increased and as the cumulative removals
increased. The color from the other columns was primarily a function of
the color in the feed to the first column. The percentage removal of
color by the four columns was almost entirely explained by the cumulative
removals of color by the four columns, i.e., the percentage removal de-
creased as the cumulative removals increased.
The TOC in the water from the fourth column was primarily a direct function
of the cumulative removals of TOC by the first and fourth columns and of
the TOC concentrations of the feed water to the first column.
The rates of removal of color by the columns were influenced as an inverse
function of the cumulative removals by the columns. The rate of removal
of TOC by the group of four columns was strongly influenced (directly)
by the TOC concentration of the feed water and to a lesser degree by the
color of the feed water, and was weakly influenced (inversely) by the
cumulative removal of color by the first column.
The variation in temperature during this period of operation was not very
great: 66% of the values fell between 81 and 67°F, with the average
being 74°F. In this range, temperature had no significant influence on
adsorption performance.
95
-------
SECTION X
OPERATING RESULTS:
PRIMARY - CARBON COLUMNS
OBJECTIVES AND DESCRIPTION OF OPERATION
Objectives
The objectives of the primary-carbon operation were to determine
whether carbon adsorption without pretreatments by bio-oxidation or
lime treatment could remove the much larger amounts of color and TOG
at high enough loadings on the carbon to give reasonable values of
carbon dosage (regeneration rates) and sufficiently low capital and
operating costs to make this sequence competitive with the other two
sequences. Preliminary economic estimates (see Section V) had indicated
that the primary-carbon sequence might be competitive with the lime-
carbon sequence. Since the concentrations of color and TOG would be
high, the rates of adsorption and the loadings on the carbon were ex-
pected to be considerably greater than with the other sequences. Other
objectives were to determine the flow velocities and contact times
needed for removal of impurities and to determine whether this sequence
of operation would give unique problems, such as excessive bio-activity
within the columns and greater rates of pressure-drop increase.
Operating Periods
The primary-carbon sequence was in operation from April through the
first week in September, 1972. Four columns were operated in series
in downflow, eight weeks at 10 gpm, and four weeks at 5 gpm. Carbon
columns were replaced regularly throughout these periods. Two columns
were operated in series for the subsequent 11-week period in downflow
at 10 gpm without any further carbon changes to observe the long-term
capacity of the system for removing color and TOG.
PRIMARY CLARIFICATION AND FILTRATION
The basin was used for primary clarification and for equalization of
impurity concentrations in the mill effluent before it was fed to the
carbon columns. Water was pumped into the basin at a rate of 15 gpm
which gave a retention time of about 8 days. As pointed out in Section
VII on Primary Clarification and Bio-Oxidation, anaerobic activity in
the basin caused an increase of low-molecular-weight compounds that are
96
-------
not absorbed by activated carbon. The average TOG content of these
compounds was 48 mg/1 in the water to the basin and 79 mg/1 in the water
from the basin, for an increase of 65%.
Even though the equalization basin provided a retention time of 8 days,
the feed to the columns varied from 1550 to 540 CU and 410 to 170 mg/l'
TOG during the period covered by these results. The rates of change of
concentrations in the feed to the columns were generally less than 100 CU
per day and 20 mg/1 TOG per day. These fluctuations in feed concentration
introduce a high degree of complexity into the correlation of operating
results with operating conditions.
In a commercial installation for treating effluent by lime treatment or
carbon adsorption, or both, effluent concentrations to the treatment
system must be held in a narrower range, and rapid changes in effluent
concentrations must be avoided through in-plant control measures or
external spill basins and equalization basins. The suspended solids in
the water to the basin varied widely - from 37 to 790 mg/1. The water
after primary clarification contained from 50 to 265 mg/1 and averaged
168 mg/1 of TSS. The turbidity of the water from primary clarification
ranged from 18 to 23 JTU.
The water to the columns was not sand-filtered prior to the carbon columns
except during the first three weeks. Insufficient data were obtained to
determine the amount of suspended solids removed by the duomedia filter
when it was being used. The pressure drop across the columns before each
backwash remained the same as when the water was filtered. Therefore,
we concluded that filtration ahead of the columns is not required if the
effluent to the columns contains less than about 200 mg/1 suspended
solids and the columns are backwashed on a schedule as discussed later
in this section. This suspended solids level may well be achievable,
depending on the mill, by grit and floating solids (bark) removal alone
or by clarification in a relatively small clarifier at high overflow rate.
REMOVALS OF COLOR AND TOG BY CARBON ADSORPTION
The results from carbon adsorption following primary clarification
(primary-carbon) are summarized in Table 24 for two periods operating
at two different flow rates (10 and 5 gpm). Detailed data for each of
the four columns are given in Table 25 for the operating period at the
higher flow rate (10 gpm). A printout of daily summaries of conditions
and results is given in Appendix E.
The product water during the 10-gpm run contained, on the average, 185 CU
and 83 mg/1 of TOG. These values are somewhat greater than may be desired
for reuse in the mill. The relationship of product concentrations to
97
-------
Table 24. SUMMARY OF PRIMARY CLARIFICATION - CARBON ADSORPTION SEQUENCE
Dates 4/25-5/20 5/20-6/22
Flow rate, gpm 10 5
Flow rate, gpm/ft2 1.42 0.71
Flow rate, vol water/vol carbon/hr 0.28 0.14
Contact time, hr 3.5 1.8
Number of fresh charges of carbon 4 3
Cummulative removal when discharged
CU/g carbon 350 360
TOC, mg/g carbon 61 70
Average dosage rate, Ib carbon per 1000 gal 20.5 28
Color removal
Feed to carbon columns, CU 791 997
Product from carbon columns, CU 185 202
Removal by carbon, % 77 80
Rate of removal by carbon, CU/(g hr) 0.69 0.46
TOC removal
Feed to carbon, TOC, mg/1 220 310
Product from carbon, TOC, mg/1 83 121
Removal by carbon, % 62 61
Rate of removal by carbon, mg/(g hr) 0.11 0.08
98
-------
Table 25. AVERAGE RESULTS FOR PRIMARY-CARBON SEQUENCE FOR PERIOD
OF APRIL 4, 1972 TO MAY 20, 1972 AT FLOW RATE OF 10 GPM
Temp, water to basin, F
to columns, F
Color, to columns, FPCU
CU
from columns, FPCU
CU
TOG, to columns, mg/1
from columns, mg/1
pH from columns
Cond., micromhos
Turbidity, JTU
Susp. solids, mg/1, from basin
from cols.
Pressure drop, psi
Color removal rate, CU/(g hr)
TOC removal rate, mg/(g hr)
Color removal, 70 to ea. col.
Color removal, % to lead col.
TOC removal, % to ea. col.
TOC removal, % to lead col.
Color remaining, 7o to lead col.
TOC remaining, % to lead col.
Color removal, difference, CU
TOC removal, difference, mg/1
Cumulative color removal, CU/g
Cumulative TOC removal, mg/g
General
a.v.
av.
tot.
tot.
tot.
tot.
data
114
76
1474
791
220
168
0.69
0.11
77
62
606
137
Lead
column
815
483
126
10.0
876
20
1.0
1.66
0.30
39
39
43
43
61
57
308
94
270
45
Second
column
588
322
110
10.0
850
18
1.0
0.60
0.05
33
20
12
7
41
50
161
16
101
16
Third
column
483
259
99
9.9
840
17
1.0
0.24
0.03
20
8
9
5
33
45
63
11
48
9
Fourth
column
368
185
83
9.4
830
17
85
1.0
0.27
0.05
29
9
17
8
23
38
74
16
21
5
99
-------
length of column is given in Figure 16. By extrapolation, the length
of column needed to reach 100 CU at 10 gpm and under the same other
conditions of operation would be about 60 ft, or about 50% more than
was used in the pilot plant runs.
It is assumed in this extrapolation that the compounds that would be
removed at the lower concentrations would have the same adsorption
properties as those already adsorbed. This assumption may not hold for
the low-molecular-weight compounds that make up a portion of the TOG.
To reach a color of 100 CU, the inventory of carbon would have to be
increased by 50% and the carbon dosage rate would have to be increased
by about 14% for the increased amount of color removed.
The lead column removed a large portion of the total color and TOC removed
by the four columns. As shown in Table 25 during the 10-gpm run, the
lead column removed 39% of the feed color, the second column removed 20%,
the third removed 8%, and the fourth removed 9%. The lead column removed
51% of the total color removed by all columns and 70% of the total TOC
removed. The possible mechanisms accounting for this large percentage
removal in the first column are given later in a discussion of the ultimate
removal capacity and mechanisms in carbon columns.
In contrast to the above large removals in the first column during primary-
carbon operation, the first of the four columns in the bio-carbon opera-
tion removed only 33% of the total color that was removed and 40% of the
TOC. The lower removals in the lead column in the bio-carbon sequence
might be due to the organic compounds being less bio-degradable after
the bio-oxidation treatment and that bio-activity was greater in the
primary-carbon columns. The removal of BOD was not determined for the
primary-carbon sequence of operation.
DOSAGE OF CARBON REQUIRED
The most reliable way to determine the carbon dosage rate with fixed-bed
columns is to continue a pilot plant operation until the number of fresh
columns added are two greater than the number of columns in series, e.g.,
six for the primary-carbon operation. The dosage rate is then found by
dividing the weight of carbon in each column by the average volume of
water treated between column changes. In this primary-carbon run at 10
gpm, we added fresh columns six times and discarded the data from the
first two column changes because these columns had been used during the
bio-carbon sequence and thus the data would not be typical of results from
the primary-carbon sequence. The time interval between the next four
column changes ranged from 92 to 160 hr and averaged 130 hr. Therefore,
the dosage rate was:
1600 Ib carbon per column change x 1000 _ 20.5 Ib
130 hr x 600 gal/hr ~ 1000 gal
100
-------
1500
10 20
30 40 50 60
LENGTH OF BED, ft
70 80
Figure 16. Color and TOG remaining versus length
of bed during primary-carbon sequence
101
-------
During the 5 gpm run, five columns were added and the average time
between changes for the last three columns added was 251 hr and the
average dosage rate was 28 Ib carbon per 1000 gal.
RATES OF ADSORPTION
Adsorption Rates During 10 gpm Run
The average rates of removal by each column for the 10 gpm run, April
25, 1972 to May 20, 1972, are given in Table 25 and are plotted in
Figure 17 as a function of average concentration of the water in each
column. The rates during the bio-carbon sequence are also shown.
The rates of removal of color and TOC appear to be linear with the
log of concentration over most of the concentration range. A discus-
sion of factors affecting adsorption rates is given in Section IX
on the bio-carbon sequence of operation.
In the case of the primary-carbon data, the average rates for the
third column were lower than those for the fourth column which is
surprising because the fourth column had a lower average concentration.
The rates of removal by primary-carbon were lower than those of the
bio-carbon sequence at concentrations below 130 mg TOC/1 and 330 CU
but were higher at the higher concentrations. The higher rates in
the lead column of the primary-carbon sequence were evidently caused
by non-adsorptive phenomena discussed later in this section.
The equation for the relationship between the rate of removal of color
and the concentration during the 10-gpm run with primary-carbon se-
quence is as follows:
Rate, CU/(g hr) = 1.83 In C - 10.2
where C = average color concentration in the columns, CU
This equation is only approximate and is s-trictly applicable only for
a color range of 300 to 700 CU, for water with the same properties
as was used in the pilot plant, and for the same other operating con-
ditions used in the pilot plant run.
The equation for TOC removal rates, which is limited to the TOC range
of 100-200 mg/1 and to the conditions used in the pilot plant run is:
Rate, mg TOC/(g hr) - 0.55 In C - 2.5
where C = average TOC concentration in the columns, mg/1
Effect of Velocity on Adsorption Rates
The effect of velocity on adsorption rates was investigated during the
bio-carbon and the primary-carbon sequences of operation. The EPA
102
-------
1000
500
o
o
H
O
o
u
w
w
200
O /
X BIO-CAKBON
BIO-CARBON
100
30
0
1 . .
0.5
COLOR
i . 1 i
RATE,
, ,
CU/
i
1.
(g
|
0
hr)
i i i
1.5
, 1 ,
0.1 0.2
TOC RATE, mg/(g hr)
0.3
Figure 17. Rate of removal of color and TOC by carbon adsorption
during primary-carbon operation at 10 gptn
103
-------
Process Design Manual for Carbon Adsorption (19) stated that several
studies have found that velocity in the range of 2-10 gpm/ft2 is not
a limiting factor in fixed-bed granular carbon adsorption. Other
investigators (9,17) found no effect in the 4-10 gpm/ft2 range.
The effect of velocity was studied at the end of the bio-carbon oper-
ation. The last two columns were used, one with a flow of 5 gpm (0.7
gpm/ft2) and the other with a flow of 12 gpm (1.75 gpm/ft2) with the
same water being fed to both. Samples of treated water were removed
at different lengths of the carbon beds to give equal retention times
of water in the columns. After four days -of operation, the conditions
were reversed so that the previous high-velocity column was then the
low-velocity column. This switching of conditions should remove the
effect of the degree of loading in each column from the comparison.
Results showed that the percentage removals and rates of removals
were the same for color removal and about 10% greater for TOG removal
in the case of the high velocity run. Therefore, this test indicated
that an increase of velocity from 0.7 to 1.75 gpm/ft was beneficial
by a slight amount.
The influence of velocity was also studied during the primary-carbon
operation by comparing results from periods of operation at 10 gpm
(1.4 gpm/ft2) and 5 gpm (0.7 gpm/ft2), as reported in Table 24. The
feed during the 5 gpm run was unfortunately 25% greater in color and
40% greater in TOG concentration. The average product concentrations
were also greater for the 5 gpm run, even though the percentage removals
of color (77 and 80%) and of TOG (61 and 62%) were about the same for
the two periods.
Since rates of removal are highly dependent on the concentration of
the impurity in the water in the adsorption beds, the rates of removal
at the two velocities were compared at the same concentrations on a
plot of rate versus average concentration in each column. The rates
of removal of color obtained at 10 gpm for April 25, 1972 to May 20, 1972
were two times those at 5 gpm when they are adjusted to the same water
concentration. The rates of removal of TOG at 10 gpm were more than
three times those at 5 gpm at the same water concentrations.
These results indicate that rates of adsorption were substantially
greater at a superficial velocity of 1.4 gpm/ft than at 0.7 gpm/ft .
This effect of velocity in the primary-carbon sequence was much more
pronounced than was observed in the bio-carbon sequence discussed
above. We conclude that commercial plants should use a velocity of
at least 1.4 gpm/ft2 and perhaps as much as 2-4 gpm/ft2 subject to
limitation by the cost of pumping.
104
-------
The pumping energy requirement and pressure drop through the columns
increase in direct proportion to the lineal velocity of flow. To
maintain the same color removal, the length of column needed'will also
by increased in direct proportion to the velocity cut will be decreased
by the greater adsorption rate afforded by the greater velocity. The
net effect of increased velocity on energy requirements will lie between
velocity to the first power and velocity squared.
The pumping energy required at 4 gpm/ft2 to remove color from 1000 to
100 CU was estimated to be only about $0.004/kgal and the total pres-
sure drop was estimated to be about 40 psi.
CORRELATION OF VARIABLES
The interactions of 12 independent variables with 18 dependent variables
during the primary-carbon operation of April 25, 1972 to May 20, 1972
at 10 gpm were examined by a computer regression analysis. The results
of this analysis generally confirmed the conclusions from the direct
analysis of the data. The 12 independent variables included temperature,
pH, and concentrations of color and TOC to the columns; and cumulative
removals of color and TOC from each of the four columns. The 18 depen-
dent variables included color and TOC concentrations from each column,
removal rates of color and TOC for each column, and percentage removals
of color and TOC from the collective columns.
This regression analysis of the data showed that the percentage removal
of color by the four columns was almost entirely explained by the (inverse)
effect of cumulative removals of the columns. The cumulative removal of
the first column had the greatest influence, primarily because this column
removed more of the total color than the other columns.
The TOC removal rates were only weakly influenced by the various conditions
(lineal velocity was constant during this operating period). The rate
was influenced by color and TOC concentrations of the feed (higher concen-
trations caused higher rates) and to a lesser extent by the cumulative
removals of the carbon (higher cumulative removals caused lower rates).
The overall color removal rates were primarily a (direct) function of
color concentration in the feed. In the second and third columns the
rate was still largely dependent on feed color concentration but was also
directly related to cumulative removals. Therefore in these two columns,
the rate increased as the cumulative removals increased, for unknown
reasons. In the first and fourth columns, the rate decreased as the
cumulative removals increased as would be expected. The rate in the
fourth column was primarily due to the cumulative removals of color and
TOC of the first column as well as that of the fourth column.
105
-------
The concentration of TOG from the fourth column was primarily a (direct)
function of cumulative removals of color and TOG by the first and fourth
columns and secondarily a (direct) function of the color and TOG concen-
trations in the feed.
The color from the fourth column was primarily a (direct) function of
the color of the feed and of the cumulative removals of TOG and color
by the second and third columns.
PRESSURE DROPS AND BACKWASHING OF COLUMNS
During essentially all of the primary-carbon runs, from April 4, 1972
to June 22, 1972, the four columns were backwashed every other day, and
the water fed to the columns from the equalization basin was not filtered
after the first month of operation. During the period of June 22, 1972
to July 8, 1972, backwashing was discontinued to determine how rapidly
the pressure drop across the columns would increase and to determine
whether a reduced frequency of backwashing could be used. The flow rate
during this 9-day test period was 10 gpm through the four columns in
series. There was no substantial increase of pressure drop across the
four columns; the final pressure drop was only 3 psi. The columns were
all backwashed and two columns were then operated at a flow of 10 gpm
and were backwashed once per week. After two weeks of operation, the
pressure drop increased to 8 psi across the two columns and to 44 psi
after four weeks. An inspection of the tops of the beds showed that the
top two feet of both carbon beds had become almost solid with particulates
removed from the water and with biological growth. The hard layer was
broken up with hand tools and with vigorous backwashing. The run was then
continued but with a backwashing frequency of once every two days. The
pressure drop per column just before backwashing varied from 1 to 30 psi,
but averaged about 4 psi.
The amount of lost production caused by backwashing amounts to 20 minutes
each time or 10 minutes per day, when backwashing every other day, or
only 0.7% of the normal daily production. The volume of backwash water
for 15 minutes of backwashing at three times the normal flow rate would
amount to 23 minutes of production per day, or 1.6% of the daily production.
However, it is expected that the water from backwashing would be reused
for backwashing after being settled or treated by coagulation to remove
most of the material displaced during backwashing. The volume of water
needed to be held for backwashing could be further reduced by backwashing
several columns in series.
106
-------
DISPLACEMENT OF TOC AND COLOR FROM COLUMNS DURING BACKWASHING
The water coming from the columns during backwashing was very black
and contained a substantial amount of color and TOC. Tests were run
to determine the quantities of color and TOC that were displaced during
backwashing of the columns. Four of the tests were during the primary-
carbon runs and one during the bio-carbon runs. In these tests the flow
of water from the column being backwashed was held constant at 30 gpm
and samples of the backwash water were taken about five times during a
backwash period of 20-60 minutes.
In one of the tests that was continued for 60 minutes, it was estimated
that a total backwashing time of 110 minutes would be required before
the backwash water would no longer contain an appreciable amount of
color or TOC and that approximately half of the total displaceable
materials were displaced in the normal backwash period of 20 to 30 min-
utes. The backwash from the lead column did not contain much greater
concentrations .of color and TOC than that from the second or third
columns. The average results from these five tests given in Table 26
show that backwashing displaced about 11% of the color and TOC that had
been removed by the carbon since the prior backwashing.
The ratio of filter paper color (FPCU) to Millipore color (MPCU) is a
measure of the size of the colloids that make up a portion of the color.
During the period of maximum color displacement by backwashing, the ratio
of FPCU/MPCU was 3.7 but decreased to 1.7 for the average backwash water.
The amount of color in the size fraction less than filter paper and greater
than 0.8 micron Millipore size was 274% of that in the "true" color range
of minus 0.8 micron Millipore filter. During the average backwashing
period this percentage decreased to 65%. These results indicate that the
color bodies displaced during the first part of the backwashing period
are much larger in size than those displaced later in the period and
that a large portion of the apparent color (FPCU) is not "true" color
(MPCU). The color bodies in the more highly colored initial fraction
of backwash water contained 18 FPCU per mg of TOC; whereas the average
backwash water contained 9 FPCU per mg of TOC. The feed water to the
columns during the primary-carbon operation contained only about 5 FPCU
per mg of TOC. This comparison of FPCU/TOC ratios indicates that the
color bodies that are deposited during normal operation and displaced
during backwashing are not only much larger than the color bodies in the
feed water but are much more intense in color per atom of carbon they
contain.
107
-------
Table 26. DISPLACEMENT OF COLOR AND TOC DURING BACKWASHING OF COLUMNS
Color TOC, Ratio
FPCUb MPClF mg/1 FPCU/MPCU FPCU/TOC
Average3 max. concentration 3674 983 210 3.7 17.5
Average concentration 1213 734 137 1.7 8.9
Amt. displaced per g carbon 8.3 3.6 0.56
Hrs. of loading removed in 4.1 9.4
backwashing
Amt. displaced in backwash-
ing as % of amt. removed
in normal operation since
last backwash 11 11
a Average of results from backwashing of 8 columns in 5 tests for 20 to
60 minutes each.
Measured after filtration through Whatman No. 2 paper filter.
c Measured after filtration through 0.8 micron Millipore filter.
NON-ADSORPTIVE REMOVAL MECHANISMS IN CARBON COLUMNS
The objective of the third operating period of the primary-carbon sequence
was to determine the residual removal capacity of the two most spent col-
umns without further carbon replacement. A potentially significant resid-
ual removal capacity of the carbon had been inferred during preceding
operations when significant removals were still being obtained from the
first or lead column (which is the most nearly spent column) just prior
to its replacement as necessitated to maintain desired quality of effluent
from the fourth column. This prolonged test of residual removal capacity
was continued 15 times the normal length of time between column changes,
with backwashing.
This run was carried out with the third and fourth columns from the pre-
vious primary-carbon runs using primary clarified water from the basin
at a flow of 10 gpm to the columns. The columns were backwashed every
two days. During this run, no fresh carbon was added. The run was contin-
ued for a total of 74 days from June 26, 1972 to September 9, 1972 until
the rates of removal appeared to have reached a constant value. After
the first 23 days, the color removal rates for both columns had decreased
significantly and were about 0.5 CU/(g hr), but the TOC rates had decreased
only to about 0.15 mg/(g hr). During the following 21 days, TOC removal
rates continued to drop, while color removal remained essentially steady.
During the final 30 days of the run, both color and TOC removal were
fairly constant and were about the same for the second column as for the
first column. The average final equilibrium removal rates were 0.5 CU/
(g hr) and 0.06 mg TOC/(g hr). As shown in Table 27 these rates were
20-30% of those in the first column during the primary-carbon run of
April 25, 1972 to May 20, 1972, also at 10 gpm.
108
-------
Table 27. COMPARISON OF REMOVAL RATES OF EXHAUSTED CARBON
WITH NORMAL OPERATION
Removal Rates
a , , Color, CU/(g hr) TOG, mg/(g hr)
Final exhaustion run 8/9 to 9/9/72, 10 gpm 0.5 o 06
Adsorption runb 4/25 to 5/20/72, 10 gpm 1^66 o!30
Exhaustion run rate as percent of adsorption
run rate 30% 20%
a For both columns
k First two columns
During the final 30 days of this run, fairly wide fluctuations in concen-
trations of the water to the columns had very little effect on removal
rates, compared to a strong dependence during normal operation. Also,
removal in both columns proceeded at equal rates, although water concen-
trations decreased from the first to the second column. In normal
operation, removal rates in the second column are substantially below
those in the first column, as seen in Table 25. By the beginning of the
final 30 day period, cumulative removals had increased to levels well
beyond equilibrium loadings estimated from a number of isotherms on
similar effluent (see Section XIII, Figures 26 and 27). Cumulative
removals actually amounted to the following multiples of equilibrium
loadings: color, first column, 3.5, second column 3.5; TOG, first
column, 1.7, second column, 1.8. Allowing for partial displacement
during backwash as discussed earlier in this section would reduce these
factors by a maxiumum of 20%.
These data and observations suggest that during this extended run,
adsorptive TOG removal had essentially ceased by the beginning of the
final 30-day period, while adsorptive color removal had ceased prior to
that time. Continued removal of both color and TOG therefore requires
explanation by other mechanisms. These mechanisms are probably also
operative during normal operation. Potential mechanisms include bio-
logical activity, filtration, and colloid destabilization.
Evidence of bio-activity was found in samples of carbon from the columns
containing high concentrations of organisms similar to those found during
bio-carbon operation discussed in Section IX. Biological mechanisms are
generally believed not to effect color removal from kraft effluent. On
the other hand, removal of COD and TOG by bio-activity in carbon columns
has been found to amount to an appreciable portion of total COD and TOG
removals in work with municipal effluents (6,16,37). The dissolved
oxygen in the column feed water during the primary-carbon operation,
109
-------
including the extended run discussed here, was always less than 1 mg/1,
hence any bio-activity must have been anaerobic, after consumption of
the restricted amount of oxygen in the top one to two feet of the first
column. The reported rates of TOG digestion in anaerobic sludge digestors
for a large number of installations were equivalent to 0.01 to 0.02 mg/
(g hr) when adjusted to the volumes of solids and the temperatures in
the carbon columns. Therefore, anaerobic activity could account for 0.01
to 0.02 mg TOC/(g hr) out of a removal rate of 0.06 mg TOC/(g hr) during
the final period, that is 17 to 33% of the total TOC removal.
Filtration undoubtedly occurs in the carbon columns (the sand-filter was
not in use during the extended run), as evidenced by the reduction in
suspended solids (see Table 25). However, it must be borne in mind that
TOC is measured on a paper-filtered (about 2 micron) basis, while color
is measured on a Millipore-filtered (0.8 micron) basis. Hence, any
explanation of color removal by filtration in the column implies the
removal of particles less than 0.8 microns in size, which are extremely
difficult to filter, or less than 2 microns in the case of TOC.
A mechanism of colloid destabilization and coagulation in conjunction
with filtration is suggested to explain the unexpected removals of color
and TOC. This mechanism is further discussed in Section XIII. Desta-
bilization may occur due to removal of stabilizing substances by adsorp-
tion or bio-activity, or by a more direct functionality of the carbon
surface. The destabilized colloidal particles coagulate and are then
more readily trapped by the carbon in the column by a filtration mech-
anism. Such filtered substances would then be subject to displacement
from the column by backwashing. The displacement by backwashing amounts
to about 20% of the color and TOC removed from solution between backwashes
during normal operation. During the extended operating phase (final 30
day period), this amount is equivalent to displacing by backwashing 66%
of the color and 100% of the total TOC removed from solution between
backwashes. During the latter period, the backwash water did contain
substantial amounts of black suspended materials, however, no quantitative
data on these were obtained.
Undoubtedly, a number of different mechanisms are at work simultaneously.
Similarly, during normal operation removals are probably increased above
strictly adsorptive removal by the non-adsorptive mechanisms in evidence
during the extended run. It may be argued that during normal operation
the rates of non-adsorptive removal are as high as those shown above, i.e.,
that 20% of the TOC removal normally observed is non-adsorptive, consist-
ing perhaps of up to 7% by bio-activity and 13% by the suggested colloid
destabilization - filtration mechanism. At the same time, 30% of the color
removal is non-adsorptive and is removed by the colloid destabilization -
filtration mechanism if bio-activity may be held (as is generally done)
not to affect color.
110
-------
SECTION XI
OPERATING RESULTS:
LIME - CARBON COLUMNS
OBJECTIVES AND DESCRIPTION OF OPERATION
The sequence of lime treatment followed by carbon adsorption (lime-carbon)
was evaluated while treating mill effluent that was passed through the
equalization basin and lime treatment at a flow of 22 gpm. The operation
of lime treatment was described in Section VIII. The lime-treated water
was fed to two carbon columns in series at a flow of 10 gpm, or 1.4 gpm/ft
The total height of carbon bed was 20 ft, which provided a superficial
contact time of 1.8 hr and a bed volume rate of 0.55 volume of water per
volume of carbon per hour. The columns were backwashed every two days.
The major objective of this lime-carbon sequence of operation was to
determine the operating performance of the carbon columns with mill efflu-
ent treated by the microlime process, as discussed in Section VIII. The
goal of the microlime treatment was to use dosages of lime sufficiently
low to provide treated water with such low dissolved calcium content that
there would be no need for carbonation. This level of dissolved calcium
was defined as about 80 mg/1 Ca, or a range of 60-100 mg/1. On the other
hand, the microlime treatment was expected to be sufficient to make any
remaining color and TOG readily removable by carbon column adsorption.
REMOVALS OF COLOR AND TOG; EFFECT OF SOLUBLE CALCIUM CONCENTRATION
The microlime-carbon treatment was shown to be capable of reducing color
to less than 100 CU and TOG to about 100 mg/1, when the level of soluble
calcium was in the 50-100 mg/1 range. The percentage removals of color
and TOC by carbon adsorption increased and the concentrations of color
and TOC in both the feed and product water from the columns decreased
as the soluble calcium concentration of the water from lime treatment to
the carbon columns increased. Removals of color and TOC appeared to be
independent of pH in the range of 7 to 12.
Final color and TOC levels obtained in the combined lime-carbon treatment
are shown in Figure 18 as a function of dissolved Ca in the column feed.
Each point represents a run consisting of consecutive daily results with-
in a fairly narrow range of dissolved calcium. Runs immediately after
column changes are identified as squares in this figure (three runs).
Ill
-------
1-4
O
CJ
I
en
§
200
r-l
"{If 150
• 100
u
8 50
0
400
300
M O
E-! ^
o
o
200
100
0
CK
O
i
Q
o
•o-
o
CT
20 40 60 80 100 120 140 160
'SOLUBLE CALCIUM IN WATER TO COLUMNS, mg/1
Figure 18. Effect of calcium concentration on removal of color and
TOG by carbon adsorption (averages from operating periods)
112
-------
The data in this figure include all runs regardless of whether or not
carbonation was used after lime treatment, using for correlation in
all cases the dissolved calcium in the column feed. As this figure
shows, both color and TOG tend to level off when more than a certain
dissolved calcium concentration is maintained, i.e., the color was about
60 CU at calcium concentrations above 80 mg/1, and TOC was about 100 mg/1
with calcium concentrations above 40 mg/1. Other factors such as service
age of the carbon and concentrations of the feed to the lime treater
caused some fluctuations from the levels stated.
Since the soluble calcium content of the water to the columns (and the
related concentrations of color and TOC) was found to have a major influ-
ence on the adsorption of color and TOC from the water, data from all of
the lime-carbon operation were grouped into ranges of soluble calcium
in the water to the columns. The operating conditions and results of
the lime-carbon operation are given by this method of grouping in Tables
28 and 29. Additional results from the carbon adsorption step are given
in Table 30. A printout of daily summaries and conditions is given in
Appendix G.
In the low-calcium range of 10-50 mg/1, which averaged 42 mg/1, carbon
adsorption reduced the color from 276 to 214 CU, or by only 22%. At
the medium-calcium range of 50-100 mg/1, which averaged 86 mg/1, good
removals of color were obtained in the carbon columns, with the feed
color reduced from 252 to 76 CU. This removal was 71% of the color to
the columns (Table 30) and was 21% of the color to the lime treater.
The total removal by the lime-carbon sequence in the medium lime range
was 91%.
Color measured after filtration through 2-micron paper filters was found
to give valuable insight into the colloidal nature and size of the color
bodies. A comparison of colors of product water after being filtered
through filter paper and through 0.8 micron Millipore filters shows that
a large portion of the apparent (unfiltered) color of the treated water
at low calcium levels was caused by colloidal color bodies or colloidal
calcium carbonate smaller than about 2 microns and larger than 0.8 micron.
TOC removal was influenced less than color removal by the calcium concen-
tration, probably because a larger portion of the TOC is made up of com-
pounds that do not react with calcium. With the medium-calcium runs,
the TOC was reduced to 100 mg/1 TOC to give a reduction of 44% of the
TOC in the feed to the columns and 28% of the feed to lime treatment.
The total removal of TOC in lime-carbon treatment was 63%.
The variation of color and TOC concentrations of water with length of
carbon adsorption bed are shown in Figure 19 for the three groups of runs.
The slope of these lines is steeper as length of column increases, thus>
in contrast to the results with the primary-carbon runs, the last (second)
113
-------
Table 28. LIME-CARBON SEQUENCE OF TREATMENT:
REMOVALS OF COLOR AND TOC
Low calcium Med. calcium High calcium
Soluble Ca to cols., mg/1 range 10-50 50-100 100-200
avg. 42 86 138
Lime dosage, Ca(OH)2, mg/1 592 692 709
CaO, mg/1 448 523 536
pH of water to columns 9.7 11.3 11.6
Color CU
to lime treatment, 861 852 937
to carbon columns 276 252 200
from carbon columns 214 76 61
TOC, filtered, mg/1
to lime treatment 225 272 275
to carbon columns 165 177 165
from carbon columns 118 100 98
% removal of color to lime trt.
in lime treatment 68 70 79
in carbon columns 7 21 15
total 75 91 94
7o removal of color to columns 22 70 70
% removal of TOC to lime trt.
in lime treatment 27 35 40
in carbon columns 21 28 24
total 48 63 64
% removal of TOC to columns 29 44 41
114
-------
Table 29. SUMMARY OF CARBON ADSORPTION FOLLOWING LIME TREATMENT:
REMOVALS OF BOD AND TURBIDITY
Soluble Ca to cols., mg/1 range
avg.
Lime dosage, Ca(OH)25 mg/1
pH of water to columns
BOD, mg/1
to lime treatment
to carbon columns
from carbon columns
70 removal of BOD to lime trt.
in lime treatment
in carbon columns
total
Turbidity, JTU
to lime treatment
to carbon columns
from carbon columns
Low calcium Med. calcium High calcium
10-50
42
592
9.7
234
212
128
9
36
45
26
52
30
50-100
86
692
11.3
166
150
130
10
12
22
24
30
12
100-200
138
709
11.6
254
235
171
8
25
33
24
50
7
-------
Table 30. CARBON ADSORPTION TREATMENT OF LIME-TREATED EFFLUENT
FLOW = 10 GPM
Low calcium
No. of days
Soluble Ca
Temperature
Hours since
pH
Color, FPCU
CU
TOC, mg/1,
BOD, mg/1
averaged
to cols., mg/1 range
avg.
from cols. °F
fresh col., 1st col
2nd col
, 1st col.
2nd col.
, 1st col.
2nd col.
1st col.
2nd col.
Turbidity, JTU
Conductivity, micromhos
7» removals
Color, % of
in columns
color to 1st col.
total
TOC, % of TOC to 1st col.
BOD removal
total
, total, %a
21
10-50
»
.
To
1000
276
165
212
52
2245
1st
8
16
42
81
1390
796
9.7
From
897
649
253
214
139
118
128
30
2nd
14
22
13
29
40
Med
. calcium
30
50-100
To
1092
252
177
150
30
3597
1st
29
19
56
86
1056
447
11.3
From
724
254
180
76
144
100
130
12
2nd
41
70
25
44
13
High calcium
64
100-200
To
914
200
165
235
50
3749
1st
28
20
138
81
1103
396
11.6
From
455
188
145
61
133
98
171
7
2nd
42
70
21
41
27
Rates of removal
Color, CU/g hr
avg.
TOC, mg/g hr
avg.
0.18 0.30
0.25
0.08 0.07
0.075
0.41 0.66
0.54
0.10 0.14
0.12
0.40 0.44
0.42
0.10 0.11
0.11
Some values are averages of only 2 or 3 determinations
116
-------
300
200
q 100 -
O
u
300
o
H
10 20
CARBON BED LENGTH, ft
30
Figure 19. Concentrations of color and TOG remaining
versus length of carbon bed
117
-------
column in lime-carbon adsorption had a greater adsorption rate than the
lead column, especially in the case of color adsorption. These curves
show that with the medium level of soluble calcium and with the other
operating conditions used in the pilot plant, the length of bed needed
to reach 100 CU is 17 ft and the length needed to reach 100 mg/1 TOC is
20 ft.
The color remaining in the water from each column as a function of oper-
ting time is given in Figure 20. The color is expressed as a percentage
of the feed color to each column to normalize to some extent the variation
of feed concentration. A similar plot of data for TOC concentrations is
given in Figure 21. These curves show that there is an immediate break-
through of nearly 50% of the TOC, and a rapid breakthrough of some of
the color as well, in the second column after it is added as the fresh
column. In the case of TOC, the water from the fresh column jumped with-
in one day to 40-50 mg/1 TOC, which was caused by the non-adsorbable
low-molecular-weight compounds in the water.
The color and TOC concentrations after each step in the lime-carbon
sequence are shown in Figure 22 for the medium-calcium runs. Note that
the color drops sharply in lime treatment as compared to a smaller per-
centage drop in TOC concentration. The slight increase of color and
TOC in the carbonator-filter step was caused by the use of carbonation
during one run of this four-run medium-calcium group. As discussed
previously, carbonation caused redissolution of floe not removed in the
lime-treater clarifier.
DOSAGE OF CARBON REQUIRED
The lime-carbon sequence was operated for a relatively long time primarily
to provide reliable information on how much water could be treated by
carbon adsorption per unit weight of fresh carbon. Since the lime-treated
feed water to the columns had low concentrations of color, it was necessary
to operate an average of 44 days before the product water increased in
color and TOC to the point that a fresh column was needed. The lime-carbon
sequence was operated for a total of 160 days, and four fresh columns of
carbon were added.
On the basis of the last three column changes, the average dosage of fresh
carbon used was 2.5 Ib per 1000 gal. The dosage requirements would be
lower if the total bed length were increased, the contact time increased,
or the feed water contained less color and TOC.
118
-------
100
en
o
CO C_>
So
o w
u w
i-*^
SJ
21
M
Crf
O O
Z i-J
M O
a u
DATE
Figure 20. Remaining color in water from columns
during lime-carbon sequence
-------
100
to
o
cn
w o
PL, O
CO O
O O
O W
o
So
7/26
DATE
Figure 21. Remaining TOC in water from columns
during lime-carbon sequence
-------
1000
60
a
o
o
H
o
v^-
I
o
800
600
400
200 U
BASIN LIME CARBONATOR FIRST SECOND
TREATER & FILTER COLUMN COLUMN
Figure 22. Color and TOG versus stage of lime-carbon
treatment for medium calcium runs
121
-------
RATES OF REMOVAL OF COLOR AND TOG
The average rates of removal of color and TOG are given in Table 30 for
each column. The rates were almost doubled in going from the low-calcium
runs to the medium-calcium runs. The rates for medium-calcium runs of
0.54 CU/g hr and 0.12 mg TOC/g hr were slightly greater than those of
the high-calcium runs, but this increase was probably due to the higher
feed concentrations during the medium-calcium runs.
At approximately equal concentrations, the rates of color removal in the
lime-carbon column sequence were much greater than those of the bio-carbon
and primary-carbon sequence, but the rates of TOG removal were much less
than those for the bio-carbon and primary-carbon runs. However, in the
comparison with TOG removal during bio-carbon, an adjustment of the
reference concentration of TOG in the lime-carbon sequence by deducting
the non-adsorbable organic solution (volatiles discussed later) results
in more nearly equal TOG removal rates at equal adjusted concentrations.
The rates of removal of color and TOG were generally greater in the second
column than in the first column even though the concentrations were much
greater in the first column. It is possible that the carbon in the first
column was subject to blinding by Ca-organics or residual high molecular
weight color bodies, the latter also being suspected of that effect in
the primary-carbon operations.
EFFECT OF OPERATING VARIABLES ON REMOVAL OF COLOR AND TOG
Effect of pH
The effect of pH at low and high levels of soluble calcium was investi-
gated in a series of runs, each lasting several days. The purpose was
to determine whether the beneficial effects of calcium on the lime-
carbon sequence were enhanced or not by the high pH that was provided
by lime treatment. It is generally recognized that the color intensity
and the solubility of the color bodies of kraft mill effluents are
increased by adjusting the pH from the normal level of 8-10 to a pH of
12. Therefore, one might expect that adsorption of color bodies by
carbon at a high pH of 12 to 12.5 would be poorer than at a lower pH.
Also, carbon adsorption of organics in general is generally thought
to be better at lower pH levels.
These pilot plant tests showed that good carbon adsorption performance
was obtained at pH 11.7, or greater, when the calcium content was great-
er than about 60 mg/1 and that adsorption remained good when the pH
was reduced to 8-9 if the calcium content remained above about 60 mg/1.
However, if the calcium content was reduced to the range of 7 to 23 mg/1
(the pH was 7-9), the carbon adsorption performance immediately became
very poor.
122
-------
A regression analysis of the data from 153 days of operation at a
fairly wide range of conditions showed that color and TOG removals
were influenced only slightly by changes in pH and that an increase
of pH tended to increase the percentage removal of color and TOG.
It therefore is apparent that pH is not a major determining factor of
adsorption performance. Particularly the high pH normally associated
with the preferred minimum dissolved calcium level of 60 mg/1 (as
discussed earlier for the microlime treatment) is not detrimental to
good performance of the carbon columns, as might have been expected.
Rather, the higher pH tends to improve column performance. Hence,
if it is desirable to reduce the pH of the water before reuse or dis-
charge, the water should be carbonated after carbon adsorption, rather
than before. To avoid precipitation of CaC03, such pH adjustment
should be taken to pH 9 or lower.
Effect of Feed Concentrations
The concentrations of color and TOG, as well as soluble calcium, in
the feed to the columns were major determinants of product color and
TOG and of rates of removal. The influences of independent variables
on the dependent variables were analyzed by means of plots of all
conditions and results as functions of operating days.
Further insight on the effects of one parameter on another were obtained
from a regression analysis which utilized a computer program. The
independent variables included in the regression analysis were concen-
trations in the feed of Ca, total inorganic carbon (TIC), pH, FPCU,
standard color, TOG, the cumulative hours of operation of the second
column since it was recharged with fresh carbon, and the cumulative
removals of color in each column. The dependent variables were color,
FPCU, and TOG from the columns, rates of removal of color and TOG,
turbidity of product water, percentage removals of color and TOG,
and the ratio of FPCU/TOCo
This regression analysis indicated that the parameters affecting color
and TOG of the product water in the order of their contribution were
color, FPCU, TOG and Ca of the feed, and hours of operation of the
second column. This analysis showed that pH and cumulative removals
were also factors, but the contributions of these variables were
already explained by the related variables of calcium concentration
and hours of operation since a new column was added. This analysis
indicated that the rates of removal were primarily functions of the
concentrations in the feed, the hours of operation since a fresh
column was added, and the related variable of cumulative removals by
the columns. The percentage removals were primarily influenced by
hours operated (and cumulative removals) and by the calcium content of
the water. The TIC (carbonate) concentration had an almost negligible
influence on column performance.
123
-------
CUMULATIVE REMOVALS BY CARBON
As discussed in Section X, it was estimated during the primary-carbon
sequence of operation that about 30% of the color and TOC removed by the
carbon was in turn eliminated from the carbon columns by the combination
of backwashing of the columns and bio-activity in the columns. These
measurements were not repeated during the lime-carbon operation, but it
is believed that a similar portion of the color and TOC was removed by
non-adsorptive mechanisms during lime-carbon operation. Therefore, the
term cumulative "removals" is used rather than "loadings" to indicate
the degree of saturation of the carbon.
The cumulative removals of color and TOC for each column of carbon are
shown as a function of operating time in Figures 23 and 24. The average
cumulative removals of the columns when discharged were 660 CU/g and 175
mg TOC/g. The slopes of these curves are directly proportional to rates
of removal. As pointed out previously, the rate of removal at any given
cumulative removal is almost entirely dependent on the concentrations
of color and TOC in the feed to that column. (Other factors affecting
rates are fixed in this case, such as flow velocity, amount of carbon
used, and temperature.) On several days the columns actually released
more color and TOC than they adsorbed, as indicated by the downturn of
the curves of cumulative removal in Figures 23 and 24. These releases
of color and TOC were caused by upsets in the lime treatment, such as
a failure of the lime feeder during the night.
It is normally expected that loadings on carbon columns will reach about
60% of equilibrium loadings established in batch isotherms. The average
loading expected in the first column on the basis of isotherms with
lime-treated water (see Section XIII) hence would have been 60% of 350
CU/g, or 210 CU/g, at the average concentration of about 220 CU in the
first column in the pilot plant. However, at 500 to 900 CU/g (Figure 23)
the cumulative removals of the carbon columns when discharged were 140
to 260% of the isotherm loadings of 350 CU/g, or 240 to 430% of the load-
ing that would normally have been expected. This difference between
isotherm loading and cumulative removal is evidently due to removals of
color by non-adsorptive mechanisms, discussed above, and in more detail
in Section X. The magnitude of the apparent non-adsorptive removals
is much greater in the lime-carbon sequence than was observed in the
normal operation of the primary-carbon sequence.
The average isotherm loading for TOC was 130 mg/1 at the average pilot
plant feed concentration of 170 mg/1, and this value was exceeded by
two of the last three columns in the lime-carbon operation (see Figure 24),
124
-------
1000 -
800 -
ro
Cn
O
U
I
o
M
53
(-1
600 -
400
200
0
Col
I
I
1
J_
1 10 20
JULY
1 10 20
AUG
1 10 20 1 10 20 1 10
SEPT OCT NOV
20
1 10
DEC
20
Figure 23. Cumulative removal of color versus
time for lime-carbon operation
-------
I-1
NJ
o
250
200
oo
00
S
8 15°
S
a
100
50
- Col
Col. B,
j_
28 10 20
JUNE JULY
1 10
AUG
20
1 10
SEPT
20
1 10
OCT
20
1 10
NOV
20
1
DEC
10 20
Figure 24. Cumulative removal of TOC versus
time for lime-carbon operation
-------
These cumulative removal curves show that the lead column when taken off-
line was removing color and TOG at a fairly good rate. Therefore, the
lead column could have been left on line for a longer period except that
a fresh column was needed to maintain low values of color and TOC in
the discharged water. In designing a commercial plant, the savings in
regeneration costs by using more columns in series must be balanced
against the greater fixed costs in providing the additional length of
carbon bed.
REMOVAL OF LOW MOLECULAR WEIGHT COMPOUNDS
A large portion of the TOC remaining after carbon adsorption was made
up of low molecular weight organic compounds, which are not readily
adsorbed by activated carbon. The average concentrations of identifiable
compounds during the lime-carbon operation are given in Table 31.
Table 31. ORGANIC COMPOUNDS IN LIME-CARBON OPERATIONS
To carbon From carbon
columns columns
Methanol, mg/1 37 37
Ethanol, mg/1 trace trace
Acetaldehyde, mg/1 19 18
Acetone, mg/1 10 14
Formic acid, mg/1 9 6
Acetic acid, mg/1 70 79
Total identified volatiles, mg/1 145 154
TOC of total identified volatiles, mg/1 60 65
Each value is an average from 8-12 analyses. The concentrations of some
of these volatiles varied with time. For example, the acetaldehyde
concentrations were 30-45 mg/1 in August and September, 1972, were 0-11
in October and November, 1972 and 11-14 later in November, 1972. As
discussed in Section VII on Clarification and Bio-oxidation, it is
believed that much of the acetic acid was formed by anaerobic activity
in the equalization basin. These volatiles apparently contributed about
65 mg/1 of the 100 mg/1 of TOC in the lime-carbon product water.
REMOVAL OF BOD AND TURBIDITY
As computed from the data in Table 29, the removals of BOD were generally
low in the carbon columns, ranging from 13 to 40% of the BOD in the feed
to the columns. The overall removals of BOD in the lime-carbon treatment
127
-------
ranged from 22 to 45% for the periods grouped by calcium levels. The
low removals of BOD can be readily understood by the relatively high
concentrations of hard-to-remove, low molecular weight compounds discussed
above, which are all bio-degradable.
Turbidity was contributed mainly by insoluble calcium-organic compounds.
As shown in Table 29, the turbidity of the water to the carbon columns
was lower in the medium-calcium runs than in the high- and low-calcium
runs. The columns removed 60% of the turbidity at the medium-calcium
content and 86% of the turbidity at the high-calcium content. A regres-
sion analysis of the lime-carbon data showed that the turbidity of the
carbon-treated water was directly related to hours of operation of the
second column since it was added as a fresh column, indirectly related
to calcium concentration (and the associated parameter of pH), and
directly related to the filter paper color of the feed water.
REMOVAL OF METAL IONS
On the basis of 16 sets of samples, the average soluble calcium content
of the water to the columns was reduced in passage through the columns
by 3 mg/1 from 80 mg/1 to 77 mg/1. The amount of total calcium removed,
including that in colloidal calcium-organic compounds or calcium carbonate,
was probably greater, but it was not determined.
Other metal ions to and from the columns were determined by analysis of
from one to six samples by atomic absorption after filtration of the
sample through a 0.8 micron Millipore filter. The average results of
these analyses are given in Table 32. Most metal ion concentrations
were affected very little by carbon adsorption. Only manganese and
potassium were removed to an appreciable extent.
128
-------
Table 32. METAL ION CONCENTRATIONS IN LIME-CARBON SEQUENCE
to lime treater
13
1.6
2.0
0.36
0.28
0.13
12
309
0.03
0.5
0.5
7.6
ConcentrationT
to columns
80
0.3
1.5
0.30
0.14
0.07
12
327
0.04
0.6
0.4
7.3
mg/1
from columns
77
0.36
2.6
0.34
0.03
0.06
3.5
_
0.03
0.6
_
_
No. values
Metal averaged
Ca 16
Mg 6
Al 5
Fe 6
Mn 6
Zn 5
K 1
Na 1
Cu 4
Ni 4
Ti 1
Si 1
PRESSURE DROP AND BACKWASHING OF COLUMNS
The lime-treated water that was used in the carbon columns was filtered
through the duomedia filter to remove as much of the suspended solids
as possible. The turbidity of lime-treated water was reduced by the filter
from about 60 to 45 JTU or by 15 JTU. The turbidity was further reduced
by the columns to 12 JTU in the medium-calcium runs and to 7 JTU in the
high calcium runs.
The columns were backwashed every two days throughout the lime-carbon
operation. The pressure drop across the first column just prior to
backwashing averaged 4 psi and across the second column averaged 2 psi.
The average pressure drop across both columns (20 ft of carbon) during
operation was about 2 psi. The pressure drop did not increase with the
length of time a given column was in service. There was no evidence of
calcium sludge deposits in the column after backwashing, as determined
by inspection of the carbon when discharged from a spent column.
During the first several weeks of the lime-carbon sequence, the columns
were backwashed with well water that had very little hardness, high
bicarbonate content, and a pH of 4-5. When the first spent column was
removed, we found that the lower two feet of the carbon was cemented
together with deposits of calcium carbonate. Evidently the high carbonate
content of the wash water formed the calcium carbonate from the calcium
in the lime-treated water which was at a pH of about 12. Thereafter,
product water was used for backwashing and no further trouble was experi-
enced with carbonate deposits.
129
-------
SECTION XII
OPERATING RESULTS:
LIME-FACET ADSORPTION
OBJECTIVES AND DESCRIPTION OF OPERATION
The FACET carbon adsorption system was under development in short-term
runs in the pilot plant prior to May, 1972. Continuous operation of the
unit was started on May 15, 1972, with a feed of lime-treated water.
After two weeks of operation, the FACET unit was shut down for the next
four months to give more attention to the lime-carbon column operation.
The FACET unit was then operated seven days per week from September 28,
1972 to December 7, 1972, with lime-treated water at 10 gpm. During
these periods of operation, techniques were perfected to give good control
of slurry concentrations in the FACET tanks and to give uniform rates of
transfer of carbon as a slurry of 10-15% solids from one tank to the
next and of discharge of spent carbon. The carbon feeder gave reliable
and consistent delivery of fresh carbon from the feeder storage hopper
to the third FACET tank (third in the direction of water flow). A flow
diagram for the FACET system is given in Figure 6.
RESULTS FROM FACET OPERATION WITH LIME-TREATED WATER
There were five operating periods of three to five days each in which
the feed water to FACET remained fairly constant and which appeared to
give results that are representative of the FACET operation. The condi-
tions and results from these periods are summarized in Table 33. The
five periods are designated as Runs A through E and are presented in
chronological order. The average temperature of the water was 85°F.
The conditions of operation that have a major effect on removals of
color and TOG include the concentration of color and TOC in the feed,
concentration of soluble calcium in the feed, carbon slurry concentration,
carbon feed (or dosage) rate, and turbidity and suspended solids of the
feedwater. The concentration of suspended solids and color in the feed-
water varied greatly as a result of unstable conditions in the lime treat-
ment step. Quite often, a color increase in the water from lime treat-
ment caused a much greater percentage increase in the color from the
FACET treatment.
The carbon transfer rates given in Table 33 under "Feed" are the rates
to the third tank, and the rates under the tank numbers are the rates
of transfer from each tank. The rate of discharge of spent carbon from
the system is that given as discharge from the No. 1 tank. All weights
of carbon, including the feed, are on an oven-dried basis.
130
-------
Table 33. RESULTS FROM FACET OPERATION WITH LIME-TREATED WATER AT 10 GPM
Run no .
Daces
Length, days
A
5/15-5/18/72
3.3
Feed or
Tank no.a
Carbon rate Ib/hr
Carbon in tank, g/100 ml
Ib
Lime to lime treater
mg CaO/1
Soluble Ca to FACET, mg/1
Color, FPCU
CU
Color removed, ?„ of feed
Color removal rate,
CU/(g hr)
TOC, mg/1
TOC removed, % of feed
TOC removal rate, mg/(g hr)
pH
Suspended solids, mg/1
Turbidity, JTU
BOD, mg/1
Total
16.3
10.3
543
600
101
1090
250
81
1.8
151
54
0.75
11.8
370
1
23
19
258
807
88
65
3.1
83
45
1.3
2
16
11
150
691
63
10
0.8
70
9
0.4
3
14
10
135
746
48
6
0.5
70
0
0
11.7
411
B
5/23-5/27/72
5
Feed or
Total
9.2
6.8
274
610
140
885
266
79
3.9
162
45
1.3
12.1
1
8.6
6.2
84
783
163
39
6.2
130
20
1.9
2
8.1
7.0
173
846
111
19
2.8
96
21
1.8
3
8.3
7.3
98
617
55
21
2.9
90
4
0.3
12.0
162
C
9/28-9/30/72
3
Feed or
Total
2.1
12.2
517
490
77
1086
264
78
2.1
168
51
0.78
12.0
25
242
1
3.6
14.5
204
952
196
26
1.7
116
30
1.23
2
3.2
12.7
179
1317
126
26
2.0
98
11
0.50
3
2.4
9.5
134
993
58
26
2.5
82
10
0.60
12.1
D
10/3-10/6/72
4
Feed or
Total
1.8
11.7
493
470
136
581
160
54
0.86
140
38
0.53
12.3
25
1
2.5
12.6
178
1055
152
5
0.23
131
6
0.25
2
2.0
12.7
179
820
95
36
1.59
101
22
0.84
3
1.4
9.7
137
.
922
74
13
0.77
87
10
0.51
11.9
250
E
10/26-11/6/72
12
Feed or
Total
2.7
14.3
605
670
153
872
157
53
0.71
158
36
0.46
12.1
40
266
1
1.7
14.0
197
1022
166
-6
-0.23
154
2
0.10
2
1.8
13.5
190
1055
100
42
1.74
132
14
0.57
3
2.4
15.5
218
741
73
17
0.62
101
20
0.71
12.0
416
146
Avg
•C, D,
E
19
Feed or
Total
2.42
13.4
567
600
137
844
174
59
0.96
156
39
0.54
12.0
1
2.17
13.7
194
1018
167
4
0.17
143
8
0.33
2
2.06
13.2
186
1047
103
37
1.75
120
15
0,62
3
2.19
13.3
188
819
71
18
0.95
95
16
0.66
12.0
280
Ratio, FPCU/CU
4.4 9.2
11
15
3.3 4.8
11
4.1 4.9
10
17
3.6 6.8
8.6 12
5.6 6.2
11
10
4.9
6.1
10
12
The results for tank 3 are for the water before it was filtered in the duomedia filter.
The averages are weighted by the days of operation of each run.
-------
SECTION XII
OPERATING RESULTS:
LIME-FACET ADSORPTION
OBJECTIVES AND DESCRIPTION OF OPERATION
The FACET carbon adsorption system was under development in short-term
runs in the pilot plant prior to May, 1972. Continuous operation of the
unit was started on May 15, 1972, with a feed of lime-treated water.
After two weeks of operation, the FACET unit was shut down for the next
four months to give more attention to the lime-carbon column operation.
The FACET unit was then operated seven days per week from September 28,
1972 to December 7, 1972, with lime-treated water at 10 gpm. During
these periods of operation, techniques were perfected to give good control
of slurry concentrations in the FACET tanks and to give uniform rates of
transfer of carbon as a slurry of 10-15% solids from one tank to the
next and of discharge of spent carbon. The carbon feeder gave reliable
and consistent delivery of fresh carbon from the feeder storage hopper
to the third FACET tank (third in the direction of water flow). A flow
diagram for the FACET system is given in Figure 6.
RESULTS FROM FACET OPERATION WITH LIME-TREATED WATER
There were five operating periods of three to five days each in which
the feed water to FACET remained fairly constant and which appeared to
give results that are representative of the FACET operation. The condi-
tions and results from these periods are summarized in Table 33. The
five periods are designated as Runs A through E and are presented in
chronological order. The average temperature of the water was 85°F.
The conditions of operation that have a major effect on removals of
color and TOG include the concentration of color and TOG in the feed,
concentration of soluble calcium in the feed, carbon slurry concentration,
carbon feed (or dosage) rate, and turbidity and suspended solids of the
feedwater. The concentration of suspended solids and color in the feed-
water varied greatly as a result of unstable conditions in the lime treat-
ment step. Quite often, a color increase in the water from lime treat-
ment caused a much greater percentage increase in the color from the
FACET treatment.
The carbon transfer rates given in Table 33 under "Feed" are the rates
to the third tank, and the rates under the tank numbers are the rates
of transfer from each tank. The rate of discharge of spent carbon from
the system is that given as discharge from the No. 1 tank. All weights
of carbon, including the feed, are on an oven-dried basis.
130
-------
Table 33. RESULTS FROM FACET OPERATION WITH LIME-TREATED WATER AT 10 GPM
Run no .
Dates
Length, days
A
5/15-5/18/72
3.3
Feed or
Tank no.a
Carbon rate Ib/hr
Carbon in tank, g/100 ml
Ib
Lime to lime treater
mg CaO/1
Soluble Ca to FACET, rng/1
Color, FPCU
CU
Color removed, 70 of feed
Color removal rate,
CU/(g hr)
TOC, mg/1
TOC removed, 7, of feed
TOC removal rate, mg/(g hr)
PH
Suspended solids, mg/1
Turbidity, JTU
BOD, mg/1
Total
16.3
10.3
543
600
101
1090
250
81
1.8
151
54
0.75
11.8
370
1
23
19
258
807
88
65
3.1
83
45
1.3
2
16
11
150
691
63
10
0.8
70
9
0.4
3
14
10
135
746
48
6
0.5
70
0
0
11.7
411
B
5/23-5/27/72
5
Feed or
Total
9.2
6.8
274
610
140
885
266
79
3.9
162
45
1.3
12.1
1
8.6
6.2
84
783
163
39
6.2
130
20
1.9
2
8.1
7.0
173
846
111
19
2.8
96
21
1.8
3
8.3
7.3
98
617
55
21
2.9
90
4
0.3
12.0
162
C
9/28-9/30/72
3
Feed or
Total
2.1
12.2
517
490
77
1086
264
78
2.1
168
51
0.78
12.0
25
242
1
3.6
14.5
204
952
196
26
1.7
116
30
1.23
2
3.2
12.7
179
1317
126
26
2.0
98
11
0.50
3
2.4
9.5
134
993
58
26
2.5
82
10
0.60
12.1
D
10/3-10/6/72
4
Feed or
Total
1.8
11.7
493
470
136
581
160
54
0.86
140
38
0.53
12.3
25
1
2.5
12.6
178
1055
152
5
0.23
131
6
0.25
2
2.0
12.7
179
820
95
36
1.59
101
22
0.84
3
1.4
9.7
137
*
922
74
13
0.77
87
10
0.51
11.9
25.0
E
10/26-11/6/72
12
Feed or
Total
2.7
14.3
605
670
153
872
157
53
0.71
158
36
0.46
12.1
40
266
1
1.7
14.0
197
1022
166
-6
-0.23
154
2
0.10
2
1.8
13.5
190
1055
100
42
1.74
132
14
0.57
3
2.4
15.5
218
741
73
17
0.62
101
20
0.71
12.0
416
146
Avg
.C, D, E
19
Feed or
Total
2.42
13.4
567
600
137
844
174
59
0.96
156
39
0.54
12.0
1 2
2.17 2.06
13.7 13.2
194 186
1018 1047
167 103
4 37
0.17 1.75
143 120
8 15
0.33 0.62
3
2.19
13.3
188
819
71
18
0.95
95
16
0.66
12.0
280
Ratio, FPCU/CU
4.4 9.2
11
15
3,3 4.8 7.6
11
4.1 4.9
10
17
3.6 6.8
8.6 12
5.6 6.2
11
10
4.9
6.1
10
12
j* The results for tank 3 are for the water before it was filtered in the duomedia filter.
The averages are weighted by the days of operation of each run.
-------
REMOVALS OF COLOR AMD TOG
Unusually low average values of color (48 and 55 CU) and TOG (70 and 90
mg/1) were obtained during Runs A and B. The good removals were probably
mainly due to the high carbon feed rates of 9.2 and 16.3 Ib/hr. At a
much lower average carbon feed rate of 2.4 Ib/hr (4.0 lb/1000 gal) during
Runs C, D, and E (19 days), the color was reduced from 171 to 71 CU for
a reduction of 59%, and the TOG was reduced from 156 to 95 mg/1, for a
reduction of 39%.
Also included in Table 33 are filter-paper colors (FPCU) because this
provides a measure of the amounts of colloidal color bodies resulting
from the FACET treatment that are removed by the 0.8 micron filter used
for standard color measurement. Most of the total (unfiltered) color
of the FACET product water was removable by filtration through Whatman
No. 2 paper or through the duomedia filter in the pilot plant. During
Runs A and B, the FACET product water after the duomedia filter had
filter-paper colors as low as 128 to 150 CU and standard colors in the
range of 30 to 60 CU. During most of the remaining periods, the product
water contained substantial amounts of colloidal color bodies that were
not removed by the duomedia filter or by filter paper. It was noted that
when the low filter-paper colors were obtained, the black color bodies
normally in the water from the last FACET tank were sufficiently large
that they would settle from the water when a sampe was left undisrupted
in a beaker for about 20 minutes. We were unable to pinpoint the condi-
tions of operation that would insure that the color bodies would always
be of the size that would be removed by the duomedia filter. A high
calcium content is believed to favor the growth or coagulation of the
larger color bodies during FACET treatment. Even though the data of
Table 33 do not show a very strong correlation between soluble calcium
concentration and product water colors, other pilot plant data indicate
that, in general, best removals of color were obtained with high concen-
trations of soluble calcium in the feedwater to FACET.
It was found that the ratio of filter-paper color (FPCU) to standard
color (CU) is a useful indicator of the size distribution of the color
bodies, and the average values of this ratio are given in Table 33. The
values of FPCU/CU increased from about 2.0 in the feed to lime treatment
to a range of 3.3-5.4 in the feedwater to FACET and to a maximum in the
third tank in the range of 10-17. This increase indicates that most of
the apparent color was in the range larger than the 0.8 micron size by
the time the water was discharged from the third tank. An experimental
program, discussed in Section XIII, indicates that some of the color
bodies in the product water that were larger than 0.8 micron were peptized
carbon, while some were coagulation products of color bodies originally
in the water. The exact contribution of each was not determined.
132
-------
The average TOG concentrations of the FACET product water during Runs A-E
ranged from 70 to 101 mg/1, and the percent removals of TOG from the
feedwater to FACET ranged from 36 to 53%. As would be expected, the
best removals of TOG occurred in Runs A and B which had high carbon feed
rates.
The values of TOG concentrations in the product water are believed to
be much higher than they would have been if it had not been for the low
molecular-weight compounds formed in the equalization basin by anaerobic
activity, which was discussed in the section on the primary-carbon
sequence (Section X).
The percentage removals of TOG and color by each stage varied widely
from one run to another. During Runs A and B with a high feed rate of
carbon, most of the color and TOG was removed in the first tank. These
results at the high carbon feed rate indicate that the color and TOG
remaining after the first stage was poorly adsorbed, even with relatively
fresh carbon. At the lower carbon feed rates of 1.8 to 2.7 Ib/hr during
Runs C, D, and E, a large portion of the total reduction of color and
TOG occurred in the second and third stages, which indicates that the
carbon feed rate was about the lowest that could be used. At the average
carbon feed rate of 2.2 Ib/hr for these runs, the dosage was 4.0 lb/1000
gal or 480 mg/1.
During Run E, on the basis of three sets of samples, the BOD was reduced
from 266 mg/1 to 146 mg/1, for a reduction of 45%.
ADSORPTION PERFORMANCE BY LIME-FACET COMPARED TO THAT BY LIME-CARBON COLUMNS
The rates of adsorption per unit weight of carbon in the system for FACET
operation during Runs C, D, and E, are compared in Table 34 to those from
microlime-carbon column operation at medium calcium concentrations. The
rate of removal of color in lime-FACET was 0.96 CU/(g hr) versus 0.59
CU/(g hr) for the lime-carbon column operation, or 1.6 times as great.
The rate of removal for TOG was 0.54 mg/(g hr) versus 0.14 mg/(g hr) for
lime-carbon columns, or 4 times as great.
The results from these two modes of operation are not strictly comparable.
For instance, the soluble calcium content of the FACET feed was higher
(137 versus 86 mg/1). The feed concentrations and the amounts of color
and TOG removed for FACET were less. The cumulative removals on the
spent carbon were less for the FACET operation, i.e., 213 versus 668 CU/g,
and 126 versus 263 mg TOC/g. The carbon dosage rate was greater for
FACET (4.0 versus 2.2 lb/1000 gal). The feed to the lime-carbon columns
was filtered through the duomedia filter, whereas the feed to FACET was
not, which put an additional load of suspended organics on the FACET
system. If the conditions in the FACET operation had been the same as
133
-------
Table 34. COMPARISON OF LIME-FACET AND LIME-CARBON COLUMN
OPERATION AT A WATER FLOW RATE OF 10 GPM
Lime-carbon Lime-FACET
Period used Med. calcium Runs C,D,E
Reference table 28a 33
Days of operation 30 19
Avg. soluble Ca, mg/1 86 137
PH 11.4 12.0
Carbon dose rate Ib DWB/1000 gal 2.2 4.0
Contact time of carbon, days 97 10
Contact time of water, hr 1.77 0.84
Carbon in system, Ib 2880 567
Mean particle size, microns 670 345
Color, feed, CU 252 174
product, CU 76 71
removed, CU 176 103
% removed 70 59
cum. removed, CU/g 668 213
rate, CU/(g hr) 0.59 0.96
TOC, feed, mg/1 177 156
product, mg/1 100 95
removed, mg/1 77 61
% removed 44 39
cum. removed, mg/g 263 126
rate, mg/(g hr) 0.14 0.54
Values involving weight of carbon have been adjusted to dry-carbon
basis.
134
-------
in the carbon-column runs, it is believed that the FACET rates would
have been lower but still substantially greater than those of the carbon-
column runs.
This comparison at equal water flow rates show that the rates of removal
with FACET were much greater, that the size of the contacting zone was
only 11% of that for the columns, and the inventory of carbon was only
18% of that for the columns. The advantages of FACET over column adsorp-
tion are that it permits changing of dose rate of carbon to meet changing
feed concentrations, it has ability to handle a feed with a high concen-
tration of filterable solids, it requires no backwashing, it will require
smaller and lower-cost equipment, it requires a smaller inventory of
carbon, and it provides a continuous flow of carbon to regeneration.
Potential disadvantages of FACET include: it might have a greater attri-
tion loss of carbon, it requires coagulation or filtration of the product
water, and it requires more automatic controls to regulate slurry concen-
trations. In a commercial unit, a greater number of stages would be used
which would permit a closer approach to true counter-current contacting,
high cumulative removals by the carbon, and lower dose rates.
ATTRITION OF CARBON IN FACET OPERATION
One of the major concerns in considering FACET as a viable commercial
process for treatment of water is whether attrition of the carbon particles
in stirred contactors would cause excessive loss of carbon as fines in
the product water. Data obtained to date indicate that loss of carbon
will constitute a minor operating cost.
Ten tests were made in the laboratory to determine attrition losses by
stirring slurries of Darco XPT in effluent water and in deionized water
for periods of 6 hr to 5 days. In most of the tests the carbon sample
was wet-screened through a 200-mesh sieve before and after the test, and
the loss in weight of +200 mesh particles was usually defined as the
attrition loss. Since it was found that the Darco XPT carbon has a high
water-soluble content of 1.2 to 3.5%, the carbon sample was first water
extracted to remove solubles. The attrition loss in the above tests was
generally less than 1% when the carbon was stirred 4-5 days and the results
calculated on weight of water-washed oven-dry carbon.
We attempted to determine attrition losses by measuring total losses of
carbon from the FACET system during the pilot plant runs by means of
weight balances over periods of about a week, but it was found that it
was very difficult to get a reliable weight balance on this system. To
get reliable data, the weight of carbon in the system at the start and end of
the test peripd was needed but could not be measured with sufficient precision.
135
-------
Another inaccuracy in making a weight balance was in measuring the rate
of removal of spent carbon with sufficient precision.
The amount of filterable solids in the product water could not be used
as a measure of attrition losses because the value of filterable solids
in the feed water (about 300 mg/1) was as great as that of the product
water, and both were difficult to determine precisely.
Some carbon was lost to the product water during the pilot plant runs
due to some of the feed carbon to the third tank not being completely
wetted which caused some particles to float and to be carried out in
the overflowing product water. The clarification zone of this tank was
later modified to prevent carry-over of fresh carbon in the product water.
136
-------
SECTION XIII
SUPPORTING LABORATORY STUDIES
A large amount of laboratory work was carried out during the construction
and operation of the pilot plant to provide information and guidance on
changes in the pilot plant and to provide answers to problems that arose.
A problem requiring the major effort was that of color increase in the
treated water during certain conditions of carbon adsorption. Other
laboratory studies included providing carbon adsorption isotherms of
various streams throughout the pilot plant operation, determining best
conditions of operation for lime treatment and carbon adsorption, and
special analyses. The findings from these studies are summarized below.
COLOR INCREASE IN PRESENCE OF ACTIVATED CARBON
Laboratory evaluations and pilot plant runs of the FACET stirred-carbon
adsorption system showed that there was generally an increase of color
of unfiltered product water even though the TOC of the water decreased.
In most cases, this apparent color was removed by filtering the water
through a No. 2 Whatman filter paper or a sand filter in the pilot plant
and was greatly reduced by filtering the sample through the 0.8 micron
filter used in determining standard color. Since the "color" produced
during contact of the water with stirred carbon was removed by the 0.8
micron filter, it was not "true" color but colloidal color. This increase
of apparent color was also noticed in preparing adsorption isotherms,
especially at low dosages of carbon of 0.2 to 0.5 g/1. The color that
was removed from the carbon columns in the pilot plant during backwashing
had much the same physical and chemical properties as that produced in
the FACET tanks.
This increase of color was studied extensively in the laboratory because
the formation of the colloidal color bodies would be expected to coat
the external surface of the carbon. This would greatly reduce the rate
of diffusion to the carbon pores and thus the rate of adsorption of the
smaller molecules on the inner surfaces of the carbon in both the FACET
and fixed-bed adsorption processes. It was desirable to learn how to
control the formation of the colloidal color in the FACET system so that
the colloids could be easily removed from the product water by sand filters
or by coagulation. It was also hoped that knowledge of the mechanism of
formation and the properties of the colloidal color would suggest improved
modes of operation of the fixed-bed and FACET carbon adsorption systems
or might lead to novel less-costly processes for coagulation and removal
of color from pulp mill effluents.
137
-------
The increase of color occurred whenever powdered or granular carbon was
stirred for 1-4 hours with pulp mill effluent or synthetic effluent
(dilute black liquor) if the ratio of initial color to external surface
area of the particles was in the range of about 5,000 to 30,000 CU per
square meter. Results of a representative experiment are shown in Figure
25. The color increase also occurred, at a much lower rate, when no carbon
was present and occurred in presence of non-carbon granular solids and
powders such as alumina, silica, diatomaceous earth, and pumice. The
increase of color occurred with each type of carbon tested: Darco S-51
powder, Darco XPT 40 x 140, Darco 20 x 40, and Westvaco WVL granular.
Most other brands of carbon exhibited the color increase at low dosages
in the standard isotherm tests.
Temperatures up to 90°C increased the rate of formation several fold over
that at 25°C. The rate of color increase was accelerated by increasing
the pH over the range of 8 to 12. Heavy metal ions (which are present
in the ash of the activated carbon and in the mill effluent) did not seem
to affect the rate of color increase. The rate of color production
decreased with time of stirring but was still significant after 4-24 hours.
An increase of solid carbon concentration in the water during stirring
over the range of 0.3 to 100 g/1 generally increased the production of
color, but the rate of increase per weight of solid was greatest at low
concentrations of solids. If the ratio of initial color to surface areas
of the solid was outside of the range of 5,000 to 30,000, there was little
increase of color.
The test conditions for maximum increase of color of pulp mill effluent
were: 5-10 g/1 of Darco 40 x 140 at 50°C for 1-4 hours. These conditions
gave increases of about 700 unfiltered CU/g of carbon and 100 paper filtered
CU/g.
The mechanisms causing this observed color increase were not pinpointed.
The results of the various studies indicate that, as might be expected,
some of the color is produced by carbon particles from the activated
carbon that were originally present as dust or that were removed by
attrition or by peptization in the presence of mill effluent. The remain-
ing increased color was evidently produced from substances in the effluent
itself.
The chemical and physical properties of the colloidal color bodies were
studied by filtering the samples through filters of 2-0.05 microns, TOC
analysis, microscopic observations, electron micrographs, ultraviolet
absorption, infrared absorption, elemental analyses, and action of solvents,
bleaching agents, and caustic. These studies showed that the color in-
crease was due primarily to formation of black color bodies larger than
1 micron. As time of stirring increased, the concentration of "true"
color passing 0.05 and 0.8 micron filters decreased, therefore there was
138
-------
50,000
CONDITIONS
100 g/1 40 x 140 MESH CARBON
25°C, pH = 10-11
COLOR MEAS. AT pH 10-11
SETTLED 20 MIN.,
UNFILTERED
rt 30,000
PAPER FILTER
0.8 MICRON FILTER
SETTLED 20 MIN.
PAPER FILTER
0.8 MICRON FILTER
I I
123
TIME OF STIRRING, hr
Figure 25. Changes in color of dilute black liquor
with time of stirring with activated carbon
139
-------
a shift from the small color bodies to the large colloidal bodies. The
fraction of color removed by filter paper and 0.8 micron Millipore filters
had very large concentrations of color per unit -of TOG 25 to 50 times
as great as that of mill effluent. The absence or presence of oxygen in
a stirred test had no apparent influence on the degree of color increase.
The color intensity of the color bodies from a color increase test was
not reduced by treatment with hypochlorite or hot caustic, and the color
was not extracted with methanol. Carbon stirred in water alone or with
detergent present did not give appreciable color increases which indicated
that attrition or peptization of the carbon were not major contributors
to the color increase. However, digestion of carbon in dilute HC1 at pH 2
caused significant color increase due to suspended black particles that
were removable by an 0.8 micron filter and suggested that peptization of
the carbon occurs under some conditions. Suspensions of activated carbon
of up to about 2 micron size were found to give a 70% detection as TOG
of the carbon actually present. Taken together with the observed high
CUrTOC ratio, this point tends to suggest that the color increase was
not from activated carbon particles.
A study of this problem by the Institute of Paper Chemistry (unpublished
report) indicated that a portion of the color from the backwashing of
carbon columns could be metal sulfides, that the color bodies after
methanol extraction had "infrared spectra" (actually IR scatter, i.e.,
no distinguishing IR adsorption bands were seen in either sample) similar
to those of activated carbon, and therefore the IPC concluded that colloi-
dal carbon from the activated carbon was the primary source of the color
of the backwash water.
Our studies have indicated that a substantial portion of the color in-
crease is due to color bodies formed from the organic matter of the water
in addition to that possibly contributed by the activated carbon. Possible
mechanisms include (1) destabilization of organic colloids by contact
with the carbon surfaces or after adsorption on the carbon (it appears
reasonable to assume that at least a portion of the effluent color bodies
is present in colloidal form, similar to lignin color in natural waters
(1,2), particularly in view of the high molecular weight determined
recently (43)), (2) formation of metal sulfides and metal-organic complexes
at the surface of the carbon that become filterable colloids, (3) oxidative
condensations to form insoluble compounds of lignin, as hypothesized by
Ganczarczyk (39) in bio-oxidation of pulp mill effluents, (4) free-radical
polymerization of lignin degradation products (42) and (5) other conden-
sation products from hydrogen bonding and other types of chemical bonds.
This study of color increase of pulp mill effluents in contact with
carbon under certain conditions has shown that the mechanisms causing
it are very complex and that methods of controlling it for improving
color removal by carbon are elusive. The study of this phenomenon is
being continued by St. Regis Paper Company.
140
-------
ADSORPTION ISOTHERMS
Equilibrium isotherms were prepared using Darco S-51, Darco 20 x 40,
and Darco XPT 40 x 140 (all pulverized to less than 325 mesh) with samples
of water from each step of treatment used in the pilot plant. Color and
TOC isotherms were prepared using 43 samples of water. Average isotherms
for each treatment step are given in Figure 26 for color and in Figure
27 for TOC. Average loadings from the isotherms for selected remaining
concentrations of color and TOC are given in Table 35. The procedure
used in making the isotherms is given in the Appendix B.
There was a considerable spread of loadings at a given color for a given
stream, probably because of differences in the chemicals present from
one sample to the next. The average loadings shown in Table 35 and Figure
26 indicate that with Darco carbon the greatest equilibrium loadings
(at given concentrations) of color were obtained with water from lime
treatment, lime-carbon, and bio-oxidation, but much lower color loadings
were obtained with water that had passed through the retention basin
and primary-carbon adsorption. The highest loadings of TOC at concentra-
tions above 80 mg/1 were with bio-oxidized water, retention basin water,
and lime treated water.
Table 35. AVERAGE EQUILIBRIUM LOADINGS ON CARBON
FROM ISOTHERMS PREPARED DURING PILOT PLANT OPERATION4
Water
Source
No. isotherms
averaged
Loading on carbon, CU/g
at 100 CU 500 CU 1000 CU
From mill ditch 7
From reten. basin 14
From P-C cols. 4
From lime trt. 7
From L-C cols. 5
From bio-oxid'n 2
From B-C cols. 1
186
87
87
179
260
120
230
345
290
207
318C
480
857
450
2600
TOC loading on carbon,mg/g
at 80mg/l
29
30
21
27
16
58
20
lOOmg/1
41
47
33
63
29
73
-
150mg/l
61
144
70
116
60
112
-
a Isotherms all with powdered Darco S-51, pulverized -325 mesh Darco
20 x 40, or pulverized Darco XPT 40 x 140.
Abreviations:
carbon.
At 200 CU.
P-C = primary-carbon, L-C = lime carbon, B-C = bio-
141
-------
1000
60
o
s
O
u
100
10
BIO-CARBON
LIME-CARBON—>>
- LIME TRTR
BIO-OXID'N
A MILL DITCH
RETENTION BASIN
PRIMARY-CARBON
I i i i 11
10
100 1000
COLOR REMAINING, CU
Figure 26. Average color isotherms after various
treatments during pilot plant operations
142
-------
00
1?
o
o
I-J
u
8
100
10
LIME TRTR
BIO-CARBON
10
BIO-OXID'N
RETENTION BASIN
-MILI DITCH
LIME-CARBON
PRIMARY CARBON
' i
1 - 1 1
ll
100
TOG REMAINING, mg/1
1000
Figure 27. Average TOG isotherms after various treatments
during pilot plant operations
143
-------
The lowest TOG loadings were with lime-carbon and primary-carbon treated
water. These average isotherms indicate that the color remaining after
bio-carbon, lime-carbon, and lime treatments is more readily adsorbed
than the TOC and could have been removed to a greater extent with greater
retention time in the carbon columns.
The color isotherms generally have two slopes. At higher concentrations
remaining after treatment the color loadings decrease rapidly with concen-
tration, but at lower concentrations the loadings decrease less rapidly.
In contrast, the TOC loadings are very sensitive to concentration remain-
ing after treatment, with the bio-treated water being the least sensitive.
The steep slopes of the TOC isotherms show that high loadings (above 30-
80 mg/g) of TOC are achieved at concentrations greater than 100 mg/1 but
the loadings at less than 50 mg/1 are very low, which indicates that it
would be difficult to reduce TOC of mill effluents to concentrations of
less than 50 mg/1, except after bio-oxidation. The steep slopes and the
very low loadings of TOC obtainable at low residual TOC concentrations
can be taken as an indication of the presence of very-hard-to-adsorb
organic molecules. Actual chemical analyses presented in Sections X and
XI show hard-to-adsorb, low molecular weight organics in the 40 to 80
mg/1 of TOC range.
For all three treatment sequences, TOC isotherms obtained on effluent
already treated with carbon show lower loadings (at given concentrations),
particularly in the higher solution concentration range, than do isotherms
of the corresponding sequences before carbon treatment. This observation
is an indication that more readily adsorbed substances are adsorbed first,
hence the solution remaining after carbon column adsorption are of a
different qualitative composition, rather than being simply equally lower
in concentration of each of the many compounds representing the measured
quantities of TOC. This observation is of interest in conjunction with
the discussions of rates of adsorption in Sections IX, X and XI. In
the bio-carbon and primary-carbon sequences, adsorption rates decrease
with decreasing concentration due to a reduction of driving force both
from the reduction in concentration in solution and from the reduction
in equilibrium loading at given concentration. On the other hand, in
the lime-carbon sequence, rates in the second column, at lower effluent
concentration, are actually higher than in the first column. This is
contrary to expectations and is perhaps an indication of extensive
blinding of the carbon surfaces in the first column from calcium-organic
deposits.
The color adsorption isotherms before and after carbon column adsorption
show an opposite relationship to that observed for TOC. A possible
explanation for this phenomenon may be found in the known (43) wide
distribution of molecular weight (up to 64,000 MW) of color bodies in
kraft effluent before carbon treatment. Removal of color bodies in the
144
-------
high molecular weight ranges from the original effluent may be predominant-
ly in the very large pores and on the external surface of the carbon,
making the majority of the smaller pores inaccessible. Once the high
molecular weight fraction is removed and fresh carbon is brought into
contact with the solution, the lower molecular weight materials (perhaps
in the 500 to 2,000 MW range) have ready access to the full range of
carbon pore sizes and adsorptive area. This hypothesis is perhaps support-
ed by the higher loadings obtained in lime treated water as compared to
non-lime treated water. Lime treatment of unbleached kraft effluent has
been found (43) to remove high molecular weight color bodies in preference
to lower molecular weight color bodies.
Isotherms were used in checking the adsorption characteristics of several
lots of Darco 20 x 40 and XPT 40 x 140 received for use in the pilot plant.
Isotherms made with pulp mill effluent showed that the granular forms of
Darco achieved slightly greater loadings of color and TOG at given concen-
trations than S-51 when all were pulverized to pass a 325 mesh sieve.
DYNAMIC VERSUS EQUILIBRIUM CONCENTRATIONS IN PILOT PLANT COLUMNS
Several tests were made in the pilot plant to determine how close the
color and TOG concentrations in the carbon columns determined during
normal operations (dynamic concentrations) approached the equilibrium
concentrations for possible use in an evaluation of measured adsorption
rates. These tests were made by shutting off the flow of water through
the columns and then removing samples of water from the base of each column
at intervals of time up to four days. After about 8 hours, the concentra-
tions had almost reached equilibrium values. These tests indicated that
the columns were operating at color concentrations 30 to 100% higher than
the equilibrium values and at TOG concentrations 30 to 80% higher.
FOAMING TENDENCY OF TREATED EFFLUENT
The effect of the various steps of treatment on the foaming of the water
by a standard foaming test was checked on 10 samples over a 6-month period.
These tests indicated that the foaming of the mill effluent increased
50 to 100% during passage through the retention basin, without aeration.
Lime treatment reduced the foaming of the feed water about 50%. Carbon
adsorption by columns or FACET removed almost all of the foaming tendency.
145
-------
EFFECT OF OPERATING CONDITIONS ON ADSORPTION OF COLOR AND TOG ON CARBON
Effect of pH
The effect of pH of the feed water on adsorption was investigated in
several isotherm tests with mill effluent adjusted to pH values of
3 to 12. Within the range of 5-12, the results were inconclusive.
Loadings varied as much as two-fold between different pH levels for
specific samples, evaluated at a given color level. Results from test
to test and color level to color level, however, were largely contradic-
tory to each other. The limited amount of testing leaves the question
of pH effect unresolved. At pH levels of less than 4, the lignin
compounds become much less soluble and "loadings" (including precipita-
tion) in the carbon increase sharply.
Effect of Temperature
A limited number of isotherms prepared at temperatures of 75° and 130°F
with Darco S-51 and Westvaco AN and mill effluent indicated that the
higher temperature gave about twice the equilibrium loadings of color
at a color concentration of 300 CU and 1 to 2.6 times as much loading
of TOG at a concentration of 100 mg/1. No comparative isotherms were
obtained to observe a possible temperature effect on bio or lime treated
mill effluent. In the pilot plant, the operating temperature range
was very narrow (74-85°F). Hence, no temperature effect was observed.
The temperature effect deserves further study, particularly in view
of the fact that mill effluent can be available as high as 130-150°F,
thus requiring no added cost for heating if the temperature effect is
to be exploited in a large installation. The plant design and cost
studies presented in Section XIV are based on the pilot plant results
obtained at 74-85°F.
146
-------
SECTION XIV
TREATMENT PLANT DESIGN STUDIES AND COST ESTIMATES
OVERALL DESIGN CRITERIA
Engineering design studies were made for full-scale plants for treating
unbleached kraft mill effluents by the lime-carbon and bio-carbon sequences
to provide water suitable for reuse in the mill. These studies were
carried out to provide a basis for comparative cost estimates and to
illustrate the methods of equipment design and operating procedures that
appear best on the basis of our pilot plant studies and results from
other effluent treatment studies.
The design and cost studies were based on the conditions given in Table
36 and water compositions given in Table 37. The treatment sequences
discussed in this section and resultant water qualities are also present-
ed in Figure 28. These conditions were chosen as typical of new and
older unbleached kraft pulp and paper mills in the south. It was assumed
that the effluents will be from mills that have catch basins for spills
and sudden discharges of highly contaminated water. With such control,
it was assumed that the variation was such that for 90% of the time the
maximum color would be less than 1320 CU and that the minimum color would
be greater than 680 CU. With this degree of variation, the TOG concen-
tration would remain in the range of 175 to 325 mg/1 90% of the time.
It was assumed that the effluents from the older and new mills would
have the same concentrations of color, TOG, and BOD, but that the older
mills would have a 50% greater volume of effluent per ton of pulp produc-
tion.
The compositions of effluent after the various treatment steps and
sequences in Table 37 are based on results achieved in the pilot plant,
particularly in terms of microlime and carbon treatment, and on results
in mill scale practice of bio-oxidation and minilime treatment. For the
sequences involving activated carbon, the target of reusable water quality
was defined as 100 CU and 100 mg/1 TOG, although better results were
achieved in the pilot plant. For the microlime-bio and minilime-bio
sequences, the final color is based on that achieved after lime treatment,
although in both commercial operations of minilime-bio the color increases
to substantially higher levels in the bio-oxidation ponds, resulting in
levels of 250 to 500 CU. Color increases were also seen in the laboratory
as discussed in Section IV.
No cost estimate is included for the lime-FACET sequence because it was
felt that the status of development achieved in the FACET system was
not sufficient to warrant an estimate of comparable precision to those
147
-------
Table 36. BASES USED FOR DESIGN AND COST STUDIES
Case
no.
1
2
3
4
4a
5
5a
6
6a
7
8
9
10
11
Age
of
mill
new
new
older
older
older
older
older
older
older
new
new
older
new
older
Mill pulp
capacity,
tons/day
800
1600
800
800
800
800
800
800
800
800
800
800
800
800
Effluent
vo lume ,
gal/ton
12,000
12,000
18,000
18,000
18,000
18,000
18,000
18,000
18,000
12,000
12,000
18,000
12,000
18,000
mgd
9-6
19.2
14.4
14.4
14.4
14.4
14.4
14.4
14.4
9.6
9.6
14.4
9.6
14.4
Treatment sequence
microlime-carbon
microlime-carbon
microlime-carbon
prim-bio-carbon
prim-bio-carbon^
microlime-bio
microlime-biob
minilime-bio^
minilime-bio
minilime-carbon
minilime only
minilime only
microlime only
microlime only
Microlime and minilime treatments have no separate primary clarification.
Estimate excludes capital cost of bio-oxidation facilities.
Table 37. AVERAGE COMPOSITION OF WATER TO AND FROM TREATMENT3
From mill to treatment
From microlime
From minilime
From bio-oxidation
From microlime-carbon
From minilime-carbon
From bio-carbon
From microlime-bio
From minilime-bio
Color,
CU
1000
300
150
1000
100
100
100
300
150
TOC
mg/1
250
150
135
175
100
100
100
70
50
BOD
mg/1
250
225
187
25
150
117
20
20
20
_pJL
10.5
11.5
8
8
8
8
8
8
8
Other Properties:
Temperature: lime-carbon = 100°F., bio-carbon and bio-lime = 90°F.
Conductivity: minilime-carbon = 1200 micromhos, microlime-carbon =
3600 micromhos.
148
-------
COST ESTIMATE
1,2,3 EFFLUENT
FLOW DIAGRAM
4,4a
5,5a
6,6a
8,9
10,11
LIME-i COo-i
T Z 1
MICRO-
LIME
CARBON
pH
* ADJUST *
\ SLUDGE
NUTRIENTS
AIR ^
• PRIMARY
BIO
»m. PflDDflM ^
"• UMKtiUN »*-
» SLUDGE
NUTRIENTS
LIME-i AIR-n
MICRO-
LIME
*_,
BIO
—
T SLUDGE
LIME-n CO? -r
I T
. MINI-
LIME
T SLUDGE
LIME-r
MINI-
LIME
*^
T SLUDGE
LIME-i
MINI-
" LIME
{ SLUDGE
LIMEn
MICRO-
" LIME
CARBO-
NATOR
*• BIO ~*~
| SLUDGE
C02 -i C02 -i
CARBO-
NATOR
pH
*" ADJUST " CARCON
| SLUDGE
C02 -T C02 -i
CARBO-
NATOR
I
pH
ADJUST
pH ^
* ADJUST
SLUDGE
SLUDGE
EFFLUENT
Color TOC
CU mg/1
100 100
100 100
300 70
150 50
100 100
300 150
Figure 28. Flow diagrams and water quality for Cost Estimates 1-11
149
-------
presented. Substantial savings are expected, as indicated in Section V,
in equipment and carbon inventory cost, partially off-set (to a lesser
degree than indicated in the preliminary estimates of Section V) by
increased carbon dosages. The lime-FACET system operating costs, including
fixed costs, are expected to be lower by 10-15% or several cents per 1000
gal.
No estimate is included for the primary-carbon system. Rough estimates
made on the basis of pilot plant results indicated costs to be well out
of line with the other potential systems, primarily because of the high
capital cost requirement as a consequence of long contact time require-
ments.
The influence of excluding the capital cost of pre-existing bio-oxidation
facilities was determined for each estimate that included bio-oxidation.
The influence of price of activated carbon ($0.27 and 0.10 per pound)
was determined for most estimates that used carbon adsorption. The
comparative costs for microlime alone and minilime alone were determined
as were the incremental costs of carbon treatment following lime treat-
ment.
DESIGN METHODS AND ASSUMPTIONS
In this part of this section, the design methods and assumptions are
discussed as they apply to each treatment step. Further information
applying to design can be found in the appendix. Detailed design calcu-
lations for Estimate No. 1 are given in Appendix J.
Primary Clarification
The design for primary clarification includes a mechanical clarifier
in which the rise rate is 700 gpd/ft , in which no coagulants are
added, and which produces a clarified water containing 50 mg/1 of
suspended solids. Also included were water lift pumps, sludge pumps,
a sludge settling tank, and sludge filters. The dewatered solids are
hauled to landfill.
Bio-Oxidation
Bio-oxidation was assumed to be provided by aerated basins that are
typical of those used by St. Regis and other southern kraft pulp mills.
It was assumed that bio-oxidation removes 90% of the BOD, 30% of the
TOC, and none of the color of the water. Nutrients are added in the
ratio of 5 Ib of N and. 2.5 Ib of P for each 100 Ib of BOD loading.
Aerators are used at a ratio of 32 pounds of BOD in the water to bio-
oxidation per hp-day.
150
-------
Microlime Treatment
The design for the microlime treatment facilities is based primarily
on the results from the pilot plant operation but also on the results
from the Interstate Paper (18) and Continental Can (14) lime treatment
plants. The lime treatment system includes all auxiliary equipment:
conveyors from the mill's lime storage, a lime slaker, and a lime slurry
feeding system that is controlled by instrumentation to provide 60-90
mg/1 of soluble calcium in the lime-treated water when adding about 520
mg/1 of CaO. Sensors are used in the lime-water mixing zone of the
reactor-clarifier to measure the soluble calcium content and to provide
a control signal to the lime-slurry feeder. Lime slurry at 10% solids
is introduced into the mill effluent just ahead of the lift pump to the
lime treater.
The lime treater is a standard reactor-clarifier with a central agitated
mixing zone to provide a long retention time of 30 minutes. A solids
concentration of 0.5 to 1.0% solids is maintained in the mixing zone.
A reaction time of 10 minutes is adequate, but the 30-minute retention
time and high solids concentration are used to dampen short-term (less
than 2 hr) fluctuations of color load. Longer-term variations in con-
centration and flow rate are handled by varying the lime dosage rate
to maintain the desired concentration of soluble calcium.
The lime treater has heavy-duty sludge rakes to handle the dense sludge.
The sludge is pumped at about 15% solids to a 12,000 gal sludge settling
and surge tank in which the sludge is concentrated to about 30%. The
sludge is dewatered with a wire filter at a rate of 100 Ib/hr dry solids
per sq ft of filter area to provide a cake of about 70% solids which is
conveyed to the mill's lime kiln feeder. The sludge was estimated to
contain, on dry basis, about equal amounts of CaCO., and calcium-organics
and a total of 43% calcium determined as CaO.
The loss of lime as soluble calcium in the treated water was estimated
to be 21% of the lime added to the water in microlime treatment but
only 2% in minilime treatment. An additional loss of 5% of the lime
added was assumed in the handling of lime sludge and filtration. This
total loss of lime was assumed to be made up by purchased lime; all
other lime used was assumed to be recalcined lime from the lime kiln.
It was assumed that the lime cake from effluent treatment is recalcined
in an existing lime kiln. The percentage of added load to the kiln
from microlime treatment was 8% and from minilime treatment was 21%
in the case of new mills having 12,000 gal effluent per ton of pulp.
As discussed below, the capital costs for lime treatment include a
proportionate share of the lime kiln cost.
151
-------
Minilime Treatment
The design for minilime treatment is based primarily on the results
reported by Spruill (14) for the Continental Can Company's minilime
treatment of 9 mgd of unbleached kraft total mill effluent. Also used
in the design calculation were results from the St. Regis - EPA pilot
plant program, from the Interstate Paper Company minilime treatment
plant at Riceboro, Georgia (18) and from the Georgia Pacific Corporation
plant for lime treatment of bleach plant effluent at Woodland, Maine
(7,8).
The minilime treatment plant includes a lime feeder, lift pumps, sludge
pumps, lime treater-clarifier, carbonator (with kiln gas), carbonator-
clarifier, sludge dewatering, and conveyor to return lime sludge cake
to the lime kiln. The lime treater and carbonator clarifier use a
rise rate of 700 gpd/ft2.
Additional information is given under microlime treatment (last para-
graph) .
Carbon Adsorption
The design for the carbon adsorption plants was based primarily on
the results from the several modes of operation in the pilot plant.
The arrangement of adsorption contactors or columns, the transfer of
carbon as a water slurry, and the carbon regeneration unit were chosen
to be similar to those used in the EPA report prepared by M. W. Kellogg
Co. (17). In that report detailed designs and cost estimates were
presented for treating by carbon adsorption 1, 10 and 100 mgd of
municipal wastewater from biological treatment. These cost data are
apparently the most reliable of the available costs on carbon adsorp-
tion and were chosen as the basis of the costs used in this report
for carbon adsorption. A process flow diagram for that plant is
given in Figure 29.
The design procedure is discussed below. Further information for the
desgin of carbon adsorption plants is given in Appendix H, relation-
ships useful in the designs of downflow carbon columns are given in
Appendix I, while a detailed example is provided in Appendix J as
part of Estimate No. 1, microlime-carbon treatment.
The total volume of carbon for a system was found by dividing the
volumetric flow rate (volume of water per volume of carbon per hour)
by the volume of water to be treated per hour. The volumetric flow
152
-------
Process Water
Oi
NOTES
1. Tank A is operating as
Primary Adsorber.
2. Tank B is operating as
Secondary Adsorber.
3. Tank C is operating on
Backwash.
4. Tank D is Idle.
5, Tank E is operating on .
Cycle to Remove Carbon.
6. Tank F is operating on ;
Cycle to Refill with
Activated Carbon.
Figure 29. Process flow diagram for FWPCA-Kellogg study (17)
-------
rate is the inverse of contact time in hours. The contact time and
volumetric flow rates were determined from pilot plant results for
the feed and product concentration of each estimate. The superficial
velocity and the cross-sectional areas of the columns were selected
from pressure-drop considerations. The combined height of columns
in series stems directly from the contact time required. It was
assumed that the rate of adsorption would be the same as that in the
pilot plant even though the rate is expected to be greater in the
plant because of the greater flow velocity selected for the full-
scale plant. The number of columns in series was selected to be at
least three (to provide adequate counter-current contacting) and to
maintain a reasonable height/diameter ratio and to keep the height
at less than about 60 ft. The number of parallel trains was determined
by the need to keep the height/diameter ratio between 1 And 2. There-
fore, the diameter of the columns was fixed by the number of parallel
trains used. In general, one additional column was provided for about
every other train to handle excessive flows and concentrations of the
feed. Two additional columns were provided, one being emptied of
spent carbon and the other being refilled with regenerated carbon.
In operation, the lead (more nearly spent) column is removed from a
train of columns when the quality of the water from the last column
exceeds a tolerable level. A column of regenerated carbon is then
added as the last column, and the carbon from the removed spent
column is transferred as a slurry to the regeneration furnace. The
regenerated carbon from the furnace is quenched in water and transferred
to an empty column. When filled, this column is held in reserve to
be used behind the last column temporarily whenever an overload to
the system causes the product water to rise above a set tolerable
level. This reserve column is put into regular service when the
color or TOG concentration in the product water increases above the
selected value and the first column of that train is removed from
service for regeneration of the carbon.
Since the function of each column progresses through the series of
columns, adequate piping and valving is provided to give the required
flexibility.
The total inventory of carbon was determined from the number and
dimensions of columns. The total inventory also included a 30-day
supply of make-up carbon.
Backwashing of the columns requires a large flow of product water
(5300 gpm in Estimate No. 1) for 15 minutes every day for the lead
columns and every other day for the other columns. To minimize the
loss of production due to this flow, the same water is used repeatedly
for backwashing columns. In the lime-carbon sequences, the water is
recycled to lime treatment after use for backwashing six columns.
154
-------
Two clarifiers are provided to remove the suspended solids from the
backwash water before each reuse. With this arrangement the loss
of product water was only 0.8%. The off-time for backwashing reduced
production by 3%, making the total loss of production 3.8%. An
equivalent procedure limits the lost production due to backwashing
in the bio-carbon treatment sequence.
The carbon regeneration unit must provide regenerated carbon at the
dosage rate, which was 1000 Ib/hr in Estimate No. 1. With an expected
loss of 5% per regeneration cycle, the make-up rate with fresh carbon
was 1200 Ib/day. At the average regeneration rate in Estimate No. 1,
one column is regenerated each 10.6 days.
The major design assumptions used in the cost studies for carbon
adsorption are given below:
Carbon Columns
Carbon Type
Flow Velocity
Excess Capacity-
Backwash ing
Filtration
Carbon Make-up
Regeneration
down-flow, fixed bed, 50% free-board (referred
to carbon bed height).
20 x 40 mesh, equivalent to Granular Darco with
bulk density of 25 lb/ft3.
with lime-treated feed water =4.2 gpm/ft ,
contact time = 1.8 hr,
volumetric flow = 0.56 v/(v hr)
with bio-treated feed water =4.0 gpm/ft ,
contact time = 4.4 hr, volumetric rate = 0.23 v/(v hr)
based on average feed rates after allowance for
off-time for backwashing = 20 to 30% of normal
flow rate or normal contaminant loadings in the
feed water.
with air for 1 min and with water for 20 min at
10 gpm/ft . Lead columns are backwashed each
day, others every other day. Backwash water is
collected, clarified, and reused in backwashing
more-contaminated columns, and then returned to
lime treater. Water used is 1% of feed and off-
time is 3% for a total loss of production of 4%
for backwashing.
no filtration is required for feed to columns.
total loss from handling and regeneration of
carbon is assumed to be 5% per regeneration.
This rate of make-up was achieved in large-scale
treatment of municipal effluent. In treatment of
pulping effluents, it is reasonable to expect
that some of the adsorbed organics will form
activated carbon, which might cause a lower loss
rate when regenerated.
spent carbon is transferred as a slurry, dewatered,
and fed to a multiple-hearth furnace. A regeneration
furnace capacity of 100 lb of carbon per day per
square foot of hearth area is assumed and 25%
excess capacity is provided.
155
-------
COST ESTIMATES
Cost estimates were prepared for plants treating the effluent from un-
bleached kraft pulp and paper mills to show which items of capital cost
and operating cost are the most critical or contribute the greatest
amount to the overall operating costs, to show the effect of price of
carbon on the costs, to show the effect of including in the capital cost
of bio-carbon systems the cost of a new bio-oxidation unit or omitting
this cost (for mills that have already amortized the unit), to compare
the costs of microlime and minilime treatments, and to compare the costs
of treating effluent from new mills and older mills (larger water usage
per ton of pulp production).
Precision of Estimates
These estimates are believed to be within 20 to 30% of the actual costs.
The costs of one estimate relative to another for the assumed conditions
are expected to be much closer - probably within _ 10%. Since any par-
ticular mill will have a different set of conditions, the cost estimates
would have to be revised to suit those conditions before a reliable
cost for that mill can be obtained.
Bases of Estimates
The bases used in the design of the plants were discussed earlier in
this section and are presented, in part, in Tables 36 and 37. Details
of the assumptions used, the design procedures used, and method of
estimating plant capital and operating costs are given in Appendix J
for Estimate No. 1.
Capital Costs -
Capital costs are based on total plant costs for existing plants or
on published costs from detailed cost studies. All costs were adjust-
ed to January 1973 by use of the Engineering News Record Construction
Cost Index.
The capital costs and some of the operating costs for minilime treat-
ment are based on those of the 9 mgd lime treatment plant installed
by Continental Can Co. in 1972 at Hodge, La., which was supported by
an EPA grant (35). The costs are the actual costs for the complete
installation including engineering and contractor's overhead and
profit. Their costs were adjusted for a lower clarifier rise rate
used in our estimates and for larger flow rates by use of 0.7 as the
exponent of the ratio of flows.
156
-------
The capital cost for microlime treatment was derived from the Continental
Can cost data with adjustments made for eliminating the carbonation and
final clarification step, for lower lime dosage, and for adding a final
pH adjustment step. No pH adjustment step was added for the microlime-
bio sequence.
The capital costs for carbon adsorption were derived from those pre-
pared by M. W. Kellogg for EPA in 1969 (17) for 1, 10 and 100 mgd
carbon adsorption plants for treating municipal effluent. Costs of
each category of equipment of the Kellogg estimate were adjusted for
the increased volumes of carbon needed and for other differences in
our estimates, as illustrated in Appendix J.
The costs for primary clarification (for those estimates using bio-
oxidation) were based on the costs for complete units for St. Regis
mills.
The capital costs for aeration basins were based on the average costs
for three installations; two St. Regis mills and a Crown-Zellerbach
mill (36). These costs were adjusted for differences in flow rate
by use of an exponent of 0.7.
Operating Costs -
The bases for operating costs are as follows:
Amortization - 6.25% of total cost of investment (TCI) per year
Repair and Maintenance - 3% TCI/yr.
Taxes and Insurance - 2% TCI/yr.
Labor - On expected manpower requirements at $5.85/hr.
Plant Overhead - 75% of labor costs.
Steam - $0.80/1000 Ib.
Fuel - $0.60/mil. Btu.
Electricity - $0.01/kWh.
Lime - Reburned - $12.40/ton CaO
Make-up - $24.23/ton CaO
Carbon make-up - $0.27/lb and $0.10/lb
Credit for treated water to replace mill make-up water = $0.06/kgal
RESULTS OF COST ESTIMATES
The capital and operating costs for new and older mills and for various
treatment sequences are given in Table 38. The new mill was assumed to
produce 12,000 gal of effluent per ton of pulp production and the older
mill was assumed to produce 18,000 gal/ton. The quality of the water from
both mills to the effluent treatment plant was assumed to be the same.
The credit for treated water is based on typical mill water cost of
$0.10/kgal less $0.04/kgal for pumping the treated water to the mill supply
tank.
157
-------
Table 38. COMPARISON OF COSTS FOR TREATING UNBLEACHED KRAFT MILL EFFLUENT
IN NEW AND OLDER MILLS BY VARIOUS TREATMENT SEQUENCES WITH CARBON AT $0.27/lb
Ul
00
Estimate No.
Pulp tons per day
Effluent, gal/ton
Effluent, mil.' gal/day
Treatment
Product water quality
Color, CU
TOC, mg/1
BOD, mg/1
Capital cost, $ million
Lime trt. & pH adjustment
Carbon adsorption
Clarifier & bio-oxid'n
Total
Operating cost, $/kgal
Fixed
Labor, o.h.,util.
Lime and chemicals
Carbon, at $0.27/lb
Total op. cost
Operating cost, $/ton pulp
Operating cost with credit of
$0.06/kgal for mill make-up water
Net, $/kgal 0.299
Net, $/ton pulp 3.58
1
800
12,000
9.6
microlime
-carbon
100
100
150
1.85
4.90
6.75
0.217
0.075
0.033
0.034
0.359
4.30
2
1,600
12,000
19.2
microlime
-carbon
100
100
150
2.90
9.00
11.90
0.191
0.064
0.033
0.034
0.322
3.85
3
800
18,000
14.4
microlime
-carbon
100
100
150
2.36
6.98
9.34
0.200
0.069
0.033
0.034
0.336
6.04
4 4aa
800
18,000
14.4
primary-bio-
carbon
100
100
20
11.92 11.92
5.09 2.39a
17.01 14.31
0.364 0.306
0.127 0.127
0.003 0.003
0.146 0.146
0.640 0.582
11.52 10.48
5 5a
800
18,000
14.4
microlime
-bio
300
70
20
2.14 2.14
2.43 0.0a
4.57 2.14
0.098 0.046
0.032 0.032
0.036 0.036
0.166 0.114
2.98 2.04
6 6a
800
18,000
14.4
minilime
-bio
150
50
20
3.68 3.68
1.82 0.0a
5.50 3.68
0.118 0.079
0.049 0.049
0.059 0.059
0.226 0.187
4.07 3.37
0.262
3.13
0.276
4.96
0.580
10.44
0.522
9.40
0.106
1.90
0.054
0.96
0.166
2.99
a Capital cost of bio-oxidation not included (for mills with existing bio-oxidation facilities)
b 12,000 gal/ton is for new mill, 18,000 gal/ton is for older mill
0.127
2.29
-------
Microlime-Carbon, Large versus Smaller New Mills
Microlime-carbon treatment of effluent from a new 800-ton mill (Estimate
No. 1) was estimated to require a capital investment of $6.75 million,
of which $1.85 million is for lime treatment and pH adjustment and $4.9
million is for carbon adsorption. The operating costs were $0.36/kgal,
or $4.30/pulp-ton, or $0.30/kgal and $3.58/pulp-ton including credit
for water. The fixed costs for amortization, taxes, insurance, and
maintenance amount to $0.22/kgal, or 60% of the total operating costs.
When the mill capacity and effluent volume is doubled (Estimate No. 2),
the capital cost is increased by 67% to $11.90 million and the operating
cost is reduced to $0.32/kgal and $3.85/pulp-ton, or $0.26/kgal and
$3.13/pulp-ton with credit for the water.
Microlime-Carbon, New versus Older Mill
The capital cost for an older mill (18,000 gal of effluent per pulp-ton)
is $9.35 million (Estimate No. 3), or 38% greater than for a new mill
(Estimate No. 1) having 12,000 gal/pulp-ton. The operating costs (before
water credit) per 1000 gal for the older mill are slightly lower than
for the new mill but are 40% greater per pulp-ton.
Primary-Bio-Carbon Sequence
The primary-bio-carbon treatment plant costs considerably more than
the plant using microlime-carbon ($17.01 million for Estimate No. 4
versus $9.35 million for Estimate No. 3). The major share of this cost
differential is caused by the substantially higher cost of the carbon
adsorption system ($11.92 million versus $6.98 million), the remainder
of the differential is accounted for by the bio-oxidation system.
Estimate No. 4a is the same as Estimate No. 4 except that the capital
cost of the bio-oxidation unit is not included in Estimate No. 4a.
Even with no capital costs for bio-oxidation, the primary-bio-carbon
sequence is 53% greater in capital cost and 74% greater in operating
cost than the microlime-carbon sequence (Estimate No. 3). One advantage
of sequences using bio-oxidation is that the BOD content of the product
water is only about 20 mg/1 as compared to about 150 mg/1 for water
from microlime treatment. It appears then for treating mill effluent
for reuse that the bio-carbon sequences will be considerably greater
in capital and operating costs than the lime-carbon sequence.
The cost of the primary clarifier is included in both Estimates Nos.
4 and 4a. To exclude the effect of an existing clarifier, the clarifier
capital cost and its share of the fixed cost must be deducted from
159
-------
Estimate No. 4a. It can also be seen that the capital cost for the
primary clarifier in Estimate No. 4a is the same as that for the micro-
lime treatment in Estimate No. 3. An existing clarifier may be used
as part of a microlime treatment system and the costs adjusted accord-
ingly.
Primary-Carbon Sequence
The pilot plant results indicated that the direct use of carbon adsorp-
tion after primary clarification would require a very long contact
time (about 5 hr) in the carbon adsorbers to reduce the color from
1000 to 100 CU. Because of the indicated high cost of a plant using
this sequence, an estimate for it was not included in these studies.
However, rough approximations indicated that the capital cost for a
9.6 mgd plant would be about $15 to $20 million and that the operating
costs would be about $0.70 per 1000 gal and $13 per pulp-ton. These
costs indicate that other sequences involving carbon adsorption would
be more economical.
Lime-Bio versus Lime-Carbon Sequence
Estimates Nos. 5 and 6 were prepared to determine whether effluent could
be treated at lower cost by microlime or minilime followed by bio-oxida-
tion, than by microlime- or minilime-carbon treatment. Microlime-bio
would produce water having a color of 300 CU (or perhaps greater if there
is an increase of color during bio-oxidation) which might make it unsuited
for reuse. However, the capital cost (including bio-oxidation capital
cost) is only $4.57 million, or 49% of that for microlime-carbon (Estimate
No. 3) and the operating cost is $0.17/kgal, or also 49% of that for
microlime-carbon. If a mill were to add microlime treatment ahead of
an existing bio-oxidation unit at no capital charge for bio-oxidation
(Estimate No. 5a), the capital costs are reduced to $2.14 million and
the operating cost is reduced to $0.11/kgal and $2.04/pulp-ton, or
$0.054/kgal and $.0.96/pulp-ton with water credit.
If minilime is used instead of microlime ahead of bio-oxidation (Estimate
No. 6), the color would be 150 CU, if we assume there will be no color
increase during bio-oxidation. (The two mills using minilime-bio-
oxidation, Interstate Paper and Continental Can, have experienced large
color increases during bio-oxidation of lime-treated water.) The capital
costs (including that for bio-oxidation) for minilime-bio are 20% great-
er than for microlime-bio and the operating costs are 36% greater.
These comparisons indicate that microlime-bio has significantly lower
capital and operating costs than minilime-bio but that the color of
160
-------
the water from microlime-bio will be greater and probably limit its
suitability for reuse applications. These comparisons also indicate
that, if water that has higher color (150 - 300 CU) but lower BOD
concentrations (35-40 mg/1) can be reused in the mill, either micro-
lime-bio or minilime-bio can provide the water at significantly lower
capital and operating costs than microlime-carbon.
Microlime versus Minilime
The microlime and minilime processes without subsequent treatment steps
are compared in Table 39 when treating effluent from an older mill
(Estimates Nos. 9 and 11) and from a new mill (Estimates Nos. 8 and 10).
Before credit for reused water, the minilime process in both cases has
a 57% greater capital cost and a 71% greater operating costs. In the
older mill, microlime treatment would cost $1.82/pulp-ton compared to
$3.11 for the minilime treatment. The microlime process has an addition-
al advantage in that it could be operated when the lime kiln is off
since the treated water does not require carbonation and the lime dosage
is much smaller. These advantages are partially offset by some dis-
advantages: microlime provides treated water having 300 CU as compared
to 150 CU for water from the minilime process, the TOG concentration
is slightly greater, the control of lime addition to maintain the soluble
calcium in a narrow range of about 70 to 90 mg/1 is more demanding than
maintaining a fixed dosage, and the greater hardness value of the product
water (about 200 mg/1 CaCOs) might limit its use in some parts of the
mill.
The combinations of microlime and minilime with carbon adsorption are
compared in Estimates Nos. 1 and 7 for an older mill. The capital cost
for microlime-carbon is 57o less and the total operating cost is 107o less
($0.50/pulp-ton), which gives microlime a slight economic margin when
used in the combined sequence to give water meeting the selected standards
for reuse.
Effect of Carbon Price on Treatment Costs
The price of carbon used in all estimates discussed thus far was $0.27
per pound, which was typical for bulk quantities of commercial granular
carbons (January 1973). Lower-cost granular carbons are now available
that have higher ash contents, and new sources of carbon, such as from
the St. Regis process of producing granular carbon from a hydropyrolysis
pulp-chemicals recovery process that is now being developed under an
EPA grant (Part II of No. 12040 EJU), indicate that lower-cost carbons
might be available in the future. To show the effect of carbon price,
161
-------
Table 39. COMPARISON OF COSTS FOR MICROLIME AND MINILIME PROCESSES
Estimate No. 11. 9. 10 8. I 7_
Pulp tons per day 800 800 800 800 800 800
Effluent, gal/ton 18,000 18,000 12,000 12,000 12,000 12,000
Effluent, mil. gal/day 14.4 14.4 9.6 9.6 9.6 9.6
Treatment microlime minilime microlime minilime microlime minilime
-carbon -carbon
Product water quality
Color, CU 300 150 300 150 100 100
TOG, mg/1 150 135 150 135 100 100
BOD, mg/1 225 187 225 187 150 117
Capital costs, $ million
Lime trt. & pH adjust. 2.36 3.68 1.85 2.91 1.85 2.91
Carbon adsorption 4.90 4.20
o\ Clarifier and bio-oxidation
M Total 2.36 3.68 1.85 2.91 6.75 7.11
Operating costs, $/kgal
Fixed 0.050 0.079 0.059 0.093 0.217 0.228
Labor, o.h., util. 0.018 0.035 0.018 0.035 0.075 0.087
Lime and chemicals 0.033 0.059 0.033 0.059 0.033 0.059
Carbon, at $0.27/lb 0.034 0.026
Total operating cost 0.101 0.173 0.110 0.187 0.359 0.400
Operating cost, $/ton pulp 1.82 3.11 1.32 2.25 4.30 4.79
Operating costs with water credit3
$/kgal 0.041 0.113 0.050 0.127 0.300 0.340
$/ton pulp 0.74 2.03 0.60 1.53 3.58 4.07
a With credit of $0.06/kgal for mill make-up water
-------
it was assumed that carbon having the same properties as the $0.27/lb
carbon would be available at $0.10/lb if produced and used on-site.
The influence of this lower-priced carbon on the costs of lime-carbon
and bio-carbon sequences is shown in Table 40.
For microlime-carbon and minilime-carbon plants (Estimates Nos. 1, 2
and 7), the lower price of carbon reduced the capital costs 6 - 9%
and the operating costs by 8 - 12% before credit for the reused water
and 9-15% after credit for reused water. In the case of the primary-
bio-carbon plant (Estimate No. 4) the lower priced carbon reduced the
capital costs by 12% and the operating costs by 21%. The greater
influence of carbon price on costs for primary-bio-carbon was due to
the much greater dosage and size of carbon adsorption units to remove
the higher concentration of color after bio treatment.
CONCLUSIONS FROM THE COST ESTIMATES
1. The microlime-carbon sequence is the lowest cost process for meeting
the product water quality standards of 100 CU and 100 mg/1 TOG for
mill reuse. The capital costs for treating 9.6 mgd from a new 800
ton/day mill are estimated to be $6.75 million, and the total operating
costs are estimated to be $0.30/kgal and $3.58/pulp-ton with credit
for the reused water.
2. Doubling the plant pulping capacity increases the capital cost by 67%
and reduces the operating costs about 12% to $0.26/kgal and $3.13/pulp-
ton with credit for water.
3. Microlime-carbon treatment for an older mill producing 50% more water
per pulp-ton but having the same pulp capacity requires a 38% greater
capital cost, and the operating costs are about the same per 1000 gal
($0.28/kgal) but the cost per pulp-ton was greater ($4.96), both after
credits for the reused water.
4. A reduction in price of carbon from $0.27/lb to $0.10/lb reduces the
capital cost for lime-carbon plants about 8%, and operating cost about
13%.
5. The capital and operating costs for primary-bio-carbon are 1.5 to 2
times those for microlime-or minilime-carbon, and the costs for primary-
carbon sequence are about 15% greater than those for primary-bio-carbon
sequence.
6. The minilime-carbon process has about 5% greater capital and 147» greater
operating costs than the microlime-carbon process.
7. For reuse of water in a new 800 ton/day mill, for applications that
can tolerate a color of 300 CU and a hardness of 200 mg/1 CaCO^, micro-
lime treatment could be used at a capital cost of only $1.85 million
and an operating cost of only $0.11/kgal or $1.32/pulp-ton, or $0.05/
kgal and $0.60/pulp-ton after credit for the water. The costs for
microlime treatment are about 40% below those for minilime alone
before credit for the water.
163
-------
Table 40. EFFECT OF CARBON PRICE ON TREATMENT COSTS
Estimate No. I 2 4 _7
Pulp tons per day 800 1600 800 800
Effluent, gal/ton 12,000 12,000 18,000 12,000
Effluent, mil. gal/day 9.6 19.2 14.4 9.6
Treatment microlime-carbon microlime-carbon bio carbon minilime-carbon
Carbon price, $/lb 0.27 0.10 0.27 0.10 0.27 0.10 0.27 0.10
Capital costs, $ million
Lime trt. & pH adjust. 1.85 1.85 2.90 2.90 2.91 2.91
Carbon adsorption 4.90 4.35 9.00 7.91 11.92 9.95 4.20 3.75
Clarifier & bio-oxid'n 5.09a 5.09a
Total 6.75 6.20 11.90 10.81 17.01 15.04 7.11 6.66
Operating costs, $/kgal
Fixed
Labor, o.h., util.
Lime and chemicals
Carbon
Total operating cost
With water credit*3
a Including capital cost of bio-oxidation
b With credit of $0.06/kgal for mill make-up water
0.217
0.075
0.033
0.034
0.359
0.299
0.199
0.075
0.033
0.012
0.319
0.259
0.191
0.064
0.033
0.034
0.322
0.262
0.173
0.064
0.033
0.012
0.282
0.222
0.364
0.127
0.003
0.146
0.640
0.580
0.322
0.127
0.003
0.054
0.506
0.446
0.228
0.087
0.059
0.026
0.400
0.340
0.214
0.087
0.059
0.010
0.370
0.310
-------
Microlime-blo and minilime-bio processes are significantly lower in
capital costs and 33 to 50% lower in operating costs than microlime-
carbon, but the color of the product water will be greater, particularly
if a color increase during bio-oxidation occurs as it does in the two
mills using lime-bio treatment.
165
-------
SECTION XV
REFERENCES
1. Black, A. P., and Christman, R. F. "Characteristics of Colored
Surface Waters", J. Am. Water Works Assoc. 55 (6):753(1963).
2. Black, A. P., and Christman, R. F. "Chemical Characteristics of
Fulvic Acids", J. Am. Water Works Assoc. 55 (7):897(1963).
3. Berger, H. F. and Brown, R. I. "The Surface Reaction Method for Color
Removal from Kraft Bleaching Effluents", Tappi 4.2 (3) :245(1959).
4. Berger, H. F., and Smith, D. R.. "Waste Water Renovation", Tappi 51
(10):37A(1968).
5. Davis, C. L. Jr. "Tertiary Treatment of Kraft Mill Effluent Including
Chemical Coagulation for Color Removal", Tappi 52 (11):2132(1969).
6. English, J. N., et_ al_. "Removal of Organics from Waste Water by
Activated Carbon", Water-1970. AlChE Symposium Series 67 (107):147
(1971).
7. Gould, M. "Lime-Based Process Helps Decolor Kraft Wastewater",
Chem. Eng. 78 (3) :55(1971).
8. Gould, M. "Color Removal from Kraft Mill Effluent by an Improved
Lime Process", Tappi 56 (3):79(1973).
9- Masse, A. N. "Removal of Organics by Activated Carbon", presented
at 156th National A.C.S. meeting (Sept. 1968).
10. McGlasson, W. G. Thibodeaux, L. J., and Berger, H. F., "Potential
Uses of Activated Carbon for Wastewater Renovation", Tappi 49 (12):
521(1966). ~
11. McGlasson, W. G. "Treatment of Pulp Mill Effluents with Activated
Carbon", NCASI Tech. Bull. 199(1967).
12. Miller, C.O.M., and Clump, C. W. "A Liquid Phase Adsorption Study
of the Rate of Diffusion of Phenol from Aqueous Solution into Activated
Carbon", AlChE Journal 16 (2):169(1969).
13. Morris, J. C., and Weber, W. J. "Adsorption of Biochemically
Resistant Materials from Solution", U.S.P.H.S., AWTR-9(1964).
166
-------
14. Spruill, E. L. "Color Removal and Sludge Recovery from Total Mill
Effluent", Tappi 56 (4):98(1973).
15. Timpe, W. G., Lang, E. W., .et .al. "The Use of Activated Carbon for
Water Renovation in Kraft Pulp and Paper Mills", presented at 7th
TAPPI Water and Air Conf., Minneapolis (June 1970).
16. Weber, W. J., Hopkins, C. B., and Bloom, R. "Physicochemical Treat-
ment of Wastewater", J. Water Pollution Control Federation 42 (!)•
83(1970). ' ~~ ~
17. Cover, A. E., and Pieroni, L. J. "Appraisal of Granular Carbon
Contacting", Report Numbers TWRC-11 and TWRC-12. FWPCA, U.S. Depart-
ment of the Interior (May 1969).
18. Anon. "Color Removal from Kraft Pulping Effluents by Lime Additon",
Interstate Paper Corp. for EPA, Final Report, Program #12040 ENC,
Grant #WPRD 183-01-68 (Dec. 1971).
19. Anon. "Process Design Manual for Carbon Adsorption", EPA Technology
Transfer (October 1973).
20. Bloom, R., .££ _§!.. "New Technique Cuts Carbon Regeneration Costs",
Enviro. Sci. and Tech. 3 (3):214(1969).
21. Berg, E. L., je_C a±. "Thermal Reactivation of Spent Powdered Carbon
Using Fluidized Bed and Transport Type Reactors", Chem. Eng. Prog.
Symposium Series 67 (107):154(1970).
22. Herbet, A. J. "A Process for Removal of Color from Bleached Kraft
Effluents Through Modification of the Chemical Recovery System",
NCASI Tech. Bull. 157 (June 1962).
23. Anon. "Design Workshop for Water Pollution Control", Environmental
Science Services (November 12, 1969).
24. Smith, R. "Cost of Conventional and Advanced Treatment of Wastewater",
J. Water Pollution Control Federation 40 (9):1546(1968).
25. DiGregorio, D. "Cost of Wastewater Treatment Processes", TWRC-6.
FWPCA. U.S. Department of the Interior (Dec. 1968).
26. Anon. "Cost of Clean Water IV-- Projected Wastewater Treatment Cost
in the Organic Chemicals Industry", FWPCA. U.S. Department of the
Interior (Jan. 1969).
167
-------
27. Anon. "Cost of Clean Water III - Industrial Waste Profiles No. 3 -
Paper Mills".FWQA Publication No. IWP-3. U.S. Department of the
Interior (Nov. 1967).
28. Eckenfelder, W. W. "Economics of Waste Treatment", Chemical Engineer-
ing 67, (18):109(1969).
29. Page, J. S. "Estimator's Manual of Equipment and Installation Costs",
Gulf Publishing Co., Houston,(1963).
30. Guthrie, K. M. "Capital Cost Estimating", Chemical Engineering 76
(6):114(1969).
31. Berger, H. E. "Development of an Effective Technology for Pulp and
Bleaching Effluent Color Reduction", presented at NCASI Meeting (Feb.
1969).
32. Davies, D. S. "Activated Carbon Eliminates Organics", Chemical
Engineering Progress 60 (12):46(1964).
33. Hager, D. G., and Rizzo, J. L. "A Chemical and Physical Wastewater
Treatment Process", presented at AlChE Meeting, Atlanta (Feb. 1970).
34. Cooper, J. C., and Hager, D. G. "Water Reclamation with Activated
Carbon, Chemical Engineering Progress 62 (10):85(1966).
35. Spruill, E. L, Private Communication Concerning Costs for Continental
Can Co. Lime Treatment Plant (July 1973).
36. Amberg, H. R. "Water Pollution Control in Pulp and Paper Industry",
Industrial Water Engineering 7 (11):26(1970).
37. Bishop, D. F., et aLt "Physical-Chemical Treatment of Municipal
Wastewater", J. Water Pollution Control Federation 44 (3):361(1972).
38. "Standard Methods for Examination of Water and Wastewater", 13th
Edition, APHA, New York(1971).
39. Ganczarczyk, J. "Fate of Lignin in Kraft Effluent Activated Sludge
Treatment", J. Water Pollution Control Federation 45 (9):1898(1973).
40. Timpe, W. G., Lange, E. W., and Miller, R. L. "Kraft Pulping Effluent
Treatment and Reuse-State of the Art", EPA R2-73-164. Program 12040
EJU (Feb. 1973).
168
-------
41. Timpe, W. G., and Lange, E. W. "Processes for Reducing the Organic
Carbon Content of Water Contaminated with Organic Compounds by
Continuous Counter Current Multistage Treatment with Activated Carbon"
U.S. Patent 3,763,040 (October 2, 1973).
42. Allan, G. G., Mauranen, P., Neogi, A.N. "Fiber Surface Modification,
Part IX, Removal of Phenolic Pollutants from Water by Oxydative
Coupling to Lignocellulosic Substrates", Paperi ja Fuu No. 6 (1971);
see also "Pollution Abatement by Fiber Modification", EPA Final Report
Program 12040-EFC (Jan. 1971).
43. Swanson, J. W., Dugal, H. S., et al. "Kraft Effluent Color Character-
ization Before and After Lime Treatment", EPA R2-73-141, Project 12040
DKD (Feb. 1973).
44. Gulp, R. L., Gulp, G. L. "Advanced Wastewater Treatment", Van Nostrand
Reinhold Co., New York, p.253(1971).
45. Timpe, W. G., and Lang, E. W. "Activated Carbon Treatment of Kraft
Mill Effluent for Reuse", Water-1973, AIChE Symposium Series 70(136):
579(1974).
46. Timpe, W. G., and Lang, E. W. "Activated Carbon and Other Techniques
for Color Removal from Kraft Mill Effluents", Procedings of the XV
EUCEPA Conference, Rome, ATICEljCA in coop. w. FAO, Rome, p.535(1973).
47. Timpe, W. G., and Lang, E. W. "Activated Carbon Treatment of Unbleach-
ed Kraft Effluent for Reuse-Pilot Plant Results", presented at TAPPI
Environmental Conference (May 1973).
48. Thibodeaux, L. J., and Berger, H. F. "Laboratory and Pilot Plant
Studies of Water Reclamation", NCASI Tech. Bull. 203 (June 1967).
49. Schoeffel, E. W., and Zimmermann, F. J. "Wet Air Oxidation of
Combustible Materials Adsorbed on Carbonaceous Adsorbent", U.S. Patent
3,386,922 (June 4, 1968).
50. Koches, C. F. "Transport Reactor with a Venturi Tube Connection to
a Combustion Chamber for Producing Activated Carbon", U.S. Patent
3,647,716 (March 7, 1972).
51. Koches, C. F., and Smith, S. B. "Reactivate Powdered Carbon", Chem.
Eng. _79 (9):46(1972).
52. Bennett, D. J., Dence, C. W., £t al. "The Mechanism of Color Removal
in the Treatment of Spent Liquors with Lime", TAPPI 54 (12)-.2019(1971).
169
-------
SECTION XVI
PUBLICATIONS AND PATENTS
PUBLICATIONS AND PAPERS PRESENTED
Timpe, W. G., Lang, E. W., et al., "The Use of Activated Carbon for Water
Renovation in Kraft Pulp and Paper MillsV, presented at 7th TAPPI Water
and Air Conf., Minneapolis (June 1970).
Timpe, W. G., "Removal of Color from Kraft Mill Effluents by Activated
Carbon and Combined Treatments", presented at the NCASI Southern-South
Central Regional Meeting, Atlanta, Ga., (July 1970).
Timpe, W. G., and Lang, E. W., "The Removal of Color from Kraft Mill
Effluents by Activated Carbon and Combined Treatments", presented at the
1971 Annual Meeting, Gulf Coast Section, TAPPI, Biloxi, Miss. (May 1971).
Timpe, W. G., "A Chemical Engineering Approach to a Closed Water Cycle
in the Kraft Industry", presented at the Ninth API - TAPPI Research
Conference, Seattle, Wash., (Oct. 1971) (presented by G. R. Webster of EPA).
Timpe, W. G., Lang, E. W., and Miller, R. L., "Kraft Pulping Effluent
Treatment and Reuse-State of the Art", EPA R2-73-164, Program 12040 EJU
(Feb. 1973).
Timpe, W. G., and Lang, E. W., "Activated Carbon Treatment of Unbleached
Kraft Effluent for Reuse-Pilot Plant Results", presented at TAPPI Environ-
mental Conference (May 1973).
Timpe, W. G., and Lang, E. W., "Activated Carbon and Other Techniques for
Color Removal from Kraft Mill Effluents", Procedings of the XV EUCEPA
Conference, Rome ATICELCA in coop. w. FAO, Rome, p.535(1973).
Timpe, W. G., and Lang, E. W., "Activated Carbon Treatment of Kraft Mill
Effluent for Reuse", Water-1973, AIChE Symposium Series 70(136);579(1974).
PATENTS AND PATENTS APPLIED
Timpe, W. G., and Lang, E. W., "Processes for Reducing the Organic Carbon
Content of Water Contaminated with Organic Compounds by Continuous Counter
Current Multi-stage Treatment with Activated Carbon", U.S. Patent 3,763,040
(October 2, 1973) (This is the so-called FACET process).
Timpe, W. G., "Process for Reducing the Organic Carbon Content and Improving
the Color of Aqueous Plant Effluents", U.S. Patent Application Serial No.
301,400 (Nov. 1, 1972).
170
-------
Timpe, W. G. , "Separation of Organic Compounds in Aqueous Condensates from
Kraft Pulping", U.S. Patent Application Serial No. 301,401 (Nov. 1, 1972).
171
-------
SECTION XVII
GLOSSARY
BOD - 5-day biological oxygen demand.
CU - Color units measured according to APHA-NCASI standard method (see
Appendix B).
FPCU - Color after filtration through 2 micron paper filter, no pH
adjustment (see Appendix B).
MPCU - Color after filtration through 0.8 micron Millipore filter, no
pH adjustment (see Appendix B).
TIC - Total inorganic carbon.
TOC - Total organic carbon.
Lime treatment - Effluent treatment based on the use of lime - see mini-
lime and microlime treatment.
Lime-carbon
Primary/lime-carbon - Terms used interchangeably for an effluent treat-
ment system involving lime treatment without prior or separate
primary clarification followed by carbon adsorption.
Minilime treatment - Effluent treatment with the minimum lime dose that
reaches near-leveling off of color at about 85% removal; dosage
approx. 1000 mg/1 CaO, leaving about 400 mg/1 dissolved Ca in
solution which must then be removed by carbonation (see Section VIII).
Microlime treatment - Effluent treatment achieving about 70% color removal;
dosage adjusted to achieve 80 mg/1 dissolved Ca, avoiding need for
subsequent removal of Ca by carbonation; dosage approx. 500 mg/1 CaO
(see Section VIII).
Primary-carbon - Effluent treatment sequence involving primary clarification
followed by activated carbon adsorption.
Bio-carbon
Primary- b io - c a rbo n - Terms used interchangeably for an effluent treatment
system involving primary clarification, biological oxidation and
carbon adsorption.
Adsorption isotherm - Relates the quantity adsorbed per unit of adsorbent
to the concentration of adsorbate at equilibrium.
172
-------
SECTION XVIII
APPENDICES
Page
A. Water Quality Standards for Specific Uses 175
Table 41: Potential Water Quality Standards for Specific 176
Water Uses
B. Procedures for Isotherms, Laboratory Lime Treatment, 179
Laboratory Bio-Oxidation and Analyses
C. Detailed Information on Pilot Plant Equipment 183
D. Printout of Daily Summaries of Conditions and Results for 185
Bio-Carbon Sequence
E. Printout of Daily Summaries of Conditions and Results for 186
Primary-Carbon Sequence
F. Printout of Daily Summaries of Conditions and Results for 187
Lime Treatment
G. Printout of Daily Summaries of Conditions and Results for 188
Lime-Carbon Sequence
H. Generalized Procedure for Designing Plant for Treating Kraft 189
Effluent in Granular Activated Carbon Columns
I. Relationships Useful in Design of Downflow Carbon Adsorption 193
Columns
J. Cost Estimate No. 1 - Micro lime-Carbon Treatment of 9.6 MGD 194
of Effluent from a New 800 TPD Unbleached Kraft Pulp and
Paper Mill
Figure 30: Flow Diagram for Microlime-Carbon Treatment of 195
Unbleached Kraft Effluent for Reuse
Major Conditions and Assumptions
Design of Lime Treatment Unit
•I QQ
Design of Carbon Adsorption Units
173
-------
Page
Capital Cost of Lime Treatment Unit 202
Capital Costs for Carbon Adsorption Unit 203
Capital Cost of pH Adjustment after Lime Treatment 204
Capital Cost of Backwash Clarification and Surge Tanks 204
Operating Costs for Lime Treatment 204
Operating Costs for Carbon Treatment 205
Total Capital and Operating Costs for Estimate No. 1 206
K. English to Metric Units Conversion Table 207
174
-------
APPENDIX A
WATER QUALITY STANDARDS FOR SPECIFIC USES
The objective of maximizing water reuse economically very likely will
require avoidance of non-essential treatment by more closely matching
treated water quality to actual use requirements.
At present, no comprehensive listing appears to be available on water
quality standards for specific water uses throughout the pulp and paper
mill. Although certain standard use and reuse patterns exist through-
out the industry, no standards have been developed for the water input
for many individual process steps. A substantial amount of reuse is
at present handled to fit specific mill circumstances, and disagree-
ment is found as to what represents suitable water. The situation is
made more complex by the fact that certain contaminant levels cannot
be defined in terms of concentration in the water feed stream alone.
If the receiving system is a water recycling system, consideration of
maximum concentration in the receiving recycle system together with
the feed and blowdown volumes is required. In other cases, a system
downstream must be considered together with the direct receiving system.
Temperature requirements must also be considered.
A beginning was made in October, 1969, to collect quality standards data
and guidelines, resulting in the data in Table 41 and notes to this table.
175
-------
Table 41. POTENTIAL WATER QUALITY STANDARDS FOR SPECIFIC WATER USES
USE
Paper machine
Headbox & wire showers low volume, high
pressure, make-up
Felt showers
Pulp mill
Brown stock washer) fresh showers 1250-1500
last drum ) contain, showers 1250-1500
White liquor system
White liquor filter (Eimco System) 100-200
Mud washer
Smolt dissolving 23)
Bleaching system
Fresh water in counter-current system 19) 19)
Woodyard 17)
Showers 0-2000gpm 30) <105 18)
Cooling and sealing water 20)
Boiler feed water
Cooling water
S T A N D A R D
Temp.
°F
135(board)3)*
103(paper)
160
160
125-200
125-200
BOD Organics
2-5 4) 3, 4)
2-5
12)
13) 13)
21)
21)
21)
Total
Color Turbid, solids
CU JTU mg/1
50-100 10-15 <250 5)
50-100 10-15
100 10)
13)
50-100
50-100
50-100
Susp. Diss.
solids solids pH
0-5 6) 5) 2)
6-10 8)
fiber- 9) 11)
free 9) 11)
16)
16)
Hardness Cl SO*" Na+
mg/1 mg/1 mg/1 mg/l
7) 2000 1,2)
7) 250 250
9)
14)
15)
22)
Turbine and evaporator 24)
Small heat exch.; seal water on 26)
pumps, bearings; air cond.; air compr.
<95
0-10
26)
<50
6-9
6 or
higher 18)
No Ca
25)
24)
.2-.5
* See notes at end of table
-------
Notes to Table 41. WATER QUALITY STANDARDS FOR SPECIFIC WATER USES
1. Corrosion is problem.
2. 2000 ppm of S0^= in white water is probably the TAPPI established
limit in conjunction with wet strength resin use. Concentration in
total white water, hence level in this make-up stream, is function
of make-up flow and S04= (incoming with pulp and A12(S04)3 and H2S04
added for pH adjustment.
3. Refining adds 5°F.
4. No rosins or resins should be present to avoid plugging up machine
showers; also, keep slime down by slimicides and residual Cl of
0.2 mg/1 (this is done anyway). ^
5. Watch in conjunction with chloride and sulfate build-up.
6. Related to turbidity specification.
7. Possibly need a Ca spec., a minimum being required for proper
functioning of sizing or wet strength additives; 40-50 ppm Ca
suggested. Ca(ECO^)2 hardness undesirable.
8. Adjust to maintain stability of water.
9. Water must be free of contaminants which would form scale in heat
exchanger or deposit on filter/washer wires. Low Ca and Fe desired.
10. Lower limits apply if pulp is to be bleached.
11. Probably 7-12. Essential to avoid lignin precipitation at this point.
12. Low molecular weight organics pose no direct problems except as
nutrients in paper machine water system, resulting in slime growth.
Could be high as "contaminated condensate" without carry-over (i.e.,
clean evaporator condensate).
13. Can be contaminated up to perhaps 1% BLS. There is a concentration
of about 2% to 4% BLS in collection system under drum. Firmer data
need to be established through tests and calculations.
14. A low Na content is desired. A certain Na limit may be calculated
from washing requirements: washing done on 1270 Na in liquid; 0.570
to 2.0% Na desired in dry mud, at 60% to 70% solids in cake.
15. Since this is the final washing stage for the lime mud, low Na content
in wash water may be even more critical than as discussed in note 14.
16. Low pH acceptable if alkali consumption can be limited.
17. Only concern is to avoid excessive foaming. Otherwise, a closed
recycle system is now contemplated for Mill B.
18. Uncertain. Consider effect of higher temperatures on extraction of
wood acids and on C0£ absorption from air. Corrosion due to low pH
is to be avoided.
19. Consider that there are three water systems to a bleach plant (batch):
a. Cold or ambient temperature water
b. 115°F water
c. 160°F water
20. In Mill A water is first used as cooling water before other uses.
21. Presence of low molecular weight organics considered not presenting
a problem.
177
-------
22. As low as possible. Chloride build-up in liquor system to be avoided.
23. Use of weak wash water is standard. Contaminated condensate (such
as clean evaporator condensate) can be used.
24. Relatively clear water, stabilized, non corrosive, non scaling.
25. Either no suspended solids, or otherwise chemically controllable.
26. Stabilized, non corrosive, non scale forming, non slime forming.
178
-------
APPENDIX B
PROCEDURES FOR ISOTHERMS, LABORATORY LIME
TREATMENT, LABORATORY BIO-OXIDATION AND ANALYSES
PROCEDURE FOR MAKING CARBON ADSORPTION ISOTHERMS(44)
1. Pulverize a representative sample of the carbon (10-20g) and screen
it through a 325-mesh sieve. Oven-dry the carbon for 3 hr at 150°C.
2. Obtain a representive sample of the wastewater to be tested, making
sure that the pH is that of the original source of the sample. In
the laboratory program (Section 4), samples were adjusted to the
average pH for a given stream if the samples were within 0.7 pH
units of the average, or not used if outside of this range.
3. Weigh out 5 amounts of carbon in disposable weighing dishes. For
pulp mill effluents, use 0.3, 0.5, 1.0, 2.5, and 5.0 g per 500 ml
of sample (0.6 - 10.0 g/1).
4. Add 500 ml of the water to be tested to a 1-liter beaker, heat to
40°C, add one of the weighed dosages of carbon, stir vigorously
(100 rpm on a Phipps and Bird gang-stirrer) for 15 minutes while
maintaining the water at 40°C. Filter about 100 ml of the water
through Whatman No. 2 filter paper using a pressure-type lab filter
(to prevent loss of volatile organics). (Isotherms, except those
on condensate streams, prepared prior to 6/71 were made at 95°C and
2 minutes, but were equivalent to those made at 40° and 15 minutes.)
5. Repeat for the other dosages of carbon with separate 500-ml quantities
of water to be tested.
6. Analyze the filtrate from each dosage of carbon for the impurity
of interest and express in mg/1 or APHA color units. Normally,
isotherms were prepared for both TOG and color.
7. For each carbon dosage, subtract the final from the initial impurity
concentration, and divide this difference by the dosage of carbon
in g/1. This number is the loading of impurity per g of carbon used.
On 3 x 3 cycle log-log graph paper, plot loading on the y-axis versus
remaining concentration on the x-axis. The isotherm is completed by
drawing a straight line through the plotted values. (Sometimes the
line has two slopes and has a dog-leg shape.) Extrapolate the line
to the initial concentration. The loading at this intercept is the
ultimate capacity of the carbon for that effluent.
179
-------
LABORATORY BIO-OXIDATION
Bio-oxidation was standardized so that approximately a 70% reduction in
BOD would be obtained. The following conditions were used: Aeration
for 40 hours with filtered air; ammonium'nitrate addition of 3 mg/1 per
100 mg/1 BOD in the feed; phosporic acid addition of 0.6 mg/1 per 100
mg/1 BOD; strained sewage addition of 25 mg/1 per 100 mg/1 BOD.
LABORATORY LIME TREATMENT
The standard laboratory lime treatment consists of lime addition,
separation of precipitate, carbonation to pH 10.5 to remove excess dis-
solved lime, separation of precipiate, and finally carbonation to pH 7.0.
A lime dosage of 5000 mg/1 CaO was used unless otherwise specified. This
is a compromise dosage and is well above the so-called minimum lime (mini-
lime) dosage of 1-2,000 mg/1 (5,14) (also called stoichiometric), but
below the so-called massive lime dosage of 20,000 mg/1 (22). The lime
dosage was slaked in 500 ml of effluent for treatment of 6-10 liter of
effluent. The slurry was poured into the effluent and mixed in a 5 gal
tank. The mixture was held under slow stirring for 5 minutes. The mix-
ture was allowed to settle and the supernatant was then filtered through
a Whatman No. 2 filter. The filtrate was placed into another tank and
bottled C02 was bubbled into the sample using a gas dispersion tube.
The sample was slowly stirred during carbonation which continued until
the pH reached 10.5. The slurry was again allowed to settle and the
supernatant was filtered through a Whatman No. 2 filter. Carbonation
was then continued down to a pH of 7.0. This treatment follows the NCASI
procedure (22).
LABORATORY PRIMARY CLARIFICATION
All samples were filtered through a Whatman No. 2 paper filter to simulate
the result of primary clarification.
ANALYTICAL METHODS
The methods used for analyses of samples in the work covered by this
report are listed below. The standard method number given refers to
the method in the 13th edition of Standard Methods (38). Methods modified
by St. Regis Paper Company are available on request.
Color - Std Method 206A in which the pH of the sample is adjusted to
7.6. and the sample filtered through a 0.8 micron Millipore filter and
light transmittance is measured at 465 nm on a Spectronic 70 spectro-
photometer, as adopted by the NCASI. Prior to 2/4/71, the NCASI adopted
180
-------
method called for filtration through 0.45 instead of 0.8 micron Millipore
filter, and light transmittance measurement at 430 nm instead of 465 nm.
In both cases, color is determined from a calibration curve using a
cobalt chloroplatinate standard, where equivalent mg/1 of cobalt
chloroplatinate is termed color units or "CU". Based on pilot plant
experience, it was found desirable to determine the parameter of "color"
not only by the industry standard APHA-NCASI method but also by modified
procedures which measure apparent color. One modified procedure involved
0.8 micron Millipore filtration prior to adjustment to pH 7.6. (Used
after lime treatment, this procedure ensures more complete removal of
suspended solids.) Another modified procedure consisted of measuring
color after filtration through a Whatman Ho. 2 filter paper and adjust-
ment to pH 7.6, referred to as "7.6 FPCU". A third modified procedure
consisted of measuring color after filtration through Whatman No. 2
filter paper without pH adjustment, noted as "FPCU". A fourth modified
procedure consisted of measuring color after filtration through an 0.8
micron Millipore filter without any pH adjustment, noted as "MPCU".
TOG and TIC - Std Method 138 using a Beckman 915 total carbon analyzer;
sample filtered through Whatman No. 2 paper filter.
jjH - Std Method 144A
Turbidity - Std Method 163A using a Hach Model 2100 turbidimeter.
Total Suspended Solids (TSS) - Std Method 224C using Whatman GF/C
glass fiber filter discs.
Conductivity^ - Std Method 154 using Yellow Springs Instrument Co.
Model 31 conductivity bridge.
BOD-5 - Std Method 219
Dissolved Oxygen (DOl - Std Method 218F using a Weston and Stack dis-
solved oxygen analyzer Model 300.
Na2S - Mercuric chloride titration method according to E. Bilberg, Norsk
Skogindustri. 11/58 470 (1958), modified by St. Regis Paper Company.
Na2Sx and Na2S203 - by unpublished methods developed by St. Regis Paper Co.
Sulfate - Std Method 156B, gravimetric method.
Chloride - Orion Instrument Company selective ion electrode method.
Metal ions - Std Method 129A using atomic adsorption with the Perkin-
Elmer Model 403 spectrophotometer.
Volatile Acids - by gas chromatographic method of S. M. Aronovic, et al,
TAPPI 54 1963 (1971) and modified by St. Regis Paper Company.
181
-------
Neutral Volatiles - by gas chromatographic method of B. G. Turner and
I. T. Van Horn, presented at TAPPI National Meeting, February 1970,
and modified by St. Regis Paper Company.
182
-------
APPENDIX C
DETAILED INFORMATION ON PILOT PLANT EQUIPMENT
1. Basin volume - 25,000 ft3 or 187,000 gal, retention time at 30 gpm =
4.3 days.
aerator - 5 hp Chemineer
clarifier - area 510 ft2, rise rate at 30 gpm = 86 gpd/ft2 =
0.06 gpm/ft2
2. Surge tank, 10,000 gal, for backwash water (normally) or surge capacity
between stages of processing
3. Lime treatment - by Potter and Rayfield
lime treater, dimensions - 10'd x ll'-6"h, 4900 gal.
rise rate at 30 gpm - 550 gpd/ft2 =0.38 gpm/ft2
retention time in mixing zone - 20 min.
retention time total = 2.7 hr.
lime feed rate for 30 gpm and 1300 mg Ca(OH)2/l =
20/lb hr.
lime feeder - Wallace and Tiernan Series A-728
volumetric feeder with 41 ft-* hopper and
50 gal mixing tank for lime slurry.
carbonator, dimensions - 7'-6"d x ll'-6"h, 2920 gal, rise rate at
30 gpm = 980 gpd/ft2 =0.68 gpm/ft*
C02 from cylinders, pH automatically controlled
(normally at 10.5)
pH adjust tank, 200 gal, C0? through controller to reduce pH to
about 8.5
4. Carbon columns - by Potter and Rayfield
four columns piped to permit any combination of series and
parallel flow up-flow or down-flow.
dimensions - 3'd x 15'h, 430 gal in bed, 795 gal total
carbon bed - 10" supported on sand-gravel plus 5' free-board at
top.
charge - 1600 Ib ICI America (Atlas Chemical) granular Darco
20 x 40 mesh.
retention time in bed at 15 gpm = 0.6 hr = 2.4 hr total.
flow yel. at 15 gpm = 2 gpm/ft2 =1.7 bed volumes/hr each col.
backwash = 30 gpm, normally every 48 hr, after air backwash.
carbon density = 25 lb/ft.
5. FACET, three stirred tanks, gravity flow of water, carbon slurry pumped
counter-current to water flow, with sloping baffle to provide
settling zone at top of tank.
183
-------
5. FACET (continued)
3'd x 5'h (straight side) 170 gal in mixing zone, 248 gal total
retention time in mixing zone at 10 gpm=17 min ea = 51 min total
at 15 gpm=ll min ea = 33 min total
rise rate at 10 gpm=1.4 gpm/ft
at 15 gpm=2.1 gpm/ftz
weight of carbon in each tank at 10% solids = 170 Ib.
operating densities = 10-20% in mixing zone.
Carbon - ICI America, Darco XPT 40 x 140 mesh.
carbon feeder - Wallace and Tiernan, Series A-728 volumetric
feeder and 90 ft^ hopper.
6. Duo-media filters - by Potter and Rayfield
before carbon columns -
4'd, 12.5 ft2 area, I1 of 0.4 x 0.6mm sand plus I1 of 1 x 3 mm
anthracite, flow velocity =2.1 gpm/ft2 at 15 gpm.
after FACET -
3'd, 7.05 ft2, I1 of sand olus I1 of anthracite,
flow velocity =2.1 gpm/ft^ at 15 gpm.
7. Instrumentation - by Foxboro
recorders -
a. 10 temperatures
b. pH of controlled streams - carbonator and pH adjust
flow rates - to lime system, to carbon columns, and to FACET
c. measured properties of sample streams = pH, color, turbidity,
conductivity. 12 sample streams that can be monitored
instrumentally, Tenor drum programmer switches flow of sample
stream to measurement sensors and controls recording of
properties on recorder.
Controlled parameters = 3 water flow rates and 2 pH values
(carbonator and pH adjust).
timer controlled blow-down of sludge from lime treater and
carbonator.
8. Parameters determined in lab analyses of samples from carbon treat-
ments - color, pH, total inorganic carbon, total organic carbon,
soluble calcium (at least daily).
volatiles and low molecular weight organics, suspended solids, BOD
(every few days).
184
-------
APPENDIX D
PRINTOUT OF DAILY SUMMARIES OF CONDITIONS AND RESULTS FOR BIO-CARBON SEQUENCE
DATE=08/15/72 TIME=16:26s 8
RR012 - WATER TR
DATE
TIME OF DAY
ATMENT PILOT PLANT FILE DUMP
-OLUMN F
CU TO BASIN PAPR
ICU/LJ MP
CU FM BASIN PAPR
MP
PAPR
MP
HAPR
MP
CU PCRT-!> HAPR
CU PORT-0
CU PCRT-2
TBC TO BASN MG/L
TOC FM BASIN
TOC PORT-0
TOC PORT-2
po BoTTTD BASN M7L
Ui BOO FM BASIN
BOO FM LAST CCL
DO FM BASIN MG/L
PH TO COLUMNS
CDND MICPC MHOS
TURB JTU TO BASN
FM BASN
PRESS DROP COLS
PRESS DROP F-l
CHARGE LB
CUM MRS OPTO
TOTAL REMOVAL *
TOC REMOVAL *
TOTAL 'REMOVAL *
BOO REMOV bASN X
OC KtM X SYS IN
COLOR % OF FEED
TQC X OF FEED
OPERATING DAY />
COLOR RATE TQT*C
TOC LOG,RATE T*C
DISK FILE LOCN
I045.
00
48 0.0
00 0.0
00 0.0
00 0.0
001000.00
00 570.00
0 0.0
_.0 0.0
1000.00 750.00
65.
910.
750.
0.
5.
0.0
0.0
lt
0.0
0.0
0.0
0.0
750.00 585.00
338.00 233.00
0.0 " '
0.0
0.0
>.0
0.0
0.0
585.00 435.00
126.00 0.0
151.00 99.00
0.0 0.
0.0 0.0
84.00 68.00
0.0 0.0
1600.001600.001600.001760
417.50 345.50 188.50 38
.0
.0
.0
.0
00
~0
82.91
34.44
62.25
0.0
0.0
15.15
0.0
0.0
0.0 0.0
19.05 16.18
0.0 0.0
0.0 0.0
65.5-
57.00
71.91
9.88
145
146
...60
28.24
45.03
0.0
71.91
9.88
147
"I*
7.2B
17.09
37.75
0.0
71.91
9.88
148
13FB
0800
15.60
0.0
... 8:8-
0.0 0.0
0.0 0.0
0.0 0.0
0.0 0.
1165.00 825..
715.00 525.01
0.0
0.0
825.0'
0.0
0.0
~-8t«—- 8:8-
0.0
0.0
0.0
3.0
0.0
0.0
0.0
0.0
0.0
0.0
675.00
675.00 540.00
345.00 259.00
0.0
0.0
0.0
0.0
540.00 428.00
0.0 0.0 0.0 0.0
119.00 85.00 60.00 53.00
0.0 0.0 0.0 0.0
0.0
0.0
0.0
7.80
0.0
0.0
0.0
0.0
iI
0.0
0.0
0.0
0.0
0.0 0.0
0.0 0.0
__2fl__aQ ___ i*. IQ ___ jL
o.o o7o o.o tr.
0.0
0.0
12.02
18. 1C
0.0
0.0
2.00
0.0
0.0
0.0
3.00
0.0
0.0
0.0
0.0
0.0
1600.001600.001600.001-760.00
463.50 391.50 234.50 84.50
520.59 271.74 132.42 35.72
479.54 232.85 113.84 20
537.73 288.38 138.47 40.49
UM COLOR LOG
63.82 29.05 12.84 2.90
14.35 4.86
74.59 35.96 16.37 4.40
74.83
28.57
63.S7
0.0
0.0
29.41
0.0
0.0
28.57
73.43
71.43
59.00
0.0
11.67-
0.0
0.0
21.C_
48.25
50.42
0.0
5.88
36.22
44.54
0.0
0.0
18.87
0.0
0.0
8.40
25.17
36.13
0.0
120.34 120.34 120.34 120.34
17.10 17.10 17.10 17.10
14FB
0800
565.72
510.00
825.00
845
0.0
0.0
0.0
0.0
0 675.0
5<>5.00
.00 675.00
620.00 450.00
0.0
0.0
0.0
0.0
0.0 0.0
0.0 0.0
675.00 525.00
0.0
0.0
0.0
0.0
525.00 450.00
"• ,00 225.00
0.0 0.0
0.0 0.0
450.00 390.00
28
92.00 0.0 0.0 0.6
99.00 65.00 50.00 30.00
0.0 0.0 0.0 0.0
0.0
0.0
0.0
8.10
— 1
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0 0.0 0.0 6.0
0.0 0.0 0.0 0.0
4.00 2.02 3.00 0.0
30.09 0.0 0.0 0.0
1600.001600.001600.001760.00
486.50 414.50 25.7.. 50 107.50
149
150
151
152
153
0.0
23.08
0.0
0.0
-Zfc£r
15.15
45.97
50.51
0.0
46.92
6.72
154
0.0 0.0
40.00 -16.67
0.0 0.0
0.0 0.0
•*&-»*
36.29
30.30
0.0
46.92
6.72
155 156
24
00
00
00
0.0
0.0
0.0
O.C
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1090.;
900. C
825.C
600.C
825.C
640. C
O.C
O.C
640.00 480.00 465.00 383.00
.00 640.00 480.00 465.00
.00 435.00 285.00 210.00
.0 0.0 0.0 0.0
0.0 0.0
0.0
0.0
101.00
105.00
0.0
0.0 0.0
72.00 55.00
0.0 0.0
49.00 0.0
0.0 0.0
0.0 0.0
8.00 0.0
0.0
0.0
3.00
37.09
0.0
0.0
0.0
0.0
0.0
0.0
2.00
0.0
78.04 37.74 16.89
72.97
31.43
60.00
61.11
0.0
23.61
0.0
0.0
0.0
9.09
0.0
0
67.97
68.57
61.00
50.81
6.85
157
0.0
50.81
6.85
158
32.81
47.62
0.0
50.81
6.85
159
0.0
50.00
0.0
0.
0.0
0.0
0.0
0.0 0.0
0.0 0.0
2.02 0.0
0.0 0.0
"S'i* 304.'05 146:31 44loO
•Hi
5.16
0.0
16.00
0.0
0.0
r.62
27.03
40.00
0.0
50.81
6.85
160
-------
APPENDIX E
PRINTOUT OF DAILY SUMMARIES OF CONDITIONS AND RESULTS FOR PRIMARY-CARBON SEQUENCE
00
0\
, ,0*T6»UAO/T2
2 • WATER TR
ATNENT PlCOT PLANT FILE DUMP
15JU
~~257o"0"?375o-~217oo~~25750'
_fl.£___(U(L.
(jro o.o
0.0
405.00
293.00
975.00 0.0 0.0
1383.001125.00 450.00
.050.00 525.JO 3^800
453.00 428.-00
3^5.00 300.00
0
.0 0
.00 450.00 405.00
0.130.aO_291*QO
0.
0 0.0 0.
1125.00 450.00 408.00
525.00 338.00 30O.QO
191.00 0.0 C.O 3.0
198.00 123.00 97.00 9*.00
.0 0.0 0.
228.00 0.0 0.0
232.00 104.00 J09.00
0.0 0.0 0.9
217.00 0.0 0.
2Z1.00 137.00 121.00
0.0 0.0
0.0 0.0
aro oa
0.0 0.
5 320.75 162.00
rfeft-fefr
^7^~I^M"I
3T8?;
25.56
49.79
0.0
5
26.04 19.38
37.93 30.1
0.0 0.0 0.3
22.22
35.19
0.0
0.00 32.19
61.99 54.75
144.00 0.
!
-------
APPENDIX F
PRINTOUT OF DAILY SUMMARIES OF CONDITIONS AND RESULTS FOR LIME TREATMENT
oo
T W.ANT - UME TREATMENT
R CARBR PHA+F BASIN
S.ft B-0 1 9
BASIN TUTOR CARBR
9.04 OiS ft*fi_
2—8
0.0 12.25
o.q
^—8:8—8t8
WATER FLOW GPM
1
0.0 12.25 0.0
0.02^.?5 Q.n
0.0 12.25 0.0
0 29.00 Q.O
COLOR FPC AT PH
MPC AT PH
750.00 690.00
608.00 396.00
1125.00 600.00
743.00 410.00
1050.001275.00
713.00 428.00
QO *£JL
QUO HICROMHOS
TURBIDITY JTU
SUU.ATJ.L.E
25.00 90.00
-fcg—Mt—M-
0.0 WQiO 0.0
.0 1193.001193.00
0.0 7632.00
7858.00
0.0 53.64 0.
11*53 p.0
*u_ i.i_
M LIME U
CU7.6 REMOVD *BA
TOD REMOVED tBA
IME H57
PROD WAT MG/L
SLUD
JULIAN
DISC FILE LOCN
0.0
0.0
-------
APPENDIX G
PRINTOUT OF DAILY SUMMARIES OF CONDITIONS AND RESULTS FOR LIME-CARBON SEQUENCE
RR01ZB - WATER TREATMENT PILOT PLANT - LIME
DATE _ 8.23 5.0 8.i* 0.<
Q ^.n
,
MRS OPER
FLOW RATE GPM
mftomtn:
MPC
oo
oo
CU TO COL
FPC
JiR£_
CU FM COL FPC
TOC"FM-§ASJI.N£—
- TO COL
FM COL
CA FM BASIN
TQ/FROW CQLS
' TIC~FM~~BASIN~
TO/FM COLS
BOD TO/FM COLS
PH F.M BASIN.. .
TO/FM COLS
CONO TO/FM MMHOS
Tl«B TO/FM JTU
PflESS DROP .CQL.S.
C.AQa2_DQS£_c!G/L_
VOLATILES MEUH
ACETALD..__
...Jpf-MMfiAbzi
CUM LOG'
LOG RATE /HR
RATE TOT *C
^ ^f 1
LOG
RA"
516.18 248.48
0.29 0.28
57.59 0.0
-fl:l*-W
LOG RATE /HR
RATE TOT #C
—REM % SYS IN
TOT REM * IN
aE&l_!_I0._£uJ..
.-JA.QJ Tt«_.|ji-
522.02 253.20
0.26 0.21
46.95 0.0
.28,33 22,87
Si.19 0.0
.1.44 ,.11 59.,
0.03 O.ov
11.58 0.0
8.57 26.6?
35.24 0."
-tti1
-W-MY"
pTCE'Lor
0.0 O.Q
—8^—8rf
0.0 0.0
0.0 0.0
_._.^j_0 OjO
0.0 0.0
§:§ 8:8
-mt88—fcH
-52577r26_
0.15 0.34
49.14 0.0
,. 18.92 _4U70
60.62 0.0
hnt.-ft-Ttf
0.28 O.?0
47.89 0.0
-&ff-W
538.4
0.19
51.64
-HtH-1
^
540.70 279.94
0.15 0.21
36.62 0.0
19.A4 26.9
46.43 0.0
542.03 283.6?
0.06 0.17
22.54 0.0
7.54 21.03
28
.54 21.03
.57 0.0
0.0
0.0
O."0~~*
0.0
0.0
0.0
0/0
0.0
0.0
0.0
O.D
h-H-^i-
1?.83 0.0
12.70 19.84
32,5-4 0.0
—i|Ip__Z2t»ia
0.0 0.0
238.00
61.00'^
-------
APPENDIX H
GENERALIZED PROCEDURE FOR DESIGNING PLANT FOR TREATING KRAFT
PULP MILL EFFLUENT BY ADSORPTION IN GRANULAR ACTIVATED CARBON COLUMNS
1. Determine the flow and quality of the water to be treated.
Normally, the plant is designed to treat the average of the daily maximum
flow rates and to treat water with the average maximum color and TOC
concentrations. Other properties of the effluent should also be deter-
mined: BOD, conductivity, settleable solids, suspended solids, pH,
temperature, etc.
Determine what concentrations of color and TOC are to be tolerated in
the treated water. The design of the carbon adsorption plant for a
given flow rate is greatly affected by the prior treatment of the effluent
water (clarification alone, lime treatment, or bio-oxidation) and by the
permissible concentrations of color and TOC in the product water from
the carbon treatment.
2. Select type of carbon.
The type of carbon is determined from extensive laboratory isotherm tests
and preferably from bench-scale column adsorption tests after a prelim-
inary selection based on previous experience, price, service provided
by manufacturer, resistance to abrasion, and other factors. The mesh
size should probably be about 20 x 40 mesh, since this size minimizes
the retention time required and yet does not cause appreciable pressure
loss at flow velocities of 1 - 4 gpm/ft .
2. Contact time and flow velocity.
Determine from pilot-plant or bench-scale column adsorption tests the
contact time that is required at flow velocities of 1 - 4 gpm/ft of
column area to meet the desired properties of the treated water. The
value of retention time dictates the volume of the carbon needed for the
plant and is the major determinant of capital costs of the plant.
4. Dosage of carbon per unit of water treated.
The dosage is the pounds of regenerated carbon that must be added to the
adsorbers (and loaded carbon that is removed) per 1000 gal of water
treated to maintain the desired quality of product water. The dosage
rate is a major determinant of operating costs since losses (make-up)
189
-------
of carbon and regeneration costs are directly related to the dosage.
The dosage is best determined from bench-scale or pilot plant adsorption
tests with at least three columns in series operated at the expected
flow-rate (gpm/ft2) until it has been necessary to add 5 fresh columns
while maintaining the desired product water concentration, or more gener-
ally 2 more columns than the number of columns in the adsorption train.
dosage rate in Ib/kgal = (Ib carbon per column) x (total fresh columns added)
total kgal of water treated
The rate can also be determined (with less reliability) from lab isotherms
with at least 5 representative samples of the water to be treated. From
the isotherms, determine average loading for the average feed concentra-
tion.
dosage rate = (feed concentration - product concentration)
0.65 (loading at feed concentration)
where: dosage is g carbon per 1 water
concentrations are in mg/1
loading is in mg/g carbon.
5. Volumetric flow rate and volume of carbon needed.
Volumetric flow rate, vol. water 1
hr x vol. carbon ~ C.T.
where C.T. = contact time, hr
volume of carbon needed, ft = (ft3 water/hr) x C.T.
= 0.125 gpm x C.T,
weight of carbon, Ib = 25 lb/ft3 x vol. of carbon
6. Design of carbon columns.
(See also EPA "Process Design Manual for Carbon Adsorption" October 1973(19))
Total cross sectional area of the columns, ft2 = (water flow in gpm)/(ve-
locity in gpm/ft2).
Diameter of columns /4(area of columns)
V 3714 -
Use multiple trains of columns in parallel flow to give height/diameter
ratio between 1 and 3.
190
-------
Total length of carbon bed, ft = 8(gpm/ft2) x C.T.
or length = (vol. carbon)/(total area)
Length of bed in each column = total length/number of columns —(minimum
of three columns) in series to provide adequate counter-current contacting.
Freeboard in columns for backwash bed expansion = 50% of carbon bed.
Extra columns for regeneration -
a minimum of 2 columns is required to function as feed and receiving
column for the regeneration furnace. Two columns can serve several
trains if provided with proper piping arrangements.
Extra columns for peak load -
- provide 10 to 30% extra columns, depending on anticipated variability
of flow and concentration. Consider separate system for pre-dampening
fluctions by storage and/or spill control.
Down-flow or up-flow of water in columns. With proper backwashing, down-
flow appears preferable.
Backwash - use velocity that expands the carbon bed 20 to 50%. This
value is about 10 gpm/ft2 for 20 x 40 mesh Darco carbon. For backwash
inlet distributor, graded gravel and sand is recommended, but other
types of carbon support are available and probably satisfactory. The
period of backwashing for each column should be about 20 minutes every
48 hours. Provide air at start of backwash to break up any clumps that
might have formed.
Weight of carbon per column, Ib = (area, ft2) x (bed height, ft) x
(density, lb/ft3)
Frequency of column change per train, days =
(wt. of carbon per column, Ib)
(dosage, Ib/kgal) (flow per train, kgal/day)
Retention time of carbon in train, days =
(wt. of carbon per train, Ib)
(dosage, lb/kgal)(flow per train, kgal/day)
7. Total inventory of carbon.
Total carbon = (no. of columns) x (volume of each) x (density of carbon)=
weight of carbon needed. Warehousing will require additonal carbon; the
amount being dependent on the purchase requirements.
191
-------
8. Make-up carbon.
The make-up carbon is equal to the loss of carbon during removal from
columns and in regeneration. This loss runs 5 - 10% of the carbon dosage
or weight of carbon that is regenerated. It is expected that the loss
will probably be near 5% for large plants treating pulp mill effluents.
9. Carbon regeneration plant.
a. Capacity, Ib/day = (carbon dosage, Ib/kgal) x (water flow, kgal/day)
b. Type of furnace - a multiple-hearth furnace is generally used.
There are three major manufacturers.
c. Operating conditions - selected by consultants or through trials
by furnace manufacturer.
192
-------
APPENDIX I
RELATIONSHIPS USEFUL IN DESIGN OF DOWNFLOW CARBON ADSORPTION COLUMNS
Superficial (empty column) velocity, ft/hr = 8(gpm/ft2)
Total cross-sectional area of columns = gpm/(gpm/ft )
Bed Height, ft
= (lineal velocity, ft/hr) (contact time, hr)
= 8(gpm/ft2) (contact time, hr)
Volumetric flow rate,by/hr = (vol water/hr)/(vol carbon) = 1
contact time,hr
- 8(gpm/ft2)/(bed height)
Bed height or length = 0.02 x flow x AC
rate
Where: Flow - water flow rate, gpm/ft2
AC = inlet - outlet concentrations in mg/1
rate = rate of adsorption in mg
g x hr
This relationship is generally applicable, but the rate must apply at
the flow rate used and over the concentration range used.
193
-------
APPENDIX J
ESTIMATE NO. 1. - MICRQLIME-CARBON TREATMENT OF 9.6 MGD
OF EFFLUENT FROM A NEW 800 TON/DAY
UNBLEACHED KRAFT PULP AND PAPER MILL
1AJOR CONDITIONS AND ASSUMPTIONS
Flow Sheet - See Figure 30
Lime Treatment
Microlime treatment is used
Dosage = 520 mg/1 CaO (approximate)
= ^
Rise rate = 700
Water quality:
Feed Product % Reduction
Color, CU 1000 300 70
TOG, mg/1 250 150 40
BOD, mg/1 250 225 10
pH 10.5 11.5
Conductivity, micromhos 1200 3600
Susp. solids, mg/1 200 250
Temperature, °F 100 100
Lime sludge dewatered to 70% solids and returned to lime kiln.
Includes capital cost of new lime kiln for the proportion of load added
by effluent treatment sludge.
Desgin basis for lime treatment includes data from Interstate Paper
Company (5,18) and Continental Can Company (14) on lime treatment of
kraft mill effluents. Capital costs for lime treatment are derived
from costs of the Continental Can Company lime treatment plant at
Hodge, La.
Carbon Adsorption
Water quality, average composition:
Reduction
Feed Product by carbon overall
Color, CU 300 100 67 90
TOC, mg/1 150 100 33 60
BO0* tag/1 225 150 33 40
pH H.5 8.0*
Temperature, °F 100 100
Susp. solids, mg/1 250 15
*Product water is carbonated to pH 8
194
-------
LIME SLAKER
MILL
EFFLUENT
KILN GAS
VO
UT
SLUDGE
FILTRATE
SLUDGE -<-
SOLIDS TO
KILN FEED
r-o-J
SLUDGE
FILTER
SLUDGE
SETTLING
TO BACKWASH
1i
CARBON CONTACTORS
BACKWASH WATER
RETURN
r^
PH
_
ADJUST
TANK
TREATED
WATER TO
MILL REUSE
MAKE-UP
WATER
SETTLING TANKS FOR
BACKWASH WATER
Figure 30. Flow diagram for microlime-carbon treatment
of unbleached kraft effluent for reuse
-------
Designed for average flow of 9-6 mgd and maximum flow of 30% greater,
or 12.5 mgd.
Designed for average concentrations above, but with excess capacity
for greater concentration of color and TOG.
Down-flow, fixed-bed, pressure carbon contactors (columns).
Air and water backwashing of columns.
Feedwater to columns is not filtered.
Product water is adjusted to pH 8 by addition of kiln gas.
Carbon used: 20 x 40 mesh, equivalent to that used in pilot plant.
Make-up (loss) of carbon per regeneration = 5%.
Contact time, volume of carbon needed, and dosage rate are based on
pilot plant results.
Capital costs for carbon adsorption plant are based on cost data given
in FWPCA Report TWRC-11 (17) prepared by M. W. Kellogg Company and in
EPA Process Design Manual for Carbon Adsorption, October, 1973 (19).
Cost of make-up carbon = $0.27/lb and $0.10/lb.
Costs for Complete Plant
Capital costs include equipment, installation, instrumentation, buildings,
engineering, contractor overhead and profit.
Time basis for cost = January, 1973 adjusted by Engineering News Record
Construction Cost Index.
Amortization = 16 years, straight line = 6.25% of total capital invest-
ment per year.
Repairs and maintenance = 3% of total capital investment per year.
Taxes and Insurance = 2% of total capital investment per year.
Plant overhead = 75% of direct labor cost.
Steam = $0.80/1000 Ib, electricity = $0.01/kwh, gas = $0.60/106 Btu.
DESIGN OF LIME TREATMENT UNIT
Lime Dosage - to maintain 80 mg/1 dissolved Ca in water leaving lime
treater. On basis of pilot plant treatment of unbleached kraft total
mill effluent, average dosage is 520 mg/1 CaO, or 0.52 x 8.34 = 4.33
lb/1000 gal, or 4.33 x 9,600/2000 = 20.8 tons CaO.
Lime Slaker and Feeder
Mill lime is fed to a lime slaking tank to maintain a constant 10%
solids slurry which is then fed through a controlled slurry feeder to
the effluent line going to the lime treater. The soluble calcium
concentration in the lime treated water is controlled by a sensor in
the lime treater which sends a signal to the slurry which is then fed
through a controlled slurry feeder to the effluent line going to the
196
-------
lime treater. The soluble calcium concentration in the lime treated
water is controlled by a sensor in the lime treater which sends a
signal to the slurry feeder to maintain the calcium concentration
at 80* 10 mg/1 Ca.
Lime Treater
A conventional reactor-clarifler is used that has an enlarged center-
well for the reaction of lime with the organic compounds of the effluent
and for reaction and growth of CaCOo crystals.
Volume of reaction zone to give 30 min. retention =
30 x 9.600.000 ,nn nnn
1 144Q - 200>000 gal.
Top area of reaction zone that has enlarged top section is estimated
to be 670 ft2, or 29 ft in diameter.
Area of clarifier section at rise rate of 700 gpd/ft2 =
9.600.000 _ 9
-2- - ~ ft
Total area of lime treater » 14,370 x 4/3.14 = 135 ft.
Lime Balance for Lime-Carbon Treatment
Conc'n. in water
Ca. mg/1 CaO. mg/1
In feed water 12 17
Lime added to lime treater 372 520
Soluble in product water 80 112
In suspended solids from lime treater3 50 70
Loss in sludge handlingb 19 26
Net recovered in sludge to kilnc 286 400
Overall recovery of CaO = 100 x 400/520 = 77%
a At 20% Ca in suspended solids, or 0.20 x 250 = 50 mg/1 Ca. This
calcium is recovered as sludge from backwashing of carbon column
which is added to the lime-treater sludge.
b Estimated loss - 5% of added lime, or 0.05 x 520 - 26 mg/1 CaO
c 520 + 17- 112 -25 = 400 mg/1 CaO
197
-------
Sludge Composition
On the basis of pilot-plant results, the sludge will consist of these
materials :
Conc'n in water
mg/1 Ib/kgal
CaC03 from 003 in feed water 343 2.85
Ca-organics 278 2.32
Susp. solids in feed water3 200
Total sludge solids produced 821 6.84
a Mainly inorganic solids
CaO in lime sludge solids to kiln = 400 mg/1 of water treated (see
above) or 100 x 400/821 = 49% CaO
Total sludge solids produced = 6.84 x 9600/2000 =32.8 tons/day
CaO in sludge = 32.8 x 0.49 = 16.1 tons/day
Volume of sludge produced at 15% solids from lime treater
6.84 x 9600 x — L_ x — Loc _ L_ = 48,200 gal/day=33.5 gpm
0.15 8.34 1.09 sp. gr.
Sludge Tank for Settling - at 12-hour retention and 30% solids =
48.200 x J.5 = 12,000 gal.
2 30
Sludge Filters - on the basis of pilot plant experience, the sludge
filters readily, presumably because microlime gives a high percentage
of CaC03 in the sludge, as compared to minilime treatment.
On the basis of the pilot-plant results, assume sludge from sludge tank
will contain 25% solids and will filter at a rate of 100 Ib/hr ft2 on
wire-drum filter.
Area of filter = 6.84 Ib/kgal x 9.600 kgal/day = 27.4 ft2
24 hr/day x 100 Ib/hr ft2
(Possibly, the settled sludge could be added to the mill's lime mud
filter).
198
-------
DESIGN OF CARBON ADSORPTION UNITS
Contact Time Required
Average pilot-plant results show that when removing color from 300 to
100 CU when the calcium content of the water was about 80 mg/1 Ca. the
length of adsorption bed at a superficial flow rate of 1.4 gpra/ft* was
19 ft and the contact time was 1.7 hr. The data showed that a slightly
shorter length of bed was required to reduce the TOG from 150 to 100
mg/1. Therefore, use a contact time of 1.7 hr plus 7% excess or 1.8 hr.
(An additional 20% excess capacity is allowed later as spare columns.)
Selections of Velocity, Pressure Drop, and Length of Carbon Bed.
Pilot plant data showed that it is desirable to use a superficial
velocity greater than 2 gpm/ft2 to provide increased turbulence and
adsorption rates. However, as velocity is increased, the length required
increased in direct relation to velocity (to maintain constant contact
time) and the pressure drop is increased by both the increased length
and greater velocity. The following tabulation shows the effect of
flow rate on pressure drop, as determined from Fig. 3-4 of the EPA
Design Manual for Carbon Adsorption (18) for the desired contact time
of 1.8 hr (0.56 V/V hr) when using 20 x 40 mesh carbon.
Flow rate, Length of bed Pressure
gpm/ft2 required, ft drop, psi
1.4 20 1.1
2.8 40 4.2
4.2 60 9.7
5.6 80 18.1
To provide a fairly low pressure drop of 9.7 psi, a flow rate of 4.2
gpm/ft is selected, which requires a total length of bed of carbon of
60 ft.
The effect of velocity on adsorption rates was not studied in the pilot
plant for the microlime-carbon sequence, while operations were carried
out at 1.4 gpm/ft2. Based on increased adsorption rates, or decreased
retention times, observed for increased flow rate in the primary-carbon
pilot plant work, it is possible that the design flow rate of 4.2 gpm/
ft2 chosen above in conjunction with contact time requirements based
on 1.4 gpm/ft2 in the pilot plant, could represent a sizeable effective
excess contact time.
199
-------
Dimensions of Adsorption Columns
Area = 6670 gpm/4.2 gpm/ft2 = 1590 ft2 total.
No. of parallel trains - use 3 to keep diameter of columns at reason-
able size.
1590 x 4
Diameter of each column - = 26 ft.
3 x 3.14
No. of columns in series per train - use 3 to provide adequate counter-
current contacting and to keep the ratio of height to diameter about 1.
Height of bed in each adsorber = 6£ = 20 ft.
3
Total height of each column, allowing 50% free board for bed expansion
during backwashing = 20 x 1.5 = 30 ft.
Arrangement of Columns
feed-
rOOO
V
Active Columns
Total columns
Active columns
Spare columns
Ready to have carbon discharged
Ready to receive regenerated carbon
Total columns with carbon
Spare columns for surge loads
inactive - being refilled
with carbon
inactive - being discharged to
regeneration
= 13
= 9
= 2
» 1
= 1
= 12
Volume of carbon per column = 3.14 x 262 x 20 = 10,600 ft3.
4
Weight of carbon per column = 10,600 x 25 lb/ft3 - 265,000 Ib.
Total weight of carbon in system = 12 x 265,000 « 3,180,000 Ib.
Excess capacity available for surge loads
= 2 columns out of 9 active columns = 22% excess.
200
-------
Total excess capacity for surge loads
22% excess columns
77, extra contact time (above)
4% lost capacity during backwash (below)
25% net excess capacity.
Column Features
Columns are field erected steel columns, tested to hold 70 psi, dished
heads, with 8-mil interior coating of epoxy-tar, or equivalent, with
gravel support for carbon, and with piping connections for backwashing
for changing the flow sequence through the columns, and for movement
of carbon as slurry to and from regeneration.
Backwashing of Columns
On the basis of pilot plant experience and results from commercial
carbon adsorption for treating municipal effluents, the 3 lead columns
are backwashed once each day and the remaining 6 columns are backwashed
every two days. Total number of columns backwashed per day = 6. During
each backwash, each column is agitated with a short backwash of compress-
ed air and then with a flow of water at 10 gpm/ft2 (5300 gpm) for 15 min-
utes.
Volume of water used per backwashing = 15 x 5300 = 80,000 gal.
Void volume of each column at 50% voids in carbon
= 10,600 ft3 carbon x 0.5 + 5300 ft3 freeboard = 10,600 ft3
= 10,600 x 7.5 gal/ft3 = 80,000 gal.
So backwashing displaces an equal volume of water from the column.
Backwash Schedule
A schedule is arranged to meet above backwashing requirements by back-
washing the No. 3 and No. 1 columns on one day and No. 2 and No. 1
columns on alternate days. One quantity of 80,000 gal of carbon-treat-
ed water is used each day; first for backwashing the No. 3 columns or
No. 2 columns, then for No. 1 columns, and then pumped to the lime
treater. The water from backwashing each column is discharged to a
second 80,000-gal tank where a coagulant is added and the suspended
solids removed by settling and the sludge is pumped to the lime sludge
tank. The clarified water is used for backwashing the next column
and is then discharged to the other settling tank. Thus two settling
tanks are needed to permit partial reuse of the backwash water.
Columns backwashed per day = 6, or 2 each shift.
Product water used = 80,000 gal/day - 100 x 80.000 - 0.8% of daily pro-
9,600,000
duction.
201
-------
Total off-time of columns for backwashing at 20 minutes per set of
columns x 2 = 40 minutes/day = 3% of time.
Total loss of production for backwashing = 3.0 + 0.8 = 3.8%.
Carbon Dosage Rate
The pilot plant runs for microlime-carbon required 2.5 Ib/kgal to
reduce color 230 to 70 CU and TOG 170 to 110 mg/1.
Since this design calls for about the same removals, use a dosage of
2.5 Ib/kgal.
Carbon added = 2.5 x 9,600 kgal/day = 24,000 Ib/day * 1000 Ib/hr.
Carbon Make-Up Rate
Most commercial carbon adsorption units experience a loss of about
5% per generation, so use this.
0.05 x 24,000 = 1200 Ib/day
Regeneration of Carbon
Average rate » 24,000 Ib/day = 1000 Ib/hr.
Hours of operation per day = 24
Furnace - use2multiple-hearth regeneration furnace with capacity of
100 Ib/day/ft , which is based on full-scale plants for treating
municipal wastewater.
Total hearth area with allowance for 20% down time (including 5% make
up) = 24,000 x 1.2 - 290 ft2
100
No. of hearths - use 6
Area per hearth = 290 ft2/6 =48.3 ft2
Diam. of furnace = v^48.3 x 4/3.14 = 8 ft.
CAPITAL COST OF LIME TREATMENT UNIT
The most reliable cost data on lime treatment is that for the Continental
Can Company plant (35) which uses minilime treatment (1000 mg/1 CaO) for
9 mgd of total mill effluent from a pulp and paper mill producing 750
tons/day- Their cost include all associated costs including capital
cost for additional lime kiln capacity, engineering, contractor's over-
head and profit, lime feeder and slaker, sludge dewatering, carbonation,
and instrumentation.
Total cost of Continental Can lime treatment plant $1,762,000
Adjusted to microlime system (lower lime rate,
no carbonation, etc.) 1,182,000
202
-------
Adjusted for lower rise rate in lime clarifier
(use 700 rather than 980 gpd/ft2)
Adjusted from 9-0 to 9.6 mgd
Adjusted from 1972 to Jan. 1973 costs by ENR
Construction Index, multiply by 1.09
Add 87=, of cost of lime kiln for sludge lime added
to mill lime kiln, adjusted to Jan. 1973 costs
Total cost for microlime treatment system
$1,261,000
1,328,000
1,448,000
232,000
1,680,000
CAPITAL COSTS FOR CARBON ADSORPTION UNIT
The cost for the carbon adsorption unit is based on the study by M. W.
Kellogg Co. for FWPCA in 1969 (17). In the Kellogg-FWPCA report the
total installed plant cost was broken down into categories. In preparing
this cost estimate, each of Kellogg costs were adjusted for the conditions
used in this estimate. For example, the cost of carbon adsorbers (44%
of the total Kellogg cost) was multiplied by 2.2, which is the ratio of
volume of carbon in this estimate to that in the Kellogg estimate. The
table below lists the categories of costs for both estimates and the
multiplying factor used in adjusting the Kellogg costs to this estimate.
The multiplying factors include, where applicable, a ratio of sizes to
selected exponents.
Adjusted Costs for Carbon Adsorption Unit
1969 costs, not including first charge of carbon
Kellogg-
FWPCA
costs,$
Foundations, etc. $ 42,000
Carbon adsorbers 670,750
Tanks, carbon transport, 103,880
carbon storage
Structural steel 20,900
Pumps compressors 65,070
Office buildings 5,400
Regeneration system 128,240
Valves, piping 268,990
Electrical, instrumentation 188,260
Paint, utilities 9,660
Conveying equipment 9,360
$1,512,510
Adjust-
ment
factor
1.5
2.2
1.5
Adjusted
to this
estimate, $
63,000
1,476,000
155,800
% of
total
55 '.7
5.9
31,400
78,100
5,400
192,400
403,500
225,900
12,600
14,000
1.2
2.9
0.2
7.2
15.2
8.5
0.5
0.5
$2,658,100 ( 100.2)
Adjust total cost above to Jan., 1973 by ENR Equipment Index (1812/1216)
= $3,961,000
203
-------
Cost of First Charge of Carbon
First Charge = 264,000 Ib/column x 12 columns - 3,170,000 Ib
Warehouse carbon for 30 days of requirements « 1200 Ib/day total
losses x 30 days = 36,000 Ib.
Total inventory of carbon = 3,206,000 Ib
Cost at $0.27/lb = 0.27 x 3,206,000 = $865,600
Cost at $0.10/lb - 0.10 x 3,206,000 = $321,000
CAPITAL COST OF pH ADJUSTMENT AFTER LIME TREATMENT
Amount of C02 needed to decrease pH from 11.5 to 8.0, as determined from
acid titrations of samples from pilot plant runs = 0.018 eq/1 or 3.3 Ib
C02/kgal.
Kiln gas needed at 8.4 ft3/lb and 25 volume-% C02 in kiln gas = 3.3 x
8.4/0.25 = 111 ft3 kiln gas per kgal.
Kiln gas needed per day = 111 x 9,600 = 1,080,000 ft3 (STP)/day.
Kiln gas at 907o utilization of C02 = 1,080,000/0.9 = 1,200,000 SCFD =
833 SCEM.
Retention time for neutralization - allow 10 minutes on basis of results
at other plants.
Volume of pH adjust tank = 10 x 9,600,000/1440 = 67,000 gal = 8,900 ft3.
Diameter of tank =^8900 x 4/3.14 = 28 ft.
Agitation on basis of carbonators at Continental Can plant = 3 at 40 hp ea.
Cost of pH adjust tank, including agitators, gas piping, installation,etc.,
(29,30,24) - $169,000.
CAPITAL COST OF BACKWASH CLARIFICATION AND SURGE TANKS
Need two 80,000-gal tanks with stirrers and conical bottoms to clarify
the washwater after each column is backwashed.
Cost including sludge pumps, installation, etc. = $70,000
OPERATING COSTS FOR LIME TREATMENT
Operating costs for recalcining lime sludge are included in the cost of
recalcined lime. <-
Make-up lime at mill purchase price =4.7 tons/day x $24.23/ton = $114/day
Calcination of lime in lime sludge at mill operating cost = 16.1 tons/day
x $12.40/ton=$200/day
Total cost of lime = $114 + 200 = $3l4/day, or $319/9600 kgal = $0.0328/kgal
204
-------
Labor - use 3 man-hours per shift at $5.85/hr, including fringe benefits=
3x3 shifts x $5.85/9600 kgal - $0.0055/kgal
Electricity - average costs for lime treatment (5,35) = $0.0080/kgal
OPERATING COSTS FOR CARBON TREATMENT
Labor - (maintenance labor is included in repair and maintenance factor
based on total plant investment)
use 1 operator per shift and 1 lab tachnician per day = 4 man-shifts/day
x 8 hr = 32 x $5.85/9600 kgal = $0.0195/kgal
Fuel at 5,700 Btu/lb carbon regenerated (19) and gas at $0.60/mil Btu =
5,600 x ID"6 x $0.60 = $0.0034/lb carbon or 2.5 Ib/kgal x $0.0034=
$0.0085/kgal
Steam - Kellogg report used 1 Ib steam/lb carbon regenerated.
cost @ $0.80/1000 Ib = 2.5 Ib/kgal x $0.80/1000 = $0.0020/kgal
Electricity - experience at large plants for treating municipal effluent
averages $0.012/kgal. Use this.
205
-------
TOTAL CAPITAL AND OPERATING COSTS FOR ESTIMATE NO. 1
Costs
At carbon cost
of $0.27/lb.
At carbon cost
of $0.10/lb.
Capital Costs. $ Million
Lime treatment
Carbon adsorption
Backwash tank
pH adjust tanks
Carbon inventory
Total cost of investment, TCI
1.680
3.961
0.070
0.169
0.866
6.746
1.680
3.961
0.070
0.169
0.321
6.201
Operating Costs. $/kgal
Amortization @ 0.0625 TCI/yr
Repairs and maint. @0.03 TCI/yr
Taxes and insurance (§0.02 TCI/yr
Labor
Plant overhead at 75% of labor
Steam
Gfls
Electricity
Carbon make-up
Total operating costs, $/kgal
0.1202
0.0577
0.0387
0.0255
0.0191
0.0020
0.0085
0.0200
0.0328
Q.Q338
0.3583
Operating costs/ton pulp (@ 12 kgal/ton) $4.30
0.1105
0.0530
0.0355
0.0255
0.0191
0.0020
0.0085
0.0200
0.0328
0.0125
0.3194
$3.83
206
-------
NJ
O
English Unit
APPENDIX K
ENGLISH TO METRIC UNITS CONVERSION TABLE
Abbreviation Conversion Abbreviation
British Thermal Unit
Cubic foot
Cubic foot
Degree Fahrenheit
Foot
Gallon (U.S.)
Gallons per day per square foot
Gallons per minute
Gallons per minute per square foot
Gallons per ton (of production)
Horsepower
Inch
Million gallons per day
Pound (mass)
Pounds per cubic foot
Pounds per square inch
Pounds per thousand gallons
Ton, short
BTU
cu ft
cu ft
°F
ft
gal
gpd/sq ft
gpm
gpm/sq ft
gal /ton
hp
in
mgd
Ib
Ib/cu ft
psi
lb/1000 gal
ton
X
X
X
0.
X
X
X
X
X
X
X
X
.X
X
X
X
X
X
0.2520 =
0.02832 =
28.32
555(°F-32) =
0.3048 * =
3.785
0.04074 =
0.06308 =
40.7
4.172
0.7457
25.4 *
3,785 =
0.4536
16.02
703.1
119.84
0.9072
kg cal
m3
1
C°
m
1
m3/m2 day
I/sec
l/min/m2
1/kkg
kw
mm
m3/day
kg
kg/m3
kgf/m2
g/m3
kkg
Metric Unit
Kilogram - calories
Cubic meters
Liters
Degree Centigrade
Meters
Liters
Cubic meter per square meter day
Liters per second
Liters per minute per square meter
Liters per metric ton (1000 kg)
Kilowatts
Millimeters
Cubic meters per day
Kilograms
Kilograms per cubic meter
Kilograms force per square meter
Grams per cubic meter
Metric ton (1000 kilograms)
* Indicates exact conversion factor
-------
TECHNICAL REPORT DATA
(f lease read Instructions on the reverse before completing)
.REPORT NO.
EPA-660/2-75-004
2.
3. RECIPIENT'S ACCESSION>NO.
4. TITLE AND SUBTITLE
ACTIVATED CARBON TREATMENT OF UNBLEACHED KRAFT
EFFLUENT FOR REUSE
5. REPORT DATE
6. PERFORMING ORGANIZATION CODE
7.AUT
E. W. Lang, W. G. Timpe, R. L. Miller
8. PERFORMING ORGANIZATION REPORT NO
EPA-660/2-75-004
9. PERFORMING ORGANIZATION NAME AND ADDRESS
St. Regis Paper Company
Research and Development Center
Pensacola, Florida
10. PROGRAM ELEMENT NO.
1BB037
11. CONTRACT/GRANT NO.
12040 EJU
12. SPONSORING AGENCY NAME AND ADDRESS
National Environmental Research Center
Pacific Northwest Environmental Research Laboratory
Corvallis,Oregon 97330
13. TYPE OF REPORT AND PERIOD COVERED
Part I - Final Report
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
is.ABSTRACT A four-year pilot plant program was carried out to investigate the techni-
cal and economic feasibility of treating unbleached kraft pulp and paper mill
effluent for reuse. Preliminary laboratory studies and cost estimates indicated
that the following treatment sequences should be investigated in the pilot plant:
1) primary clarification, carbon adsorption; 2) lime treatment, carbon adsorption;
3) primary clarification, bio-oxidation, carbon adsorption.
Water of reusable quality can be provided from unbleached kraft effluent by several
combinations of treatment utilizing activated carbon. Unbleached pulping effluents
typically contain about 1000 color units, 250 mg/1 TOC, and 250 mg/1 BOD. Reusable
water quality as defined in this study is 100 color units and 100 mg/1 TOC. The
most economical treatment is the microlime-carbon process that utilizes low dosages
of lime and clarification followed by carbon adsorption in down-flow granular
carbon beds. Capital cost for treatment by this process of 9.6 mgd of unbleached
kraft effluent from an 800-ton-per-day mill was estimated to be approximately
$6.7 millions. Operating costs, inclusive of capital depreciation, were estimated
to be $0.30 per 1000 gal and $3.58 per pulp-ton, including credit for the reused
water. Carbon adsorption in continuous counter-current stirred contactors was
found to have promise of lower operating cost and substantially lower capital costs
as compared to adsorption in fixed beds.
17.
KEY WORDS AND DOCUMENT ANALYSIS
a.
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
COSATI Field/Group
Industrial Waste Treatment, Activated
Carbon Treatment, Chemical Removal, Bio-
chemical Oxygen Demand, Industrial
Wastes, Pulp Mills, Cost Estimates,
Pilot Plants, Effluents, Calcium
Carbonate
Pulp Mill Effluent,
Color Removal, Tertiary
Treatment, Water Reuse,
Unbleached Kraft Waste-
water Treatment, Micro-
lime-Carbon Process,
Wastewater Treatment,
Tnhal Organic
05D
Ores
iS (TTtli
8. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (ThisReport/
Unclassified
21. NO. OF PAGES
207
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
Form 2220-1 (9-73)
U.S. GOVERNMENT PRINTING OFFICE: I975-698-576 I\S2 REGION 10
------- |