EPA-660/2-74-008
February 1974
Environmental Protection Technology Series
Color Removal And Sludge Dispos
Process for Kraft Mill Effluents
Office of Research and Development
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and
Monitoring, 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
' , '^v • •
This report has been assigned to the ENVIRONMENTAL
PROTECTION TECHNOLOGY 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-660/2-74-008
February 1974
COLOR REMOVAL AND SLUDGE
DISPOSAL PROCESS
FOR
KRAFT MILL EFFLUENTS
By
Edgar L. Sprtrfll, Jr.
Project 12040 DRY
Program Element 1B2037
Project Officer
Dr. Richard L. HW
Office of Research and Development
Environmental Protection Agency
1600 Patterson St.
Dallas, Texas 75201
Prepared for
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
MASJUHGION. D.C. 20460
For «al» by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 90402 - Price $1.85
-------
EPA Review Notice
This report has been reviewed by the Office of
Research and Monitoring, 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.
-------
ABSTRACT
A treatment plant, removing color by lime addition and recovering
sludges, has been treating over 80% of the effluent of an unbleached
kraft mill for one year. Using up to 1,100 mg/1 of CaO, with normal
mill fiber loss as a precipitation aid, average color reduction was
80% for all-kraft effluent. At upper range of lime dosage, when
residual dissolved Ca was above 400 mg/1 as CaO, color removal was
85 - 93%. When mill production included 33 - 40% NSSC hardwood pulp,
color reduction averaged only 65%.
About 12% BODc reduction was observed, and average TOC reduction was
nearly 40%. The chief negative factor is need for emergency protection
against alkaline impact on secondary treatment and receiving stream.
Following centrifuge dewatering, sludge incineration has had minimal
impact on kiln operation; there were some adverse effects on lime
quality. Lime recovery was 93%. Mill kiln capacity must be increased
about 25%.
Primary clarification and sludge disposal are included in the process.
Operating costs, exclusive of capital factors, are estimated at $0.50-
$0.80 per ton of paper, or 5.5
-------
TABLE OF CONTENTS
SECTION PAGE
I CONCLUSIONS 1
II RECOMMENDATIONS 3
III INTRODUCTION 4
IV DEVELOPING A PLANT DESIGN 10
V EQUIPMENT PERFORMANCE AND CHANGES 50
VI PROCESS PERFORMANCE 65
VII ECONOMIC CONSIDERATIONS 79
VIII ACKNOWLEDGEMENTS 82
IX REFERENCES 83
X GLOSSARY 85
XI APPENDICES 87
-------
FIGURES
PAGE
1 PROCESS FLOW SHEET - EARLY VERSION 9
2 EFFECT OF TIME AND LIME DOSAGE ON COLOR (KRAFT) 11
3 EFFECT OF TIME AND LIME DOSAGE ON COLOR (WITH NSSC) 12
4 EFFECT OF LIME DOSAGE ON SETTLING RATE 16
5 EFFECT OF REACTION TIME ON SETTLING RATE 17
6 UNIT AREA FOR SLUDGE THICKENING 21
7 EFFECT OF CARBONATION pH ON DISSOLVED CALCIUM 26
8 EFFECT OF COLOR SLUDGE DEWATERING 29
9 FLOW SHEET WITH CENTRIFUGE 36
10 FLOW SHEET FOR PLANT DESIGN 40
11 GAS DIFFUSER 44
12 EMERGENCY FLOW SCHEME FOR EFFLUENT 49
13 CENTRIFUGE FLOW MODIFICATION 61
14 FLOW SAMPLER FOR pH PROBE 64
15 DAILY DATA WORK SHEET (SAMPLE) 68
16 SOLUTION CaO VERSUS COLOR REMOVAL (KRAFT) 70
17 SOLUTION CaO VERSUS COLOR REMOVAL (WITH NSSC) 71
18 SOLUTION CaO VERSUS TOC REDUCTION (KRAFT) 73
19 SOLUTION CaO VERSUS TOC REDUCTION (WITH NSSC) 74
20 COLOR - TOC RELATIONSHIP (KRAFT) 75
21 COLOR - TOC RELATIONSHIP (WITH NSSC) 76
VT
-------
TABLES
No. Page
1 Effect of Variables on Color Removal (with NSSC) 13
2 Effect of Added Black and Brown Liquor on Color Removal
Efficiency 13
3 Pilot Plant Color Removal Data 14
4 Effect of Time and Lime Dosage on Calcium Precipitation 15
5 Effect of Time and Lime Dosage on Calcium Precipitation
(with Added Liquor) 15
6 Five-Variable Statistical Study of Process 18
7 Estimated Amount of Color Sludge 19
8 Color Sludge Thickening Data 20
9 Effect on Filtering Rate of Ratio of Mud Solids to
Sludge Solids 22
10 Effect of Sludge Concentration of Filtration Rate 23
11 Average Filtration Rates 24
12 Efficiency of Pilot Plant Carbonate Clarifier 27
13 Efficiency of Color Removal from Pulping Area Effluent 27
14 Pilot Centrifuge Data, Factory 30
15 Pilot Centrifuge Data, Factory 31
16 Pilot Centrifuge Data, Mill Laboratory 32
17 Pilot Centrifuge Data, Mill Laboratory 33
18 Pilot Centrifuge Data, Mill Laboratory 34
19 Treatment System Up-Time 67
20 Color Reduction Before and After Lime Recovery 69
vii
-------
SECTION I
CONCLUSIONS
A treatment process, employing lime in concentrations below the solubil-
ity limits of calcium hydroxide in water, has been developed for removal
of colored matter from unbleached kraft mill effluent and for recovery
of the lime and precipitated organic substances. A full-scale plant em-
bodying this process has been designed and built at Continental Can Com-
pany's Hodge, Louisiana, mill and operated since August 18, 1971. Oper-
ation through August 31, 1972 is described in this report, and the fol-
lowing conclusions are drawn:
1. Effectiveness of the Process
- Addition of lime to combined, unclarified kraft pulp and
paper mill effluent, at rates below 1100 mg/1 as CaO, can
produce average color reduction of 80% on a continuous,
mi 11-scale basis.
- If system reliability can insure uninterrupted lime feed
at rates which maintain a residual dissolved calcium con-
centration of at least 400 mg/1, color reduction will ex-
ceed 85% and range to above 90%. The required lime addition
will normally be about 1000 mg/1 as CaO. (High sodium alka-
linity may reduce this performance.)
- A reduction of about 12% in 6005 is realized, and removal
of solid-phase organic material is virtually 100%. How-
ever, BODij may actually increase if lime feed is inadequate
or intermittent.
- Where NSSC materials contribute more than half of the color,
efficiency of color removal may be reduced 15% or more.
- Sedimentation of calcium carbonate after carbonation is
adversely affected by incomplete removal of colored sub-
stances and is influenced by other factors not fully eluci-
dated in this study.
- Precipitated color bodies, together with the usual "primary
treatment" sludge, can be disposed of by incineration in a
lime kiln, coincidentally with recovery of the lime utilized
in the treatment.
2. Design and Performance of Equipment
- Basic concepts and criteria used for design of most equip-
ment for the project have proved valid.
1
-------
- Kiln gas flow rate provided for carbonation must reflect
minimum C02 concentration which is likely to occur, and
must include the demand imposed by sodium alkalinity of the
effluent.
- Rake torque imposed by the carbonation clarifier is below
expectations based on other calcium carbonate slurries. A
greater area and retention time allowance for this clarifier
may be justified.
- Centrifuge dewatering of sludges requires careful choice and
application of equipment, but the centrifuge appears to re-
present the method of choice for this operation.
3. Effects on Kraft Mill Operations
- For kraft mills treating 10,000 - 15,000 gallons of effluent
per ton of pulp, lime kiln requirements are increased about
25%.
- Where filler clays and similar insolubles are avoided, lime
contamination will remain within tolerable limits.
- Requirements for carbon dioxide supplied by lime kiln stack
gas impose a constraint upon the scheduling of interruptions
in kiln operation, as do lime supply and sludge disposal.
- Holding ponds or other emergency alternatives are required
to prevent alkali damage to secondary treatment operations
(or to receiving stream) in case of system failure at the
carbonation stage.
4. Operating Costs
- To treat 13 million gallons of effluent per day from a 1,500
tons-per-day, integrated kraft pulp and paper mill, operating
costs will be approximately $0.50 per ton of paper (assumed
fuel cost, 48£ per million Btu). Costs of depreciation, in-
surance and taxes are not included.
- The functions of conventional primary clarification and final
sludge disposal are provided within the stated costs of the
process.
-------
SECTION II
RECOMMENDATIONS
There is evidence that, after color precipitation with lime, the final
effluent may extract color from natural organic materials during and
after secondary treatment. This phenomenon was outside the scope of
this project, but the mechanisms and magnitude of such color effects
on receiving streams should be explored.
The effect of lime kiln stack gas on color, total organic carbon and
8005 of effluent needs further study.
The effects of the organic components of color sludge on the heat re-
quirements for calcining have not been adequately measured. Study might
properly include direct establishment of accurate heat and material bal-
ances, as well as examination of pyrolysis products in the exit gases
from the kiln.
It should be ascertained whether moderate further increases in lime dose
have significant effects on reduction of TOC and 8005.
-------
SECTION III
INTRODUCTION
The Kraft Effluent Color Problem
The spent liquors from wood pulping by alkaline methods, such as the
kraft process, are very dark colored; indeed they are commonly referred
to as "black liquor." Further delignification procedures involving al-
kalis, such as caustic extraction steps in bleaching, also yield highly
colored solutions. The kraft process characteristically provides for
separation, concentration and combustion of spent liquor to recover soda
chemicals and heat value; however, existing technology does not prevent
some loss, and effluents from kraft mills are distinctly colored.
Discharge of substantial ratios of untreated mill effluent into streams
has long been recognized as dangerous to fish and other aquatic life.
The adverse effects are largely alleviated by adaptations of the con-
ventional processes of "primary" and "secondary" waste treatment. Al-
though such treatment may be very effective in removal of suspended
solids and reduction of BOD, these procedures are almost totally in-
effective in reducing color of the effluent.
Color contributed by paper mill effluent is at least esthetically un-
desirable, and it may make the stream unsuitable for such uses as muni-
cipal water supply or certain recreational purposes. Residual organic
substances, refractory to bio-oxidation, may represent a significant COD
(or TOC) value.
Previous Work
In 1952, an investigation and review revealed that much work had been
done to develop means for reducing color of pulping effluents, but that
no practical methodology had resulted, and little technical literature
on the subject existed. (!) It was found that color could be substan-
tially removed from solution by a number of agents, including alum, min-
eral acids, ferric sulfate, lime, barium aluminum silicate, activated
silica, and various heavy metal salts, and by combinations of these
agents. However, the costs of such treatments, applied to the great
volume of water discharged by a pulp and paper mill, were very high,
and disposition of the residues presented further problems for which
there were no attractive solutions.
This review marked the beginning of a strong and sustained effort,
supported by the paper industry, to develop technology for elimination
or reduction of color from chemical pulping and bleaching effluents.
Efforts were largely directed toward use of lime, both because that
material is low in cost and because it offers possibilities for recovery
-------
and re-use in the pulping operation. The ensuing program produced a
number of reports showing that lime could yield a good degree of color
reduction and presenting several attempts at practical recovery of the
lime. (2,3,4,5,6,7)
The combination of the lime treatment research program of the N.C.S.I.
(National Council for Stream Improvement, more recently changed to
N.C.A.S.I., adding "Air and . . ."), is a patented process fully re-
ported in their Technical Bulletin No. 157. (8,9) That process was
specifically designed for bleachinq system effluents (especially efflu-
ent from caustic extraction stages). The method involves use of very
large amounts of lime (typically about 20,000 parts, w/w, per million
parts of effluent), and the colored substances are said to be deposited
on suspended, solid-phase calcium hydroxide. The color laden calcium
hydroxide is sedimented in a clarifier and withdrawn as a heavy sludge.
The treated effluent is carbonated to remove soluble lime, converting
it to calcium carbonate, which is then removed and mixed with the cal-
cium hydroxide sludge. The mixed sludge is filtered and is then used
to causticize sodium carbonate in the kraft mill liquor-processing sys-
tem. Color bodies are dissolved into the kraft cooking liquor and ul-
timately arrive with the spent liquor at the recovery furnace, while
the lime is conventionally reburned and re-used.
The N.C.S.I. scheme, called "The Massive Lime Process" for evident rea-
sons, has the advantages of providing a high degree of color removal,
a large reduction in COD, a substantial reduction in 6005 (at least from
bleachery wastes), and a means for disposing of the separated color bod-
ies and recovering the lime. Its chief disadvantages are: the very
large amount of lime required and the substantial changes in operation
of the kraft causticizing system. Since no more lime can be employed
than is needed for causticizing, the volume of water which can be treated
is limited to 3,000 to 5,000 gallons per ton of kraft pulp made.
The Massive Lime Process has been studied under EPA grant 12040 DYD, for
which a comprehensive report has been prepared. (10)
Considerable work has been done by at least one investigator to develop
a color removal process based on commercial alum. (11) Although good
color reduction has been reported and a sludge recovery procedure has
been outlined, economic feasibility does not appear to have been devel-
oped.
The use of activated carbon, especially as a "polishing" treatment to
remove the last traces of color (12,13) has been studied, and the possi-
bilities for re-use of the reclaimed and renovated water has been dis-
cussed. (14)
Color Precipitation by Below-Saturation Lime Concentrations
Technical personnel at the Hodge, Louisiana mill of Continental Can
-------
Company, Inc. maintained active contact, through corporate membership
in N.C.S.I., with the N.C.S.I, color removal research efforts. After
several years consideration of the problem of adapting the Massive Lime
Process to their need to treat a rather large volume of effluent, a
laboratory program was begun exploring lime and other precipitants.
Data developed by Berger and others of N.C.S.I, was subjected by them
to analysis which led them to the conclusion that lime precipitation of
color, as they had practiced it, proceeds by a mechanism wherein adsorp-
tion of color bodies on solid calcium hydroxide is an essential factor.
(15,16) .
Observations made at Hodge led to the hypothesis that soluble calcium
hydroxide could effect color precipitation at satisfactory rates by
utilizing other solid-phase surfaces for deposition. It was further
reasoned that such a material, or combination of materials, might confer
needed thickening and dewatering properties.
The ensuing program involved addition of various solids, co-precipitation
of solids, use of flocculation aids, and a variety of mechanical and chem-
ical modifications of treatment.
In the course of the study, it was found that, in the presence of the
amounts of fiber "fines" commonly escaping paper mill reclaim systems
and passing into the sewer, color could be rapidly precipitated with
lime additions well below the normal water solubility of calcium hydrox-
ide. The economically fortuitous availability of this material led to
intensive study of the properties of the precipitates as related to the
precipitation conditions and the implications of various flowsheet possi-
bilities.
It was observed that samples representative of total kraft mill effluent,
when treated with 1,000 to 1,500 mg/1 of CaO, exhibited settling rates
more rapid than simple primary clarification of the original effluent.
The precipitates could be thickened to volumes suitable for further de-
watering by vacuum filtration or centrifugation. Direct vacuum filtra-
tion proved unpromising. However, when these same precipitates were
mixed with recausticizing sludge (calcium carbonate) in proportions
approximating the normal availability of each, acceptable filtration
properties were found. Formation rates from 35 Ibs./sq. ft./hr. to
several times that value were observed, depending upon the amounts of
precipitated lignin and of fiber in the mixture.
Centrifuge dewatering efforts were made, using a laboratory centrifuge
with 50ml glass bottles. Because of the elastic nature of the fiber-
containing sludge, and because minimum cake volume could not be observed
or measured under dynamic conditions, no consistent or conclusive data
were obtained.
A small (2 g.p.m.) pilot system, providing continuous lime addition,
sedimentation and carbonation, was constructed to test the validity of
-------
bench-scale data against the variations of effluent quality, and to
compare continuous settling with laboratory batch tests. The results
indicated consistent color reduction of kraft effluents in the range of
80 - 90%. The presence of NSSC (hardwood) process effluent reduced this
performance, but at reasonably low ratios, the effect was not serious.
Sedimentation experience seemed to support previous estimates of per-
missible rise rates around 1.0 g.p.m./ft.2
Project Definition
Progress of the research work at Hodge had been shared with the Execu-
tive Secretary of the Louisiana Stream Control Commission. From an
early date, the possibility of a treatment facility to handle total (or
nearly so) mill effluent was envisioned, to deal with the special color
considerations of Hodge mill discharge into a low-flow stream.
The cost of so expensive a venture with unproven technology was forbid-
ding. Since the most serious problems lay in the realm of integrating
the system with mill lime processing, a system large enough to establish
practicality -- but too small to serve the mill — seemed even less at-
tractive. Since a full-scale treatment facility should have wide impli-
cations in the advance of industrial waste treatment technology, a con-
ference was arranged between representatives of Continental Can Company
and the Federal Water Pollution Control Administration (later, Federal
Water Quality Administration, and then incorporated into the Environ-
mental Protection Agency). The possibility of a demonstration grant for
technical studies, with some participation in construction costs, under
the FWPCA program or technology development, was the subject of this meet-
ing and others to follow.
The ensuing discussions, and the agreement of the Continental Can Company
corporate management to undertake the financial commitment, led to the
offer and acceptance of a demonstration grant with the following plan and
objectives:
Beginning with the tentative flow sheet shown in Figure 1, laboratory
and pilot plant work would be extended and refined, and a facility would
be designed and built, capable of treating the entire process effluent
of the Hodge mi 11.
Upon completion of construction, the system would be placed in opera-
tion, and for a period of one year studies would be carried out and
operational effort would be made to:
1. Demonstrate effectiveness of the process in reducing color
of pulp and paper mill effluents.
2. Demonstrate practicability of the process as a means for
disposal of the fibrous sludge normally separated in paper
mill primary clarification systems.
7
-------
3. Determine effectiveness of the process in reducing oxygen
demand of pulp and paper mill effluents.
4. Assess the effects on lime calcining of using a lime kiln
to incinerate the combined sludges (including fiber) from
the lime treatment system.
5. Develop economic parameters related to the process.
It was agreed that a continuous study program would be maintained through-
out the demonstration period to record related variations of water papa-
meters (pH, color, BOD, COD or TOC, calcium), sludge properties and lime
kiln operations. Interpretations would be developed and the study expan-
ded or modified to gain broader understanding in pertinent areas.
8
-------
<£>
LIME KILN
LIME
STORAGE
CLARIFIER
CAUSTICIZING LIME. MUD
I
I
coz
l_
1
CLA
1<>^
•"I
CLARIFIER ' CORRECTION
CAR8ONATOR
TREATED
WATER
L/ME TO
KILN
BELT FILTER
PROCESS FLOW SHEET - EARLY VERSION
Figure 1
-------
SECTION IV
DEVELOPING A PLANT DESIGN
Bench Studies
A renewed program of bench-scale study of the process was instituted to
confirm the applicability and effectiveness in treating the range of
effluent variations to be expected in plant operation; to define more
closely the limits of the process; and to provide data for engineering
the system.
Color Removal Efficiency. A series of jar tests was conducted to exa-
mine the effect of the lime dosage and reaction time on color removal
efficiency. The results are plotted in Figure 2. It appears that at a
dosage of 1,000 mg/1 Ca(OH)2 and greater, reaction time is not important.
The benefit of adding amounts of lime greater than 1,000,mg/1 is small,
in this case. The original color of this effluent was 300 units, which
is fairly low, and the source is a combined screen room and paper machine
sewer.
A similar series was conducted with additions of kraft black liquor and
NSSC brown liquor to simulate total mill effluent of an integrated mill.
The mixture had a color of 550 units, representing the equivalent of a
mill with good pulp washing and low recovery losses. Treatment with
lime at several concentrations and reaction times was employed to gen-
erate the data plotted in Figure 3. It is apparent that at any level
of lime dosage, the efficiency of color removal is lower than in the
previous series. The effect of reaction time still appears to be small.
Using 1,500 mg/1 Ca(OH)2 (or 1,135 mg/1 CaO), an 85% color reduction
was observed. It will be noted that this curve does not appear to extra-
polate to zero on the abscissa; this is largely due to carbonates in the
effluent, which precipitate some of the calcium ion.
Another series of tests were made using effluent obtained while the mill
was producing and using some NSSC as well as kraft pulp. The color of
the effluent was 2,820 units, which is in the high range, and the pre-
dominant contribution of color was from the semi-chemical process. At
a lime (calcium hydroxide) dosage of 1,600 mg/1, a 68% color removal was
achieved, and at that dosage increasing the reaction time from 30 minutes
to 60 minutes had no effect. These data are shown in Table 1.
10
-------
70O
90
80
70
ih60
o
OJ
or 40
030
10
REACTION TIMES
X O MINUTES
D
A
V
O
15 MINUTES
30MINUTES
SO MINUTES
MINUTES
Zoo ^oo eoo
LIME ADDITION ,
80O TOGO 12.00
/L. as CaO
Hoo
EFFECT OF TIME AND LIME DOSAGE ON COLOR
(KRAFT)
Figure 2
11
-------
TOO
9o
5
UJ
O
U.
80
70
o 50
ui
g
8 30
REACTION TIMES
X O MINUTES
O JS
D 30
A GO MINUTES
v 90 MINUTES
200 -f-00 600 800 1000 7200
LIME ADDITION ,
as
CaO
EFFECT OF TIME AND LIME DOSAGE ON COLOR (WITH NSSC)
Figure 3
12
-------
TABLE 1
EFFECT OF VARIABLES ON % COLOR
REMOVAL EFFICIENCY (WITH NSSCj
Treatment Lime Addition as Ca(OH)2
Time, min 1.200 mg/1 1,600 mg/1
30 18.4 68.1
60 41.5 68.0
Several trials were made in which black liquor and brown liquor were
added to a kraft screen room and paper mill effluent. "Normal" (N)
amounts of these liquors were estimated, to represent proportional waste
inputs of these materials from all mill sources. Twice-normal amounts
of black and brown liquors were added to determine their effect. In
these series, the effluent samples were treated with 1,500 mg/1 Ca(OH)2
for 30 minutes. The results are presented in Table 2. They indicate
that about 90% color removal is achieved on effluent to which no additions
were made. Addition of 2N black liquor had little effect on the color
removal efficiency. However, the data in sets "B" and "C" suggest that
the color of the added brown liquor is considerably more difficult to
remove, and that black liquor may actually assist in removal of NSSC
color.
TABLE 2
EFFECT OF ADDED BLACK AND BROWN LIQUOR
ON % COLOR REMOVAL EFFICIENCY
Samp!e
A
A
B
B
C
C
All of the work described above was done by "jar-test" techniques. In
addition, a number of runs were made with the 2 g.p.m. pilot plant re-
ferred to in Section III. One series yielded the data shown in Table 3.
The pilot plant was operated with 20 minutes retention in the reactor and
rise rates of about 1.0 g.p.m./ft.2 in the clarifiers. Retention time in
each clarifier was also about 20 minutes. For this series, feed effluent
was stored in a tank, so that feed for each run would be constant in com-
position. The series contains some feed samples darkly colored due to
the presence of brown liquor. It appears that color reduction efficien-
cies in excess of 80% are possible if sufficient lime is added.
13
Black
Liquor
none
2N
none
2N
none
2N ,
Brown
None
88
86
92
90
90
89
Liquor Added
2N
85"
82
57
79
49
59
-------
TABLE 3
PILOT PLANT COLOR REMOVAL DATA
Input Output % Color mg/1
Color Color Removal CaO Added
850 90 90 1,000
900 60 v 93 1,000
4,000 2,300 42 1,000
4,750 825 83 2,000
1,950 300 85 2,100
475 215 55 1,000
Calcium Distribution. The purpose of the carbonator and second clarifier
(see flow sheet, Figure 1) is to recover lime which remains soluble in
the discharge from the first clarifier. It is relevant, therefore, to
ask how much lime is available for recovery and how much is precipitated
in the first clarifier.
Calcium analyses were made on the supernatant liquid from a number of
the jar tests described above. Calcium data from the tests mentioned
above in reference to Figure 2 and Figure 3 are presented in Tables 4
and 5. These data show that the percentage of the calcium precipitated
in the reaction is increased when pulping waste liquors are added to the
effluent at any level of calcium dosage. Further, they indicate that as
the lime dosage is increased, the percentage of the calcium precipitated
also increases. In the proposed range of lime dosage, roughly half of
the calcium will remain in the effluent leaving the first clarifier.
In any interpretation of the above data, it should be noted that the sol-
uble carbonate content of the effluent will react with lime to form cal-
cium carbonate, thus reducing the Ca(OH)2 concentration available to
promote color-reduction mechanisms. Since the carbonate content before
lime addition was not determined, and carbon dioxide absorption was not
rigorously avoided during the jar-test procedures, conclusions drawn from
the data should not be over-extended.
Despite the reservations just stated, one clear conclusion is evident
(in addition to the approximate fraction of lime in solution): the
color-precipitation reaction is not strictly stoichiometric with respect
to the calcium removed from solution.
14
-------
TABLE 4
EFFECT OF TIME AND LIME DOSAGE
ON CALCIUM PRECIPITATlM
Calcium in Supernate, mg/1 as Ca
Lime Dosage, mg/1 Ca(OH)2
1.000 1.500
~29T~ ~37l~
Time
Min. 500
T~ T3T
15 148 328 416
30 168 316 434
60 260 292 432
90 138 320 400
Average as
Ca(OH)2 312 574 762
% of Ca
Precipitated 37.6 42.6 49.2
2,000
"2BT"
500
460
528
468
892
55.4
TABLE 5
EFFECT OF TIME AND LIME DOSAGE ON CALCIUM
PRECIPITATION (KITH ADDED LIQUOR)~
Calcium in Supernate, mg/1 as Ca
% of Ca
Precipitated
500
TTF
136
104
148
104
226
54.8
Lime Dosage
1,000
mg/1 Ca(OH)2
1,500
260
272
244
296
481
51.9
324
368
340
412
638
57.5
2,000
""SET
400
456
484
452
797
60.2
Settling of Color Sludge. The effluent sample treatments used to
generate the data for Figures 2 and 3 were also used to determine set-
tling rates. The settling rate information is presented in Figure 4
and Figure 5. All of these rates were greater than 3.5 inches per min-
ute. On this basis, it appeared that earlier projections of a clarifier
rise rate of 1.0 g.p.m./ft. were amply justified. The adequacy of this
design rate was also confirmed by observation of the pilot plant opera-
tion.
15
-------
z
2
DC
ul
D-
to
UJ
X
vj
lu
vD
Z
_J
K
UU
00
X-REGULAR EFFLUENT
0-UQUOR ADDED
I
^oo
-------
X-RE6ULAR EFFLl/ENT
O-LIQUOR ADDED
-x
-o
ZO 4-0 60
REACTION TIME
80
MINUTES
loo
EFFECT OF REACTION TIME ON SETTLING
Figure 5
17
-------
Statistical Study. A half-replicate of a 25 factorial experiment was run
to test the effects of reaction time, lime dosage, temperature, the in-
fluence of brown liquor, and the influence of black liquor. The response
measured was color reduction. The data are presented in Table 6.
TABLE 6
FIVE - VARIABLE STATISTICAL STUDY OF PROCESS
Temp.
QF.
100
120
100
120
100
120
100
120
100
120
100
120
100
120
100
120
The analysis was done by a statistical specialist in Continental Can
Company's Central Packaging Research Laboratory in Chicago. She reported
the following results:
1. There was no detectable effect of temperature on time
of reaction.
2. Increased lime addition improved the efficiency of color
removal (95% confidence).
3. As the amount of black liquor was increased, the efficiency
of color removal was increased (90% confidence).
4. As the amount of brown liquor was increased, the efficiency
of color removal decreased (99% confidence).
The last two conclusions confirm strongly the earlier observations that
the color of black liquor is much more easily removed than the color of
brown liquor.
Sludge Quantities. The amount of sludge from the color precipitation
(first) clarifier must be known, to insure provision of adequate dewater-
18
Ca(OH)2
mg/1
800
800
1,600
1,600
800
800
1,600
1,600
800
800
1,600
1,600
800
800
1,600
1,600
Time,
Min.
30
30
30
30
90
90
90
90
30
30
30
30
90
90
90
90
Black
Liquor
N
N
N
N
N
N
N
N
3N
3N
3N
3N
3N
3N
3N
3N
Brown
Liquor
3N
N
N •
3N
N
3N
3N
N
N
3N
3N
N
3N
N
N
3N
% Color
Reduction
44.4
63.2
80.0
72.8
56.8
55.0
32.8
92.0
84.6
74.1
80.2
86.3
61.2
75.4
89.7
85.8
-------
ing and kiln capacities. Although estimates had been made on the basis
of theoretical treatment of data already obtained, further data were
obtained by a direct empirical approach. This was done using 50-gallon
batches of effluent to which known amounts of lime were added. The
sludge was allowed to settle and separate, and a dry weight was deter-
mined. The results are shown in Table 7. It is seen that the amount of
sludge is roughly proportional to the amount of lime added, and that at
a lime dosage of 1,200 mg/1, about 5 tons of sludge can be expected from
each million gallons of effluent. The amount roughly doubles when the
lime dosage is doubled. From this information normal sludge solids were
projected on the basis of the 1,200 mg/1 data, and peak loads on the basis
of 2,400 mg/1 level.
f
Obviously, sludge solids values will be affected by variations in fiber
content of the effluent.
TABLE 7
ESTIMATED AMOUNT OF COLOR SLUDGE
Sample Date
11/06
11/06
11/07
11/07
11/07
11/09
11/09
11/09
11/10
11/10
11/11
11/11
11/14
CaO, mg/1
1,200
1,200
1,200
1,200
1,200
2,400
2,400
2,400
2,400
2,400
2,400
2,400
1,200
Sludge,
Tons Per Day
55
52
59
59
62
105
138
122
118
120
135
139
78
Trials made with 50-gallon samples.
Calculated to total effluent of 12 million gallons per day.
Thickening of Color Sludge. Removal of as much water as possible from
the color precipitation sludge prior to filtration or centrifugation is
advantageous because it will reduce the design size of these expensive
devices. Moreover, if filtration is integrated with that of causticiz-
ing sludge, excessive filtrate volume would increase soda losses in
proportion to the amount of filtrate not usable as weak wash.
The response of color clarifier sludges to the action of a picket-type
thickener was next studied. If such a thickener should prove effective,
it was felt that either the sludge could be fed to a separate thickener,
19
-------
or the rake mechanism of the first clarifier could be designed to produce
a thickening action. The second of these alternatives seemed the more
attractive, since it would be less expensive. Suppliers were approached
at this time to solicit suggestions concerning suitable mechanisms for
this purpose.
In thickening sludges of this type, the critical dimension is the area
of the sludge blanket, rather than the depth. "Unit area" (U.A.) is
defined as the area in square feet required to thicken a sludge flow of
one ton per day from a given starting consistency to a desired final
consistency. The units of "U.A." are square feet per ton per day.
A series of laboratory thickening tests was run using sludge produced in
the 2 g.p.m. pilot plant. One group was run using a lime dosage of 2,400
mg/1 CaO, and another using 1,200 mg/1 lime. Resultant unit areas re-
quired to achieve 10% solids with 2,400 mg/1 group are shown in Figure 6.
It can be seen that to thicken from 2% to 10% solids will require a unit
area of about 40 square feet per ton per day. Since it has been esti-
mated that there would be about 10 tons per day of sludge produced at
this lime addition rate, a total thickening area of 400 square feet per
daily million gallons effluent would be necessary. This is approximately
half the area which would be provided by the most rapid clarifier rise
rate which was contemplated.
The analysis of the 1,200 mg/1 lime addition group did not yield such
neat figures. The reason for the scatter was not determined with cer-
tainty, but the data still appeared useful. (Variable fiber content may
have been involved, and variable carbonate content would affect the ac-
tive lime concentration, especially at the lower lime dose.) The data
are shown in Table 8. The highest unit area shown is 154 square feet
per ton per day. Since 5 tons of sludge per million gallons are estima-
ted, a total thickening area 770 sauare feet per daily million gallons
is required for this worst case. (A median value around 60 square feet
per ton per day is indicated.) A capability prediction of 10% solids
content for thickened sludge seemed reasonable.
TABLE 8
COLOR SLUDGE THICKENING DATA
Sludge Number
5
5
6
6
8
8
24
25
35
35
*Unit Area (U.A.
Initial Solids /
2.2
4.4
3.2
6.3
2.9
5.8
2.1
1.9
2.0
1.0
), sq. ft./t.p.d.
I Unit Area Required*
67
52
45
34
47
27
99
154
104
114
to obtain 10% final solid:
Efflu-
ent treated with 1,200 mg/1 CaO, except No. 5 treated with 800 mg/1.
20
-------
T5
Q_
3-40
20
10
X
SLUDGE No
O - 9
-------
The calculations used to arrive at the conclusions in Table 8 were based
upon the technique discussed by Rich (17). The data were also analyzed
by a staff specialist of a major manufacturer of clarification equip-
ment, and he arrived at essentially the same conclusions.
Sludge Filtration. A program of study was directed to the fi 1 terabi1ity
of mixtures of primary clarifier sludge and causticizing lime mud. Mix-
tures of lime mud and effluent sludge were subjected to filter leaf tests.
The filtering cycle used employed half of the cycle time for filtration,
a quarter of the time for drying and a quarter of the time "dead." Fil-
ter leaf tests usually were made using 20, 30, and 40-second filtering
durations. When other times were used, they are noted.
The importance of the relative amount of lime mud and sludge is illus-
trated by the data in Table 9. For this series, a quantity of effluent
was treated with 800 mg/1 CaO. The resulting sludge then was mixed with
several amounts of lime mud taken at about 35% solids from the mill's
mud washer. Lime mud alone will filter at the rate of 300 - 350 pounds
per hour per square foot. Examination of the data in the table shows
that when mixed with sludge, the filtering rates are considerably slower
and that the rate is strongly influenced by the proportion of lime mud
to sludge.
For most of the testing program, 50-gallon batches of effluent were
treated with appropriate amounts of lime, and the sludge produced was
mixed with a proportional amount of lime mud. In practice, the sludge
from 50 gallons was mixed with 2.5 pounds (O.D. basis) of lime mud.
This mixture represents the conditions of a mill using 175 tons per day
of CaO for causticizing, and treating 12 million gallons of effluent
daily.
TABLE 9
EFFECT ON FILTERING RATE OF
RATIO OF MUD SOLIDS TO SLUDGE SOLIDS
Mud/Sludge Ratio* Filtration Rate**
3.0 29.1
4.9 54.2
7.0 95.2
*Lbs. mud solids added to sludge from 100 gallons effluent.
**(Lb.)/(Hr.)(sq.ft.), at 30 second formation time.
Some work was done to demonstrate the effect of the solids content of the
mud-sludge mixture on the filtration rate. Two sets of data illustrating
the effect are shown in Table 10. It is apparent that as the solids con-
tent is increased, the filtration rate is improved, and it is concluded
that an effort should be made to provide as much mechanical dewatering
capacity as possible in the system prior to filtration in order to mini-
mize the work required by the filters. Picket thickeners would seem
suitable devices for this purpose. These data also illustrate the marked
22
-------
Solids Before
Filtration, %
10.0
10.0
14.3
14.3
9.3
9.3
9.3
12.8
12.8
12.8
Formation Time
Seconds
18
36
18
36
30
45
60
30
45
60
Formation Rate
(lb.)/'(Hr.)(sq. ft.)
44.0
65.5
59.9
82.2
73.4
59.0
51.0
114.0
91.7
82.0
improvement in filtration rate brought about by using a faster cycle.
The negative aspect of this expedient is the thinner cake produced; if
a belt filter were required, separation of cake from the filtering belt
would become a problem if the cake became too thin and flexible.
TABLE 10
EFFECT OF SLUDGE CONCENTRATION ON FILTRATION RATE
Day
10/10
10/10
10/10
10/10
10/17
10/17
10/17
10/17
10/17
10/17
There are presented in Table 11 average sludge/mud filtration rates for
the days indicated. The data are arranged in order of increasing fil-
terability, rather than in chronological order. Where available, the
solids content of the mixture prior to filtration is shown. These data
are applicable for a drum at 50% submergence, using a 30-second formation
time. On the average, if a 20-second formation time were used, the fil-
tration rates would be about 20% higher. It should be noted, too, that
the initial solids content of the sludge mixtures average about 15%. It
should also be remarked that thickening studies indicate the possibility
that the solids content of material going to the filter might be as high
as 25%, in which case higher filtration rates may be expected. It is
apparent from the data that there is considerable day-to-day variation in
the filtration rates. The reason for this variation was not known; per-
haps it is related to the amount of fiber fines in the white water. Be-
cause of this variability, and because experimental data may never reveal
the full range of possible conditions, considerable judgment must be ex-
ercised in sizing a filter for this application.
In Table 11 are also presented moisture data on the filter cake after
filtration. An average of 48% solids is noted. When lime mud alone is
filtered by the same technique, the filter cake is at about 60% solids.
It is felt that in practice, the solids content of the mud/sludge cake
will be about 50 - 52%, which is slightly higher than the laboratory
data indicate, but still considerably lower than normal lime mud.
Carbonation and Carbonate Settling. Recovery of soluble calcium after
the initial clarification next received some study. The flow scheme in-
23
-------
TABLE 11
AVERAGE FILTRATION RATES
Mud/Sludge Mix Filtration Rate, Filter Cake,
Day % Solids (lb.)/(hr.)(sg. ft.) % Solids
10/19 --— 31 48
10/21 32 45
10/14 14.0 33
10/22 13.0 49 45
10/16 54 52
10/20 15.3 57 50
11/10 58 50
11/07 15.4 59-1/2 47
11/09 64 49
10/23 66 46
11/06 10.4 67-1/2 49
11/09 12.2 71 47
11/08 14.5 72 47
11/06 19.1 103 48
10/18 106 45
10/17 123 50
Average 48%
Formation time - 30 seconds
volves reacting the soluble calcium hydroxide with the carbon dioxide
from lime kiln stack gas; the resulting calcium carbonate is then re-
moved by sedimentation, dewatered, and calcined. The most important
questions recognized at this point were:
1. Design of gas-liquid transfer system to effect reaction of
the carbon dioxide.
2. Determination of the restraints establishing the extent of
reaction to provide optimum recovery Of calcium.
3. Mechanical considerations leading to most rapid and complete
sedimentation of the calcium carbonate precipitate.
The first question seemed reasonable amenable to established chemical
engineering knowledge and methods. Calcium carbonate scale formation
and foam generation were recognized as problems.
The pH of the solution was taken as the readily measurable parameter by
which to measure progress of the reactions:
Ca(OH) + C02 ' > CaCQ3 + H20
CaCOs + H20 + C02 » CaH2 (C03)2(Calcium bicarbonate)
24
-------
The pilot plant carbonator was run at different carbon dioxide flow
rates to vary the pH of the carbonated effluent leaving. The carbon-
ated effluent was centrifuged and an analysis made for dissolved cal-
cium. The results are presented in Figure 7. They show a minimum
dissolved calcium content at about pH 9.6. At lower pH levels, the
dissolved calcium increases because of bicarbonate formation. Gener-
ally, low levels of dissolved calcium were observed over the pH range
8.5 - 10.5.
These pH values are notably lower than those reported for the "massive
lime" study, (9) where a minimum solubility was indicated above pH 11.
It is probable the difference is due to the fact the NCSI study dealt
with bleach extraction wastes high in sodium alkali. In the presence
of dissolved lime, this alkali is virtually all NaOH. Carbonation
yields carbonate, and then bicarbonate -- both soluble. The carbonate
from this source enters into the solubility product which determines
solubility of calcium carbonate. It is to be expected, therefore, that
the pH for minimum CaC03 solubility will vary with soda alkalinity of the
starting effluent.
It was noted in the NCSI research that there seemed to be an optimum
pH at which the calcium carbonate flocculated to provide good settling.
Numerous observations were made in the present study, both of jar test
and pilot plant runs; none led to a reproducible flocculating condition.
There were scattered, transient evidences that raised hopes briefly,
however.
The pilot plant system was operated to seek information on recovery
at the post-carbonation clarification step. The Hodge pilot plant
clarifier had severe limitations for this purpose. There was almost
no opportunity for floe formation, and mean retention time was less
than 20 minutes. The total up-flow path was only about two feet, and ,
the calculated rise rate was about 1.0 g.p.m. per square foot.
The pilot plant was operated using a lime slurry pumping rate a little
above 1,000 mg/1 CaO while carbonation was carried out at three differ-
ent pH levels. The amount of calcium entering and leaving the clarifier
was measured, so that the clarifier efficiency could be measured. The
results are shown in Table 12. About 70% of the calcium entering the
clarifier was recovered; the best result was obtained at pH of 8.5, but
the spread of results was small enough to raise a question whether they
validly indicate a significantly preferable pH.
Considering the crude design of the pilot plant clarifier, this effi-
ciency didn't seem bad. Since a similar process in municipal softening
systems is operated routinely without difficulty, it was felt that 90%
recovery in a properly designed, full-scale unit was not an unrealistic
expectation. Because more than half the feed calcium would be recovered
in the first clarifier, the indicated overall recovery would amount to
about 95%.
25
-------
1GO
O
no
100
I 80
o
iij
a
60
40
20
11 10 9 8
CARBONATION pH
EFFECT OF CARBONATION pH ON DISSOLVED CALCIUM
Figure 7
26
-------
TABLE 12
EFFICIENCY OF PILOT PLANT CARBONATE CLARIFIER
Carbonation Efficiency
PH %
TO 66.6
9.5 67.6
8.5 70.6
The calcium concentration to clarifier = 554 mg/1 as CaO.
Treatment Suitability of Various Wastes. Samples from various branch
sewers of the mill effluent system were collected to evaluate their
relative color contributions and to provide a basis for determining
which should not be (or need not be) treated by the proposed system.
It was concluded that effluent from bag plant, power plant, water treat-
ment and causticizing areas need not be treated. Tall oil plant sewer
waste was judged undesirable because of the occasional release of a
batch of sodium sulfate brine. At high sulfate concentration, calcium
sulfate will be precipitated, reducing the calcium hydroxide available
for color precipitation, and contaminating the kiln product when the
sludge is calcined. If a holding pond were available to distribute the
impact over a suitable dilution of other effluent, a different choice
might be made.
During examination of the various component subdivisions of sewer flow,
it was noted that no tests were recorded in which the raw waste was
primarily from a waste liquor source not modified by papermaking pro-
cesses. Samples from the sewer of the digester and washer area were
treated at two lime dosage levels. The results are shown in Table 13.
They show that, on these generally dark samples, the treatment was able
to remove in excess of 80% of the color, even at relatively low levels
of addition.
TABLE 13
._
FROM PULPING AREA EFFLUENT
Color Removal Efficiency. %
Effl. Color, Lime AddedLime Added
Day NSSC* APHA Units 760 mg/1 1,520 mg/1
1T708 "Yes" 27BT3857886.7
11/10 No 1,500 66.7 91.7
11/11 Yes 3,500 71.4 78.6
11/12 No 1,950 76.4
11/13 No r- 3,190 81.8 91.9
*0n "Yes" days, about 3Q% of pulp produced was NSSC (remainder, kraft).
27
-------
Centrifuge Dewatering. Although the accumulated data indicated that an
acceptable dewatering of color sludge could be accomplished by mixing
with causticizing mud washer underflow and filtering with a belt filter,
two disadvantages were recognized:
1. Soda removal from causticizing mud would be hindered,
resulting in higher soda input to the kiln.
2. The belt filter requires closer attention than con-
ventional, precoat-type mud filters.
Solid-bowl centrifuge dewatering of combined sludges had be,en considered
earlier as an alternative to belt filtration, but was deemed unfeasible
because of a tendency to leave lighter portions of the solids (e.g., pre-
cipitated color bodies) in the liquid discharge. Separate centrifuging
of the first clarifier sludge had not received much serious consideration
because of the low solids levels obtained in laboratory trials.
A fresh examination of process material balances, as well as the promise
of better adaptability to remote operation control, led to further con-
sideration of this technique. It became apparent that, since caustici-
zing mud is commonly discharged at 65% solids from a conventional precoat
filter, the separate concentration of first clarifier sludge to slightly
over 20% solids would be comparable to the belt filter dewatering of com-
bined sludges to 50% solids. Figure 8 shows the effect of increased
sludge dewatering. Pilot-plant samples of sludge were prepared for test-
ing by centrifuge manufacturers.
Two 5-gallon samples were tested by one manufacturer in specialized test
apparatus with discouraging results. Their data indicated an assured
solids concentration of only 15%, with appreciable solids carry-over in
the centrate liquid.
A 55-gallon sample sent to another company was processed through a very
small commercial type solid-bowl continuous centrifuge with quite promis-
ing results. Discharge cake solids contents of 20 - 25% were achieved;
solids removal, as high as 98% was found feasible by use of a flocculant.
The data from this series of trial runs appear in Tables 14 and 15. Ar-
rangements were made to rent a similar centrifuge both to conduct further
tests and to prepare sludge for other experimental purposes.
A rented centrifuge was delivered to the mill, where pilot lots of sludge
had been prepared for processing. The Sharpies Model 600 centrifuge has
a normal volumetric throughput maximum of about 6 gallons per minute.
Standard rotational speed is 5,000 r.p.m., representing a relative cen-
trifugal force of about 2,100 times gravity. Operating variables include:
. 1. Pond depth.
i
2. Speed differential of internal conveyor screw.
28
-------
•££0
z
y
Of
111
I
tn
O
•400
350
CAUSTICIZING FILTER
@ 68 % SOLIDS
COLOR SLUDGE CAKE
CAKE
AS SHOWN
COM PAR 150 N-
COMBINED SLUDGES
BELT HLTfR @ <5O %
18 ZO 22 24
SOLIDS, COLOR SLUDGE CAKE
EFFECT OF COLOR SLUDGE DEWATERING
Figure 8
29
-------
3. Feed slurry solids concentration.
4. Feed slurry flow rate.
5. Additives to aid flocculation and drainage.
The Model 600 is a small unit -- a choice made necessary by the diffi-
culty in preparing large quantities of representative sludge. The inlet
ports are small, and a small feed pump is required. As a result of these
dimensions, (together with the fibrous nature of the sludge) high sludge
concentrations were difficult to pump, and flow rate was difficult to
adjust.
TABLE 14
Test No.
CENTRIFUGE TRIALS (FACTORY)
1
8
Feed:
Code
Rate, #/Hr.
% Insols.
Insols, #/Hr.
Centrate:
Rate, #/Hr.
% Insols.
Insols, #/Hr.
Cake:
Rate, #/Hr.
% Water
Water, #/Hr.
% of Solids
Unsedimented
Flocculant:
Sol 'n, gph
Dose, #/T, s/s
RCF x g.
Pond
Convr. diff.
A A
322 736
A A
1315 1411
A
2400
B B B
432 1171 2215
3.25 3.25 3.25 3.25 3.25 2.93 2.93 2.93
10.4 23.9 42.6 45.9 78.0 12.7 34.3 65.0
240 618 1120 1210 2160 390 1065 2040
.005 .027 .157 .189 .695 .088 .387 .600
0.01 0.17 1.80 2.30 15.0 0.34 4.10 12.3
82 118 195 201 240 42 106 175
87.2 80.0 79.1 78.2 73.8 70.5 71.5 69.7
71.6 94.3 154 157 177 29.6 75.8 122
0.10 0.71 4.20 5.00 19.2 2.70 12.0 19.0
zero
2100
3
50rpm
20rpm
30
-------
TABLE 15
Test No.
Feed:
Code
Rate, #/Hr.
% Insols.
Insols, #/Hr.
Centrate:
Rate, |/Hr.
% Insols.
Insols, f/Hr.
Cake:
Rate, #/Hr.
% Water
Water, #/Hr.
% of Solids
Unsedimented
Flocculant:
Sol'n, gph
Dose, f/T, s/s
RCF x g
Pond
Convr. d-iff., rpm
CENTRIFUGE TRIALS (FACTORY)
9 10 11 12
1347 1347
.037 .040
0.50 0.54
15.0
6.4
2100'
4
20
1336 1345
0.040 .177
0.54 2.4
183 183 194 185
79.0 79.5 80.2 80.2
145 145 156 148
1.3 1.4 1.4 6.2
7.50 3.75 1.50
3.2 1.6 0.6
13
2770
.146
4.1
290
74.5
216
5.3
7.50
1.6
14
C
1530
2.55
39.0
C
1530
2.55
39.0
C
1530
2.55
39.0
C
1530
2.55
39.0
C
3060
2.55
78.0
C
3060
2.55
78.0
2754
.088
2.4
306
75.3
230
3.1
75.0
3.2
As shown by Tables 16, 17, and 18, adjustment to comparatively low pond
depth and differential speed resulted in cake densities in the range of
22 - 25%. Excellent recovery percentages were obtained with the use of
a polymeric flocculant, and fairly satisfactory ones were achieved with-
out such additives. It was considered likely that part-time use of a
flocculant would suffice to prevent excessive buildup of "fines" in the
system.
The potential advantage of separate sludge dewatering to such levels, as
compared to the belt filter, was estimated by the following calculations:
Basis of calculations:
Causticizing mud solids
Color clarifier solids
Attainable % solids, conventional
mud filtration
367 tons/day
63 tons/day
65 - 70%
31
-------
TABLE 16
co
ro
Run No.
Centri f uge :
Diff. , rpm
Pond
Torque
Effl. gpm
Solids, %:
Feed
Cake
Effl uent
Recov. , %
Flocculant,
*/T, D.S.
1 2
51 51
3 3
1.0 1.5
0.80 3.32
2.55 2.47
24.0 23.1
0.15 0.49
94.1 80.0
none ^
CENTRIFUGE TRIALS (MILL LAB)
3456
51 51 51 51
3 3.5 3.5 3.5
1.2 0.5
2.1.4 0.50 0.84 2.14
4.00 2.77 2.84 2.98
22.9 26.3 17.3 18.5
0.74 .048 0.19 0.39
81.4 98.4 93.4 86.0
7 8
51 51
3,5 3.5
1.0
3.36 3.80
2.87 2.84
19.9 21.4
0.52 0.52
84.0 83.6
9 10
51 51
3.5 4
1.2
1.14 0.83
3.14 3.23
20.4 6.53
0.26 0.27
93.0 56.1
11
51
4
2.00
3.24
6.94
0.59
89.3
RCF x g
2100
-------
TABLE 17
CENTRIFUGE
14
51
3.5
1.2
15
51
3.5
16
51
3.5
2.0
TRIALS
17
51
3.5
1.2
(MILL LAB)
18 19 20 21 22 23
51 51 51 51 51 51
3.5 3.5 3 3 3 3
Run No. 12 13
Centri fuge:
Diff., rpm 51 51
Pond 4 3.5
Torque 1.5
" Effl. gpm 2.2 0.89 1.33 2.50 4.35 0.87 4.00 2.40 0.75 2.14 3.16 4.92
Solids, %:
Feed 3.24 5.42 5.52 5.52 5.52 6.96 7.17 7.17 5.13 5.13 5.24 5.22
Cake 6.92 16.9 17.9 19.9 22.5 14.0 19.4 19.1 18.7 16.3 20.2 19.5
Effluent 0.57 0.46 0.61 1.31 1.48 2.18 2.00 2.00 0.26 0.91 0.95 1.35
Recov., % 89.7 94.2 92.2 81.7 78.5 73.1 0.5 80.5 96.4 86.8 86.7 80.0
Flocculant,
#/T, D.S. none
RCF x g 2100
-------
TABLE 18
CO
Run No.
Centrifuge:
Diff., rpm
Pond
Torque
Ef f 1 . gpm
Solids, 55:
Feed
Cake
Effluent
Recov., %
Flocculant,
#/T, D.S.
RCF x g
CENTRIFUGE TRIALS (MILL LAB)
24 25 26 27 28 29 30
30.6 30.6 30.6 30.6 30.6 30.6 30.6
3333333
0.77 1.30 4.00 1.59 0.72 0.86 1.46
'
4.73 4.82 4.68 4.66 4.75 4.80 4.88
20.5 20.0 25.0 24.1 22.3 25.4 25.2
0.56 0.31 0.86 0.32 .014 .029 .020
98.9 95.2 84.6 93.8 99.8 99.6 99.7
none > 0.97 1.70 0.97
2100 L
31 32 33 34
30.6 30.6 30.6 30.6
3333
2.00 4.00 2.00 1.50
-
5.0 4.61 4.68 5.02
22.2 23.5 22.3 25.6
.047 .497 .017 .018
99.4 91.2 99.5 99.9
0.82 0.73 1.44 1.81
-------
Attainable % solids, combined mud
and sludge, belt filter 50%
Attainable % solids, color sludge,
centrifuge 22 - 25%
Water Content:
430 tons combined sludge, 50% solids 430 tons water
(a) 367 tons caust. mud, 65% solids 197.6
(b) 367 tons caust. mud, 70% solids 157.3
(c) 63 tons color sludge, 22% solids 223.4
)
(d) 63 tons color sludge, 25% solids 189.0
(a) plus (c) 421.0
(a) plus (d) 386.6
(b) plus (c) 380.7
(b) plus (d) 346.3
It can be seen that reduction in water to be evaporated in the lime kiln
may range from 18,000 to 167^000 pounds daily.
It was felt that the belt filter would be much more difficult to adapt
to remote control and minimum manpower requirements. Moreover, main-
tenance considerations appeared to favor the centrifuge over the belt
filter. A flow sheet incorporating the centrifuge alternate is shown in
Figure 9.
Sludge Calcining. During consultations with rotary kiln manufacturers
concerning special design and operational considerations for sludge
calcining, a warning was voiced. Engineers for one manufacturer noted
that small increases in free lime content of lime kiln feed have caused
serious problems of ball and ring formation in a number of kilns.
The question of whether the primary sludge of this process contains appre-
ciable Ca(OH)2 had received much attention. Solubility considerations
would seem to cast doubt upon such occurrence, because the surrounding
liquid is not saturated with calcium hydroxide. However, the sludge will
provide hydroxyl ions to causticize a solution of sodium carbonate, and
(as has been previously noted) the calcium content of the sludge does not
follow a definite stoichiometric relationship to identifiable calcium-
precipitating anions.
Discussions of the problem with engineers of two kiln manufacturers raised
the following questions, among others:
!
1. Will the associated organic materials, co-precipitated with
the calcium and hydroxyl ions, alter the effect, compared
to calcium hydroxide?
2. Will the fiber content serve to reduce the balling and
ringing tendency?
35
-------
CARBON DIOXIDE
JU "
KJLN
MILL EFFLUENT
CO
FIRST
CLARIFIER
SECOND
CLARIFIER
CARBONATOR
TO KILN «*
CAU5TICIZ./NG
MUD
PH
ADJUSTMENT
DECOLORIZEP
EFFL^EfVT TO
OXIDATION PONDS
FLOW SHEET WITH CENTRIFUGE
Figure 9
-------
3. Are there operational correctives which would allow
successful calcining in the presence of free lime?
The first two questions attracted no firm opinions, and'the third elicited
disagreement among the experts.
Laboratory calcining had established the satisfactory chemical composi-
tion of lime made from the process sludge, but such procedures offered
little guidance to physical behavior in a rotary kiln. Further confer-
ences with a kiln maker dealt with the possibility of pilot-scale calcin-
ing as a means of resolving the questions. A pilot-plant kiln, I1 dia-
meter by 10' long, was available which would require some 50 pounds per
hour of feed (recausticizing sludge with proportional color-removal
sludge). It was felt provision should be made to run up to 20 hours.
With our small-scale facilities, and the problems of thickening, the
accumulation of sufficient color-removal sludge was tedious, but the
availability of the small, solid-bowl continuous centrifuge made it
possible to concentrate it properly.
A quantity of the centrifuged sludge containing about 200 pounds of solids
was mixed with causticizing sludge filter cake containing about 1,500
pounds of solids. This mixture was subsequently calcined, with encour-
aging results. This operation was conducted and observed by a group ex-
perienced in interpretation of miniature kiln performance. Their report
was that a slight tendency toward ring formation was evident, but that
the rings were rather fragile. Their opinion was that, in a comparatively
large kiln, stable and troublesome rings were not likely.
Meanwhile, alternative flow sheet modifications were developed to elimin-
ate the free lime before calcining, if such a course should become neces-
sary. The first clarifier sludge would be carbonated before dewatering.
In the case of centrifuge dewatering, the centrate might be used for lime
mud washing if color release presented barriers to recycling to the clari-
fier.
Color Changes after Lime Treatment. As the project was approaching the
point of major construction commitment, there were several warnings of
"color reversion" or "color pick-up" possibilities. The warnings were
investigated, and several experiments were conducted. The following
conclusions were reached:
1. Colored substances are removed and incinerated by the
process; this color cannot return.
2. Due to soda alkalinity, color can be extracted by effluent
from woody and other plant materials in lagoons or streams.
3. Biological processes may increase or decrease color of
effluent after lime treatment. Some quite varied obser-
vations were made, inviting further study for which time
was not immediately available.
37
-------
4. Although the effects noted in (2) and (3) might
affect ultimate color effects in receiving waters,
they do not negate the effects of the treatment.
Published Information. Work done in the bench and pilot phases of the
project has been the subject of a published paper. (18) The theoretical
considerations upon which the project conception was based has received
support in studies directed by Dence and Luner. (19)
Plant Design
Before entering into specifics of unit designs, several basic decisions
were necessary to establish the working flow sheet. It was determined
that:
1. The sludge produced in the color precipitation step would be
dewatered by a continuous, solid-bowl centrifuge. The oper-
ational and economic considerations bearing upon this choice
have been noted previously in the discussion of the centrifu-
gation studies.
2. A separate reaction vessel would not be used for color precipi-
tation. Instead, lime would be added in the feed pipe to the
clarifier center flocculation well, which would be designed for
maximum reaction time.
3. A sewer conduit from the mill would deliver waste water to a
lift station which would pump to a first clarifier which would
be at such a level that gravity flow would suffice for further
transportation of the main waste stream. There had been earlier
hopes that gravity flow could be maintained throughout, but the
only feasible sites presented unacceptable sewer problems.
4. To provide for a then-current need for an average daily capacity
of 12 million gallons and a 50% reserve capability for peak rates
and for possible changes, the principal system units would be
designed for a maximum flow of 12,500 gallons per minute.
5. Provision would be made for removal of trash and grit from the
waste water before adding lime. Previous experience with primary
clarification of effluent had pointed out the frequent appearance
of such foreign materials as sticks, rags, and pieces of string,
rope, and gasketing. These had caused serious problems in a simple
clarifier system; the complex system now proposed would be far more
vulnerable to debris. Moreover, the centrifuge would present a
special sensitivity to gritty, abrasive materials.
6. Use of lime, whether freshly purchased or re-calcined, would be
common to causticizing and waste treatment, without separation
or distinction.
*
»
38
-------
7. Calcium carbonate sludge, obtained by carbonation of the
first clarifier effluent, would be mixed and filtered with
the recausticizing "lime mud", using conventional kraft
mill equipment.
8. Carbonation for lime recovery would be accomplished by use
of lime kiln stack gas after "scrubbing" for dust removal.
Upon reviewing these decisions, along with the laboratory and pilot
plant work, the flow sheet was drawn in the form shown in Figure 10.
As a basis for preparing detailed cost estimates and for soliciting
bids, equipment needs were defined and specified. The following sum-
mary deals with equipment in process flow sequence, rather than in the
chronological order of actual consideration.
Grit and Trash Removal. Earliest consideration of the removal of trou-
blesome solids from the raw waste had envisioned a traveling bar screen
followed by a small basin for collection of grit and other small, dense
solids. Further consideration of specific materials found in the efflu-
ent soon revealed need for more complex design. It could be seen that
a bar screen would not protect adequately against strings, bits of rope,
or slender pieces of wood, all of which would then pass through pumps,
control valves, and the entrance ports of a centrifuge. Although expected
grit volume was not large, it was apparent that other materials, such as
woody rejects from cleaners and screens, and water-logged bark would also
settle and would comprise sufficient volume (at least occasionally) to
demand a well-designed removal and disposal system. When a centrifuge
of very high centrifugal force was chosen for sludge processing, it was
noted that grit content of the sludge might be a prime factor in mainten-
ance costs; obviously, effective grit removal assumed added importance.
After consideration of several devices, our choice finally settled upon
a revolving disc screen. Specifications were prepared for a unit 14 feet
in diameter with screen plates of Type 304 stainless steel, with 1/2 inch
diameter holes. A solids-collection drag conveyor was required to gather,
dewater, and deliver solids to a container, and a spray system was pro-
vided to flush debris from the screen onto the conveyor.
The grit settling capacity was calculated upon a minimum capability, at
peak flow, to remove all grit above about 65 mesh. A chamber 10 feet
wide (bottom sloped to 5-foot recessed channel) and 60 feet in effective
length was provided, with a normal water depth of 7 feet 6 inches. Con-
struction details include an overflow weir in the grit channel and a dis-
charge weir after the disc screen, so proportioned that differential level
across the disc screen cannot exceed about 18 inches. (Because of the
large screen radius, drastic forces might otherwise develop in case of
screen "blinding.")
The grit collector is designed to rake a bottom channel 5 feet wide,
with buckets approximately 8 feet apart, and at a speed not exceeding
8 feet per minute. The mechanism is specified to be capable of pulling
39
-------
MILL
EFFLUENT
COLOR
CLARIFIER
I-
CO2
GRIT
CHAMBER
CARBONATiOH
BASIN
COLOR SLUDGE
CENTRIFUGE
COLOR. SLUDGE
STORAGE TANK
LIME N\UD
CARBONATION
CLARIFIER
CAUST\C MOD
STORAGE TANK
VACUUM
FILTER.
I
A
OUTLET
BASIN
TO K»LN
FLOW SHEET FOR PLANT DESIGN
Figure 10
-------
through a 9 Inch grit accumulation at the bottom of the grit channel,
dewatering and elevating the sediment to discharge into a container.
To minimize wear, neither the screening system nor the grit removal mech-
anism is operated constantly. Operation of the screen and its discharge
conveyor are initiated whenever head loss across the screen exceeds 2
or 3 inches, and the ensuing timer-controlled cycle is set for a minimum
of 3 minutes duration. The grit conveyor operates under an adjustable
timer control, commonly set for 10 minutes running time each 60 - 120
minutes.
Lift Station. To collect accepted effluent from the disc screen, a pump-
ing pit 18 feet in diameter by 17 feet deep was designed. Construction
consisted of a circle of sheet piling, which was excavated deep enough
for a ballasting bottom of mass concrete.
Pump specifications provide three pumps, with 16 feet tube length, each
with capacity for 9,000 GPM at 30 feet head and 6,500 GPM at 45 feet head.
These are intended to accommodate peak flow with two units running, thus
allowing one as a spare in case of failure or routine maintenance of any
unit.
The pumps, through check valves, discharge into a manifold connecting to
a 24-inch pipe. It is in this pipe that lime slurry is mixed with the
raw waste.
Lime Slurry System. Lime for the treatment process is withdrawn from a
"day bin" which is shared with the kraft causticizing system. A screw
conveyor with manually-controlled, variable-speed drive regulates lime
flow into a fixed-speed transfer conveyor (also screw type) which delivers
into the slaker inlet. The variable-speed conveyor has a screw 14" in
diameter with a speed range of 1 - 3 RPM. The transfer conveyor screw is
also 14 inches in diameter.
The lime slaker is a. renovated unit originally used for kraft liquor
causticizing-. It has a reaction bowl 9 feet in diameter by 7 feet deep,
supplied with a 15 HP paddle-blade mixer. The grit classifier is of the
oscillating rake type, with a 2 HP motor.
The slaker overflows into a slurry tank 12 feet in diameter by 10 feet
deep. The slurry is kept tn suspension by a 7.5 HP vertical mixer.
Two slurry pumps (one a spare) are provided to deliver the lime suspen-
sion from the slurry tank to the raw waste lift pump discharge. The
pumps are specified as rubber-lined (replaceable liners) for resistance
to abrasion. The required performance capability is 180 GPM at 55 feet
head, with a maximum pump speed of 1,200 RPM. The pumps discharge through
a pipe about 1,100 feet long, and no flow throttling is provided. Water
is supplied to the slurry tank to maintain a constant level; thus, the
pump delivers full discharge pressure, maintaining a velocity which should
41
-------
preclude deposition on the pipe walls. An automatic flushing (water)
cycle is initiated whenever the pump is de-energized.
Waste Clarifiers. Since design criteria for both clarifiers were based
on maximum rise rates of 1.0 GPM per square foot, it was possible to de-
sign them for identical internal mechanism. (In addition to maintenance
advantages, it would be possible to assure one primary clarifier in case
of failure of the other unit.)
The mechanisms were specified for a clarifier 135 feet in diameter, having
15 foot side water depth. A flocculation zone equivalent to a center well
of 40 feet diameter was required. Thickening-type mechanisms capable of
developing 7% (w/w) minimum (first clarifier) sludge consistency, based
on pilot plant solids quality data, were specified. Since shear proper-
ties and maximum consistencies,of both sludges represented first-of-a-kind
design problems, it was determined that torque capabilities should be the
highest competitively available in standard manufacture. This proved to
be a continuous rating of 1,200,000 ft-lbs., with peak load design of
1,800,000 ft-lbs. Two long (full clarifier radius) and two short arms
(25% minimum radius) were designed for sludge raking. Sludge is raked
into an annular central well, within which full raking was required. A
torque indicator, with transmitter for continuous, remote recording of
rake torque, was included in the specification. Side or top entry of
raw waste feed was specified; the final choice was for over-the-top entry,
with feed pipe suspended below an access walkway bridge.
The clarifier basin designs were for concrete construction. Bottom slope
to the center was one inch per foot. Clear liquid discharge at the upper
periphery is through 66 equally-spaced, submerged orifices 6 inches in
diameter. An integrally-cast, external, concrete collection trough, 3
feet wide and 6 feet deep, delivers the effluent to a discharge box.
Sludge withdrawal is accomplished through two stainless steel pipes cast
in concrete beneath the clarifier floors. Because of the clarifier ele-
vation, the color clarifier discharge pipes (8 inches nominal diameter)
are essentially straight, and the ground-level pumps are only slightly
above the elevation of the inner slope of the clarifier floor. The car-
bonation clarifier is slightly lower; the sludge pipes (6 inches diameter)
lie parallel to the clarifier bottom and have slight bends up to pumps at
ground level (clarifier water level is 12 feet above pump suctions).
Carbonation System. To provide for proper contacting of color clarifier
effluent with kiln stack gas (source of carbon dioxide), a tank 30 feet
in diameter by 12 feet deep was designed. Water from the color precipi-
tation clarifier enters just below the water level of the carbonation
tank and flows out near the bottom.
Proposals were considered for one, three, and four agitators. The final
choice was for four 40 HP units with turbine-type impellers driven at
85 RPM. Considerations leading to this selection included; a more desir-
able subdivision of gas entrance points, ability to withstand temporary
42
-------
outage of a unit, and avoidance of sidewall baffles which would affect
static pressures at entrance and exit openings.
It was considered probable that impellers of these agitators would accu-
mulate deposits of calcium carbonate, and various expedients were consi-
dered for protection. Highly-polished surfaces, various alloys, elasto-
meric coatings, and repellent materials such as the fluorocarbon polymers,
were among the proposals. None was found to give enough promise to justify
the cost, and, since there seemed to be no serious corrosion problems,
ordinary steel was specified.
Pilot plant experience and reports of related experience elsewhere had
indicated that perforated-pipe distribution of carbonation gas would re-
sult in closure of the pipe openings by scale formation. It was also
felt that gas pipes entering through the tank bottom would be subject to
occasional back-up of water which might tend to be trapped, increasing
back-pressure. (Required gas pressure is a substantial factor in cost
of supplying carbonation gas.) The final design of the gas diffuser is
sketched in Figure 11. Each diffuser discharges two 36-inch curtains of
gas bubbles beneath one of the agitators. Stainless steel supply pipes,
6 inches in diameter, enter from the top of the tank. The diffusers con-
sist of half-cylinders of 6 inch radius, 36 inches long, with 1-inch, 60°
serrations on the long edges.
Outlet Basin. The carbonation clarifier discharges into a basin about 21
feet by 27 feet, and 12 feet deep. More kiln gas is added to adjust pH.
Horizontal, perforated pipes are used for gas diffusion; as pH drops
below 9.0, calcium carbonate solubility increases, so scaling should not
be a serious problem. The basin also receives any overflow from the color
clarifier or carbonator. From the basin, water can be discharged over
adjustable weirS to either the secondary treatment system or to a holding
pond. Provision is made for vertical pumps by means of which the "de-
colorized" water can be returned to the mill for selective re-use.
Sludge Pumps. To withdraw sludge from the color clarifier, specifica-
tions were based upon a sludge of 7% consistency in which fiber comprised
20% of the total solids weight (dry basis). Stated pump requirement was
for a horizontal centrifugal pump capable of handling a 10% solids content,
delivering 300 GPM at 170 feet total discharge head, with a maximum pump
speed of 1,200 RPM.
For carbonation clarifier sludge, a sludge consistency of 10%, comprised
essentially of calcium carbonate, was assumed. A capacity of 100 GPM
at 120 feet total discharge head was required, with a maximum pump speed
of 1,200 RPM.
Changes were made in both pumps after start-up, as will be explained
later. The changes were dictated by the composition and properties of
the sludges.
Color Sludge Storage. To receive sludge from the color clarifier, a 30
43
-------
GAS DIFFUSER
Figure 11
-------
foot diameter tank 20 feet deep was provided. The tank was identical to
one specified for causticizing lime mud, and similar 15 HP agitation mech-
anism was provided. (Supplementary agitation was added later, as will be
noted.) This tank provides surge capacity between the clarifier and the
centrifuge.
Color Mud Pump. To deliver color sludge to the centrifuge, a rubber-lined
centrifugal pump, similar to those for lime slurry, was specified. The re-
quired rate was 185 GPM at 80 feet discharge head.
Color Sludge Centrifuge. Considerations in centrifuge requirements have
been discussed previously, especially as relates to dewatering capability.
Specifications were based upon observations which indicated that optimum
input consistency of sludge might be about 8%. Stated solids capacity was
70 tons per day. Automatic protective devices, to protect against torque
overload and other probable sources of equipment hazard, were specified.
As a consequence of the required degree of dewatering, the only supplier
offering to meet the specifications was one who offered a high "G-force"
unit. Relative centrifugal force was 1,880, at a bowl speed of 2,300 RPM.
The backdrive spindle of the planetary gear system rotating the internal
conveyor was provided with an eddy-current electric brake which permits
variation of conveyor differential speed by a ratio of two to one (approx-
imately 25 to 50 RPM). The protective alarm system includes a signal to
discontinue feed and/or provide flushing at a preset level of conveyor
torque, and trip-out for low oil pressure, low oil flow (either bearing),
high bearing temperature (either bearing), excessive torque, low brake-
cooling water pressure, or high brake-cooling water outlet temperature.
A fault detector display panel indicates trip-out cause.
The centrifuge solids are delivered by a screw conveyor into the lime
kiln feed screw hopper, along with discharge cake from the kraft causti-
cizing mud filter. Filtrate flows to the lift pump station to be repro-
cessed.
Carbonation Sludge Recovery. Piping was arranged to deliver the recovered
calcium carbonate underflow from the carbonation clarifier to the causti-
cizing "mud" storage tank. This would permit the combined calcium carbon-
ate sludges to be dewatered by the lime mud filter, which was generously
sized to provide the needed capacity to prepare the materials for calcin-
ing.
The Lime Kiln. Design of the lime kiln was based upon the expected re-
quirements of the color removal system, combined with those for liquor
recausticizing for about 600 tons of kraft pulp production. Kiln speci-
fications required a capacity of 235 tons of lime per day at 85% avail-
able calcium oxide. Conditions for this performance included 45.2%
sludge solids, and a total dry solids load, including fiber and recir-
culated dust load, of 422 tons per day. The indicated moisture load was
well above the expected level, to insure adequate chain section provision.
The kiln size was set at 290 feet by 12 feet inside shell diameter, with
45
-------
6 inch refractory lining. The successful bidder provided a 58 foot chain
system, weighing 130,700 pounds. Included was a venturi scrubber for
stack gas, rated at 99% dust recovery, and an induced draft fan designed
to provide a 25 inch (water) pressure, while handling 101,500 ACFM at
400bF.
Kiln Gas Blower. To supply lime kiln stack gas to the carbonation system,
specifications were drawn for a compressor having a capacity of 4,000 ACFM
at 160°F, at a discharge pressure of 10 psig. =Gas is received from the
discharge of the venturi scrubber system at a pressure of 1.0 atmosphere.
Since scrubber specifications indicated a CaC03 content of only 0.08%
by weight, bids were offered by manufacturers of rotary lobe-type compres-
sors, who represented their units as suitable. These units were priced
lower than the "water-piston" type compressors, which the development en-
gineers had considered most suitable for the service.
Attention is called to subsequent findings concerning both the equipment
performance and the required gas volume.
Instrumentation. The needs for process instrumentation are largely in-
fluenced by three important considerations. First, the system is an added
responsibility of the operator of another process, and the demands upon
his time should be limited. Second, much of the equipment is relatively
remote from the normal location of operating people. Third, the system
is ponderous and slow with respect to most functions and responses. The
last factor mitigates considerably the sophistication which might other-
wise be necessitated by the first two.
Control of the grit conveyor and disc screen system by a timer and differ-
ential level sensor, respectively, has already been explained.
A level recorder in the control room, together with a "high level" annun-
ciator light, inform the operator concerning lift station performance.
The same level measurement is used to actuate a pump discharge valve so
as to maintain a constant level in the pit, maintaining pace with the
amount of effluent arriving.
Lime feed to the slaker is controlled by a manually-set speed control
governing the feed conveyor; the speed is monitored by an indicating
tachometer on the control panel. Water supply to the slaker is regulated
by an indicating flow controller. Recording thermometers are provided for
water input and slurry overflow from the slakerj temperature rise .is a
measure of calcium hydroxide concentration of the product slurry. (Con-
version of one pound of CaO to Ca(OH)2 releases 486 Btu. Thus, production
of a 10% calcium hydroxide slurry is accompanied by a temperature rise of
about 380F.) (20)
?
The lime slurry tank is provided with a level indicator-controller. The
controller actuates a water valve to maintain a predetermined level in
the tank (usually about 6 feet). Low water level will activate a relay -
46
-------
to shut off the slurry pump. (The slurry pumps are also interlocked to
prevent function if all lift pumps are inactivated.) Upon trip-out of
the slurry pump, a switch and timer provide flush-out of the slurry pip-
ing.
Torque exerted upon the rake of the color clarifier is recorded at the
control panel. A separate limit switch is connected to an annuciator
light to signal "high torque." Identical signals are provided at the
carbonation clarifier, and torque for both units is recorded on the same
2-pen instrument.
At the carbonation tank, pH measurement is indicated locally and trans-
mitted to a recorder-controller at the control panel. A signal from the
controller actuates the control valve admitting kiln gas to the carbon-
ation system. Both pH and valve position ("% open") are recorded on the
2-pen recorder.
Constant pressure in the kiln gas supply line, as well as protection
(pressure relief) of the compressor, are provided by a pressure controller
and vent valve. A control panel station permits adjustment of pressure
and observation of valve position.
Control of pH at the outlet basin is facilitated by pH instrumentation
substantially identical to that for the carbonation vessel.
Control and recording of sludge flow from each clarifier to storage is
provided. In each system, a magnetic flow meter signal is recorded, and
control action is supplied to a valve designed for precise flow control.
Because of the plugging hazard inherent in the sludge properties, a paral-
lel valve circuit with full pipe diameter, pneumatically-actuated plug
valve is installed; this unit can be positioned by a manual control on
the main panel.
The storage tank for color clarifier sludge is equipped with level mea-
surement. This measurement is recorded, and connections to the annuncia-
tor panel activate alarms to warn of high or low levels. A recorder to
display torque loading of the tank agitator paddles was planned, but a
suitable sensing element was not available.
Flow of sludge to the centrifuge is controlled and recorded by instrumen-
tation at the main panel. To provide maximum uniformity of flow to the
centrifuge, a concentrically constricted, elastomer-sleeved valve ("sphinc-
ter valve") was desired. Vendors were hesitant, lest the sludge prove too
abrasive for this type valve, so a stainless steel V-ball unit was used.
Torque load required to drive the internal conveyor to the centrifuge is
continuously recorded. The signal for this measurement is an electrical
measurement of "percent excitation" of the eddy-current brake system pre-
viously ventioned. A manually-set percent-excitation value (commonly 60 -
65%) will serve to interrupt sludge feed and flush the centrifuge, and at
47
-------
95% of brake rating, the centrifuge motor would trip out. Further details
of protective devices have been noted in description of the centrifuge.
Emergency Provisions. To provide for short-term malfunctions of the sys-
tem, including equipment failures or capacity overloads, as well as for
shutdowns for changes, the arrangement shown in Figure 12 was designed.
Flow in excess of pump capacity, or total flow if pumps are shut down,
can overflow to a holding pond. Unacceptable product can be diverted to
the same pond, as well as the contents of the clarifiers if they must be
emptied for maintenance activities.
48
-------
-£>
VO
HOLDING
POND
OVERFLOW
GRIT
BASIN
f
RAW WASTE
DRAINS
COLOR REMOVAL
PROCESS
TO
STREAM
BIO-
OXIDWON
SYSTEM
EMERGENCY FLOW SCHEME FOR EFFLUENT
Figure 12
-------
SECTION V
EQUIPMENT PERFORMANCE AND CHANGES
Grit and Trash Removal
The grit conveying system, as delivered by the vendor, had several obvious
operational shortcomings. Joints in the conveyor drag flight guides were
vulnerable to firm hanging against the ends of conveyor drags. At the
downstream extremity, these guides lacked the clearances necessitated by
the lateral movement permitted by the chain tracking variations. Shape
of these members encouraged jamming by sticks or large chips in the efflu-
ent.
During the early hours of trial running with low water levels (to facil-
itate visual operation), these deficiencies were identified. The parts
were trimmed, shaped and bent to obtain suitable spacing and self-clearing
characteristics.
The conveyor flights are bolted to conveyor chain lugs, and a number of
loosened bolts were noted during early weeks of operation. All nuts were
welded to the bolts to eliminate the obvious hazard involved.
After a few weeks of operation, conveyor drag flights exhibited notice-
able wear at the points of contact with steel wear strips in the bottom
of the grit basin. A welded overlay of hard alloy was applied, and wear
has been reduced to an acceptable level. It appears that several years
may elapse before these drag members must be replaced.
Subsequent performance has been largely trouble-free. On occasion, the
over-night collection of settled material has exceeded considerably the
capacity of the 6 cu. yd. refuse container. Most of this material would
have been harmless to the process system. The most common type of mater-
ial has been woody fragments which have been cut into granules (rather
than defibered) by pulp refiners. Even some defibered pulp settles out
under the conditions required for grit sedimentation. Good in-plant
operation has usually been sufficient to keep the volume of settled ma-
terial below troublesome levels.
The disc screen, after some initial problems in correcting a badly-formed
seat for the rubber seal strip, has performed satisfactorily. The timing
control was designed to provide a minimum 3-minute cycle for the screen
and a longer period for the carry-off conveyor. It was found that, after
the screen stopped, head loss across the screen would sometimes reach
switch-on point before the carry-off conveyor had completed its cycle.
Under such conditions, the timer system would not function until the man-
ual stop-start was actuated. An electrical solution to the problem was
found. However, it is felt the longer cycle for the conveyor is super-
flous, since the quantity of material on the conveyor would be insuffi-
50
-------
cient (after three minutes of screen rotation) to cause difficulty, even
under freezing conditions.
The disc screen has exhibited adequate hydraulic capacity, the one-half
inch drilling has provided the needed system protection, and the quantity
of debris collected has been so small as to present no problems.
Lift Station
The pumping station, as engineered, seems to be an adequate design. How-
ever, experience with the pumps has emphasized the importance of stringent
and detailed specifications when low-bid acceptance is mandatory. In par-
ticular, the pumps have required excessive maintenance attention to shaft
seals and to lower bearings. The seals have been extensively modified to
counteract their tendency to loosen by turning in the direction of pump
rotation.
Lime Slurry System
The lime slaking and feeding system started without difficulty and oper-
ated as expected. At first, the temperature rise of water from input to
output did not seem to correspond to the expected relation between volu-
metric flow rate and weight of lime added. Analyses of slurry samples
yielded results which corresponded to the temperature rise but not to the
lime found in the receiving process; however, reproducibility of these
data was poor. After several weeks of operation, it was discovered that
the water flow meter scale was incorrect. The indicating scale was lin-
ear, while the instrument characteristic was logarithmic; also, full-
scale flow was twice the maximum scale indication.
When the correct scale was used, agreement was found between temperature
rise and hydrated lime content. For the reaction,
CaO + HgO » Ca(OH)2
heat release is 486 Btu per pound of CaO reacted. (20) Thus, slaking
lime to produce a slurry containing 10% calcium hydroxide (7.6% CaO)
will lead to a temperature rise of about 38°F. If twelve million gallons
of effluent per day (8,333 g.p.m.) are to be treated with 1,000 mg/1 (w/w)
of CaO, the required slurry would be about 98 g.p.m. at a temperature rise
of 38°F. It should be remembered that temperature differential reflects
CaO actually slaked, not lime feed to the slaker.
Control of lime feed to the slaker has been affected by two types of
mechanical problems. The first was choking of the feed conveyor with
metal debris. The chief source of such material was broken and detached
buckets from the lime bucket elevators, which were not adequate to the
demands of the heavy, hot-lime service, due to specifications which were
not sufficiently stringent in writing and/or interpretation. The other
problem was performance of the variable-speed control on the feed screw.
This control functioned by shifting effective diameters of pulleys in a
51
-------
belt-driven speed reducer. Failures occurred in the small motor which
controls the drive ratio. Two solutions to the latter are: (1) use of
a simple, manual-crank adjustment, or (2) a different or heavier-duty
speed controller (both more expensive). Experience with this system
suggests the cheaper, manual system. Only rarely have feed rates been
changed more than once during an 8-hour operating shift.
During the early months of operation, little difficulty resulted from
omission of any device for heating the slaker water supply. This might
be partially due to the frequent availability of recovered process water
at about 140°F.; however, water at 80°F. was also used part of the time.
Later, periods occurred when slaking the incomplete and substantial
amounts of lime were rejected by the classifier rate. It was usually
confirmed by other evidence that lime reactivity was low on these occa-
sions; commonly, the lime appeared "overburned", with hard, glazed pel-
lets. Since such conditions will arise occasionally, it was found de-
sirable to add a steam injection heater ahead of the water temperature
measurement. By so controlling the heater that final slurry temperature
is kept below boiling, the temperature rise can still be used to monitor
lime feed. (Obviously, at higher temperatures, heat loss will increase.)
There had been concern that the long lime slurry delivery pipe would be
plugged by solid lime deposits, since such experiences have been reported
from other milk-of-lime pumping systems. However, the maintenance of
good flow velocity has apparently produced the designed effect; no plug-
ging of the main line has been observed. Deposition has occurred at
three points in the slurry system: in the gravity overflow pipe from
slaker to slurry tank, in the short suction piping to the slurry pumps,
and at the point of slurry introduction into the mill effluent. The
first two points represent low-velocity conditions (one pump is normally
idle, with slurry in its suction pipe), and the third is a point at which
calcium carbonate formation occurs. In all three, the periods between
clean-outs has generally been three to five months. Piping has been
arranged to facilitate quick punching out of the deposits. On one occa-
sion, when there was a prolonged problem with low-reactivity lime, the
classifier compartment of the slaker became plugged.
Color Clarifier
The "primary" or "color" clarifier mechanism functioned with virtually
no difficulty, and with no outage for malfunction, throughout the 15
months service covered by this report. (It was used for 3 months as a
conventional "primary" clarifier for mill effluent before start of the
year of color removal demonstration.) This was true despite a much
higher degree of sludge thickening and amount of sludge retention than
contemplated in design. On one occasion, the torque switch, which was
set low at about 1,000,000 ft.-lbs. torque, was tripped due to heavy
sludge load. After about two hours, during which the torque switch was
reset, the mechanism re-started and remained in service.
Distribution and flocculation in the input compartment ("center well")
52
-------
was satisfactory. At the close of the demonstration period, head loss
measurements indicated some fouling of the inlet pipe, but the unit had
not been shut down for inspection.
The clarifier shell, with discharge ports, collection launder, and over-
flow control dams, displayed satisfactory hydraulic capacity and control.
Some fouling of discharge ports had required cleaning on two occasions.
The material involved in the fouling was soft (easily removed) and ap-
peared to be related to periods of inadequate lime dose, and to NSSC
components and dried foam from the carbonation system. (This foam is
discussed elsewhere.)
Carbonation Tank
When the carbonation tank was filled and the gas flow began, the four
agitators were started, and the drive motors were found to be severely
overloaded. The supplier of the agitators had warned that the motors
would overload if the agitators were run in ungassed water, but should
be within rating at normal gas flow. Trials were made with various gas
flow rates to individual gas diffusers, and no condition was found which
effected any considerable reduction in motor load. Each agitator had six
blades which were bolted to a hub. By removing each alternate blade, the
unit was left with three symmetrically located blades. With this change,
the motor load was found to be within a few percent of full horsepower
rating. Little difference was found between gassed and ungassed oper-
ation. Because of this last observation, it was possible to eliminate
an electrical interlock between agitator motors and gas blower.
After the first three days of operation (interrupted at this point for
other reasons), it was found that three of the four gas diffusers had
broken loose at the welded "legs" which had secured them to the tank
bottom. This occurrence is an indication of the violent turbulence
which exists in the carbonator. Each diffuser was then welded to four
new legs made of 3" x 3" heavy angle iron. The legs were full-perimeter
welded to the tank bottom, and each was double-welded along a 3-inch
contact with the diffuser. No further trouble was experienced with these
anchors.
After about 50 - 60 days of actual operation, the interior of the carbon-
ator was inspected. The agitator blades were covered with a hard scale
of calcium carbonate which ranged up to more than one inch thickness.
The diffusers were also coated with about one-half inch of scale along
the lower edges, so that the serrations in the metal edges were largely
rounded over. Scale on the agitator's shaft (steel) and gas supply pipe
(stainless steel) was rather thin and showed evidence of intermittent
flaking. The vertical tank wall had a deposit about 3/4" thick, which
was soft enough to be scraped away with a thumb nail. Scale on the agi-
tator blades and diffuser edges was hammered off in about twenty minutes.
It was subsequently found desirable to repeat this removal about once
per month. Other deposits in the tank proved self-limiting and have
caused no serious trouble.
53
-------
The tank was originally provided with two round vent stacks of 24" dia-
meter. Soon after operation began, it became apparent that the exit
gas velocity was sufficient to entrain any froth, with scant opportunity
for water showers to break the bubbles. Substituting two 30" square
stacks (approximately doubling the cross-sectional area) improved the
situation, but at times foam emission was still a nuisance. The possi-
bility of foam problems had been foreseen prior to construction, and
there had been discussion of a mechanical foam breaker or a foam tower
as potential devices to deal with foam. The expelled foam was observed
to be heavy, wet and laden with calcium carbonate. A temporary expedient,
which would give opportunity to examine behavior of the foam, was decided
upon. A horizontal, 20" diameter pipe was run from the lower portion of
one vent stack (total height,.8 feet) across the outer wall of the color
clarifier. Water level was more than a foot below the top of the clari-
fier, providing capacity for a substantial volume of foam. It was found
that the foam spread and decayed rapidly in contact with air. Any cal-
cium carbonate or other solid matter was trapped and recovered by the
clarifier. Although the partially-dried foam residue was unattractive
in appearance, an adequate foam disposal was provided, and the usage has
been continued.
After about four months service, a blade-mount stub on one agitator hub
broke, apparently as a result of metal fatigue. A similar breakage was
experienced the following month; this time, the break occurred at night,
and before it was discovered, the resulting vibration wrecked the reduc-
tion gear housing. One additional hub failure was recorded during the
first year of operation. During this period, there were two gear replace-
ments due to broken gear teeth.
Toward the close of the year of operation covered by this report, some
limitations in hydraulic flow capacity of the carbonator were observed.
One observation indicated a change in the system; another resulted from
increased volume handled, revealing a latent problem. In the first in-
stance, it was found that scale formation in the discharge pipe was re-
stricting outflow from the carbonator. The other case involved input
flow: at higher throughput, an unexpectedly high liquid level was de-
veloped in the color clarifier outlet compartment. It appears that de-
sign was based upon liquid entry at the surface of water (specific gra-
vity = 1.0), whereas in operation the liquid is gassed to a density of
about 0.8 and rises some two feet higher. (Introducing flow at bottom
of the carbonator would obviate the effect of gassing.)
Gas Blowers
The carbonation tank is supplied with lime kiln stack gas by a Rbots-
Connersville blower (or compressor -- output pressure is 5 - 10 psig)
rated at 4,000 ACFM. Within the first several days of operation, this
unit became the object of two serious concerns.
The first concern related to suitability of the equipment to the service
demands. Bearings and seals required replacement at this time. Although
54
-------
examination indicated the failure was due to corrosion damage while the
unit was idle between manufacture and start-up, several maintenance pro-
blems could be seen. Dust and scale accumulations fapparently calcium
carbonate, chiefly) were already evident, and the seals seemed quite
vulnerable to the effect of such solid material. Replacement of seals
requires almost complete dismantling of the machine, and we have found
it preferable (based on outage time) to move the entire unit to the plant
machine shop.
i
Other seal failures have occurred, signalled by the appearance of water
in the oil circulating system, with subsequent trip-out of the oil pres-
sure switch. On one occasion, one of the head castings was cracked by
pressure exerted by dust accumulations. There have also been problems
re-starting the blower after it has been shut down for a short time; this
was attributed to scale accumulations on the rotor lobes and housing.
(Cooling reduced clearance below that required for free rotation.) Occa-
sional injection of light fuel oil into the blower intake has reduced the
scaling problem somewhat.
The second concern about the blower had to do with capacity. Even though
effluent throughput during early operation was far below design flow rate,
the gas supply was barely sufficient for carbonation at peak demand. This
raised questions about the adequacy of design specifications.
To test the needs versus performance of the system, chemical analyses of
gas and liquid flows were studied. During the early months of operation,
it was found that the kiln gas was being diluted by air intrusion through
a defective labyrinth seal between the kiln and the feed-end housing. The
result was a C02 concentration of only 9 - 12% (dry basis). When the kiln
seal.was repaired, in the closing days of 1971, C02 concentration rose to
a level of 16 - 17% (at minimum ratios of combustion air to the kiln).
Exit gases from the carbonation varied in C02 concentration, depending
upon pH of the liquid phase; however, the concentration was normally below
1.5% for pH above 10.0.
Carbon dioxide demand involves one factor which was not weighed adequately
in design considerations. Much of the sodium content of raw waste is pre-
sent as sodium-organic materials or sodium bicarbonate, which are involved
in calcium precipitation reactions which produce sodium hydroxide, which
in turn consumes carbon dioxide to reach normal carbonation pH; even more
is needed for neutralization prior to bio-oxidation.
Review of the calculations of required kiln gas flow rate indicates three
contributors to a low estimate. Soda alkalinity was one. Variations of
calcium alkalinity above average estimates were not provided for. The
third, and most serious, appears to be the assumption of kiln gfas composi-
tion based upon fuel oil (carbon.-hydrogen ratio of 1:2) instead of natural
gas (carbon:hydrogen ratio of 1:4); no provision was made for variations
from optimum sludge loading or fuel:air ratios.
Arrangements have been made to replace the blower with two units of the
55
-------
"water piston" type, having a combined capacity of 6,000 ACFM. The
choice of type is based upon generally lower maintenance requirement
and up-time as well as a lower vulnerability to dust content of the
gas. In the latter regard, it should be noted that the normal dust
content is very low because of the excellent venturi scrubber; however,
momentary interruptions of scrubber function will entail substantial
dust loadings.
One further note on the kiln gas supply system has to do with the pres-
sure requirements for carbonation. The system was designed for a working
pressure of 10 psig, with the expectation that normal back-pressure of
the system would approach 80% of this value. Actual performance has in-
volved full-flow pressure of about 5 psig. The cloud of gas bubbles
above the gas diffusers, coupled with the flow induction of the turbine
agitators, appear to result in discharge conditions more favorable than
the calculated static pressure of the carbonation vessel.
Carbonation Clarifier ,
The second clarifier has functioned with no mechanical difficulties and
no maintenance attention except routine lubrication. It had been ex-
pected that considerable torque would be developed if sludge withdrawal
were not so controlled as to avoid large accumulations of sludge. It
was assumed that the calcium carbonate produced in the carbonation step
would behave similarly to that from kraft mill recausticizing; caustici-
zing "mud" will settle to a dense cake which (at consistencies around
50% solids) can immobilize clarifier mechanisms. On the other hand, the
sludge collected in the carbonation clarifier has not readily thickened
beyond 25% consistency4 and has poured from a sample container after
sitting overnight. As a result of the sludge properties, torque devel-
oped by the clarifier drive mechanism has never exceeded 15% of the de-
sign rating.
Clarification performance of the unit has been less uniformly satisfac-
tory. Losses of calcium carbonate in the clarifier effluent have ranged
from 2% to 70%. The highest values have been associated with the pre-
sence of organic compounds high in ratio of NSSC-to-kraft origin, or
with inadequate lime treatment. However, not all poor performance has
been readily explainable in this manner. Efforts to improve retention
by sludge recirculation -- either into the carbonator or into the clari-
fier influent — have not yielded significant results. Flocculation
tests with polyelectrolytes by "jar test" techniques, using samples from
the clarifier center well, have not provided a solution to the problem.
Visual observation of these tests does not indicate that the problem
would have been eliminated by a moderate reduction in design rise rate.
Some scaling of the input pipe to the clarifier has been noted. It is
believed this deposition is due largely to the frequently marginal com-
pleteness of carbonation, due to inadequate carbon dioxide supply.
56
-------
Outlet Basin
The outlet basin receives the output flow from the carbonation clarifier,
and the connecting pipe has caused some minor annoyance which suggests
an improvement in piping design. Liquid level in the outlet basin is
more than ten feet below that at the clarifier outlet; from the clarifier,
the pipe extends several feet horizontally and then the effluent passes
through a standard pipe elbow bend into a vertical pipe to another ell
which directs flow into the basin below normal water level. Such a pipe
configuration entrains large amounts of air and causes considerable splash-
ing and may produce foam.
A chief function of the basin is to provide post-carbonation of effluent
to reduce pH to a level compatible with subsequent bio-oxidation or in-
plant re-use. The perforated pipe gas diffuser functioned well for a few
weeks while enough C02 was provided for adequate carbonation plus pH ad-
justment. When carbonation became marginal, scaling of holes in the pipe
began. It is felt the diffuser will be generally satisfactory when the
gas supply problem is resolved.
Sludge Pumps
The pumps originally purchased to handle color clarifier underflow sludge
were Morris Model 3JC14, a high-head sludge pump represented as capable
of handling 1-1/4 inch spherical solids. When the clarifier (with the
pumps) was put in service as a conventional primary clarifier, before the
start of lime treatment of the installation of the disc screen, the pump
intake quickly clogged with bark, wood slivers and pulp. A Morris 4 HS
pump was borrowed to substitute for one of the original pumps and worked
well in this interim, moderate-head service. The 4 HS pump as an impeller
deeply recessed from the head (a style often described in the industry as
a "trash pump"), providing high tolerance to rough solids, with a sacri-
fice of horsepower efficiency. The working head rating is 140', compared
with 170' for the original pump.
When the entire system was completed and lime treatment begun, the re-
maining 3JC14 pump was again tried. When sludge consistency rose to a
desirable range, recurrent choking at the pump suction was observed. The
difficulty no longer seemed to be with coarse debris, but with rheological
properties of the normal process sludge, which offered excessive flow
resistance at the restricted inlet opening and led to cavitation at the
impeller. On the other hand, the Model 4 HS pump worked quite satisfac-
torily, even though the sludge consistency averaged (contrary to earlier
expectations) above 15% solids for monthly periods and exceeded 20% for
several consecutive days. Often, the sludge required shaking to empty
it from a 4-inch diameter cylindrical sample cup. (The pumping distance
is over 650' through 6" pipe, terminating through a 4" magnetic flowmeter
and 3" control valve. Intake is through about 90' of 8" pipe, with about
19' static suction head.)
57
-------
The 1-1/2 JC14 pumps provided for the carbonation clarifier were also
the source of some problems. When sludge accumulated in the clarifier,
consistency would increase to about 20% solids; under this condition,
the pumps would frequently fail to pump. Only by dilution could flow
be maintained, although the concentrated sludge appeared very fluid;
apparently the slurry is somewhat dilatant, and effective viscosity
rises sharply at high shear rates at the restricted intake orifice of
the pump. A 3JC14 unit removed from color sludge service was substituted
for one of the pumps and has been used satisfactorily whenever sludge con-
sistency was high.
The rubber-lined Galigher pumps which deliver color sludge to the centri-
fuge have worked reasonably well. The pumping system, which includes
flow-control instrumentation with V-ball control valve, has been subject
to flow-rate oscillations which are not desirable in a centrifuge oper-
ation. It is uncertain whether pump or valve is chiefly responsible; it
has been suggested that a positive, progressing-cavity pump might provide
better performance. However, there have not been enough difficulties to
justify a change.
Carbonation Sludge Disposition
When lime treatment was begun and initial sludge pump problems resolved,
some carbonation sludge was pumped to the recausticizing mud tank. There
was no clear evidence that the new sludge altered filtration character-
istics of the "mud" (nor certainty that it did not). Due to impurity
load already in the system, in combination with causticizing equipment
difficulties, the precoat filter was already unable to handle the existing
load. The new liquid volume simply added to system losses.
To relieve this situation, at least temporarily, the carbonation sludge
was diverted to the color clarifier input pipe. As expected, there was
no noticeable change in clarification or color reduction. The change did
involve an increase by 50 - 100% in the insolubles mass imposed upon the
centrifuge, and the sludge thickened to a greater density. Torque load-
ing on the color clarifier rakes was not markedly increased, and the cen-
trifuge was able to tolerate the total sludge load. Since the arrangement
proved workable and filter limitations continued, the change has remained
in use.
Color Sludge Storage
The agitator mechanism of the color sludge storage tank was soon found to
be much less effective in the sludge suspension than an identical unit
handling recausticizing mud. When the tank was more than half full, a
watery surface layer appeared, and when the tank was held 75% filled, with
no addition or withdrawal, the torque increased until the drive mechanism
overloaded. Even at low levels (30 - 40% full), solids caked along the
lower walls of the tank, and performance of the withdrawal pumping system
(to centrifuge) gave evidence of dense masses occasionally breaking loose.
.58
-------
After a consultation with mixing specialists, a supplementary, side-
entering propeller mixer was installed at a level which did not interfere
with the radial arms of the original agitator. The new mixer fully loaded
a 20 HP motor. Located almost directly above the outlet pipe, the mixer
has largely eliminated trouble from sludge lumps (although some deposits
form on other areas of the tank periphery), and the tank can be filled
without agitator overload or appreciable stratification of sludge.
Centrifuge
At start of operations, dewatering effectiveness of the Sharpies solid-
bowl centrifuge surpassed expectations. Since pilot-plant work indicated
that every effort would be required to attain 25 - 28% consistency in the
discharge cake, a low pond setting was employed. At some sacrifice of
liquid retention time (expecting a consequent loss of centrate clarity),
this setting provides a maximum drainage time for cake as it emerges from
the liquid and is transported (by conveyor scroll action) up the "beach"
comprising the conical discharge section of the centrifuge bowl. The
first days of operation yielded cake at more than 40% solids.
After less than three days of operation, a centrifuge universal joint
broke. Two universal joints connect the eddy-current brake with the
differential spindle of the conveyor planetary gear box. (By controlled
restraint of spindle rotation, the eddy-current brake regulates speed
differential between the centrifuge bowl and the internal screw conveyor.)
With failure of the universal joint, conveyor action ceased. The failure
was not immediately detected, so sludge feed continued, with water being
expelled while sludge solids accumulated within the bowl. Because of
operator inexperience in detecting and diagnosing this type of occurrence,
and failure to initiate prompt flushing and cleaning action, the conveyor
became so tightly jammed that a screw flight was bent before removal was
finally effected. During the next few weeks of intermittent system oper-
ation, two more universal joints failed, and there were a number of cen-
trifuge trip-outs because of transient peaks in conveyor torque. By
means of a drive pulley change, the centrifuge was slowed from 2,450 to
2,300 RPM; plate dams were changed to increasd "pond" level within the
bowl; and a 5-second time delay was introduced into the high-torque trip-
out switch wiring. With these changes, together with improved operator
skills, satisfactory operation was soon established and continued for
over two months before another universal joint failure.
During the following five months, universal joint failures gradually grew
in frequency. It was noted that the difficulties were related to periods
of high pulp fiber content in the sludge. Representatives of the centri-
fuge manufacturer, who worked closely with project personnel, had analyzed
the problem as a "chattering" effect, involving rapid variations in con-
veyor torque, which result in metal fatigue. This conclusion was dra-
matically confirmed when a direct torque-arm restraint eliminating the
cushioning effect of the eddy-current brake, was temporarily installed on
the spindle shaft. Violent (and frighteningly loud) vibration quickly
tripped the safety release.
59
-------
The manufacturer's engineers now concluded that consolidation of the cake
in the centrifuge bowl was leading to a situation in which large areas
were alternately seizing and releasing as a unit, instead of rolling up
locally along the edges of the conveyor scroll. They arranged to ex-
change the original conveyor for a new one with an altered scroll config-
uration. The effect of the change was a dramatic improvement. Torque
indication dropped substantially and improved in uniformity. During the
following three months, which concluded the demonstration period here
reported, there were no further mechanical failures attributed to this
problem.
The original feed tube design provided with the centrifuge was a double-
wall stainless steel unit which would permit use of flocculating additives
(such use was never found necessary). There were several occasions of
brittle fracture of these tubes. Substitution of a simplified tube design
of mild steel seems to have avoided further failures.
The numerous protective circuits provided for the centrifuge (previously
listed) have been the source of frequent interruptions of centrifuge oper-
ation. Some of the interruptions have been legitimate warnings of hazards,
such as an unreliable cooling water supply pressure to serve the eddy-
current brake. The high-torque trip was found to be responding to tran-
sient peaks for which its response was not helpful; a time delay provided
a better control. However, there have been many spurious alarm signals.
Trouble-shooting efforts have not clarified all of the problems; it appears
that much can be attributed to a sophisticated and sensitive system oper-
ated in a humid and dirty environment, Relocation and protection of com-
ponents is planned.
One control need which was not provided was a protective response in case
conveyor action is interrupted, either by brake failure or by mechanical
failure in the back-drive train (such as a universal joint). A planned
addition is a relay to discontinue sludge feed whenever torque approaches
zero.
One of the most commonly activated protective systems has been one to
respond to rising torque load to interrupt sludge feed and inject flush
water to prevent reaching the trip-out point. Since sludge feed rate is
automatically controlled, an over-riding flow interruption distorts con-
trol action so that, when flow resumes, a brief period of cyclic over-
response occurs. This has often triggered a repetition of the high-torque
condition. To replace the use of a blocking valve to stop flow, a new
piping arrangement has been designed to switch flow to a recirculation
line; thus constant flow rate is maintained through the control loop. The
new flow system is shown in Figure 13.
Centrifuge performance includes, in addition to cake dewatering, the
clarity of the centrate. Since centrate is returned to the raw waste
input, solids are not lost. However, if large proportions of the most
troublesome solids are discharged with the centrate, they may be repeat-
edly recirculated until they accumulate to a quantity which cannot be
60
-------
-M-
COLOR
SLUDGE
TANK
CENTRIFUGE FLOW MODIFICATION
Figure 13
61
-------
contained. It was therefore a matter of some concern when it was first
observed that centrate consistency was sometimes more than 25% of the
feed sludge consistency. At best, it was feared that polyelectrolyte
flocculation aids would be required. However, no serious problem
has arisen. Solids recovery by the centrifuge varied from 70% to
98% in a rather random time frame, with no evident cyclic pattern.
Successful sludge dewatering was one of the most critical engineering
risks undertaken in embarking on the project. It now seems that the
centrifuge application was a sound concept, and that persistent effort
and cooperation by supplier and user have achieved a creditable
working application.
Lime Kiln
Introduction of centrifuge sludge cake into the lime kiln has produced
no evident problems except the obvious load limitations due to the
moisture demands upon fuel and capacity. When the waste treatment
sludge addition is started or stopped, there are no conspicuous
differences in appearance of the product. When the kiln has been
operated with this sludge alone for several hours, there have been no
unusual observations. There have been a variety of problems in kiln
operation, but all seem independent of whether waste treatment sludge
was being processed. The original chain system installed in the kiln
appears to have been more than adequate; it has now been shortened
by burning back at the hot end. There has been some feeling that dust
recirculation is higher with the effluent treatment sludge, but no data
have yet shown clearly that this is so.
Instrumentation
In general, the instrument scheme has been found conceptually valid with
the exception of some details of the centrifuge protective provisions
and the disc screen control, both of which have been discussed in
connection with the operation of the respective process equipment units.
One deficiency might be noted in the sizing of the control valve for
sludge flow from the color clarifier to the color sludge storage tank
(FV-014 on the instrument loop sheets). This 3" valve body with 2"
trim has a very high pressure drop for a lengthy piping system which
already presents a difficult pumping load, and there is no severe demand
for narrow flow deviation. A standard 4" V-ball control valve would seem
preferable and should avoid the frequent choke-ups which have required
use of the 6" by-pass valve, which is too large for convenient manual
control. (However, it has worked!)
One disappointing area of performance has been the pH instrumentation.
Although some of the problems were due to calcium carbonate scaling of
electrodes in the carbonation control system (see below), the pH measur-
ing circuits themselves have performed rather poorly. The problems
62 '
-------
include standardization drift, non-linear response, and generally erratic
behavior. In addition there has been frequent non-agreement between pri-
mary (transmitter) and remote indications.
As previously noted, calcium carbonate scaling -- of electrodes and of
sample piping -- have been a problem in the carbonation pH system. The
problem has been controlled by piping which can be easily, quickly and
regularly rodded free of deposits. The electrode assembly, provided
with flexible cable, can be readily lifted out of the chamber so that
the electrodes can be cleaned with hydrochloric acid squirted from a
squeeze-bottle. The sampling configuration is shown in Figure 14.
63
-------
pH ELECTRODE
ASSEMBLY
CLEAN-OUT CKPS
CARBONATION TANK
FLOW SAMPLER FOR pH PROBE
Figure 14
64
-------
SECTION VI
PROCESS PERFORMANCE
General Operating Considerations.
The Continental Can Company mill at Hodge, Louisiana produces unbleached
kraft pulp (primarily pine) and a smaller amount of neutral sulfite semi-
chemical (NSSC) hardwood pulp with "cross-recovery" whereby spent NSSC
liquors are processed in the kraft recovery system. During the demonstra-
tion period covered by this report, paper machine output averaged about
750 tons per day. Typically, NSSC corrugating medium (79% NSSC, 21%
kraft) was produced at about 300 tons per day during three periods each
month, totaling 10 days. Thus, a month might include 8 days (rarely over
3 consecutive days) of 30 - 40% NSSC pulp consumption, 17 days (seldom
over 7 consecutive days) of all-kraft production, and 5 days involving
transition. Because of the varied product mix of the mill, daily pro-
duction varied widely between 600 and 900 tons.
Effects of NSSC pulping upon effluent properties has been accentuated
by the fact that, while kraft pulp washing performance at the Hodge mill
was well above industry average, the NSSC washing facilities were much
less effective. (Improvements were under construction, but not completed
during this period.) As a result, NSSC contributions to effluent color
were disproportionately large, compared to tonnage ratios. The varied
production rates of the mill (noted above) also result in variations in
the concentration of kraft components of the effluent.
Mill water use averaged around 12 million gallons daily (8,300 g.p.m.).
Throughput of the color removal system was typically 6,000 - 7,000 g.p.m.
Wastes from power plant, causticizing area, bag manufacturing and by-
products processing were excluded from plans for the system. Some waste
water from the evaporator and recovery furnace area was not treated be-
cause needed sewer changes had not been completed. Liquid contents of
each of the clarifiers is about 1,600,000 gallons above the settled
sludge, so that the theoretical total detention time of the two clari-
fiers was around eight hours. However, in these deep clarifiers, the
characteristic liquid travel is far from "plug" flow, and this fact will
have sybstantial effects on concentration/time gradients through the
system.
After some early attempts to establish a program of sampling to follow
time lag through the system, a uniform daily sampling procedure (based
on convenience) was adopted. Raw waste, and the overflow effluents from
the two clarifiers were sampled by three continuous samplers from which
24-hour composites were gathered at the start of each working day. The
raw waste was sampled by a 75ml dipper which was energized at flow-
proportional intervals; this type sampler was chosen .to assure true sam-
pling of fiber content and to minimize interference of trash. The clari-
fier outputs were sampled with diaphragm-type metering pumps. "Grab"
65
-------
samples of sludges, lime and lime slurry were gathered at the same time,
as were process instrument data.
The operational project, as just outlined, offers some impressive ad-
vantages, and also presents some limitations for conducting a technical
evaluation study. Among the advantages:
1. The plant's operational and waste disposal programs were
firmly committed to the project, assuring motivation for
successful operation.
2. The day-to-day variability of the effluent tested the
capabilities of the process under a very wide variety of
conditions.
3. Unpredictability of some process demands placed many problems
in the hands of operating crews revealing adaptability to
"real world" resources.
4. The scale of the operation and the length of the demonstration
period eliminated many of the questions commonly inherent in
experimental operations.
Among the limitations encountered:
1. It was not possible to develop fixed experimental programs
under constant conditions to afford ideal statistical bases
for evaluation. Indeed, the variability of conditions
necessitates much selectivity in data handling to determine
some of the desired relationships. The time lag and mixing
effects (for example) necessitate consideration of the
previous day's parameters in evaluating an item of output
quality data.
2. Capacity of equipment, and the necessity of handling all mill
output, sometimes limit the exploration of significant ranges
of key variables.
3. Because of the size and cost of some equipment components,
system alterations require much time. Even with the most
expeditious handling, procurement of some large items
may require several months.
4. Some transient input parameters cannot be evaluated because
the response peaks are so greatly attenuated by in-process
mixing.
Start-Up and Operation.
Full-scale lime treatment was begun on August 18, 1971 and continued
66
-------
in promising fashion for three days before it was interrupted for more
than a week by a mechanical failure (which has been discussed elsewhere)
Reasonably continuous operation was established in very early October.
Except for 10 days for holidays and repairs and alteration work in
December, and 7 days lost in January because of lime kiln problems,
general system up-time has been above 93%. 'The great majority of the
lost time has been due to problems with lime kiln, lime conveyors and
(to a lesser degree) kiln gas compressor.
During operation of the system, lime feed to the slaker has been
maintained almost 93% of the time. The lime interruptions involved in
this calculation ranged from several minutes to a few hours, and they
were largely due to conveyor problems and to clogging at the outlet of
the lime supply bin. The record by months is shown in Table 19.
During all of this operating period a comprehensive monitoring program
was maintained. A sample of the daily data sheet is pictured in Figure
15. Monthly tabulations of raw data are presented in Appendix D. The
test methods are found in Appendix C. A preliminary summary of the
early months of operation has been published previously. (21)
TABLE 19
TREATMENT SYSTEM UP-TIME
Lime Feed, % of
Month Flow Up-Time, % Flow Up-Time
October 1971 (29 days) 100 91
November 100 96
December (21 days) 100 96
January 1972 (24 days) 86 83
February ' 99 95
March 85 88
April 91 93
May 82 97
June 95 93
July 91 96
August 95 91
Effluent Effects.
Reduction of effluent color, as measured in daily composite samples,
has varied widely, ranging up to above 95% reduction. Although some
of the occasions of poor color removal have not been associated with
any recognized process aberration, most of the data can be explained
by identifiable factors. The two most conspicuous factors have been
inadequate lime addition and the presence of a high ratio of color
derived from NSSC process wastes. There have been a few occasions when
color sludge appeared in the overflow of the color clarifier and was
67
-------
MILL OPERATIONS DIVISION of
CONTINENTAL CAN COMPANY, INC.
HOD3S, LOUISIANA
COLOR REMOVAL DATA
Date
9"/"
PH
Color
TOG
Alk.
CaO
Na*
S3
Flow
Haw 'waste
s,z
840
2.2>J
c/Mo
i
3 XX
11, °i
l.Lt<\
To Final Primary
Carbonator Clarifier Centerwell
/£.5L 1C. 4 J&.Z
$2 /06s-
/37
/6
W6//a3o too/230
3^i, IC>% JO4O
pnm Na^SOf^ 7'«^" weq/1
Fiber, Hi* DS I RAW
Turb
From
Carbonator
10.$
340
WASTS, Hrs.j
Outlet
Basin
9-f
&3 Spindle RPM /flgjQ Fiber ^
Kick-offs & V«'hy
UyJS FrSD: Hour s ^JLJ. Avg . Flov /S'j. Ft . Stor . ? AM •_, Sewered?
SLUD3S SOLIDS: FoedJJ^S. CakejS'.l $1 Ash.Sy.Oj8. Centrate^- i.J. Scrub Water32, g/1.
pH RECORD3KS: Carbonation pH /Q.£. Valve Pos.'yoJa. Hours over 11 Z. i below 9.5.V
Outlet Basin pH /fi^,, Valve Pos. Q ..
«.4
E1.S 6/?2a
DAILY DATA WORK SHEET (Sample)
68
Figure 15
-------
partially re-dissolved. There appear to be a larger number of days on
which lime feed was interrupted one or more times and, although the
average lime feed was adequate, some of the color clarifier output dur-
ing the day had received lower dosages.
Average color reduction for all days of system operation during the last
eleven months of the demonstration period was 70%. If only the days
of 100% kraft production (i.e., when no NSSC effluent was involved) are
considered, the average color reduction was 80%.
Direct measurement of lime addition to the system was not precise enough
for proper evaluation, so the selected indicator of lime treatment has
been the soluble calcium content of the liquid overflowing the color
clarifier. Indeed, theoretical considerations support the conclusion
that this parameter is the prime determinant of color precipitation. Upon
examining the data representing 40 days on which color reduction was above
88%, it was observed that, in virtually every case, the soluble calcium
at the color clarifier had been 400 mg/1 or more. Separate tabulations
of data for kraft effluent and effluent including NSSC wastes were plotted
to show the relation between percent color reduction and calcium ion con-
centration. A most-probable curve was then drawn for each set. These
curves are shown in Figure 16 and 17.
Throughout the demonstration period, it has been noted that the average
color of water samples taken at the color clarifier outlet has been lower
than that of samples taken at the carbonation clarifier outlet. For the
total period, this difference has amounted to an increase of about 25% in
color after the first clarifier. The effect on this color difference
on calculated color removal efficiency, for both all-kraft and NSSC-
containing effluent is shown in the monthly averages listed in Table 20.
TABLE 20
COLOR REDUCTION BEFORE AND AFTER LIME RECOVERY
All-Kraft With NSSC
Month Beforea Afterb Beforea Afterb
October 83.3% WM 70.0% 64.8%
November 85.1 82.1 61.7 58.6
December 85.7 83.1 64.3 57.9
January 78.0 73.5 67.7 64.1
February 89.6 82.1 69.2 64.0
March 84.6 76.6 73.6 67.7
April 87.9 79.4 78.6 70.7
May 82.0 79.2 73.6 63.2
June 88.6 85.9 71.2 65.5
July 86.1 84.0 75.8 68.1
August 83.4 74.3 71.3 66.8
aSampled at outlet of color clarifier.
bSampled at outlet of carbonation clarifier.
Percentage based on raw waste color.
69
-------
fOOI-
o
t~
o
ZD
O
LLJ
CL
60-
O 40
200
3oo
DISSOLVED
CaO
SOLUTION CaO versus COLOR REMOVAL (KRAFT)
Figure 16
70
-------
•joo
2
o
n
OJ
u
100
200 300
DISSOLVED CA, ^/JL CaO
SOLUTION CaO versus COLOR REMOVAL (WITH NSSC)
Figure 17
71
-------
There has been much speculation about the cause of this color increase.
Possibilities suggested have included:
1. Difference in chemistry of pH adjustment.
2. Sample storage at different pH and calcium levels.
3. Effect of unknown component of kiln gas.
The first possibility has been ruled out, but the cause is still not
known. Further work is planned.
The reduction in total organic carbon (TOC) content of the effluent
has appeared subject to the same factors as color reduction. When per-
cent reduction is plotted against soluble calcium, there is much more
scatter of data points. The scatter is believed to reflect a variable
ratio of lignin to saccharides in the effluent. The low molecular weight
soluble saccharides are largely unaffected by lime treatment. Curves re-
presenting data plots for all-kraft and NSSC-containing effluents are
shown in Figure 18 and Figure 19. The data suggest that increased sol-
uble calcium concentrations beyond 400 mg/1 CaO will effect a greater
proportional reduction in TOC than in color. Average reduction in TOC
for eleven months was 37%.
In correlating data on color and TOC, it was assumed at first that graphs
relating the two would pass through zero on both co-ordinate axes. On
this basis, it appeared the data yielded different slope lines for efflu-
ents with NSSC, as compared to all-kraft. However, closer examination
indicated an intercept on the TOC axis at a value between 50 and 100 for
the average of the data compiled in this project. Treating the data
according to this hypothesis yielded an equation:
TOC = K + 0.19 (Color Units),
which would apply to both groups. This positive, minimum value for TOC
seems to reflect the fact that some lignin is precipitated by alum in
kraft papermaking. Graphs reflecting project data for TOC and color
are shown in Figure 20 and Figure 21.
The effect of lime treatment on BODs was one which did not bear out the
promise of bench scale and pilot studies. A reduction of about 30 - 35%
in BODs nad been expected, but no more than about 12% has been indicated
by plant data. Meaningful information is not available for the full
demonstration period, since sample collection during early months was
found to be accompanied by enough bio-oxidation before testing to distort
the results.
Fiber content of the raw waste ranged upward from 125 mg/1 {24-hour
average), and averaged about 300 mg/1. No levels were encountered which
appeared to have any adverse effect on color removal.
72
-------
100
8o
§
u
13
Q
UJ
40-
0
o
.0*0
I
I
I
too
DISSOLVED
300
Ca MS/L as
SOLUTION Cad versus TOC REDUCTION (KRAFT)
Figure 18
73
-------
40
z
o
h-
^0
I
I
fOO 200 3OO
DISSOLVED Ca
400 50O
L as Ca0
SOLUTION CaO versus TOC REDUCTION (WITH NSSC)
'Figure 19
74
-------
,500
I
U
O
2 oo
ICQ
I
jooo
<20oo
COLOR, APHA UNITS
.COLOR - TOC RELATIONSHIP (KRAFT)
Figure 20
75
-------
•JOOO-
300
(BOO
o
H
2.00
JQOO 2,000 3®°° 4000
COLOR, APHA UNITS
COLOR - TOC RELATIONSHIP (WITH NSSC)
Figure 21
76
-------
Clarification, Carbonatlon and Sludge Handling.
Clarity of color clarifier effluent has been excellent when adequate
lime has been applied to kraft wastes. NSSC effluents sometimes have
a slightly turbid appearance, although they are free of settleable solids,
The only evidences of solids carry-over have been due to accumulation of
excessive sludge volume in the clarifier. These occasions have resulted
from sludges which did not thicken to the usual degree; the low rake
torque which resulted was misinterpreted as indicating low sludge level,
and sludge withdrawal rate was not adequate.
Sludge withdrawn from the color clarifier has averaged above 14% solids
and has ranged from 9% to above 22%. The highest consistencies have
been associated with high ratios of calcium carbonate, while the lowest
consistencies have usually been related to high fiber ratio. Some of
the variation has not been explained by such simple explanations, and
there appear to be some more complex factors affecting sludge bulk and
also influencing the rake torque developed at a given consistency.
Except for the mechanical limitations noted in the previous section,
carbonation has proceeded in accordance with expectations. Sampling
difficulties have limited the precision of estimates of gas absorption
efficiency. However, it appears that the efficiency has been above 85%
for pH values above 10.0. Efficiency will increase with pH. Automatic
control of pH has required some careful instrument adjustment to compen-
sate for the equilibration time of the system. However, control to
within +.0.2 pH units has been accomplished whenever proper instrument
performance and steady kiln operation (uniform gas composition) were
achieved.
Carbonation clarifier effluent clarity has been variable. Calcium
losses in the clarifier overflow have ranged from 20 mg/1 to over
200 mg/1 as CaO; in the later months of operation, the average loss has
been slightly above 100 mg/1 as CaO, about twice the content of the raw
waste. The most important adverse factors contributing to higher losses
have been high quantities of NSSC components, and low lime dose. It
appears that all-kraft effluent with adequate lime treatment would halve
this loss. It is also believed that uniform pH control would assist
settling of the calcium carbonate, although no pH has been established
as best for,this purpose.
Since underflow sludge from the carbonation clarifier is delivered to
the color clarifier input, consistency is not critical, and a high
pumping rate is normally maintained. Thus, there is usually no sub-
stantial sludge accumulation in the clarifier except in case of pump
failure. As noted elsewhere, the limiting settled consistency has been
about 25% solids, with surprisingly good fluidity. (See discussion in
Section V.)
I
Centrifuge discharge cake has averaged about 33% solids over the full
demonstration period. Since installation of the new design conveyor
77
-------
screw, the average consistency has been 35%, with a range of 30% to
40%. Centrate discharge has ranged from 1% to 7% solids, with an average
consistency of 3.5%. At the average values of 14%, 35% and 3.5% for feed
sludge, cake and centrate, respectively, the solids recovery of the cen-
trifuge is 83.3%. It had been believed that, if average percent recovery
fell so low, a retention aid would be required to avoid occasionally ex-
cessive accumulation of small-particle material. However, no additives
have been used at any time.
Effects on Lime Burning.
The greatest uncertanties which had been felt about the effect of color
sludge on lime burning were on two points: the possibility of forming
rings or balls in the kiln, and the physical characteristics of the lime.
In both cases, no effects have been noted. The kiln has formed ballst
but the most serious were when no color sludge had been introduced. The
product lime does not look different; there is no evident difference in
fragility of pellets; reactivity in the slaker does not change when
color sludge is added or omitted.
Although it seems certain that lime impurities are contributed by the
color sludge, the amounts have been less than those involved in changes
in the kraft mill system (specifically, changes in green liquor dregs
handling). At most, the impurity level has risen to slightly over 10%
(defined as the sum of acid insolubles and RO^)- There does seem to be
some adverse affect on filtration of causticizfng mud, but problems
with green liquor quality have complicated evaluation.
The impact of color sludge on kiln capacity and fuel use has (at worst)
not exceeded the calculated difference based on water content of the
sludge. Precision of data on lime output has not been good enough to
indicate whether the organic matter in the sludge makes a useful fuel
contribution. However, observations at the time of fuel (natural gas)
trip-outs have shown flame in kiln regions which should represent effec-
tive heat input.
Note:
A U. S. Patent (No. 3,639,206) has been issued, covering the process
involved in this project. (22)
78
-------
SECTION VII
ECONOMIC CONSIDERATIONS
Capital Cost.
Capital costs related to the color removal process may be divided
into three categories: (a) process facilities used exclusively for
carrying out the process, (b) cost of a proportion of lime-burning
and related system for incineration and lime recovery, and (c) other
costs, including land, sewer additions and alterations, holding reser-
voirs, access roads, etc. We shall deal here with only the first two.
The color removal process facilities are considered to include all
effluent handling from inlet piping to the grit basin to the discharge
piping of the outlet basin, including overflow and by-pass structures;
and further to include lime supply and recovery, beginning at the lime
feed conveyor to the color removal slaker and ending at the certrifuge
discharge conveyor. An approximate subdivision of cost is as follows:
Lime Input $ 35,602
Inlet Control; Grit & Trash System 121,319
Lift Station 55,170
Color Clarifier 274,335
Carbonation 95,994
Carbonation Clarifier 277,082
Outlet Basin 40,880
Sludge Storage & Dewatering 135,627
Instruments & Controls 115,015
Process Piping 256,278
Motor Controls 24,805
Electrical Wiring & Lighting 80,146
Pipe Bridge 51,547
Painting & Misc. 38,122
Spare Parts 59.393
Total Direct 1,661,315
Engineering, Plans & Specifications 87,825
Construction Supervision 12,421
Total Cost - Color Removal $1,761,561
(Excluding Lime Kiln Cost)
It is economically significant that the process performs the function
of the conventional primary clarifier and provides complete and final
sludge disposal by incineration. The color clarifier is actually
smaller than the primary clarifier which would be needed for the same
effluent. Thus, a new mill would require no other expenditures for these
purposes, and an existing mill might adapt an existing clarifier to serve
in the color removal process.
The lime kiln requirement to meet peak needs for effluent treatment amounts
to at least 25% of total kraft mill lime supply. The cost of the 12' by
79
-------
290' kiln totalled $2,667,000, Including conveying and storage equipment.
The effluent treatment share of cost would amount to $533,391. If addi-
tional calcining capacity necessitates an additional kiln, costs will be
much greater.
Operating Costs.
During the demonstration period, the volume of waste treated was slightly
below the acticipated flow rate, because of water use economies and
because it was not possible to transfer some desired flows to the sys-
tem. Cost factors have been developed for a flow rate of 9 million
gallons per day and calculated for an average production of 750 tons
of paper per day. The chief cost components are: lime loss, fuel use,
electric power consumption and maintenance.
Lime loss has been calculated from the difference in calcium concen-
tration between input and output water, and from sludge lost to sewer.
During the last 7 months of the demonstration period, average lime loss
was 85 tons per month. The data indicate that effluent losses have
been high because of NSSC effects which have been previously discussed.
There have been some sludge losses to sewer; with operational improve-
ments, these losses should become rare. Costs have been calculated on
the basis of purchased quicklime make-up at $22 per ton. Lime loss of
85 tons per month is equivalent to 7.5 pounds per ton of paper, or 630
pounds per million gallons. The costs are $0.0825 per ton, or $6.93 per
million gallons.
Average operating electrical horsepower has been about 1,000, for which
the cost of electricity is estimated at $100 per day. This amounts to
$0.13 per ton of paper.
Fuel requirements attributable to color sludge processing have been
equivalent to 17,500 million Btu per month. This calculates to 0.77
million Btu per ton, or 65 million Btu per million gallons of effluent.
If a value of $0.48 per million Btu is assigned, the above requirements
amount to $0.37 per ton of paper, or $31.12 per million gallons.
Maintenance costs are estimated at $50,000 per year; this is equiva-
lent to $0.19 per ton, or $15.63 per million gallons.
Summarizing, operating cost (not including depreciation, taxes or
insurance) under these particular conditions would be:
Per Ton Per Million Gallons
Lime Make-Up $0.08 $ 6.93
Fuel 0.37 31.12
Electric Power 0.13 11,11
Maintenance 0.19 15.63
Total WT77 $64.79
Obviously, these costs will be much affected by mill tonnage and water
80
-------
use. To illustrate, we have assumed a mill expansion to 1,500 tons per
day with water economies to yield a total water use of 15 million gal-
lons daily, of which 13 million gallons requires lime treatment. Pro-
cess balance and pulp washing are assumed adequate to avoid the NSSC
effects on calcium carbonate recovery which applied to the previous
case. Lime make-up needs are taken as 5,400 pounds per day. Treatment
dosage and fiber loss are assumed to be in the same ratio to effluent
volume as in the first case, resulting in a fuel demand of 25,280 mil-
lion Btu per month (0.56 million Btu/ton). Electrical load is increased
by use of an additional kiln gas blower and a second centrifuge which
must operate at least half time, so that operating load is 1j350 HP. Total
maintenance cost is essentially unchanged.
The resulting cost, based on the same unit prices as in the previous
case are:
Per Ton Per Million Gallons
Lime Make-Up $0.04 $ 4.57
v Fuel 0.27 31.12
Electric Power 0.09 10.38
Maintenance 0.10 11.75
Total 3OO $57.82
It should be noted that the fuel pricing has a substantial effect on
the total cost.
81
-------
SECTION VIII
ACKNOWLEDGEMENTS
During a period of more than four years, many Continental people
contributed to the successful completion of the project. Among
them:
- Mr. S. Bruce Smart, Paper Group Executive Vice-President,
headed the Continental management team, including Mr. E. A.
Henry {Division General Manager) and Mr. J. G. Lee (Division
Mgr. of Mfg.), which provided the leadership for a large com-
mitment of corporate funds in a pioneering effort.
- Mr. C. C. Kunz, Plant Manager at Hodge, by insistent assertion
of project priority, assured the needed team effort despite
competing demands for time.
- Mr. W. Leroy Coker, Assistant Plant Manager of the Hodge mill,
served as project manager during the major portion of the period.
- Mr. Edgar L. Spruill conceived the project, participated in the
plant design and construction, and provided technical super-
vision of the operation and evaluation program.
- Mr. Fred Turner, research technician, performed much of the basic
laboratory and pilot plant work which defined the project.
- Dr. John Schulz supervised process studies for the final plant
design.
- Mr. Bobby Sammons supervised the departmental operating crews.
- Mr. Ross F. Miller co-ordinated the major efforts of plant
design and construction.
The advice and support of Mr. Herbert Berger, of the National Council
for Air and Stream Improvement, is appreciatively acknowledged.
The efforts of Mr. Dick Moll, Pennwalt Corporation, in helping
establish satisfactory centrifuge dewatering are noted with thanks.
The encouragement and help of Mr. Robert A. Lafleur, Executive Secretary
of the Louisiana Stream Control Commission, is gratefully acknowledged.
The support of the project by the Environmental Protection Agency,
and the assistance provided by Mr. George Webster, Dr. James D.
Gallup, Mr. Robert Hi Her and the Grant Project Officers, Mr. George
Putnicki and Dr. Richard Hill, are acknowledged with sincere thanks.
82
-------
SECTION IX
REFERENCES
1. Moggio, W. A., "Experimental Chemical Treatments for Kraft Mill
Wastes," Tech. Bull. No. 50, National Council for Stream Improve-
ment, New York, N. Y. (1952).
2. Moggio, W. A., "Kraft Mill Wastes Research," Tech. Bull. No. 57,
National Council for Stream Improvement, New York, N. Y. (1953).
3. Moggio, W. A. and Freeman, L., "Treatment of Calcium-Organic Sludges
Obtained from Lime Treatment of Kraft Mill Effluents, Part I," Tech.
Bull.. No. 62, National Council for Stream Improvement, New York,
N. Y. (1953).
4. Moggio, W. A., "Color Removal from Kraft Mill Effluent," TAPPI
XXXVII, (1955) 564-567.
5. Moggio, W. A., "Development Studies on the Removal of Color from
Caustic Extract Bleaching Effluent by the Surface Reaction Process,
Part I," Tech. Bull. No. 107, National Council for Stream Improve-
ment, New York, N. Y. (1958).
6. Berger, H. F. and Brown, R. I., "The Surface Reaction Method for
Color Removal from Kraft Bleachery Effluents," TAPPI XLII. (1959)
245-248.
i'
7. Berger, H. F., "Removal of Color Bodies from Bleachery Effluents
Through Modification of Kraft Recovery Process," Tech. Bull. No. 131,
National Council for Stream Improvement, New York, N. Y. (1960).
8. Berger, H. F., Gehm, H. W., and Herbet, A. J., "Decolorizing Kraft
Waste Liquors," U. S. Patent 3,120,464 (February 4, 1964).
9. Herbet, A. J., "A Process for Removal of Color from Bleached Kraft
Effluents Through Modification of the Chemical Recovery System,"
Tech. Bull. No. 157, National Council for Stream Improvement, New
York, N. Y. (1962).
10. "Evaluation and Demonstration of Massive Lime Process for Color
Removal;" Project No. 12040 D4D, Office of Research and Monitoring,
Environmental Protection Agency, January, 1972.
11. Fuller, R. R., "Color Removal from Kraft Effluents," Southern Pulp
and Paper Manufacturer, September 10, 1971. (Also U. S. Patent
3,627,679.)
83
-------
12. Fuchs, R. E., "Decolorization of Pulp Mill Bleaching Effluents Using
Activated Carbon," Tech. Bull. No. 181, National Council for Stream
Improvement, New York, N. Y. (1965).
13. McGlasson, W. G., Thibodeaux, L. J., and Berger, H. F., "Potential
Uses of Activated Carbon for Wastewater Renovation," TAPPI XLIX,
(1966) 521-526.
14. Berger, H. F., "Evaluating Water Reclamation Against Rising Costs
of Water and Effluent Treatment," TAPPI XLIX, (1966, No. 8) 79A-82A.
15. Thibodeaux, L. J., Smith, D. R., and Berger, H. F., "Wastewater
Renovation Possibilities in the Pulp and Paper Industry," Chemical
Engineering Progress Symposium Series, No. 90, Chemical Engineering
Progress, 64:178.
16. Smith D. R., M.S. thesis, Louisiana State University, Baton Rouge
(1968).
17. Rich, L. G., "Unit Operations of Sanitary Engineering," John Wiley &
Sons, New York.
18. Spruill, E. L., "Color Removal from Paper Mill Waste," Proceedings
of the 25th Purdue Industrial Waste Conference, p. 761. "Paper Mill
Waste; Treatment for Color Removal," Water and Sewage (Industrial
Wastes supplement), 118:(No. 4)IW/15 TJJIrTh7A"pH 1, 1971T
19. Bennett, D. J., Dence, C. W., Kung, F. -L., Luner,, P., and Ota, M.,
"The Mechanism of Color Removal in the Treatment of Spent Bleaching
Liquors with Lime," TAPPI 54(12):2019 (1971).
20. McDonald, R. G. (Ed.), "Pulp and Paper Manufacture," Second Edition,
Vol. I, p. 533.
21. Spruill, E. L., "Color Removal and Sludge Recovery from Total Mill
Effluent," TAPPI 56(4).
22. Spruill, E. 1., U. S. Patent 3,639,206 (Feb. 1, 1972), assigned to
Continental Can Company, Inc.
84
-------
SECTION X
GLOSSARY
ACFM - Actual cubic feet per minute; volume (including water vapor)
at the existing pressure and temperature.
Black Liquor - Spent liquor from wood pulping by the kraft process,
including dissolved organic matter and other products of cooking liquor
reactions. (The organic content comprises about half the original dry
weight of wood.)
BOD - Biochemical oxygen demand; BODc refers to measurement of demand
iuring a five-day period under specified conditions. (See Appendix C.)
Brown Liquor - Spent liquor from wood pulping by the NSSC process.
Carbonatlon - Treatment with carbon dioxide. In this paper, refers
especially to neutralization of alkalinity and/or precipitation of
calcium carbonate with carbon dioxide.
Caustic Extraction - A processing stage, in pulp bleaching practice,
wherein lignin previously treated with chlorine or other oxidants is
solubilized with alkali (usually sodium hydroxide) and separated from
the pulp.
Causticizing - (Sometimes "recausticizing") - Conversion of sodium
carbonate to sodium hydroxide by reaction with lime (N32C03 + CaO +
H20 = 2 NaOH + CaC03); especially applied with reference to sodium car-
bonate recovered by incineration of kraft black liquor.
Clarifier- A device, usually continuous in operation, in which
suspended matter is separated from a liquid by sedimentation, per-
mitting withdrawal of a clear liquid a relatively concentrated
slurry of the solid matter.
COD - Chemical oxygen demand, generally considered to represent the
oxygen required to convert most organic substances to carbon dioxide,
water and other fully oxidized compounds; commonly measured by di-
chromate consumption. (See Appendix C.)
Kraft - A widely-used wood pulping process employing sodium hydroxide
with sodium sulfide to dissolve lignin. Also called "sulfate process,"
because sodium sulfate is used as a make-up chemical; carbon char in
black liquor incineration reduces the sulfate to sulfide.
Mud (or "Lime Mud") - A common kraft pulping usage to designate the
concentrated calcium carbonate slurry separated in the causticizing
process. Used in this report to help distinguish from sludges or slurries
generated by the waste water treatments described.
85
-------
mg/1 - Mllligrams per liter; for dilute acqueous solutions, this value
is approximately identical to parts per million by weight.
NSSC - Neutral sulfite semi-chemical, a pulping process employing sod-
ium sulfite with (usually) sodium carbonate; hardwoods are commonly
pulped at 70 - 80% yields for production of corrugating medium.
Recovery System (kraft pulping) - A process wherein kraft black liquor
is concentrated by evaporation, burned in a furnace to recover heat
value of organic substances (for steam generation) and chemical value
of inorganic materials (which are ashed, dissolved and causticized to
make fresh "cooking" liquor).
TOC - Total organic carbon, a measure of the carbon contained in the
organic compounds in a waste water, for a given waste water source,
this measurement may provide a useful index to chemical oxygen demand,
or otherwise constitute a measure of pollution potential. This para-
meter is susceptible to a standard instrumental procedure. (See Appen-
dix C.)
86
-------
SECTION XI
APPENDICES
Page No.
A. Equipment List for Demonstration Plant 88-91
B. Instrument Loops - Color Removal Plant
- List of Instrument Loops 92-94
- Instrument Loop Diagrams 95-118
C. Sampling, Analytical and Testing Methods
- Tests on Liquid Effluents 119
- Tests on Lime and Sludges 119-120
D. Monthly Data Tabulations, Typical Month 121-124
87
-------
APPENDIX A
EQUIPMENT LIST
00
00
Equipment No.
0420
0421
0422
0430
0431
0440
0441
0010
0020
0030
0040
0021
0031
0041
0050
•%
0060
0061
0062
Name of Equipment
Grit Collector, with Chain Drive
Motor
Speed Reducer
Disc Screen, Chain Drive &
Reducer
Motor
Conveyor, Screen Discharge
w/dri ve
Motor
Raw Waste Lift Station
Raw Waste Lift Pump #1
Raw Waste Lift Pump #2
Raw Waste Lift Pump #3
Motor, Lift Pump #1
Motor, Lift Pump #2
Motor, Lift Pump #3
Color Clarifier/Thickener
Clarifier Mechanism
Motor, Rake Drive, 41
Motor, Rake Drive, #2
Mfr. & Model
Jeffrey Mfg. Co.
(Included)
(Incl.) Falk 6K24
Jeffrey Mfg. Co.
(Steel exc. 31655
screen pi.)
(Included)
Jeffrey Mfg. Co.
(Included)
Field Erected
Sheet pile &
concrete
Goulds Pump, Inc.
VIM-14
Goulds Pump, Inc.
VIM-14
Goulds Pump, Inc.
VIM-14
General Electric
General Electric
General Electric
Shell field erected
concrete
Eimco Corp.
CXXT Extra
Heavy Duty
General Electric
General Electric
Capacity, HP, etc.
3 HP @ 1800 RPM
14' 0 w/0.5" holes
1 HP @ 1800 RPM
1 HP @ 1800 RPM
18 0, 17' deep
6500 GPM
-------
CO
to
Equipment No.
0070
0080
0071
0081
0100
0110
0111
0120
0130
0121
0131
0140
0141
0142
0150
0160
0170
0180
0181
Name of Equipment
Color Clarifier Sludge
Pump #1
Color Clarifier Sludge
Pump #2
Motor, Sludge Pump #1
Motor, Sludge Pump #2
Color Sludge Storage Tank
Tank Mechanism
Motor
Color Mud Pump #1
Color Mud Pump #2
Motor
Motor
Centrifuge
Motor
Eddy-Current Brake
Filtrate Hopper
Sludge Hopper
Centrate Head Tank
Lime Feeder with v.s. Reducer
Motor
Mfr. & Model
Morris Men. Works
3JC14 (V-belt
drive)
Morris Mch. Works
3JC14 (V-belt
drive)
General Electric
General Electric
Rothschild Boiler
& Tank Works
Eimco Corp.
Type BRM
General Electric
Galigher 2VRG200
2 x 2-1/2, belt-
dri ven
Galigher 2VRG200
2 x 2-1/2, belt-
driven
General Electric
General Electric
Pennwalt; Sharpies
P-5400
General Electric
(Spec. Winding)
Rothschild Boiler
& Tank Works
Rothschild Boiler
& Tank Works
Rothschild Boiler
& Tank Works
Screw Conveyor Corp.
(Incl. with reducer)
Capacity. HP, etc.
300 GPM (3 170'
(500 GPM @ 40')
300 GPM @ 170'
(500 GPM @ 40')
40 HP @ 1800 RPM
40 HP £ 1800 RPM
30 0 x 20' Steel
15 HP 9 1800 RPM
185 GPM (P 80'
185 GPM e 80'
15 HP @ 1800 RPM
15 HP @ 1800 RPM
"x"
285 GPM
200 HP @ 1800 RPM
40 HP
14" 0 x 16'
5 HP (P 1800 RPM
-------
ID
o
Equipment Ncr.
0200
0201
0210
0211
0212
0220
0230
0240
0250
0251
0260
0270
0261
0271
0280
0290
0300
0310
0400
0291
Name of Equipment
Lime Slaker Conveyor with
Reducer
Motor
Lime Slaker, with Rake
Classifier
Agitator, Motor
Classifier Motor
Vent Stack
Grits Chute
Lime Slurry Tank
Agitator, Slurry Tank
Motor
Lime Slurry Pump #1
Lime Slurry Pump #2
Motor
Motor
Carbonation Tank
Agitator #1, Carbonation Tank
Agitator #2, Carbonation Tank
Agitator #3, Carbonation Tank
Agitator #4, Carbonation Tank
Motor, Agitator #1
Mfr. & Model
Capacity, HP, etc.
Screw Conveyor Corp. 15" 0 x 16'
Falk 323 EX II
(Included above) 3 HP @ 1800 RPM
Dorr-Oliver No. 7
General Electric
General Electric
Rothschild Boiler
& Tank Works
Rothschild Boiler
& Tank Works
Rothschild Boiler
& Tank Works
Cleveland Mixer Co
Heavy Duty Model AL
General Electric
Galigher Vac-Seal
2 x 2-1/2
Galigher Vac-Seal
2 x 2-1/2
General Electric
General Electric
Rothschild Boiler
& Tank Works
Chemineer, Inc.
MDP-400-721
Chemineer, Inc.
MDP-400-721
Chemineer, Inc.
MDP-400-721
Chemineer, Inc.
MDP-400-721
General Electric
324 T
15 HP @ 1800 RPM
2 HP
-------
Equipment No.
0301
0311
0401
0320
0321
0322
0330
0340
0341
0342
0350
0360
0351
0361
0370
0380
0381
0382
4000
6000
6010
Name of Equipment
Motor, Agitator #2
Motor, Agitator #3
Motor, Agitator #4
Compressor, Kiln Gas with
Lube Oil System
Motor, Compressor
Motor, Oil Pump
Carbonation Clarifier
Clarifier Mechanism
Motor, Rake Drive, #1
Motor, Rake Drive, #2
Carbonation Sludge Pump #1
Carbonation Sludge Pump #2
Motor, Sludge Pump #1
Motor, Sludge Pump #2
Outlet Basin
A/C Unit, Motor Control Room
Motor, Compressor
Motor, Fan
Motor Controls
Instruments & Process Controls
Instrument Panel, Control Room
Mfr. & Model
General Electric
324 T
General Electric
324 T
General Electric
324 T
Roots-Connersvi11 e
1421, RGVS
General Electric
(with Compressor)
Tank, field-erected
concrete
Eimco Corp., CXXT
Extra heavy duty
General Electric
General Electric
Morris Mch. Works
1-1/2 OCL4 (Belt
drive)
Morris Mch. Works
1-1/2 JCL4 (Belt
drive)
General Electric
General Electric
Concrete, field-
erected with gas
diffuser piping
Carrier Corp.
(Included)
(Included)
General Electric
Foxboro Company
Foxboro Company
Capacity, HP, etc.
40 HP @ 1800 RPM
40 HP @ 1800 RPM
40 HP @ 1800 RPM
4000 ACFM
250 HP @ 900 RPM
1 HP (3 1800 RPM
135' 0 15' side-
wall depth
1,200,000 ft.-lbs.
working torque
10 HP @ 1800 RPM
10 HP (3 1800 RPM
100 GPM @ 120'
100 GPM (3 120'
10 HP (3 1800 RPM
10 HP (3 1800 RPM
21' x 27' x 12' D.
10 HP
1-1/2 HP
-------
APPENDIX B
INSTRUMENT LOOPS
Loop 001 Level, Raw Waste Pump Station
LIC-001 Controller, 0 - 15 ft.; Proportional & Reset
LR-001 Recorder; 4 in. strip; 0-15 Scale
LT-001 Transmitter; 0 - 180 in. I-^O; output 10-5 MADC
LV-001A Control Valve; 24" Butterfly; 316 SS Trim; Cyl. Op. w/
positioner
LV-001B Control Valve, alternate flow - Sec. Clarifier; same as
LV-001A
LAH-001 Annunciator - High Level
Loop 002 Flow, Raw Waste to Color Clarifier
FE-002 Orifice Plate, radius taps; 50" we at 15,000 GPM
Loop 003 Speed Control, Lime Feeder
HS-003 Switch
SC-003 Speed Control
Loop 004 Speed, Lime Feeder
SI-004 Indicator, 0-30 RPM
ST-004 Transmitter
Loop 006 Dilution to Slaker
FI-006 Indicator, 0 - 200 GPM
FV-006 Control Valve, 2.5" CI Globe; w/positioner
FE-006 Orifice Plate, flange taps; 100" we at 200 GPM
FT-006 Transmitter
FC-006 Controller
FAL-006 Annunciator - Low Flow
Loop 007 Temperature, Slaker Dilution
TE-007 Resistance Bulb, 0 - 300°F.
TR-007 Recorder, 4" strip; 0 - 300; 2 pen (with TR-008)
Loop 008 Temperature, Slaker
TE-008 Resistance Bulb, 0 - 30QOF.
TR-008 Recorder, 4" strip; 0 - 300; 2 pen (with TR-007)
Loop 009 Lime Slurry Tank Level
LT-009 Transmitter, 0 - 120" water
LIC-009 Level Controller
LV-009 Control Valve, 2" CI Globe; diaphragm operator
FV-009 Flush Valve, Slurry Line; 2" DI Ball; Elec. 2-pos. operator
Loop 010 Color Clarifier Rake Torque
XT-010 Torque Transmitter
92
-------
XR-010 Torque Recorder, 4" strip; scale 0-2 (with XR-011)
XS-010 Clarifier Rake Switch
XAH-010 Annunciator, High Torque
Loop Oil Carbonation Clarifier Rake Torque
XT-Oil Torque Transmitter
XR-011 Torque Recorder, 4" strip; scale 0-2 (with XR-010)
XS-011 Clarifier Rake Switch
XAH-011 Annunciator, High Torque
Loop 012 Carbonation pH
AE-012 pH Element, Glass/calomel
AIT-012 Transmitter, pH, 0 - 14
AIC-012 Controller, with proportional and reset functions
AV-012 Control valve, 8", 316 SS Butterfly; diaphragm oper., w/
positioner
AR-012 Recorder, pH, 0 - 14; 4" strip (2 pen, with ZR-012)
ZR-012 Recorder, valve position; with AR-012
Loop 013 Outlet Basin pH
AE-013 pH Element, 014
AIT-013 Transmitter, pH, 0 - 14
AIC-013 Controller, with proportional and reset functions
AV-013 Control Valve, 3", 316 SS Butterfly; diaphragm oper., w/
positioner
AR-013 Recorder, pH, 0 - ,14; 4" strip (2 pen, with ZR-013)
ZR-013 Recorder, Valve Position, with AR-013
Loop 014 Color Clarifier Sludge Flow
FE-014 Magnetic Flow Meter, 4", 304 SS Tube, Teflon Liner
FIC-014 Flow Controller, 0 - 300 6PM; Proportional & Reset Functions
FR-014 Recorder, 0 - 300, 4" strip; 3 pen, with FR-015 & 018
FT-014 Transmitter, 0 - 300 GPM; 10-50 MADC
FV-014 Control Valve, ball, 3" body, 2" SS Trim; Diaph. Op. w/
positioner
FV-014A Control Valve, ball, 6", 316 SS Trim; cyl. operator
FAL-014 Annunciator, low flow
Loop 015 Carbonation Clarifier Sludge Flow
FE-015 Magnetic Flow Meter, 2", 304 SS Tube, Teflon Liner
FIC-015 Flow Controller, 0 - 150 GPM; proportional & reset
FR-015 Recorder, 0 - 150; 4" strip; 3 pen, with FR-014 & -018
FT-015 Transmitter, 0 - 150 GPM; 10-50 MADC
FV-015 Control Valve, 3" ball, 316 SS Trim; diaph. op.; w/
positioner
FV-015A Control Valve, 4" ball, 316 SS Trim; cyl. operator
HIC-015A Controller, hand loader, 0 - 100%
Loop 017 Color Sludge Storage Tank Level
I-T-Q3Z Transmitter, 0 - 240" water; 10-50 MADC
LR-017 Recorder, 0-20 ft.; 4n strip
93
-------
LAH-017 Annunciator, high level
LAL-017 Annunciator, low level
Loop 018 Flow, Color Sludge to Centrifuge
FE-018 Magnetic Flow Meter, 4"; 304 SS tube, Teflon liner
FIC-018 Flow Controller, 0 - 200 6PM; proportional & reset
FR-018 Recorder, 0 - 200; 4" strip; 3 pen, with FR-014 & -015
FT-018 Transmitter, 0 - 200 GPM; 10-50 MADC
FV-018 Control Valve, 4" ball, 2" SS trim; diaph. op., w/
positioner
Loop 019 Annunciator
ANN-019 Annunciator, 4 wide x 3 high; horn & flasher
Loop 020 C02 Compressor Relief
PIC-020 Controller, 0-15; proportional & reset
PT-020 Transmitter, 0-15 psig; 10-50 MADC
PCV-020 Control Valve, 6" Butterfly, 316 SS; diaphragm oper., w/
positioner
Loop 021 Torque - Centrifuge
XR-021 Recorder, 0 - 100%, 4" strip
XT-021 Transmitter 0 - 50 mv to 10-50 MADC
XAH-021 Annunciator, high torque
Loop 023 Slaker Agitator Motor
QA-023 Annunciator, agitator stopped
Loop 026 Flush, Color Sludge Pump and Lines
FV-026 Control Valve,'!" ball, 316 SS Trim; diaphragm oper., w/
solenoid, 3 ways
FV-026A Control Valve, 10" ball, 316 SS trim; cyl. oper., w/hand
jack
HS-026 Switch
Loop 027 Flush, Centrifuge
FV-027 Control Valve, 3" globe, 316 SS trim; diaph. oper., w/
solenoid
HS-027 Switch
Loop 028 Carbonation Sludge Recirculation
FV-028 Control Valve, 3" ball, 316 SS Trim; cyl. oper.; w/
positioner
HIC-028 Controller, hand loader, 0 - 100%
94
-------
ES.
E.5. nia VAC)
BY-PASS TO
. CLARIF.
*•
TO COLOR
CLARIF.
LEVEL, RAW WASTE PUMP STATION
Loop 001
-------
PE-002
RAW WASTE PUMP
STATION EFFLUENT
NOTE: QFUFiCE FOR FUTURE RECORDING
RAW WASTE FLOW TO COLOR CLARIFIER
Loop 002
-------
LIME FEEDER
G7.3-OT8O
— — \\
LIME FEEDER SPEED CONTROL
Loop 003
-------
00
LIME
DAY BIN
27.8-OOEO
PESJER
GT.3 -O18O
LIME
FEEDER
SPEED
Loop
004
-------
VO
<£>
VZ 304 S3. TUBING
167.3-006-PI
'/£" 304 55. TUBING
|£7.3-00€>-P2
IG73-OQ6-A1
A.S. C P5I6)
C ES.C118 YAC.)
DILUTION WATER
TO SLAKER
67-3 - O210
SLAKER
DILUTION WATER
FLOW
Loop
006
-------
E.5.018 VAC.) *
o
o
R/I
^ E.S. 018 YA.C)
DILUTION WATER TO
SLAKER ,
TEMPERATURE, WATER TO SLAKER
Loop 007
-------
R/I
5 E.S. 018 VAC.)
SLURRY FROM LIME
SLAKER 673-OZ1O
TEMPERATURE, SLURRY FROM SLAKER
Loop 008
-------
r
I 67.3-00
<> FLUSH WATER
TO PR|. CLARlF.
LEVEL, LIME SLURRY TANK
Loop 009
-------
E.S. (MS VAC.) V-
o
CO
FWE
\
:* TORQUE
4 E S. (63- 95 Y.D.C, )
FWE
NOTE: SEE EIMCO DWGS. 7276 IC4
7Z76IC5
COLOR CLARIFIER RAKE TORQUE
Loop 010
-------
OCR
\ O1Q
PWE
X=TORQUE
4 E.S. (63-95 V.D-C.)
FWE
NOTE: SEf EIMCO DVVGS. 72.7&tC^
CARBONATION
CLARIFIER
RAKE
TORQUE
Loop
01
1
-------
E.S. (118 VAC.)
-------
E.S. 018 V.A.C) V--
o
en
pH
67.3-013-Afl
A-S. (40 PS»G)
1> E.S-018 VAO
VAC.)
OUTLET pH
Loop 013
-------
t ^, __^_ ^_ -
E.S. 018V.A.C) ES. ms VA.C.)
NOTE: /A/STALL BY' PASS |
FV-OI4A A80VE FV-01-^
VA.C.)
-------
e.s. me VA.C.)
O
00
ij
ill
18 V.AC.)
J61.3 -Q)5"-AE|
SLUDGE FROM CARB. CLAR1F.
TO MUD STORAGE TANK
CARBONATION SLUDGE FLOW TO STORAGE
Loop 015
-------
.S. ma
-4 E. S. (63-95 VD.C.
o
-------
1
E.S.
4 E.S. (118 tfAC.)
I67-3-Q18-A1
T
*
COLOR SLUDGE STOR.
TK, TO CENTRIFUGE
FLOW, SLUDGE TO CENTRIFUGE
Loop 018
-------
HIGH LEVEL
RAW WASTE
LIFT STATION
HIGH TORQUE
CENTRIFUGE
*1
LOW FLOW
DJLUTION
TO SLAKER
HIGH TORQUE
CO4.&R
CLARIF. RAKE
Hl&H TORQUE
CENTRIFUGE
*Z
MOTOR STOPPED
SLAKER
AGITATOR
HIGH TORQUE
CARS- CLARIF.
RAKE
HJGH TORQUE
COLOR SLUDGE
TANK AGITATOR
LOW FLOW
PR I. CLAR\F.
SLUDGE
HIGH LEVEL
COLOR SLU06ET
STORAGE TAWK
LOW LEVEL
COLOR SLUDGE
STORAGE TANK
SPARE
ANNUNCIATOR ENGRAVING
Loop 019
-------
VENT TO STACK
167.3-020-A1
A.S. H-O PSD
r
^ E.S. I H8 VAC)
COZ COMPRESSOR
67.3 -03ZO
COMPRESSED GAS TO
CARBONATION BASIN
4 AFTER CLAR1F
C02 COMPRESSOR RELIEF CONTROL
Loop 020
-------
<*>
COLOR SLUDGE
FROM
STORAGE PUMPS
PUT.
X-TORQUE
FWE
\ \ \ \ \ \ \
\ \ \ \ \ \ \
COLOR SLUDGE
CENTRIFUGE
EQ. No. OHO
CENTRIFUGE HIGH TORQUE ALARM
Loop 021
-------
AGITATOR
LIME SLAKER
EQ. NO. OE10
SLAKER AGITATOR STOPPED ALARM
Loop 023
-------
DILUTION WATER TO SLAKER
SERVICE N* 734
PRESSURE, SLAKER DILUTION WATER
Loop 025
-------
67.3-026-A1
A.S.<
FLUSH WATER
TO CENTRIFUGE
SPECIFIED ON PANEL DWG
COLOR SLUDGE
PUMPS
COLOR SLUDGE
STORAGE TK.
67-3 -O!00
COLOR SLUDGE PUMP & LINES FLUSH
Loop 026
-------
[6T3-027-A1
AS. (
*)
EV-027
FLUSH WATER
- 027
OLOR SLUDGE TO CENTRIFUGE
CENTRIFUGE FLUSH
Loop 027
-------
00
FROM CARBONATE SLUDGE
PUMPS TO CARBONATION BASIN
-* A.S.
CARBONATION SLUDGE RECIRCULATION
Loop 028
-------
APPENDIX C
SAMPLING, ANALYTICAL AND
TESTING METHODS USED
Tests on Liquid Effluents
1. Alkalinity was determined by a modification of the standard
methods to adapt to the high alkalinity and color encountered
in some of the samples. Titrant acid was 0.1 normal, end-
points were taken at pH 8.8 and 4.25, using a standard poten-
tiometric instrument. Sample size was adjusted to obtain
titrations between 10 and 25 ml. Results were calculated to
mg/1 as CaC03.
2. Biochemical Oxygen Demand (BOD5) was determined by the azide
modification described in "Standard Methods for the Examination
of Water and Wastewater," 12th Edition, and dissolved oxygen
was determined by the probe method, using a 451 temperature-
compensated probe with electric stirrer.
3. Calcium was determined by EDTA titration, using Chrome Black T
indicator. Samples were acidified with a few drops of HC1;
highly colored samples were then treated briefly with activated
carbon and filtered; one ml of cone. NH*OH was added to obtain
suitable pH. (Samples from the color cTarifier were obtained
from clear, supernatant liquid, since dissolved calcium was
desired.)
4. Color determinations were made after filtering the sample through
an 0.8 micron membrane filter, diluting as needed to bring below
500 color value, adjusting pH to 7.6, and reading absorbance
with a Spectronic 70 spectrophotometer with round, 19mm cu-
vettes, cgmparing with a curve developed against a standard
APHA platinum-cobalt solution.
5. Total organic carbon (TOC) was determined with a Beckman 915
analyzer using the procedures described in the EPA manual,
"Methods for Chemical Analysis of Water and Wastes," 1971,
p. 221.
6. Sodium was determined by an Instrumentation Laboratories
! Model 143 flame photometer and reported as equivalent
in mg/1.
Tests on Lime and Sludges.
1. Solids content of sludges, including centrate liquid, were
determined by weighing about 25 grams into an aluminum, dis-
119
-------
posable dish, drying overnight at 105°C. and reweighing.
Dissolved solids are included in this measurement, commonly
amounting to about 0.1% to 0.2%. Because of the solids ranges
involved in this study, the error was considered acceptable.
(For more dilute slurries, a parallel determination of dissolved
solids would be required for correction.)
2. Available lime was determined by a modified Scaife method,
whereby:
(a) 1.42 grams of finely ground lime is weighed into a 500ml
volumetric flask and a duplicate sample into a 400ml beaker.
(b) 200ml of distilled water is added to each and boiled for
5-10 minutes.
>(c) The contents of the beaker are titrated with 1.0 Normal
HC1 and the acid volume noted. 5ml less of the same acid
is added to the volumetric flask, which is then diluted to
the mark and settled.
(d) From the clear supernate in the flask, 200ml is withdrawn
and titrated against 0.2 Normal HC1.
2 (ml 1.0 N acid) + (ml 0.2 N) = % Available CaO
3. Causticizing value of lime is determined as follows:
(a) Dissolve 190 grams of soda ash in 1,800ml of distilled
water in a 2-liter beaker and bring to a boil.
(b) Cool slightly, and slowly add 100 grams of the lime
sample. Boil for 15 minutes, adding water to maintain
the level in the beaker.
(c) Remove the beaker from heat, stir briefly, reverse
stirring motion to stop swirling, and allow to settle.
Record time to reach 50% of depth.
(d) When well settled, pipette 5ml of supernatant liquid
into a flask and titrate with 1.0 N HC1 to the phenol-
phthalein and methyl orange end-points ("P" and "M").
(e) Calculate:
% Causticizing Value = 2 P"M
0.01M
120
-------
APPENDIX D
MONTHLY DATA TABULATIONS (TYPICAL MONTH)
1. WATER QIUUTX MIA
«*«' Jo/rtF,
"'«"'
DCPT
AUMLDOTI
m
P-H
TO CARB'N
P-M
OUT
P-M
COLOR
JB_
TO
CAHB'M
OOT
T.O.C.
_ffi_
OPT
COD
HI/OPT
FINAL
EFFLUQtT
TDS
(13)
KMhOh
DEKAND
M!SSL
SDLPATB
IB/OW
JA.
^ft
«- tie
*/«*o
_io_
-0-^.gP
To -
±i
••<^o
>f-j.ao
IfS-
^£_
/ft/
/gJ~
0 - f±/>
160 -A A a
ife
L- fflg
770 -
^
y/p
i
1 o
IAI
/n
16
3fe
zt
4,3.0 - lfO
-30-.
""
yjTP
13 O
II IO
zU
7e -&tg
^JO
& /A - 12. D
-JLTO
140
IAO
o -a.au
i -.370
t&
I4D
ro
O-A/O
/ZT
HHi
IIO-tK>O
'
^go -i
fn-S-tO
J /J*
7J/J -
to-tlo
IAD
•760
SSI
418
0 -3-
•Jfe- IIto
la -J.
/jrt/o
J« e
fttf
J. 6 -SJ-O
^AJ-
/£*,
333
g-
-JSj^
y^<> -^
/4^Q
"To"
ae
LtZi
^A4L
o - AaO
ItO -
Aid
It
-------
ro
COLOR REMOVAL PROCESS DATA
RAW WASTE
KLLLIOK
GALLONS
-
.2.3..
7, c/
tf-l-
e. .
7. «?
o
•7.4-
?.J
Color
«. i
/a.<
O..C
VA.J
^.J
/a. i
IA.O
Jl.T
//tf
l/.f
SA.O
a.o
IA -0
>!•£>
J.C
//f
n.i
Carb'n
//. c
ic.S
ID.tl
lt.«
n-o
//.a
• i.f
f-
^t
/O.
J±£
1. /
Jt.'J
11.3
Otttlat
Ca+ «£ CaO
Raw
Wast.
VO
To
Carb'n
To
5atlai_
lt.
JLS6
Afk
-3«*
t*jf
/so
too
>Z2-
tAS
Il
//O
KRAFT
Paper
Tons
NSSC
Cor. M«d.
Tons
VI
Iff
eii-
110
Aft
jj2_
fttf
4&Z
fOtf
131-
^Kr
A»
i
•&&-.
25
26
29
30
32
33
34
38
36
3T
38
-------
ro
CO
SCREW
SM
10
_/!_
jr. UMS RED AKD CKNTHTOS DATA
MONTH
SLAKER
£8L
.££_
qj
*f
3£
-3JCL.
^
^-
&c
^JT
5LC_
J1S_
-------
ro
KILN OPERATION BATA
' ' *«•'
DEPt
SLUDGE FILTER COS
H*** A BH
4 SOLIDS I
6,1,0 I f.&O
&
»
Free CaO
, SCI
ii.C
£.£ '• ,CO;S
a.7
fe/.o
5"7.C A.SJ"
,p^5'
-J2.
;i i 5i& i :
M.7
^.e
SS.S
AVAIL.
CaO. i
7.1.
Ti.O
80.0
s/.c
KILN PRODUCT ANALYSIS
CAUST'G
EFF'Y.4
sem,.
RATE
> ACID
IHSOL.
> LOSS
ON ION.
LDIB
LOSS.
LBS.
n>
GRITS
DISCARD
CU.FT.
.^aoji.
GAS
US3D
MCF
;aj-i
AS-iO
it^g
s-lio
AIJ.O
SCRD3
WATER
GPL S.S.
¥7,0
. *•
0
-"-\
-------
SELECTED WATER
RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
1. Rept ft No.
3. Accession No.
w
4. Title
COLOR REMOVAL AND SLUDGE DISPOSAL
PROCESS FOR KRAFT MILL EFFLUENTS
7. Author(s)
5. Report D&tt
6,
S. Performing
Report No.
Edgar L. Spruill, Jr.
9. Organization
10. Project No.
12040 DRY
Continental Can Company, Inc.
Mill Operations Division
Hodge, Louisiana 71247
11. Contract/Grant No.
H3. Type ef Report and
) Period Covered
12. Sponsoring Organization
IS. Supplementary Notes
Environmental Protection Agency report number, EPA-660/2-7^-008, February
'16. Abstract
A treatment plant, removing color by lime addition and recovering sludges, has been
treating over 80% of the effluent of an unbleached kraft mill for one year. Using up to
1,100 mg/1 of CaO, with normal mill fiber loss as a precipitation aid, average color
reduction was 80% for all-kraft effluent. At upper range of lime dosage, when residual
dissolved Ca was above 400 mg/1 as CaO, color removal was 85-93%. When mill production
included 33-40% NSSC hardwood pulp, color reduction averaged only 65%.
About 12% BODg reduction was observed, and average TOC reduction was nearly 40%.
The chief negative factor is need for emergency protection against alkaline impact on
secondary treatment and receiving stream.
Following centrifuge dewatering, sludge incineration has had minimal impact on
kiln operation; there were some adverse effects on lime quality. Lime recovery was 93%.
Mill kiln capacity must be increased about 25%.
Primary clarification and sludge disposal are included in the process. Operating
costs, exclusive of capital factors, are estimated at $0.50-$0.80 per ton of paper, or
5.5$ to 6.5$ per thousand gallons, depending on fiber losses and water usage.
17a. Descriptors
*Pulp and Paper Industry, *Waste water treatment, *Chemical Precipitation,
*Lime, *Color, *Sludge Disposal, Physical Properties, Centrifugation,
Dewatering, Incineration ..'Ultimate Disposal, Costs.
17b. Identifiers
*Lime Treatment, *Color Removal, Kraft Effluent,
Kraft Sludge Disposal
17 c. COWRR Field & Group
IS. Availability
19. * Sf'ittrity Class..
(Keport)
20. Security Class. •
•&«) . •.. . • '
21. #d. of "
Pages
22. Price
t
Send To:
WATER RESOURCES SCIENTIFIC INFORMATION CENTER
U.S. DEPARTMENT OF THE INTERIOR
WASHINGTON. D. C. 2OZ4O
Abstractor
Institution
------- |