EPA-600/2-77-034
January 1977
Environmental Protection Technology Series
LEATHER TANNERY WASTE MANAGEMENT THROUGH
PROCESS CHANGE, REUSE AND PRETREATMENT
Industrial Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2 Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY 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.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-77-034
January 1977
LEATHER TANNERY WASTE MANAGEMENT THROUGH
PROCESS CHANGE, REUSE AND PRETREATMENT
by
James M. Constant!n
George B. Stockman
Pfister and Vogel Tanning Company
Milwaukee, Wisconsin
Grant S-801037
Program Element 1BB037
ROAP/Task. No. 21 BAA/49
Project Officer
W. L. Banks
U.S. Environmental Protection Agency
Region VII
Kansas City, Missouri 64108
Prepared for
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Industrial Environmental Research
Laboratory, U.S. Environmental Protection Agency, Cincinnati, Ohio, and
approved for publication. Approval does not signify that the contents
necessarily reflect the views and policies of the U.S. Environmental
Protection Agency, nor does mention of trade names or commercial products
constitute endorsement or recommendation for use.
ii
-------
FOREWORD
When energy and material resources are extracted, processed, converted,
and used, the related pollutional impacts on our environment and even on our
health often require that new and increasingly more efficient pollution con-
trol methods be used. The Industrial Environmental Research Laboratory -
Cincinnati (IERL-CI) assists in developing and demonstrating new and improved
methodologies that will meet these needs both efficiently and economically.
The title of this report aptly describes the contents - "Industrial
Waste Water Management in Tanneries through Process Change, Reuse and Pre-
treatment." The authors provide details for design of a sulfide oxidation
unit to treat unhairing liquor, a chrome recovery unit, a protein precipita-
tion process and methods of water conservation. This work is a major contri-
bution to the state of the art regarding lowering pollution loadings from
tanneries. The authors worked in close harmony with the project officer and
have provided many details of tannery pollution characteristics not previous-
ly recorded. Either the authors or project officer can provide further
information on the subject matter contained herein.
David G. Stephan
Director
Industrial Environmental Research Laboratory
Cincinnati
m
-------
ABSTRACT
Pfister & Vogel Tanning Company discharges process wastes to Milwaukee's
Sewage Treatment/Fertilizer Manufacturing Plant. At a daily production level
of 4,350 cattlehides, wastewater volume average 1.42 mgd, containing 17,300 Ib
BOD-5 (54,600 Ib COD); and the 126,000 Ib of total solids analyzed 38,500 Ib
of suspended solids. Total nitrogen averaged 4,070 Ib. These wastes also
contained 7,370 Ib of Oil and Grease, 828 Ib of sulfide, 14,700 Ib of alkalin-
ity, and 1,340 Ib of total chromium. While the incremental pH varied from 3.3
to 12.7, composited samples averaged 9.8. Subprocess waste fractions have
been characterized.
This project demonstrates pollution reductions for a large side leather
tannery through process changes, recovery, reuse, and pretreatment. Tradi-
tional tanning operations were replaced by a straight-through process. Salt-
cured hides, loaded dry to a hide processor (cement-mixer) are not removed
until chrome tanning is completed, thereby effecting a 50% reduction in beam-
house-tanyard effluent volumes. Additional reductions in trivalent chromium,
sulfides and Oil and Grease have been demonstrated. Investigations into
protein recovery, and hydrocyclonic separation of solids from selected wastes
have been initiated.
Work included in this report covers research on pilot and bench scale
units as well as in full plant scale operation. Trivalent chromium was
removed from waste streams and re-used in subsequent tanning operations. Sul-
fide removal was accomplished through plant scale catalytic oxidation. Oil
and Grease level in the total plant effluent was reduced by process
modification.
This report is submitted by Pfister & Vogel Tanning Company in fulfill-
ment of Grant Number S-801037, under partial sponsorship of the Environmental
Protection Agency. Work was completed in March 1976.
-------
CONTENTS
Soc_ti_qn_ Pago
I Introduction 1
II Conclusions 3
III Recommendations C
IV Background 7
V !-'aste Surveys 9
VI Chrome Recovery Feasibility Studies 13
VII Chrome Recovery System Design 70
VIII Chrome Recovery System-Full Scale Operation 31
IX Chrome Recycle-Matched Pairs Leather Tests 92
X Hydrogen Sulfide Control 97
XI Sulfide Oxidation in Unhairing Wastes-Feasibility Studies. . 103
XII Sulfide Oxidation System Design . 120
XIII Sulfide Oxidation System-Full Scale Operation 133
XIV Recovery of Crude Protein from Oxidized Unhairinn Waste. . . 146
XV Removal of Solids from Soak Wastes 153
XVI Generation of Oil & Grease and Chrome During Blue Stock
Mringing 156
XVII Gravity Separation of Oil & Grease from Ceamhouse Hastes . . 159
XVIII Recovery of Lime from Spent Lime Liquors 162
XIX Acknowledgments 166
XX References 167
-------
[lumber
1
2
3
4
5
6
7
3
9
10
n
12
13
14
15
FIGURES
Hide handling steps for traditional hair-save process. . . .
Effect of pll on supernatant chromium concentration
Effect of Malcolyte 677 on the settling of flAOM-precipitated
chromium hydroxide
Effect of Nalcolyte 677 on the settling of Ca(nn)2
precipitated chromium hydroxide
Dissolved air flotation apparatus
Settling of lime-precipitated chromium hydroxide
Suspended solids vs chromium for lime-precipitated
chromium hydroxide.
Graphical determination of the rate of subsidence and
thickener retention time
Effect of chromium concentration on design volumetric
loadinq
Effect of suspended solids concentration on desiqn
volumetric loadino
Effect of suspended solids concentration on
retention time
Effect of feed suspended solids concentration on
underflov/ concentration
Effect of susnended solids on unit area
Buchner funnel test apparatus
Effect of various alkalis on the filtration of
chromium hydroxide
Pane
8
13
22
23
26
29
30
31
37
33
39
40
41
45
4C
10 Filter leaf test apparatus 4
o
-------
Number Page
17 Effect of temperature on the filtration of lime-
precipitated chromium hydroxide 58
13 llydrocyclonic separator 66
19 Flocculation of chromium hydroxide 78
20 Chrome Recovery System 80
21 Bench-scale oxidation of sulfides in unhairing wastes. . . . 106
22 Oxidation of sulfides by paddle aeration 112
23 Determination of field transfer rates 119
24 Determination of field transfer efficiency 125
25 Determination of mean oxyqen saturation value 126
26 Sulfide Oxidation System 132
27 Full-scale sulfide oxidation curve 138
28 Rate curve for full-scale sulfide oxidation 139
29 Full-scale sulfide oxidation rate curve - potassium
permanganate catalyst 140
30 Titration curve for acidification of oxidized
hair-burn liquor 148
31 Effect of pi! on total solids for acidification of
oxidized hair-burn liquor 149
32 Effect of pll on sludge volume for acidification of
oxidized hair-burn liquor 150
33 Flocculent settling of lime in relime liquor 164
vn
-------
Number
1
2
3
4
5
C
7
0
9
10
11
TABLES
EFFECT OF PROCESS CHANGE ON UASTEWATER LOADINGS
WASTEWATER LOADINGS FOR LEATHER MANUFACTURING SUBPROCESSES .
TYPICAL ANALYSIS OF CONCENTRATED SPENT CHROME
EFFECT OF flaOH DOSAGE ON PRECIPITATION OF CHROMIUM
HYDROXIDE
EFFECT OF Ca(OH)? DOSAGE ON PRECIPITATION OF CHROMIUM
HYDROXIDE - PART I
EFFECT OF Ca(0!l)2 DOSAGE ON PRECIPITATION OF CHROMIUM
HYDROXIDE - PART II
EFFECT OF NH^OH DOSAGE ON PRECIPITATION OF CHROMIUM
HYDROXIDE
EFFECT OF Na2C03 DOSAGE ON PRECIPITATION Oc CHROMIUM
HYDROXIDE
EFFECT OF VARYING DOSAGES OF NALCOLYTE 677 ON THE
SETTLING OF LIME-PRECIPITATED CHROMIUM HYDROXIDE ....
CHROMIUM AND SUSPENDED SOLIDS CONCENTRATIONS FOR THE
SETTLING OF LIME-PRECIPITATED CHROMIUM HYDROXIDE ....
SURFACE AREA REQUIREMENTS FOR THE THICKENING AND
Page
9
11
1
-------
iNtHiiber Pane
15 EFFECT OF VARIOUS ALKALIS Oil FILTER YIELD AND FILTRATION
RATE FOR CHROMIUM HYDROXIDE 51
16 FILTER LEAF TEST RESULTS - PART I 53
17 FILTER LEAF TEST RESULTS - PART II 54
18 EFFECT OF VARIATION IN DRUM SPEED OH THE VACUUM FILTRATION
OF LIME-PRECIPITATED CHP.OMIU?'. HYDROXIDE - PART 1 55
19 EFFECT OF VARIATION IN DRUM SPEED ON THE VACUUM FILTRATION
OF LI'1E-PRECIPITATED CHROMIUM HYDROXIDE - PAPT II .... 56
20 EFFECT OF VARIATION IN SUSPENDED SOLIDS CONCENTRATION ON
THE VACUUM FILTRATION OF LIME-PRECIPITATED CHROMIUM
HYDROXIDE 59
21 EFFECT OF TEMPERATURE ON THE FILTRATION OF LIVE-
PRECIPITATED CHROMIUM HYDROXIDE
22 ACIDIFICATION OF°CHRO'UUM HYDROXIDE FILTER CAKE - PART I. . 61
23 ACIDIFICATION OF CHROMIUM HYDROXIDE FILTER CAKE - PART II . 61
24 PILOT-SCALE VACUUM FILTRATION OF CHROMIUM HYDROXIDE .... 63
25 REAPPORTIONMENT OF OIL AND GREASE DURING THE FILTRATION
OF CHROMIUM HYDROXIDE 64
26 CYCLONE SEPARATION OF CALCIUM SULFATE FROM SPENT CHROME
LIQUOR AT VARYING ORIFICE DIAMETERS - PART I
27 CYCLONE SEPARATION OF CALCIUM SULFATE FROM SPENT CHROME
LIQUOR AT VARY inn ORIFICE DIAMETERS - PART II 68
23 DISTRIBUTION OF CHROMIUM It! HIDE PROCESSOR HASTES 71
29 ALTERNATIVE LEVEL? OF CHROME CAPTURE 71
30 SURFACE AREA FOP SLUDGE THICKEN I Hr- - PAP.T I 75
31 SURFACE AREA FOR SLUDGE THICKENING - PART II 75
32 SURFACE 1P.EA FOR SLUDGE THICKENING - PART III 75
33 FILTER AREA FOR CHROMIUM HYDROXIDE 7C
34 CHROME RECOVERY SYSTEM PERFORMANCE - LIME PRECIPITATION
PART I 33
-------
Number Page
35 CHROME RECOVERY SYSTEM PERFORMANCE - LIME PRECIPITATION
PART II 85
36 CHROME RECOVERY SYSTEM PERFORMANCE - SODA ASH
PRECIPITATION 89
37 CHROME RECOVERY SYSTEM OPERATING COSTS - SODA ASH VS
LIME PRECIPITATION 90
38 PROPORTIONS OF DISSOLVED SULFIDE PRESENT AS H2S 98
39 OXYGEN UPTAKE FOR SULFIDE OXIDATION 109
40 SULFIDE OXIDATION BY PADDLE AERATION Ill
41 EFFECT OF CATALYST DOSAGE ON SULFIDE OXIDATION BY
PADDLE AERATION . . . . 114
42 SULFIDE OXIDATION BY SURFACE AERATION - PILOT SYSTEM 115
43 CUMULATIVE CAPTURE OF SULFIDES 121
44 ALTERNATIVE LEVELS OF SULFIDE COLLECTION AND TREATMENT. ... 121
45 OPERATING COSTS FOR VARIOUS SULFIDE OXIDATION METHODS .... 130
46 EFFECT OF CATALYST DOSAGE ON SULFIDE OXIDATION 135
47 TREATMENT OF UNHAIRING WASTES BY PROTEIN PRECIPITATION. ... 151
48 SIEVE ANALYSIS OF SAND FROM SOAK LIQUOR 154
49 EFFECT OF WRINGING ON BLUE STOCK ANALYSIS 157
50 CHROME AND FATS DISCHARGED FROM BLUE STOCK WRINGING 158
51 EFFECT OF FLOTATION ON THE OIL AND GREASE CONTENT
OF VARIOUS WASTE FRACTIONS 161
52 FLOTATION OF PICKLE/SOAK WASTEWATER MIXTURES 161
-------
SECTION I
INTRODUCTION
Approximately 60% of the leather produced in the United States is tanned
in plants located in municipalities(l). The combined treatment of industrial
and sanitary wastes offers a more efficient reduction of industrial wastes,
and minimizes duplication of treatment operations. Most municipal tanneries
discharge their process wastewaters to municipal sewerage systems without
extensive pretreatment. With the enactment of the Water Pollution Control
Act Amendments of 1972, Publicly Owned Treatment Works (POTW's) must now
establish a system of User Charges as well as a program for Industrial Cost
Recovery to qualify for Construction Grants. The concentrations of pollu-
tants in tannery wastes, as measured by BOD and suspended solids, are many
times those found in domestic wastes and User Charges for most tanneries have
or will become a severe financial burden.
The Tanning Industry has been making a concerted effort over the past
ten years to develop treatment methods for their wastes. Numerous demonstra-
tion grants have been awarded to individual plants both by the Federal Water
Quality Administration and the Environmental Protection Agency for the pur-
pose of evaluating the technical and economic feasibility of various treat-
ment methods. The emphasis of the studies carried out to date has primarily
been directed toward treatment of effluents as discharged. Only limited
efforts have been reported thus far in the U.S. to achieve the desired
results through tanning process changes and by reuse and treatment of the
wastes in-plant.
Most urban tanneries are unable to apply demonstrated wastewater treat-
ment technologies because of the prohibitive space requirements. The only
options available for many plants, therefore, will be to reduce effluent
loadings through conservation of water and processing chemicals, by process
changes and by pretreatment of selected high strength waste fractions.
Chromium, for example, is a limited as well as a politically-sensitive
resource. It is forecast that increasingly difficult prices and availability
are forthcoming, making reuse of spent chrome liquors especially important.
The basic objective of this project has been to demonstrate process
modifications and wastewater reuse and treatment methods which are techni-
cally and economically feasible in effecting major reductions in the total
waste load discharged from a large side leather tannery.
The methodology developed through this project, therefore, will be
applicable to a large segment of the Tanning Industry. Concomitantly, it
-------
should also have a salutary effect on the total waste management effort for
those municipalities 1n which tanneries are located. Additionally, those
tanneries which are in a position to employ the more traditional treatment
procedures will be provided with alternate methods for achieving the desired
and/or required water pollution control objectives.
Measurements have been made and reported in units familiar to and used
by the tanning industry. Conversion to the International System of Units
(SI) mav be made by reference to the "Metric Practice Guide," AST!! E 330-76.
-------
SECTION II
CONCLUSIONS
This work has demonstrated the feasibility of several in-house waste-
water management techniques including process changes, recovery and reuse
procedures and treatment methods, which mitigate pollution loading.
Hide processors were installed and the traditional hair-save methods of
leather production were replaced by a "straight-through" hair-burn process
in which soaking, unhairing, bating, pickling and tanning is accomplished
within a single processing unit. Besides eliminating numerous hide handling
steps, this process change effected a 50% reduction in beamhouse-tanyard
effluent volume from an average 107 gallons per side to 54 gallons per side.
While much of this water was eventually restored as "sewer-flushes," an over-
all 38% reduction in 1970 effluent volumes has been observed, primarily due
to an intensive water conservation campaign. Consistent with the 50% reduc-
tion in lime usage, the pH was lowered from 11.0 to 9.8 with a corresponding
31% reduction in fixed suspended solids. A 71% increase in sulfide loadings
was noted due to the required increase in sharpeners for hair pulpinq. Like-
wise, total Kjeldahl nitrogen, volatile suspended solids and total volatile
solids loadings increased substantially by 79%, 46% and 32%, respectively.
Slight increases in Oil and Grease, chloride, BOD-5, COD, total solids,
suspended solids and dissolved solids were also recorded. The substantial
increase in total chromium loadings from 2.36 to 5.18 lb/1000 Ib equivalent
green-salted hides is partially due to a designed increase in chrome offered
as well as reduced exhaust efficiency in hide processors - dictated
ultimately by product quality considerations.
The unit processes and unit operations which could be incorporated into
a chrome recovery scheme were investigated and demonstrated using conven-
tional laboratory and pilot-scale procedures. Significant process/equipment
design criteria rely heavily on the demonstrations that the choice of alkali
for chromium hydroxide precipitation dramatically effects solubility and
sludge dewatering characteristics. Hydrated lime was the most effective
precipitating agent followed by caustic soda, aqueous ammonia and soda ash.
Only lime-precipitated chromium hydroxide reacted favorably with flocculants
to form a large, sturdy, settlable floe. An anionic polyelectrolyte,
Nalcolyte 677 (Nalco Chemical Company) was optimum at 20-30 ppm.
Dissolved air flotation was ineffective at thickening chromium hydroxide
sludges at an air-to-sol ids ratio of 0.025-to-l.
Gravity settling of lime-precipitated, floculated chromium hydroxide,
however, proved highly effective.
-------
The design volumetric loading for chromium hydroxide thickening was
established at 400 gpd/ft2 for the most concentrated chrome-bearinn waste
fraction generated. The effect of dilution of spent chrome (with wash waters)
on design thickener surface area is offset by increased settling rates. Most
importantly, within the 2,330 to 6,590 mg/1 Cr concentration range, unit
areas remain constant at approximately 12 ft2 per ton of dry solids per day.
Effective dewatering of thickened chromium hydroxide sludge was demon-
strated. Hydrated lime and soda ash-precipitated sludges filtered best.
Pilot rates of 6.6 and 7.9 gal/ft2-hr, respectively, with corresponding dry
solids rates of 2.1 and 1.6 Ib/ft2-hr were observed.
Production-scale chrome recovery trials demonstrated the recovery of
chromium from the most concentrated tanyard wastes generated from eleven (of
eighteen) production units daily. Overall, end-of-the-pipe reductions in
total chromium are calculated at 37% from an average 5.2 Ib to 3.3 Ib
Cr/1000 Ib equivalent green-salted hides. Of the chrome captured, treatment
efficiencies in excess of 98% reduction were observed.
Matched sides comparative leather tests revealed no significant differ-
ence in leather quality when reacidified chrome was used to replace 23% of
the normal production chrome offering. No upper limit for this replacement
has yet been determined.
Oxidation of sulfides in unhairing wastes was also investigated in detail
to determine the most effective process/system design. Manganese sulfate
catalyzed air oxidation was shown to be most economical.
The optimum catalyst dosage fell within the 0.05 to Q.15 Mn+'l"/S=range.
Of the several types of aeration equipment investigated, surprisingly
enough, traditional paddle vats were found highly effective, and should be
considered a viable alternative when such equipment can be diverted from
normal manufacturing uses. The preferred method was identified as surface
aeration using floating aerators. Applicable criteria for the specification
of such equipment include wastewater volume, sulfide content, catalyst and
power costs and available space and time. The sulfide oxidation efficiency
of a typical 2 hp floating surface aerator in pilot-scale operations was
approximately 1.87 times the manufacturer's oxygen transfer efficiency rating.
On a production-scale, a 10 hp floating surface aerator nave complete
oxidation of sulfides in 6,400 gallons of unhairing wastes (including concen-
trated hair-burn liquors plus half of the wash water from two unhairinn runs)
from an initial concentration of 2,830 mg/1 in 2.75 hours. This is equivalent
to a sulfide oxidation aerator efficiency of 5.83 Ib S=/hp-hr, or 1.71 times
the manufacturer's rated oxygen transfer efficiency. A 98% reduction in sul-
fide was achieved after only 2.1 hours of aeration. The observed sulfide
oxidation aerator efficiency was 2.02 times the rated oxygen transfer effi-
ciency at 6.9 Ib S=/hp-hr.
Additional bench-scale investigations identified all tannery waste frac-
tions as highly contaminated with Oil and Grease. Segregated spent chrome
-------
and pickle liquors may be treated by simple gravity flotation (1 hour
retention) to effect 95 and 89% reductions, respectively. On a production-
scale, a 10,000 gallon "batch" of spent chrome liquor was treated by gravity
flotation for a 92% separation. An additional 7% of the Oil and Grease was
recycled with the recovered chrome while only slinhly more than 1% was
actually sewered. Flotation for "removal" of Oil and Crease was ineffective
for soaking, unhairinq, relime, and bating liquors.
Detailed sampling and analyses of individual beamhouse and tanyard
waste fractions revealed unhairinq wastes as the most highly contaminated.
This waste fraction (excluding relime) contributes 19%, 33%, and 30% to the
total effluent BOD, COD and suspended solids loadings, respectively. Crude
protein can be precipitated from unhairing waste to effect a major reduction
in these contaminants. In bench-scale experiments, oxidized hair-burn liquor
was acidified to generate a rapidly settling protein precipitate. After 60
minutes settling, COD reductions up to 79% were observed. The optimum pH
for precipitation was within the 0.9 to 5.0 range, while the minimum sludge
volume obtained was within the 1.0 to 3.9 pH range.
-------
SECTION III
RECOMMENDATIONS
Several of the process/equipment specification and design criteria
developed during this project have been adopted in production-scale chrome
recovery and sulfide oxidation systems. Further work must be undertaken to
optimize these installations. In particular, the effect of around-the-clock
operation of the chrome recovery system must be studied to determine
"continuous" chromium and Oil and Grease removal efficiencies. Likewise,
operation of the sulfide oxidation system must be coordinated with typical
production scheduling to demonstrate optimum sulfide reductions.
The reductions in overall tannery chromium and sulfide loadings due to
these wastewater management techniques have been calculated. However* fur-
ther sampling and analyses will be required to characterize the more subtile
interactions among these "changed" waste-streams which are currently only
hypothesized.
Because of the demonstrated wide differences in the effects of various
alkalis on chrome recovery efficiencies, additional alkali studies including
calcium carbonate, sodium carbonate, magnesium oxide and magnesium carbonate
must be investigated along with alternative coagulants and flocculants.
Flotation by in situ generation of C02 should be pursued further as a
possible alternative to present chromium hydroxide thickening techniques.
Additional work is also warranted to determine the effectiveness of flo-
tation for treating segregated tannery waste fractions, as well as combined
tannery effluents.
The aeration requirement for the batch oxidation of sulfides in unhair-
ing wastes has been determined using a catalyst dosage within the apparent
optimum range. Additional work is required to define the catalyst dosage -
aeration interactions outside of this catalyst range. These interactions
will determine the optimum design for sulfide oxidation.
The hydrocyclonic separation of calcium sulfate from reacidified chrome
is a major process innovation in the chrome recovery system. However, in
light of the unspectacular removal efficiencies observed, various operating
parameters including flow, pressure and orifice diameter should be studied
to "optimize" this separation. Also, this unit operation should be investi-
gated as a tool for the treatment of suspended solids in other waste
fractions.
-------
SECTION IV
BACKGROUND
Historically, at Pfister & Vogel, leather was manufactured by the tradi-
tional hair-save method with paddle or drum soaking, paddle liming and
mechanical unhairing followed by drum bating, pickling and tanning. As
Figure 1 illustrates, the process was a lengthy, labor-intensive one,.requir-
ing eight hide handling steps for prefleshed stock and nine handling steps
for unfleshed stock. This process generated 107 gallons of wastewater per
side processed to the blue condition. During 1972 a major process revision
was made to improve material handling, process time and space requirements.
This conversion consisted of the installation of a straight-through,
hair-burn process in which soaking, unhairing, bating, pickling and tanning
is accomplished sequentially within a single processing unit. For prefleshed
hides, handling requirements were reduced to only loading the hides into the
processing unit and final blue piling. This new process generated only 54
gallons of wastewater per side processed to the "blue." Twelve hide proces-
sors were installed to handle approximately two-thirds of Pfister & Vogel's
production.
These processing units are equipped with recirculating/drain pumps
Introducing a new dimension to in-house wastewater management techniques.
Process wastes from the .units can now be discharged as discrete waste frac-
tions at distinct points within the plant with a greater degree of hydraulic
control. Individual waste fractions can be segregated through a manifold
arrangement for recycle of treatment before discharge to common tannery
sewers.
This new process technology effected a marked decrease in wastewater
volume. A 50% reduction in effluent volume from beamhouse and tanyard opera-
tions was observed. The programmed 50% reduction in lime consumption drasti-
cally reduced the mean alkalinity of the total tannery effluent. Also, a
substantial Increase 1n the quantity of sharpeners (sodium sulfide and/or
sodium sulfhydrate) was required by the switch to the "hair-burn" approach.
Thus, the potential for the formation of toxic hydrogen sulfide gas in the
plant sewers and inside the sewage screening facility was greatly increased.
Numerous safety and alarm measures were adopted and sewer flushes were
installed to augment sewer flows, provide needed dilution, reduce retention
times and reduce this hazard. The increased flow to meet these safety and
housekeeping measures unfortunately offset most of the designed reduction, in
process effluent volume.
-------
UNFLESHED STOCK
PREFLESHED STOCK
LOAD SOAKS
SOAK PILING
FLESHING
LOAD LIME PADDLES
LIME PULLING
MECHANICAL
UNHAIRING
WEIGH-OFF PACKS
_v
LOAD TAN DRUMS
BLUE PILING
LOAD SOAKS
SOAK PILING
LOAD LIME PADDLES
LIME PULLING
MECHANICAL
UNHAIRING
WEIGH-OFF PACKS
LOAD TAN DRUMS
BLUE PILING
Figure 1: HIDE HANDLING STEPS FOR
TRADITIONAL HAIR-SAVE PROCESS,
-------
SECTION V
WASTE SURVEYS
The effect of Pfister & VogeTs conversion from a paddle soak - paddle
lime - drum bate, pickle, tan-hair-save process to a "straight-through" hair-
burn process on effluent loadings is shown in Table 1. It should be noted
that numerous process modifications made concomitant with the change-over
were dictated by the need for adjustments in leather properties. These modi-
fications contributed significantly to the ultimate hair-burn process waste
profile. A 38% reduction in 1970 effluent volumes was observed from 8,960
gallons per 1000 Ibs of equivalent green-salted hides charged to the process
to the current level of 5,550 gallons per 1000 Ibs of hides. Only a portion
of this reduction, however, is attributable to the process conversion.
While beamhouse/tanyard process waste volumes were reduced 50% from 3,970
gallons .per 1000 Ibs in 1970 to the current 1,980 gallons per 1000 Ibs, an
additional 1,330 gallons of water per 1000 Ibs had to be put back into the
system as a safety and housekeeping requirement to insure adequate flushing
of internal sewers.
TABLE 1. EFFECT OF PROCESS CHANGE ON WASTEWATER LOADINGS
Parameter
Flow
PH
BOD - 5 (20°C)
COD
Oil & Grease
Chloride as Cl"
Alkalinity as CaCOs
Calcium as Ca++
Sulfide as S=
Kjeldahl N as N
Ammonia N as N
Total Solids
Suspended Solids
Dissolved Solids
Total Fixed Solids
Total Volatile Solids
Fixed Suspended Solids
Volatile Suspended Solids
Chromium as Cr
Loading, lb/1000 Ib
Hair-Save
8,960
10.8-11.2
62.4
185
22.0
106
-
23.6
1.87
3.77
3.29
451
136
315
338
113
64
72
2.36
green-salted hides*
Hair-Burn
5,550
8.9-10.7
66.9
211
28.4
119
56.8
_
3.20
15.7
4.64
486
149
337
340
146
44
105
5.18
* Except for flow in gallons/1000 Ib and pH which is unitless.
-------
Most of the observed flow reduction was due to an intensive water conser-
vation program. A major element of this program was the incorporation of a
gravity screening and water recirculation system for the wet collection of
buffing dust. This recycle alone effected a wastewater reduction of nearly
one quarter million gallons per day while reducing wastewater effluent load-
ings by an estimated 200 Ibs of suspended solids per day in the form of
screened and separated, wet buffing' dust.
The data presented in Table 1 are calculated loadings of various waste-
water parameters per 1000 Ibs of equivalent green-salted hides. The actual
Pfister & Vogel rawstock mix consists of varying proportions of green-salted
cattle hides and prefleshed, brine cured cattle hides dependent upon relative
quality, price, availability and finished leather requirements. The observed
raw stock mix ranged from 22% to 100$ prefleshed hides on the sampling days
and all loadings have been adjusted to an equivalent green-salted hide weight
basis to provide a common denominator. The equivalent green-salted weight
was extrapolated from Pfister & Vogel's production experience which indicates
that each pound of prefleshed brine cured hide is equivalent to 1.28 Ibs of
conventional, green-salted hides. Effluent samples were composited (in pro-
portion to flow) from 15-mlnute-incremental grab samples taken immediately
downstream from tbe 5/32" rotary wastewater screen, over a twenty-four
hour period - and were'analyzed according to Standard Methods. Flow was
measured by a "pump-on-time" totalizer connected in parallel with flow-
calibrated industrial sewage pumps. These data confirm the expected reduc-
tion in pH and fixed suspended solids consistent with a 50% reduction in
lime usage. Likewise, the marked increase in sulfide from 1.87 to 3.20 lb/
1000 lb 1s due to the required increase in sharpeners for hair pulping. The
total solids increased only slightly while there was a considerable change
1n the proportions of fixed and volatile residues. Most notably, total vola-
tile solids and volatile suspended solids increased 32% and 46%, respectively,
while total Kjeldahl nitrogen was boosted by 79% to 15.7 lb/1000 lb. This
process conversion more than offset any reduction due to buffing dust screen-
ing with net increases in BOD-5 and total suspended solids of 7% and 14%,
respectively. A slight increase in the COD-to-BOD ratio from 2.96:1 to
3.15:1 was observed.
The effluent loading due to each "lower tannery" subprocess wastewater
discharge was determined for a typical hide processor run in which prefleshed,
brine cured hides were used as rawstock. Twenty-four discrete process dis-
charges were sampled covering the most concentrated process floats through
the most diluted wash waters. After determination of the volume and strength
of each waste fraction, the appropriate loadings were grouped and the total
pollution profile attributable to the unit processes of soaking, Hm1ng,
bating, pickling and tanning were calculated. These data are compiled in
Table 2.
10
-------
TA3LE 2. UASTE'.'ATER LOADINGS FOR LEATHER MANUFACTURING 5U3PROCESSES
Subprocess wastewater loadinos
Soak Limo Relime
.'!!
JOD(20°C)
COD
Oil & Grease
Chloride as Cl~
Alkalinity as CaCC>3
Acidity as CaC03
Sulfide as 5=
iCjeldahl ;l as N
Ammonia N as ri
Total Solids
Susp. Solids
Dis.Solids
Tot. Fixed Sol .
Tot. Vol. Sol.
Fixed Suso. Sol .
Vol. Susp. Sol.
Chrom'un as Cr
9.8-
10.1
9.9
13
6.0
33
7.8
0.54
-
S3
12
77
69
19
5.0
6.6
11.6"
12.2
15
75
4.0
45
31
4.8
5.1
-
120
34
33
86
31
17
17
"
12.3-
13.1
5.6
13
1.6
9.4
33
1.2
1.3
-
41
13
28
35
5.8
11
2.2
"
in lb/1000 Ib equiv. nreen-salted hides*
Sub-
Bate Pickle Tan Total Post-tan
3.9-
9.1
5.0
6.3
2.0
2.9
3.9
0.067
5.9
4.6
44
8.8
35
20
24
4.5
4.2
"
2.7
1.1
3.6
0.42
7.6
1.1
0.005
0.25
-
20
0.6C
19
16
4.0
0.27
0.38
"
2.9-
3.4
3.3
13
5.2
15
10
0
0.89
-
92
2.6
39
75
17
0.47
2.2
4.3
40
120
19
110
76
11
6.1
14
4.6
400
70
330
300
100
38
32
4.8
6
56
4
4
0
0
59
14
44
23
34
0
14
1
.5
.7
.4
-
.38
.19
.38
.6
24 hr
Composite
67
210
28
120
57
3.
16
4.
490
150
340
340
150
44
100
5.
2
6
2
* except pH which is unit!ess
-------
The COD/BOD ratios for the various lower tannery process wastewaters
ranged from 1.28:1 for bating wastes to 4.99:1 for unhairing. Also, a ratio
of 8.61:1 for post-tan operations including blue stock wringing, coloring,
retan, fatHquoring arid finishing was observed. This exceptionally high
COD/BOD ratio 1s probably due to the presence of spent vegetable tannins from
retan operations.
All of the processing steps contribute significant quantities of BOD,
COD, Oil & Grease, chloride, nitrogen and residues to total effluent loadings.
Unhairing wastes are the most highly concentrated, however,'and contain most
of the sulfide together with an appreciable amount of alkalinity.
The contaminants are so widely distributed throughout all of these pro-
cess waste fractions that no segment may be overlooked as insignificant. For
example, the pollutants generated per thousand pounds of equivalent green-
salted hides from soaking, unhairing and reliming each individually exceeds
the EPA's Best Practicable Effluent Limitations for subcategory one tanneries
(c.f. 40 CFR Chapter 1, Subchapter N: Part 425 Subpart A, 39 FR 12960,
Apr. 9, 1974) [since remanded] for pH, BOD, TSS and Oil & Grease. Bating
process wastes exceed BPCT limits for BOD, Oil & Grease, and TSS: picklinn
wastes exceed the pH limitations; and tanning wastes exceed the limits for
pH, 011 & Grease and total chromium. Every processing step, except pickling,
exceeds the "Best Available Technology Economically Achievable" (1983) limits
for total Kjeldahl nitrogen, while only the soaking and tanning fractions do
not exceed the 1983 limitations on sulfide.
Finally, the characteristics of the combined wastewater generated from
all tannery operations after blue piling (including blue stock wringing but
excluding beamhouse and tanyard wastes) are also depicted in Table 2. These
wastes which contribute roughly 40% to the total tannery effluent volume
Independently exceed 1977 limits for BOD, TSS, total chromium and 011 &
Grease and also exceed the 1983 limit for total Kjeldahl nitrogen. It is
obvious that while certain process waste fractions have been identified as
more highly contaminated than others, no one process discharge can be singled
out for treatment, recovery or reuse under the expectation of "eliminating"
a specific, major wastewater contaminant from the final effluent. As the
unhairing wastes and chrome tanning wastes are among the most highly contami-
nated, treatment of these fractions has received first priority. Sulfide
oxidation and chrome recovery are discussed later.
The Oil & Grease component of our total tannery effluent was further
characterized. Aliquots of a typical 24-hour effluent sample taken down-
stream from our rotary screen and composited 1n proportion to flow were
treated according to the procedure for Grease outlined in Standard Methods
ed. 13., Part 209A. The residues obtained were further analyzed by Infrared
spectroscopy and computerized gas chromatography/mass spectrometry. No
hydrocarbons of the type expected from petroleum products were present.
The extracted "Greases" were Identified as animal fats and/or vegetable oils.
12
-------
SECTION VI
CHROME RECOVERY FEASIBILITY STUDIES
A review of the literature shows that while the recovery and reuse of
chrome tanning agents 1s Indeed feasible, there 1s little laboratory or pilot
data available which can be used for designing chrome recovery systems.
The purpose of this study was to develop, through laboratory and pilot-
scale Investigations, pertinent design data for the unit processes and unit
operations employed In the recovery of chromium from spent chrome tanning
liquors.
LITERATURE REVIEW
The recovery and reuse of chromium tanning agents has been reported by
many Investigators. Harnly (2) reported the successful recycle of chrome
during World War II by precipitation with caustic soda followed by decanta-
tlon, acidification and restrennthenlng. Benrud (3) reported seven serial
reuses of chrome liquors by refortification. Miller (4) reported the reuse
of chrome over a period of six weeks through replenishment with dry chromium
salt, sulfurlc add and sodium chloride with no significant build-up of sul-
fates. Davis and Scroggle (5) also demonstrated on a pilot scale that several
commercially prepared chrome tans could be reused at least 13 times by float
adjustment and restrengthenlng. They reported that the recycle, resulting
in a 25% reduction in chrome usage, did not adversely affect either the sub-
jective or physical properties of the resultant leather.
Additional work by Davis and Scroggie (6) showed that the species of
chromium complexes present in tannino liquors, throughout the range of molecu-
lar weights and charge distributions as measured by gel filtration chromaton-
raphy, were taken up by hide. Exhaust liquors from these commercial chrome
tannages were found to exhibit the same distribution of chromium species as
the original chrome liquors.
According to Petruschke (7) the most promising alternative to reuse
through replenishment is precipitation with alkali followed by filtration,
washing and redissolving of the recovered chromium hydroxide. Hauck (8)
reported that calcium hydroxide, sodium hydroxide, ammonium hydroxide, sodium
bicarbonate and soda ash had been used to precipitate chromium hydroxide from
spent chrome; and that the selection of alkali should depend on a number of
factors including the optimum pH for precipitation. He also suggested that
a rotary vacuum drum filter might be effective in dewatering chromium
hydroxide sludges.
13
-------
Young (9) demonstrated that chromium hydroxide slurry could be concen-
trated by pressure filtration. To optimize this pressure filtration, 1t was
suggested that the sludge be thickened by settHnq prior to filtration, with
a flocculating agent added to enhance settling rates. An anionic polymer
appeared to give the best results.
',-,'eber (10) calculated that recovery of chrome by 1on exchange would be
prohibitively expensive because of the blockage of exchanoe columns and high
regeneration costs.
For precipitating chromium hydroxide from spent chrome liquors, Weber
concluded that Hme was the "best" alkali. He also noted that "mechanical
dehydration" was effective 1n dev/atering chromium hydroxide sludges.
Das et a]_. (11) used sodium carbonate for precipitatino chromium hydrox-
ide. The precipitate was filtered and the filter cake dissolved in hot sul-
furic acid. The basicity of the reacidlfied chrome was then adjusted for
subsequent reuse. Abramovich (12) reported that "fresh" calcium hydroxide
was more effective in precipitating chromium hydroxide than were alkaline
beamhouse effluents. Sodium hydroxide and ammonium hydroxide were used by
Bianchi (13) to precipitate chromium salts from spent chrome solutions.
After filtration, the precipitate was washed and boiled in dilute sodium hy-
droxide followed by dilution and a second filtration. The recovered chromium
was dissolved in sulfurlc acid and used in the preparation of new tanning
baths.
The literature review leads the authors to conclude that the current
practice of chromium recovery from spent chrome tanning liquors would benefit
Immensely by a systematic evaluation of the design parameters applicable to
chrome recovery systems.
It has been demonstrated, for example, that several alkalis can be used
successfully to precipitate chromium hydroxide from spent chrome tanning
liquor, however, little is known about the effect of alkali type on the unit
processes and unit operations employed. Likewise, little is known about the
effect of alkali type on the flocculatlon and settling of chromium hydroxide.
It is suggested 1n the literature review that vacuum filtration might be
used to dewater precipitated chromium hydroxide, however, filtration rates,
which are essential for sizing such equipment have not been reported.
Thus, the major objective of this phase of the project was to investi-
gate the unit processes and unit operations which could be Incorporated 1n
a chrome recovery scheme; and to establish through conventional laboratory
and pilot-scale procedures representative data for the design of a chrome
recovery system.
Typical values of significant waste parameters for the spent chrome
liquor employed in this study are presented in Table 3. These values repre-
sent the concentrated spent chrome waste fractions and do not include subse-
quent wash waters.
14
-------
In practice, washes are commonly used after tannlnn to remove Insoluble
matter such as bits of flesh and fats and nreases, as well as soluble salts
from the surface of the tanned hides. The wash water volume may be as much
as five times that of the original float volume. The deslqn of any chromium
treatment or recovery process must take Into account this variation 1n chrome
concentration of the wastes generated, I.e., concentrated liquors followed by
large volumes of dilute wash waters.
TABLE 3. TYPICAL ANALYSIS OF CONCENTRATED SPENT CHROME
Parameter Concentration, mg/1*
Chromium (as Cr) 6,260
Acidity (as CaCOa) 15,000
pH 3.30
011 & Grease 6,310
Chlorides (as CT) 18,200
BOD-5 (20°C) 4,920
COD 18,900
Total Sol Ids 116,200
Total Fixed Solids 92,220
Total Suspended Sol Ids 3,760
Fixed Suspended Sol Ids 840
Total Kjeldahl N (as N) 1,110
* Except pH which 1s unltless.
PRECIPITATION OF CHROMIUM HYDROXIDE
Chromium 1s a transition element, and 1n Its 3+ oxidation state shows a
strong tendency to complex with a wide variety of Ugands to form coordination
compounds - most of them highly colored. Amonq the simplest of the Cr3+ com-
plexes Is CrfHgO)3,'1" or hexaquochromlum (III). This violet colored complex 1s
a weak add which readily undergoes hydrolysis. Hydrolysis proceeds as the
acidity of a solution of Cr(H20)g"f 1s decreased until above pH 5, the grey-
green, hydrated Cr3* hydroxide precipitates.
[Cr(H20)6]3+ + H20 -
[CrOH(H20)5f+ + H20 =
[Cr(OH)2(H20)4r + H20 - [Cr(OH)3(H20)3](s)
Commonly the third hydrolysis product 1s denoted as Cr(OH)3 (chromium hydrox-
ide) or Cr203 (chromic oxide).
The precipitation of hydrated Cr3+ as the hydroxide is usually expressed
by the simpler notation:
Cr3+ + 30H- = Cr(OH)s(s)
15
-------
The solubility product, KSD, for chromium hydroxide 1s reported by Hognes
ejt al_. (14) to be: H
» 7xlO'31
•J
i 3+
The degree of hydrolysis of [Cr(H20)(J 1s termed basicity. Basicity
is a ratio calculated from analytical values for the chromium concentration
and the tltratable acidity. According to the American Leather Chemists
Association Method CIO, basicity 1s calculated from the following formula:
A - B
Basicity (percent) * Kx 100
where
A = milHequlvalents (me) of Cr?03 per liter
(ALCA Method Cl)
B = tltratable acidity 1n mllHequivalents
per liter (ALCA Method C5)
Thus, theoretically, the acidic species [Cr(H20)s] has a basicity of
0% and the precipitated Cr(QH)3(H20)3 has a basicity of 100%. Between these
extremes, any degree of basicity Is attainable.
In "pure" solutions, the concentration of Cr3"*" is a function of the
hydroxyl 1on concentration and therefore pH. At acidic and neutral plls the
solubility of Cr3+ 1s attributable to a common 1on effect. THvalent chro-
mium, however, 1s amphoterlc so that Cr{OH)3 dissolves 1n the presence of
excess hydroxyl 1on to form anlonic chromium complexes.
The optimum pH for the precipitation of chromium hydroxide from spent
chrome tanning liquor was demonstrated by varying dosages of sodium hydroxide
added to a concentrated spent chrome solution.
The standard jar test procedure was followed using a six-place variable
speed mixing apparatus and six 1000 ml Pyrex beakers. To each of the six
beakers 500 ml of spent chrome at a concentration of 5800 mg/1 Cr was added.
The acidity of this solution was 300 me/1. A theoretical dosage of 150 me
of sodium hydroxide was required to precipitate the chromium hydroxide
stoichiometrlcally. Therefore, the dosages of 2N sodium hydroxide added for
neutralization were varied from 110 me to 190 me corresponding to a range of
832-125% of the tltratable acidity as determined by the ALCA Method C5.
After the alkali addition, the six samples were diluted to final volumes of
600 ml. The samples were mixed at 100 rpm for two minutes followed by an
additional 20 minutes of slow mixing at 30 rpm and then were allowed to settle
under quiescent conditions.
There was no noticeable settling within the first four hours. After
24 hours, however, the precipitated chromium hydroxide had settled enough
to withdraw a sample of supernatant for analysis. The effect of alkali
dosage on pH, chromium concentration of the supernatant and on sludge volume
1s illustrated in Table 4.
16
-------
The pH values from 8.55 to 9.70 He within the expected optimum range
for the precipitation of chromium hydroxide. Notably, the correspondlnq
alkali dosages ranging from 91% to 117% of the stolchiometric requirement
for 100% basicity yielded supernatant chromium concentrations which did not
differ significantly.
To determine the relative effectiveness of different alkalis to precipi-
tate chromium hydroxide from spent tanning liquors the same procedure used
for caustic soda was repeated using hydrated Hrne, aqueous ammonia and soda
ash, respectively. The comparative effectiveness of these alkalis 1s Illus-
trated 1n Figure 2.
The calculated amounts of calcium hydroxide were suspended in 100 ml of
distilled water and added to each of six 500 ml spent chrome samples followed
by five minutes of rapid mixing at 100 rpm and 20 minutes of slow mixing at
30 rpm. Alkali dosages were varied from 85% to 140% of the stoichlometric
equivalent required for 100% basicity. To one sample, after the addition of
lime and mixing, 30 mg/1 of an anionic liquid polymer, Nalcolyte 677, was
added as a 0.1% solution followed by a five minute rapid mix and 20 minutes
of slow mixing.
TABLE 4. EFFECT OF flaOH DOSAGE ON PRECIPITATION OF CHROMIUM HYDROXIDE
NaOH
Dosage, me/1
250
275
300
325
350
375
% NaOH
83
91
100
108
117
125
pH
8.25
8.55
8.85
9.38
9.70
10.10
Supernatant
Cr,mn/l
4.74
2.44
2.74
2.50
2.38
5.35
Sludne.ml
420
430
450
440
435
420
The spent chrome samples used 1n the lime-precipitation study initially
contained 6690 mg/1 of chromium (385.5 me/1) and had an acidity of 343 me/1.
A Ca(OH)2 dosage of 343 me/1, therefore, was required to increase the basic-
ity to 100%.
The effect of hydrated lime on the precipitation of chromium hydroxide
is summarized in Table 5.
Chromium hydroxide sludge settled faster with hydrated Hrne when other
precipitating agents were used. The sludge volume was smaller and the con-
centration of chromium in the supernatant liquid was lower. The lowest
supernatant chromium concentration obtained after 24 hours of quiescent
settling was 0.10 mg/1 when 140% alkali was used. The supernatant chromium
concentration at 100% alkali was 0.90 mg/1. The Improved separation at the
higher alkali level was achieved, however, at the cost of Increasing the
sludge volume by nearly 10%.
17
-------
10.0
10.5
Figure 2: EFFECT OF pH ON SUPERNATANT CHROMIUM
CONCENTRATIONS.
-------
The polymer addition dramatically improved the settling of the suspended
chromium hydroxide such that a final effluent concentration of 0.16 mg/1 Cr
after 24 hours could be expected with 100% alkali addition.
TABLE 5. EFFECT OF Ca(OH)2 DOSAGE ON PRECIPITATION
OF CHROMIUM HYDROXIDE - PART I
Ca(OH)2
Dos age, me/ 1
209
274
343
404
470
343*
% Alkali
60
80
100
120
140
100
PH
6.45
7.68
8.55
8.98
9.48
8.45
Supernatant
Cr,mg/l
1,100
414
0.96
0.50
0.10
0.16
SI udge ,ml
250
300
345
375
375
350
* 100% alkali dosage plus 80 mg/1 of Nalcolyte 677
From Figure 2, it is apparent that the pH range employed in lime-
precipitation did not demonstrate the increased solubility of chromium at
higher pH values predicted from the amphoteric nature of these chrome salts.
Therefore, a second series of lime-precipitation runs was conducted using
Ca(OH)2 dosages ranging from 85% to 350% of the stoichiometric requirement
for precipitation. The spent chrome employed in these latter runs which also
contained 6,690 mg/1 of chromium but an acidity of 265 me/1 was screened
(0.040" openings) and skimmed to effect an approximate 90% reduction in
Oil & Grease. As shown in Table 6, after 24 hours of quiescent settling,
supernatant chromium levels ranged from 37.8 mg/1 at pH 7.3 to 0.10 mg/1 at
pH 11.9. The chromium solubility increased to 0.24 mg/1 at pH 12.7. The
optimum pH for lime-precipitation was shown to lie within the 8.7 to 12.7
range.
When aqueous ammonia was used as the precipitating agent in the range
of 73% to 125% of the stoichiometric requirement, the optimum precipitation
of the Cr (OH)3 occurred at a pH of 7.82. The chromium concentration in the
supernatant liquid after 24 hours of quiescent settling was 8.84 mg/1.
TABLE 6. EFFECT OF Ca(OH)2 DOSAGE ON PRECIPITATION
OF CHROMIUM HYDROXIDE - PART II-
Ca(OH)2
Dosage, me/1
224
264
330
396
660
924
% Alkali
85
100
125
150
250
350
PH
7.30
7.58
8.11
8.70
11.90
12.70
Supernatant
Cr.mg/1
37.8
11.3
1.30
0.59
0.10
0.24
SI udge, ml
320
300
320
400
430
430
19
-------
The difference observed 1n the optimum pH.for caustic soda and ammonium
hydroxide probably results from the formation of soluble chromium-ammonia
complexes at the higher pH values.
The spent chrome samples contained 5,800 mg/1 of chromium and had an
acidity of 300 me/1. The alkali dosage therefore required to raise the
basicity to 100% was 300 me/1. The results of the ammonium hydroxide precip-
itation of chromium hydroxide from spent chrome tanning liquor are shown in
Table 7.
TABLE 7. EFFECT OF NfyOH DOSAGE ON PRECIPITATION
OF CHROMIUM HYDROXIDE
NH40H
Dosage, me/ 1
220
260
300
340
380
% Alkali
73
87
100
113
126
pH
6.38
7.40
7.82
8.10
8.28
Supernatant
Cr,mg/l
137
11.5
8.84
9.98
11.9
Sludge, ml
410
420
435
440
440
Sodium carbonate was demonstrated to be the least effective precipita-
ting agent. The spent chrome used for precipitation by soda ash contained
6,180 mg/1 of chromium and had an acidity of 298 me/1. Dosages of Na2C03
were varied from 75% to 137% of the alkali requirement. As shown 1n Table B,
the optimum dosage was between 87% and 100% alkali, corresponding to superna-
tant pH's of 8.30-8.48. The supernatant chromium concentration within this
dosage range varied from 40.5 mg/1 to 41.5 mg/1,
TABLE 8. EFFECT OF ^Oh DOSAGE ON PRECIPITATION
OF CHROMIUM HYDROXIDE
NAgCOs
Dosage, me/1
224
260
298
335
372
410
% Alkali
75
87
100
112
125
137
pH
8.12
8.30
8.48
8.60
8.75
8.91
Supernatant
Cr,mg/l
77.5
41.5
40.5
45.6
46.2
56.8
Sludge, ml
400
410
440
450
450
445
Several important conclusions can be drawn regarding the precipitation
of chromium hydroxide from spent chrome tanning liquors. These studies demon-
strated that solubility is highly dependent on pH, and that the optimum pH
varies depending on the alkali used. Interestingly, also we found that the
optimum dosage for each of the alkalis used was 100% of the acidity as mea-
sured by the ALCA Method C5.
20
-------
An additional finding was that the type of alkali used to precipitate
the chromium hydroxide has a significant effect on the concentration of Cr3+
found in the supernatant liquor. The order of effectiveness for removal of
chromium hydroxide was determined to be calcium hydroxide> sodium hydroxide>
ammonium hydroxide> sodium carbonate.
Finally, Nalcolyte 677 proved to be an effective flocculating agent
especially when hydrated lime was used for precipitation. This polymer
effected immediate flocculation and substantially faster settling as well
as improved supernatant clarity and lower supernatant chromium concentrations
after 24 hours. The polymer only slightly improved settling characteristics
when alkalis other than lime were used. The effectiveness of. lime as a coag-
ulant is probably related to its ability to provide a nucleus of high charge-
to-area ratio within the pH range employed. Also, a denser floe is probably
formed because of the co-precipitation of insoluble calcium sulfate.
COAGULATION AND FLOCCULATION OF CHROMIUM HYDROXIDE
Under optimum conditions for the precipitation of chromium hydroxide
from spent chrome liquors, an extremely fine, slow settling, amorphous,
blue-green precipitate is formed. Factors contributing to the slow rate
of settling probably include repulsive, electrostatic forces at the surface
of the chromium hydroxide particle, and the high affinity of CrfOH), for
water. Thus, aggregation into dense, settleable particles is inhibited.
Since the chemistry of Al3-f is in many ways similar to that of Cr3+ and
since small amounts of aluminum would not interfere with anticipated reuses
of the chrome, it was hypothesized that aluminum sulfate was the best coagu-
lant that might be used to "destabilize" the chromium hydroxide suspension.
Further, trivalent aluminum, in accordance with the Schultz-Hardy Rule (15)
was expected to be a far more effective coagulant than a mono or divalent ion.
Studies using a coagulant to improve suspended solids removal were con-
ducted. Six 500 ml samples containing 5,800 mg/1 of chromium were neutral-
ized with 2,N NaOH to pi I 9.0. The samples were mixed at 100 rpm for five
minutes then 20 minutes at 30 rpm. Varying amounts of a fresh solution of
A!2(504)3 - 18 H20 containing 5,000 mg/1 of A13+ were added and the pH was
re-adjusted to 9.0 with sodium hydroxide. After adjusting the volume to
600 ml, the samples were rapidly mixed for 5 minutes, slow-mixed for another
20 minutes and then allowed to settle. Al^+ concentrations up to 500 mg/1
gave no significant change in the rate of settling of the precipitated
chromium hydroxide.
Further attempts to agglomerate the chromium hydroxide were made using
two anionic polyelectrolytes (Nalcolyte 677 and Rohm and Haas A-10) and a
cationic polymer (Buckman Bubond 60). Addition of 80 mg/1 of Nalcolyte 677
to sodium hydroxide-precipitated chromium hydroxide at pH 9.0, definitely
improved the rate of settling, whereas the other polymers did not produce
a noticeable change in the settling rate. The effect of the addition of
Nalcolyte 677 is illustrated in Figure 3. Slighjtly improved floe formation
resulted in significantly improved settling.
21
-------
ONO POLYELECTROIYTE
D80mg/l NALCOLYTE 677
350
2 4 6 8 10 12 14 16 18 20 22 24
TIME, hr
Figure 3: EFFECT OF NALCOLYTE 677 ON THE
SETTLING OF NaOH-PRECIPITATED
CHROMIUM HYDROXIDE.
22
-------
350
ONO POLYELECTROLYTE
D80mg/l NALCOLYTE 677
30 45
TIME, min
60
Figure 4: EFFECT OF NALCOLYTE 677 ON THE
SETTLING OF Ca(OH)2-PRECIPITATED
CHROMIUM HYDROXIDE.
23
-------
The addition of polyelectrolytes to the ammonium hydroxide, or soda ash-
precipitated chromium hydroxide did not Improve the settleablHty of these
sludges. By contrast, the addition of 80 mg/1 of polyelectrolyte to lime-
precipitated chrome resulted 1n the Immediate formation of a large, sturdy,
rapidly settling floe. The effect of Nalcolyte 677 on the settling of Hme-
predpltated chromium hydroxide 1s shown 1n Figure 4.
The optimum dosage of Nalcolyte 677 for the coagulation of a spent
chrome Hqufr containing approximately 5,000 mg/1 of Cr was determined. A
calcium hydroxide dosage which gave 100% basicity was added to six 500 ml
allquots of chrome and the volumes were adjusted to 600 ml with distilled
water. After 5 minutes of rapid-mixing and 20 minutes of slow-mixing, vary-
ing amounts of a 0.5% solution of Nalcolyte 677 were added. The effect of
polymer dosage on settling is shown 1n Table 9.
The optimum dosage of Nalcolyte 677 for the flocculation of lime-
precipitated chromium hydroxide at pH 9.0 and a chromium concentration of
4,830 mg/1 was found to be 20-30 mg/1.
TABLE 9. EFFECT OF VARYING DOSAGES OF NALCOLYTE 677 ON THE
SETTLING OF LIME-PRECIPITATED CHROMIUM HYDROXIDE
Sludge Volume
Nalcolyte 677 Dosage, mg/1 After 15 Minutes, ml
10 480
20 425
30 390
40 390
60 390
80 390
24
-------
FLOTATION OF CHROMIUM HYDROXIDE
Flotation, a unit operation, commonly used to separate suspended solids
from a liquid phase (16), is accomplished by introducing fine bubbles of air
into the waste to be treated. As bubbles adsorb on the surface of particu-
late matter, air-solid c^nglomsrates are formed which have a net specific
gravity less then that of the liquid phase. This net buoyant force causes
the solid particles to float to the surface where the scum collects and is
removed by skimming.
The most commnn type of flotation for suspended solids separation uses
dissolved air. Air 1s dissolved in the waste by subjecting a fraction of
recycled clarified effluent to extreme turbulence under a pressure of several
atmospheres. The "aerated" recycled waste is then mixed with raw waste at
atmospheric pressure. Under this "reduced" pressure, the mixture is super-
saturated with air which 1s released from the mixture in the form of minute
bubbles. The rising bubbles attach to solid matter and carry the suspended
material to the surface of the flotation chamber while clarified effluent
is withdrawn from the bottom of the chamber.
Dissolved air flotation 1s claimed to offer several advantages over
conventional solids separation and thickening equipment. Some of the advan-
tages Include:
1) Reduced space requirements
2) Lower capital Investment
3) More concentrated sludge
4) Higher surface loading rates
In general, however, when suspended sol Ids levels exceed 1%, other separation
methods will likely be more economical (17).
Dissolved air flotation was investigated as an alternative to sedimenta-
tion for the thickening of chromium hydroxide sludges precipitated from spent
chrome tanning liquor. The experiment was designed to establish the follow-
ing pertinent design parameters:
1) Flotability of chromium hydroxide
2) Rise rates
3) A1r-to-solids ratio
4) Recycle, ratio
5) Retention time
Batch flotation tests were conducted using the apparatus shown schemati-
cally in Figure 5. Chromium hydroxide sludge was placed in the flotation
25
-------
COMPRESSED AIR
PRESSURE REGULATING VALVE
PRESSURE GUAGE
/
±® ^
PRESSURE TANK
AIR
WATER
RELIEF
VALVE
--FLOTATION
COLUMN
Figure 5: DISSOLVED AIR FLOTATION
APPARATUS.
26
-------
chamber which consisted of an acrylic column having a five inch inside diam-
eter and a height of four feet. An injection port was located two inches
from the bottom of the column to introduce pressurized recycle water. A
thirty gallon pressure tank was employed for the aeration of recycle water.
Compressed air was injected into the bottom of the pressure tank and continu-
ously bled from the top, maintaining a violent turbulent condition within the
tank. Pressure within the tank was maintained at 50 psi with a pressure
regulating valve located on top of the tank.
In all runs, tap water was employed as the recycle fraction, and was pre-
aerated for a minimum of twenty minutes to insure that the dissolved air con-
centration approached saturation. The injection of aerated recycle into the
flotation column was regulated to give thorough mixing of the two fractions
within one to two minutes.
Twelve inches (one gallon) of a soda-ash precipitated chromium hydroxide
sludge containing 4,520 mg/1 of Cr3+ at an optimum pH of 8.5 was placed in
the flotation column. The addition of 150% of recycle failed to effect any
flotation. In subsequent runs, the addition of up to 600% recycle to one-
half gallon of sludge failed to produce an acceptable flotation of the soda-
ash precipitated chromium hydroxide.
Similarly, a caustic soda-precipitated sample containing 5,100 mg/1 of
Cr3+ at an optimum pH of 8.8 was tested. Addition of up to 1,000% recycle
failed to effect flotation. Hhile flocculation could be enhanced slightly
by the addition of flocculants, there was no change in the flotation results
at the 10 to 1 recycle ratio.
According to Eckenfelder and O'Connor (18), tap water saturated under
50 psi at room temperature has the potential to release 5.8 x 10~4 pounds of
air per gallon when returned to atmospheric pressure. Assuming that the air
release was equivalent to that demonstrated by Eckenfelder, the caustic soda-
precipitated chromium hydroxide which contained 28,000 mg/1 of suspended
solids was subjected to an air-to-solids ratio of:
A = 5.8 x 10-4 Ib/gal x 2.5 gal
B 28,000 mg/1 x 2.21 x IQ-o ib/mg x 3.78 1/gal x 0.25 gal
A = 0.025 Ib air/lb solids
If
Therefore, if the dissolved air flotation of caustic soda-precipitated
chromium hydroxide in spent chrome tanning liquors is possible, air-to-solids
ratios in excess of 0.025 Ib air/lb solids must be employed. At a recycle
ratio of 1,000%, however, the design requirements are prohibitive. It was
concluded, therefore, that dissolved-air flotation of chromium hydroxide is
not a viable alternative for solids separation in the recovery of chromium
from spent tanning liquors.
A second alternative for the flotation of phromium hydroxide from spent
chrome tanning liquors employs in situ generation of carbon dioxide gas.
27
-------
A 200 ml sample of spent chrome containing 2500 mg/1 of chromium was
adjusted to pH 8.7 with 2N soda ash. To this mixture 150 mg/1 of Nalcolyte
677 was added. The addition of aluminum sulfate equivalent to 350 mg/1 of
A]3+. lowered the pfl to 7.0. The mixture effervesced and flotation of chro-
mium hydroxide occurred. Flocculation was excellent, althouqh the removal
efficiency was judged to be poor as the underflow retained the green tint
characteristic of hydrated, trivalent chromium salt solutions.
The bubbles of carbon dioxide generated during the addition of an
acidic aluminum sulfate solution apparently adhere to the floe particles
in the same manner as in dissolved-air flotation.
Finally, a third alternative for the flotation of chromium hydroxide
by precipitate flotation has been demonstrated by Bhattacharyya ejt aj_. (19)
Chromium removals of 97% efficiency from chromium (III) hydroxide suspensions
containing 0.93 minimoles of Cr per liter (48 mg/1 Cr) were obtained by the
addition of sodium dodecylsulfate, an anionic surfactant, at a mole ratio of
surfactant-to-chromium of 1 to 10.
While these results indicate that precipitate flotation of chromium
hydroxide from extremely dilute spent chrome by an anionic surfactant such
as sodium dodecylsulfate may be possible, the high concentrations of chromium
present in "normal" spent chrome liquors (commonly 5000-7000 mg/1) prevent
direct comparison.
SETTLING OF CHROMIUM HYDROXIDE
Laboratory studies proved conclusively that calcium hydroxide used in
conjunction with Nalcolyte 677 gave the most efficient precipitation of
chromium from spent tanning liquors.
To produce reliable data, useful for the design of a basin for settling
lime-precipitated chromium hydroxide, further investigations were carried
out 1n a 5 Inch diameter x 48 inch high acrylic column. The settling experi-
ments were designed to determine the effect of chromium and suspended solids
concentration on the design overflow rate for clarification and sludge feed
rate for thickening.
For each run an appropriate volume of concentrated spent chrome was
diluted to 3.0 gallons and the temperature was adjusted to 90°F. The pH of
the solution was raised to 9.0 with hydrated lime. The chromium hydroxide
slurry was then transferred to the acrylic column. Nalcolyte 677 was added
by injecting the polymer through a 1/4" glass tube to the bottom of the
acrylic column at a dosage of 50 mg/1. Air was injected through the same
tube for two minutes to effect a rapid-mixing. Slow-mixing was achieved by
gently stirring the mixture for another 5 minutes. At this point, the liquid
height in the acrylic column was slightly greater than 30 inches. While stir-
ring, with the mixture fully suspended, a sample was withdrawn from a sampling
port 12 inches from the bottom of the column, and the liquid was drained to
the 30 inch level. The floe was allowed to settle under quiescent conditions.
After 90 minutes, a sample of the supernatant liquid was withdrawn from a port
28
-------
10 20
30 40 50 60
TIME, min
80 90
Figure 6: SETTLING OF LIME-PRECIPITATED
CHROMIUM HYDROXIDE.
29
-------
CO
Q
_J
CO
Q
LU
Q
Z
UJ
CL
CO
ZD
CO
50000
40000H
30000
20000
10000
0
1000 2000 3000 4000 5000 6000 7000
CHROMIUM, mg/l
Figure 7: SUSPENDED SOLIDS VS CHROMIUM
FOR LIME-PRECIPITATED
CHROMIUM HYDROXIDE.
30
-------
3a&-Co=49,200mg/l
Qu=87,000mg/l
tu=67min
40 50 60 70 80 90
TIME, min
Figure 8: GRAPHICAL DETERMINATION OF THE
RATE OF SUBSIDENCE AND
THICKENER RETENTION TIME.
31
-------
at the 24 inch level. Likewise, a sample of the sludge was taken after
decanting the supernatant liquid. All three of these samples were analyzed
for chromium and for suspended solids.
The dependence of rate of settling on suspended solids concentration and
therefore on chromium concentration was clearly demonstrated. As shown in
Figure 6, an increase in suspended solids concentration from 3,180 mg/1 to
49,200 mg/1, which corresponds to chromium concentrations of 650 mg/1 and
6,590 mg/1, respectively, effected a substantial decrease in the settling
rate of the chromium hydroxide.
Settling in the most dilute chrome sample is best described as discrete
particle settling. The interface between the sludge and supernatant liquid
was poorly defined until the floe had subsided well below the 15 inch level.
The second most dilute sample also resembled discrete particle settling,
although a well defined interface did develop at a column height of about
25 inches. More concentrated samples formed a well defined interface between
supernatant and sludge which is characteristic of the hindered type of zone
settling.
Analysis of the samples of the feed slurries, supernatant liquids and
sludges for each of the settling runs yielded the concentrations of chromium
and suspended solids shown in Table 10.
These settling tests revealed that dilution of spent chrome liquor
before lime-precipitation and settling results in a decrease in the suspended
solids and chromium concentrations of the sludge. It should be noted, how-
ever, that as a result of this dilution the ratio of suspended solids to
chromium found in both the precipitated feed slurry and in the sludge also
decreased significantly. As shown in Figure 7, the plotting of suspended
solids vs. chromium yields a non-linear curve. Thus, while the concentration
of chromium found in the feed-precipitate decreased from 6,590 mg/1 to
650 mg/1, the suspended solids decreased from 49,200 mg/1 to 3,180 mg/1 for
a 35% decrease in the suspended solids-to-chromium ratio from 7.47 to 4.90.
The areas required for the clarification and thickening of lime-
precipitated chromium hydroxide were determined by graphical analysis of
the settling curves in Figure 6 as described by Rich (20). Figure 8 demon-
strates through this graphical technique that an underflow concentration (Cu)
of 87,400 mg/1 could be achieved with a feed suspended solids concentration
of 49,200 mg/1 and a retention time of 67 minutes. Therefore, for every
1,000 gallons of daily waste flow the area required for thickening would be:
A - ^u - (0.0928 cfm) (67 min) _ ? ^ 2
_____ -..sun
32
-------
where:
A = area 1n ft2
q = flow 1n cfm
tu= retention time in minutes
Z0= depth of column in feet
A thickener to concentrate the slurry to a sludge containing 87,400 mg/1,
therefore, would require a surface area of 2.50 ft2 per 1,000 gallons/day.
From Figure 8 the initial rate of subsidence was calculated by determin-
ing the slope of the settling curve near time zero. At a feed suspended
solids concentration of 49,200 mg/1 an initial rate of subsidence of 0.0252
ft/min was observed. Therefore, for every 1,000 gallons of daily waste flow
the area required for clarification would be:
A = go = (437 gal/day)
u (1,440 min/day) (7.48 gal/ft3) (0.0252 ft/min)
A = 1.61 ft2
where:
A = surface area in ft2
q0= liquid overflow rate in cfm
u = rate of subsidence in ft/min
The area required for clarification was significantly less than that
required for thickening. The sludge obtained from clarification should,
therefore, be considerably less concentrated than that obtained from thicken-
ing. Maximum concentration is important both from the standpoint of optimiz-
ing sludge dewatering equipment and subsequent reprocessing of the chromium
hydroxide. Thus, the area determined for sludge thickening was used for the
design of a settling basin to concentrate chromium hydroxide.
Surface area requirements for each of the six dilutions are illustrated
in Table 11 along with other relevant design parameters.
The decreases in chromium concentration by dilution effected a marked
increase in the rate of subsidence of the precipitated chromium hydroxide.
A tenfold decrease in the chromium concentration from 6,590 mg/1 to 650 mg/1
resulted in a fiftyfold increase in the settling rate of the resulting
33
-------
CO
TABLE 10. CHROMIUM AND SUSPENDED SOLIDS CONCENTRATIONS FOR THE
SETTLING OF LIME-PRECIPITATED CHROMIUM HYDROXIDE
Cr
6,590
4,650
3,490
2,230
1,330
650
Feed
SS
49,200
33,340
21,700
12,020
7,160
3,180
SI
Cr
12,700
9,700
7,700
5,580
5,040
4,610
udge
SS
87,400
64,120
44,500
31,600
25,220
20,960
Supernatant
Cr
3.74
4.50
2.98
0.84
4.36
1.96
SS
151
120
126
162
114
129
All units mg/1.
-------
u>
en
TABLE 11. SURFACE AREA REQUIREMENTS FOR THE THICKENING AND CLARIFICATION
OF LIME-PRECIPITATED CHROMIUM HYDROXIDE AT VARIOUS CHROMIUM AND
SUSPENDED SOLIDS CONCENTRATIONS (PER 1000 GALLONS OF DAILY WASTE
FLOW)
Cr> mg/1
6,590
4,650
3,490
2,230
1,130
650
SS, mg/1
49,200
33,340
21,700
12,020
7,160
3,180
tu, m1n
67
44
30
16
14
8.5
q0» cfm
0.0475
0.0516
0.0546
0.0646
0.0736
0.0895
clarification
u,ft/min area, ft2
0.0252
0.0549
0.0985
0.390
0.568
1.04
1.61
0.940
0.565
0.168
0.130
0.082
thickening
area, ft2
2.50
1.64
1.12
0.60
0.52
0.32
-------
chromium hydroxide. Likewise, the same decrease in chromium concentration
effected an 88% decrease in the design sludge thickening area requirement,
from 2.50 ft2 to 0.32 ft2 for each 1,000 gallons per day of feed volume.
The significance of these thickening results are more obvious when
expressed in terms of volumetric loadings in gallons of feed per day per
square foot of thickener surface area. As shown in Table 12, the design
loading 1s Increased from 400 gpd/ft2 to 3,120.gpd/ft2 when the chromium
concentration of the feed slurry is decreased from 6,590 mg/1 to 650 mg/1.
This effect of chromium concentration on the design volumetric loading
of a sludge thickening unit is illustrated in Figure 9.
The single most important parameter for the design of a chromium hydrox-
ide sludge thickening unit is the suspended solids loading. The relationship
found to exist between suspended solids concentration and the design volu-
metric loading for chromium hydroxide sludge thickening is shown in Figure
10. Likewise, the relationship between feed suspended solids, retention time
and underflow concentration are shown 1n Figure 11 and Figure 12.
From the practical point of view of collecting and treating spent chrome
tanning liquors, the dilutions employed 1n these settling studies represent
concentrations which can be found in a typical spent chrome discharge and the
discharge of final wash waters. The Influence of the dilution of a spent
chrome liquor on several parameters Including suspended solids concentration,
design volumetric loading, total feed solids, and unit area (thickener sur-
face area required per ton of solids) 1s illustrated 1n Table 13.
The effect of the increased volume due to the dilution of spent chrome
liquor upon design surface area is offset by the observed increase in set-
tling rates. These data reveal that the net change .in surface area require-
ment is related to the decrease 1n total feed solids with increased dilution.
Within the hindered settling-suspended solids range as chromium was diluted
from 6,590 mg/1 to 2,230 mg/1, a decrease in total feed solids of approxi-
mately 23% from 408 Ib/day to 295 Ib/day was noted (Figure 13) while the unit
area remained constant at an average of 12.1 ft2/ton-day. The net effect of
this dilution was a 29% decrease in the surface area requirement for settling
from 2.50 ft2 for 1,000 gpd at a chromium concentration of 6,590 mg/1 to
1.77 ft2 for approximately 3,000 gpd at a chromium concentration of 2,230
mg/1.
VACUUM FILTRATION OF CHROMIUM HYDROXIDE
In tannery practice, chrome liquors are commonly used at stock concentra-
tions ranging from about 1.0 to 2.0 pounds of 0^03 per gallon (0.68 to
1.37 pounds of Cr per gallon). The highest concentration of lime-precipitated
chromium hydroxide sludge demonstrated after bench-scale thickening contained
only 0.106 pounds of Cr per gallon, whereas the most dilute contained only
0.038 pounds per gallon. Sludge concentrations using other alkalis for pre-
cipitation were even lower. Further concentration is necessary for subse-
quent processing of the recovered chromium salt.
36
-------
3500-
1000
UJ 500
Q
0
1000 2000 3000 4000 5000 6000
CHROMIUM, mg/l
7000
Figure 9: EFFECT OF CHROMIUM
CONCENTRATION ON DESIGN
VOLUMETRIC LOADING.
37
-------
3500
Q.3000
CD
0
5
<
o
cc
h-
111
O
2500
2000
1500
1000
w
g 500
0
O
10000 20000 30000 40000
SUSPENDED SOLIDS, mg/l
50000
Figure 10: EFFECT OF SUSPENDED SOLIDS
CONCENTRATION ON DESIGN
VOLUMETRIC LOADING.
38
-------
80-
c
E
*? 40-
10000 20000 30000 40000 50000
SUSPENDED SOLIDS, mg/l
Figure 11: EFFECT OF SUSPENDED SOLIDS
CONCENTRATION ON RETENTION
TIME.
39
-------
900001-
80000
70000
_ 6000°
^
o>
E
3 50000
O
40000
30000
20000
10000
0
10000 20000 30000 40000 50000
FEED SUSPENDED SOLIDS, mg/l
Figure 12: EFFECT OF FEED SUSPENDED
SOLIDS CONCENTRATION ON
UNDERFLOW CONCENTRATION.
40
-------
251
201
05
"O
i
C
o
*-»
•^
CM
LU
QC
<
h;
z
D
151
o1
-e-
o
10000 20000 30000 40000
SUSPENDED SOLIDS, mg/l
50000
Figure 13: EFFECT OF SUSPENDED SOLIDS
ON UNIT AREA.
41
-------
f\S
TABLE 12. EFFECT OF CHANGE IN CHROMIUM CONCENTRATION ON THE VOLUMETRIC
SURFACE LOADING OF A THICKENER DESIGNED TO CONCENTRATE
LIME-PRECIPITATED CHROMIUM HYDROXIDE
Chromium, mg/1
6,590
4,650
3,490
2,230
1,130
650
Area*, ft2
2.50
1.64
1.12
0,60
0.52
0,32
Volumetric loading,
gpd/ft?
400
61Q
893
1,670
1.920
3,120
* Area per 1000 gal/day of daily waste flow
-------
TABLE 13. EFFECT OF DILUTION ON SEVERAL SIGNIFICANT PARAMETERS FOR
THE DESIGN OF A CHROMIUM HYDROXIDE SLUDGE THICKENER
Volume,
9Pd
1,000
1,420
1,390
2,950
5,820
10,100
Chromium
mg/1
6,590
4,650
3,490
2,230
1,130
650
SS, mg/1
49,200
33,340
21,700
12,020
7,160
3,180
Design
Volumetric
Loadinq
gpd/ft2
4^0
610
8^3
1,670
1,920
3,120
Area, ft2
2.50
2.33
2.12
1.77
3.03
3.24
Total Feed
SS.lb/day
408
394
342
295
347
267
Unit n.rea,
ft2/ton-day
12.2
11.8
12.4
12.0
17.5
24.4
-------
Dewatering of chromium hydroxide sludrjes by filtration has been reported
by several researchers (8, 9, 11, 13). Pressure filtration offers the
advantage of high pressure differentials, while vacuum filtration is more
amenable to continuous automated operation and can usually be employed at
lower capital costs.
The feasibility of vacuum filtration of chromium hydroxide sludges was
demonstrated through a series of bench-scale laboratory tests. Initially,
Buchner funnel tests were carried out on a spent chrome liquor precipitated
with several different alkalis. The apparatus employed in the Buchner tests
is shown in Figure 14.
The basic filtration equation presented by Rich (21) is applicable to
the Cu'chner filter test.
t/v = . V +
2S2Ap Sap
where t = time, seconds
V = volume of filtrate, ft3
u = liquid viscosity, (Ib mass)/(ft-sec)
w = weight of dry cake solids per unit of
filtrate, (Ib force)/(ft3)
R = specific resistance, sec2/lb mass
Rf = resistance due to filter media,
(Ib force )( sec2 )/(lb mass) (ft2)
S = area of filter cake, ft2
Ap = permanent pressure drop due to friction,
Ib force/ ft2
A plot of t/V vs V is a straight line with slope b and intercept a
where:
. pwR .
b = :-— — and
2S2Ap
a = A
SAP
Therefore, when the pressure differential, viscosity, filter area and solids
content are known, the specific resistance can be obtained from the slope of
the filtration curve.
44
-------
SAMPLE
_9cm BUCHNER
FUNNEL
VACUUM GUAGE
7} VALVE
VACUUM
-RUBBER STOPPER
.PYREX GRADUATED
CYLINDER
Figure 14: BUCHNER FUNNEL TEST APPARATUS
45
-------
E
o
0
O
laOH
NH4OH
Na2CO3
.*-
Ca(OH)2
Ca(OH)2
+
POLYELECTROLYTE
20 40 60 80
V, ml
100
120 140
Figure 15: EFFECT OF VARIOUS ALKALIS ON
THE FILTRATION OF CHROMIUM
HYDROXIDE.
46
-------
Once the specific resistance has been determined, the filter yield can
be predicted by the equation:
B =
where B = filter yield, (Ib force)/(ft2-sec)
Y = fraction of cycle time attributed to cake
formation
tc = time required for one complete filter
cycle, seconds
Thus, the specific resistance is directly proportional to the slope of the
straight line plot of t/V vs V and the filter yield is inversely proportional
to the square root of the specific resistance.
In the Buchner tests, spent chrome was precipitated using an optimum
dosage of four different alkalis. The fifth sample was a lime precipitated
chromium hydroxide to which 30 mg/1 of Nalcolyte 677 was added. After
neutralization each of the samples contained 5,120 mq/1 of chromium.
A 150 ml sample of each of the sludges was transferred to the Buchner
funnel for filtration through K'hatman No. 42 filter paper at a pressure
differential of 20 inches of mercury. Filtrate volume was measured at 30
second intervals, and the time at which the filter cake began to crack was
recorded. The plots of t/V vs V for the several alkalis used are shown in
Figure 15. These curves indicate that filtration was markedly affected by
changes in the type of alkali used for precipitation.
The values for specific resistance (R ) as calculated from the tabulated
values for w and b are shown in Table 14. Pressure differential was main-
tained at 20 inches of mercury and each run was conducted at room temperature.
The filter had an area of 7.75xlQ-2 ft2 and the viscosity for each sample was
approximately 5.95x10"^ Ib mass/ft-sec (0.886 centipoise). Values for the
weight of dry cake solids per unit of filtrate (w) were calculated from the
corresponding values for feed suspended solids and cake solids and a feed
density of 66.2 lb/ft3 using the equation:
w =
pg/n(
0-xc)/xc
47
-------
VALVE
FILTER
LEAF
SLUDGE
VACUUM
GUAGE
VACUUM
/ FILTER FLASK
Figure 16: FILTER LEAF TEST APPARATUS.
48
-------
where p = liquid density, Ib mass/ft3
g = acceleration of qravity, ft/sec2
g = Newton's law conversion factor,
c 32.17 (Ib force)/(lb mass)
x = weight fraction of suspended solids
in unfiltered sludge.
x. = weight fraction of suspended solids
in filter cake.
The calculated values of filter yield are shown in Table 15. These
data confirm that filter yield is highly dependent upon the type of alkali
used for precipitation. Thus, by changing from caustic soda precipitation
to lime precipitation a 203% increase in filter yield from 0.639 Ib/ft2"hr
to 1.94 Ib/ft2"hr was effected. Likewise, the filtration rate was increased
100% from 2.40 gal/ft2~hr for the caustic soda-precipitated chrome to 4.80
gal/ft2~hr for the lime-precipitated chrome. These increases are due in part
to the nearly 50% increase in the feed suspended solids content from 28,000
mg/1 for caustic soda to 41,500 mg/1 for lime. The resultant filter cake
solids increased from 2.5g to 6.4g (dry basis).
The most significant effect of changing the type of alkali was that the
area requirements calculated for rotary vacuum drum filtration differ drasti-
cally. For example, to filter 1,000 gallons/hour of a lime-precipitated
sludge containing 41,500 mg/1 of suspended solids a filter area of about
185 ft2 would be required. If the chromium hydroxide were precipitated by
caustic soda, however, the area requirement would be nearly 400 ft2. Thus,
choice of alkali can result in a difference in the design surface area for
the vacuum filtration of chromium hydroxide from spent chrome liquors of
greater than 100%.
To confirm the results obtained in the Buchner tests and to predict
filter yield and filter area requirements more closely, a serjes of filter
leaf tests were carried out on the same sludges used in the Buchner tests,
The filter leaf apparatus is shown in Figure 16.
A 400 ml aliquot of each of the sludges was transferred to 1,000 ml
pyrex beakers. The filter leaf was fitted with a stainless steel wire gauze
over which a Whatman No. 42 filter paper was placed. A vacuum was applied
and the filter leaf was positioned beneath the sludge surface. After 150
seconds of cake formation at a vacuum of 20 inches of mercury the leaf was
removed and inverted for an additional 150 seconds of dewatering. On a
rotary vacuum drum filter this is equivalent to a cycle time of 400 seconds
at 37.5% submergence with 25% of the cycle attributed to the cake discharoe.
The filtrate was collected for each run and its volume measured. The feed
volume for each run was measured by difference. The filter cakes obtained
were weighed both wet and after overnight drying to obtain filter yields and
cake moistures.
49
-------
TABLE 14. EFFECT OF VARIOUS ALKALIS ON THE FILTRATION
OF CHROMIUM HYDROXIDE
Alkali
NaOH
NH4OH
Na2C03
Ca(OH)2
Ca(OH)2
+ polymer
b,
sec/ml *•
0.0282
0.0150
0.0086
0.0077
0.0067
Cake solids,
25.7
25.9
22.8
31.6
41.2
Feed SS,
mg/1
28,100
25,800
31,000
41,500
41,200
w. .
lb/ft3
1.99
1.82
2.16
3.03
2.92
Rx
9 '
secVlb mass
3.20xlOn
1.86X1011
8.95xl010
5. 72x1 010
5.17xl010
-------
TABLE 15. EFFECT OF VARIOUS ALKALIS ON FILTER YIELD
AND FILTRATION RATE FOR CHROMIUM HYDROXIDE
Alkali
NaOH
NH4OH
Na2C03
Ca(OH)2
Ca(OH)2
+ polymer
B (filter yield),
Ib/ft2-hr
0.639
0.799
1.28
1.94
1.92
Ap = 20 in Hg
* = 0.375
tc = 400 sec
y = 0.886 x 6.72 x 10~4 Ib/ft-sec
T = 25°C
Filtration Rate
gal/ft2-hr
2.40
3.22
4.43
4.80
4.92
-------
The results from the filter leaf tests in Tables 16 and 17 confirm that
the choice of alkali is indeed a critical factor in the vacuum filtration of
chromium hydroxide from spent chrome tanning liquors. Values found for
filter yield in these filter leaf tests were slightly higher than the yields
predicted in the Buchner tests. Also, filtration rates and sludge feed rates
were found to be higher.
The cakes formed by filtration of the caustic soda and ammonium hydrox-
ide-precipitated sludges v/ere each approximately l.Gmm thick and had a
pasty consistency which made separation of the cake from the filter paper
difficult. The 3.2rnm cake formed from the lime-precipitated sludqe was
separated without difficulty. The soda ash-precipitated sludqe yielded a
cake which was thicker and easier to discharge than the other sludges but
had the highest moisture content at 77.2% i^O and had the greatest weight
per unit volume of sludge filtered.
The filter leaf test simulates the operation of a rotary vacuum drum
filter, whereas the Buchner test requires calculation of filter yields from
the specific resistance of the various sludges. The authors, therefore,
concluded in the face of the observed test result differences that the filter
leaf data are the more germane, and should be used in the design of a vacuum
filtration system.
It can be concluded from these data that the order of filterability of
chromium hydroxide sludges from the various alkalis is: hydrated lime> soda
ash> aqueous ammonia> caustic soda.
Several additional factors which can effect filtration include viscosity,
pressure differential, suspended solids concentration of the sludge and cake
formation time.
In order to test the effect of variation 1n cake formation time (subse-
quently translatable to drum speed) on filtration rates, a series of filter
leaf tests were conducted in which form time was varied from 44 seconds to
220 seconds (simulating drum speeds of 0.5 rpm to 0.1 rpm, respectively, at
37.5% submergence). A lime-precipitated chromium hydroxide sludge floccu-
lated with Nalcolyte 677, which contained 45,560 mg/1 of suspended solids
and 7,880 mg/1 of chromium was filtered in each run at a temperature of 70°F
and under a vacuum of 20 inches of mercury.
The results displayed in Tables IS and 19 demonstrate that an increase
1n drum speed increased the rate of filtration and therefore the filter
yield. Thus, an increase in drum speed from 0.1 rpm to 0.5 rpm effected an
increase in sludge feed rates by over 100% from 6.15 oal/ft2-hr to 13.4 gal/
ft^-hr. However, there was a corresponding decrease in the cake thickness
from 7.2mm at 0.1 rpm to 2.8mm at 0.5 rpm.
As the cake thickness decreased, cake separation from the filter became
more difficult. In practice, therefore, the optimum drum speed would be the
highest speed which can be utilized without sacrificing good cake discharge.
The vacuum filter leaf tests indicated that filter cake formed from the fil-
tration of lime-precipitated chromium hydroxide sludge should be
52
-------
TABLE 16. FILTER LEAF TEST RESULTS - PART I
en
CO
Alkali
NaOH
NH4OH
Na2C03
Ca(OH)2
Ca(OH)2
+ polymer
Feed SS,mg/l
28,100
25,300
31 ,000
41,500
41 ,200
Filtrate vol . ,ml
87
130
153
158
163
Feed vol. ,ml
95
140
173
175
180
'.•Jet cake wt. ,g
9.58
12.56
24.40
19.50
19.14
Cake moisture, %
74.3
74.1
77.2
68.4
58.8
-------
TABLE 17. FILTER LEAF TEST RESULTS - PART II
Alkali
NaOH
NH4OH
Na2C03
Ca(OH)2
Ca(OH)2
Filter yield,
0,588
1.02
1,60
1,77
1.16
Filtration rate,
g3l/ft2«-hr
2,93
4.47
6,26
5,45
5,60
Feed rate,
gal/ft^hr
3,22
4,81
1,70
6,01
6,V3
polymer
-------
en
en
TABLE 18. EFFECT OF VARIATION IN DRUM SPEED ON THE VACUUM FILTRATION
OF LIME-PRECIPITATED CHROMIUM HYDROXIDE - PART I
Drum
speed,
rpm
0.500
0.333
0.250
0.200
0.167
0.125
0.100
Form
time,
sec
44
66
88
no
132
176
220
Dewater
time,
sec
48
72
96
120
144
192
240
Crack
time,
sec
21
16
20
25
22
30
35
Filtrate
volume,
ml
109
149
173
197
215
250
274
Cake
thickness,
mm
2.8
4.0
4.8
4.8
5.6
6.4
7.2
Wet cake
weight,
g
23.0
28.8
30.9
36.0
41.2
50.6
53.6
Dry cake
solids,
%
27.4
26.2
27.7
26.8
25.6
23.4
27.4
Area of filter leaf » 7.75xlO"2ft2
-------
en
TABLE 19. EFFECT OF VARIATION IN DRUM SPEED ON THE VACUUM FILTRATION
OF LIME-PRECIPITATED CHROMIUM HYDROXIDE - PART II
Drum speed,
rpm
0.500
0.333
0..250
0.200
0.167
0.125
0.100
Filter yield,
Ib/ft^-hr
6.03
4.82
4.09
3.7.0
3.37
2:84
2.79
Filtrate volume,
gal/ft2-hr
1,2.2
10.. 2
a. 85
8:06
7.. 35
6.. 38
5.59
Feed volume,
gal/ft2-hr
13.4
11.2
9.74
8. -87
8.. 09
.7.02
•6.3-5
-------
satisfactorily discharged at a thickness as low as 4.8mm.
To determine the effect of the suspended solids concentration in the
feed on the vacuum filtration of chromium hydroxide sludges, the filter leaf
tests were repeated at 0.2 rpm and 0.1 rpm using a lime-precipitated sludge
which contained 30,100 mg/1 of suspended solids and 5,250 mg/1 of chromium.
The results of these runs which are shown in Table 20 demonstrate that while
cake thickness decreased substantially from 4.8mm to 3.2mm at 0.2 rpm and
from 7.2mm to 5.6mm at 0.1 rpm, the lower feed solids effected an increase
in filtration rates from 8.06 gal/ft2-hr to 9.10 gal/ft2-hr and from 5.59
gal/ft^-hr to 6.54 gal/ft2-hr, respectively. The corresponding filter yields
dropped from 3.70 Ib/ft2-hr to 2.41 Ib/ft2-hr at 0.2 rpm and from 2.79 lb/
ft2-hr to 1.95 Ib/ft2-hr at 0.1 rpm with the decrease in feed suspended
solids.
A fourth variable which was tested for its effect on filtration was
temperature. The variation in filtration rates with temperature is attrib-
uted to the change in viscosity. The results of a series of Buchner tests
conducted on a lime-precipitated sludge containing 45,560 mg/1 of suspended
solids and 7,880 mg/1 of chromium under a vacuum of 20 inches of mercury are
shown in Figure 17 and Table 21. By increasing the feed temperature from
66°F to 110°F at a drum speed of 0.125 rpm and at 37.5% submergence, the fil-
ter yield was increased 16.7% from 2.34 Ib/ft2-hr to 2.73 Ib/ft2-hr. Like-
wise, feed rates and filtration rates were increased by 16.7%.
As these data indicate, there is a distinct advantage to processing
spent chrome by precipitation and filtration while it is warm. Where the
temperature may drop to 70°F or lower, it may even prove economical to pre-
heat the sludge by steam injection or other means prior to filtration.
ACIDIFICATION OF CHROMIUM HYDROXIDE FILTER CAKE
The widest applicable success of any chrome recycle process depends upon
being able to produce a basic chromium sulfate which resembles closely the
stock chrome solution being used in the tanning process.
The acidification of recovered chromium hydroxide filter cake with con-
centrated H;?SQ4 yields chrome liquors of significantly variable characteris-
tics depending upon the type of alkali used for precipitation.
Screened spent chrome samples (200 ml) containing 4,300 mg/1 of chromium
and an acidity of 220 me/1 were precipitated to 100% basicity with the
various types of alkali and diluted to 250 ml each. After filtration through
a Buchner funnel, the filter cakes were recovered and dissolved in concen-
trated H2S04 to pH 2.5 to 3.3. As shown in Tables 22 and 23, the redissolved
cakes showed considerable variation with respect to the parameters of speci-
fic gravity, chromium concentration, and residues.
Most significantly, the suspended solids concentration for lime-
precipitated chrome was drastically higher than for sodium hydroxide, ammo-
nium hydroxide or sodium carbonate. Calcium stilfate was the principal
57
-------
i.o-
0.8
I 0.6
:>"
-•>.
•»-•
0.4
0.2
0
60
80
66°F
180
100 120 140 160
V, ml
Figure 17: EFFECT OF TEMPERATURE ON
THE FILTRATION OF LIME-
PRECIPITATED CHROMIUM
HYDROXIDE,
58
-------
TABLE 20. EFFECT OF VARIATION IN SUSPENDED SOLIDS CONCENTRATION ON THE
VACUUM FILTRATION OF LIME-PRECIPITATED CHROMIUM HYDROXIDE
01
vo
ss,
mg/1
45,560
30,100
45,560
30,100
Drum
speed,
rpm
0.200
0.200
0.100
0.100
Wet cake
weight,
g
36.0
25.3
53.6
37.6
Dry cake
solids,
%
26.8
24.9
27.4
27.2
Cake
thickness,
rnrn
4.8
3.2
7.2
5.6
Filter
yield,
Ib/ft2-hr
3.70
2.41
2.79
1.95
Filtrate
Feed
volume, volume,
gal/ft2-hr gal/ft2-hr
8.06
9.10
5.59
6.54
8.87
10.0
6.15
7.19
-------
TABLE 21. EFFECT OF TEMPERATURE ON THE FILTRATION
OF LIME-PRECIPITATED CHROMIUM HYDROXIDE
Temp,
Op
66
82
92
no
P.
cp
0.96
0.83
0.75
0.61
b,
sec/ml2
0.00461
0.00412
0.00345
0.00339
o Rl'
secz/lb mass
3.05xl010
3.15xl010
2. 99x1 010
3.60xl010
B,
1b/ft*-hr
2.34
2.48
2.70
2.73
Filtrate
volume,
gal/ft2-hr
5.57
5.91
6.43
6.51
Feed
vo 1 ume ,
gal/ft2-hr
6.12
6.40
7.07
7.16
w = 3.14 lb/ft3
-------
TABLE 22. ACIDIFICATION OF CHROMIUM HYDROXIDE FILTER CAKE - PART I
CTl
Alkali
NaOH
NH4OH
Na2C03
Ca(OH)2
Volume of
redissolved
chrome, ml
14.1
13.4
19.3
16.4
Cr
Sp. Gr. pH mq/1
1.212 2.80 57,000
1.192 2.50 61,000
1.203 3.30 43,000
1.244 3.00 49,600
Ib/gal
0.475
0.508
0.358
0.408
Alkali
NaOH
NH4OH
Na2C03
Ca(OH)~
TABLE 23. ACIDIFICATION
Total solids,
mq/1
373,000
420,000
312,000
441 ,000
OF CHROMIUM HYDROXIDE FILTER CAKE - PART II
Total
dissolved Suspended
solids, solids,
mg/1 mg/1
346,900 26,100
392,300 27,700
292,200 19,800
279,000 162,000
Cr in
Suspended
solids,
5.4
4.3
4.2
1.6
-------
component of these suspended solids. It should be noted that the chromium
component of the suspended solids was substantially lower for the lime-
precipitated chrome than for other alkalis indicating that there is no gross
coprecipitation of chrome with calcium sulfate.
Another important difference is related to the relatively wetter filter
cake obtained for the filtration of soda ash-precipitated chromium hydroxide,
and therefore the greater volume of redissolved chrome.. Thus, soda ash pro-
duced a redissolved chrome which was substantially lower in chromium concen-
tration (at 43,000 mg/1) than did the other alkalis. Likewise, total solids
and suspended solids were considerably lower.
It was further noted that although the total and suspended solids con-
centrations for lime-precipitated chrome were substantially higher than when
other alkalis were used, the total dissolved solids of 279,000 mg/1 were
lower than for other alkalis. Therefore, if the calcium sulfate could be
conveniently removed, precipitation of this insoluble salt may prove benefi-
cial in that it significantly reduces the non-chromium fraction of dissolved
salts present in the re-acidified liquor.
Finally, it is important to point out that these recovered chrome
liquors are substantially lower in chrome content than most commercially
available tanning agents. Therefore, in order to produce a "standard" tan-
ning concentration of 1 Ib/gal of 0203 (0.685 Ib Cr/pal) it would be neces-
sary (in Pfister & Voqel production) to blend the recovered chrome liquor
at a 5 to 4 ratio with a stock chrome liquor of 1.5 Ib 0303 per gallon
(1.03 Ib Cr/gal).
PILOT FILTRATION STUDIES
While lab-scale Buchner tests and filter leaf tests can be used to pre-
dict optimum filtration conditions and filter area requirements, there are
several important factors which must be evaluated on a pilot-scale before
full-scale filtration equipment can be specified. These factors include
the type of cake discharge mechanism required (i.e., belt, string, scraper or
roller discharge, etc.) and filter medium to be employed. Likewise, it is
imperative to establish whether blinding of the filter medium might be
encountered in full-scale operations. Blinding must always be anticipated
in the filtration of wastes containing substantial concentrations of oils,
fats and greases.
In pilot-scale filtration studies, a T wide by 3' diameter rotary
vacuum drum filter was employed. The filter was fitted with a string dis-
charge and a Nylon multi-filament fabric ("Feon" No. MMH-W4IO-SE7) with an
air permeability of 5.2 cfm/ft2 @ 0.5" H20. Total filtering area equalled
9.4 ft2.
Test runs on the pilot filter confirmed the filter area calculations
for separation of precipitated chromium hydroxide from spent chrome liquors,
as predicted by the laboratory, bench-scale tests. Likewise, during 40 hours
of "non-continuous" operation of the filter, the string discharge mechanism
62
-------
and Nylon filter cloth performed satisfactorily with no sign of wear. The
animal fats and greases present in our spent chrome liquors did not blind
the filter media.
Approximately 100 gallons of spent chrome liquor containing 4,660 mg/1
of chromium was precipitated at pH 9.0 with a slurry of hydrated lime and
then was flocculated with 30 mg/1 of Nalcolyte 677. The precipitated sludge
contained 38,440 mg/1 of suspended solids. During the filtration run, drum
submergence was maintained at 37.5% and drum speed at 0.125 rpm.
As shown in Table 24, at 68°F and under a cake-form vacuum of 22" Hg
the sludge was filtered at a rate of 6.0 gallon of filtrate /ftz-hr. Filter
cake was discharged at an average thickness of 3.2mm and an average moisture
content of 63%.
TABLE 24. PILOT-SCALE VACUUM FILTRATION OF CHROMIUM HYDROXIDE
Hydrated lime Soda ash
pH 9.0 8.7
flocculant 30 ppm Nalcolyte 677 none
suspended solids 38,440 mg/1 23.740 mg/1
temperature 68°F 706F
feed rate 6.6 gal/ftz-hr 7.9 gal/ft2-hr
filtration rate 6.0 gal/ft2-hr 7.3 gal/ft2-hr
cake thickness 3.2 mm 5.6 mm 0
filter yield (wet) 5.7 Ib/ft2-hr 6,8 lb/ft^-hr
cake moisture 63% 76.4%
filter yield (dry) 2.1 Ib/ft2-hr 1.61b/ft2-hr
drum speed = 0.25 rpm
cake form vacuum = 22" Hg
drum submergence = 37.5%
Soda ash precipitation (with 1 Ib/gal soda ash solution) of another
100 gal fraction of the spent chrome liquor yielded a sludge containing
3,890 mg/1 of chromium and 23,740 mg/1 of suspended solids at a pH of 8.7.
At 0.125 rpm, 37.5% submergence and under a cake-form vacuunrof 22" Hg, this
sludge was filtered at 70°F at a rate of 7.3 nal filtrate/ft2-hr correspond-
ing to a feed rate of 7.9 gal/ft2-hr, with a filter yield of 1.6 Ib dry
solids/ft2-hr. Cake was discharged at an average thickness of 5.6 mm and
contained 76.4% moisture.
During both runs the discharge of cake from the filter by the string
mechanism was complete and no significant blinding of the Nylon filter
medium was observed.' The 3.2 mm cake thickness was judged to be the minimum
which can be tolerated for lime-precipitated chromium hydroxide. The 5.6 mm
observed for soda ash-precipitated sludge, however, is thicker than the
63
-------
optimum for cake discharge.
Thus, it 1s apparent that the soda ash-precipitated sludge with its
substantially lower solids content can be filtered at a faster volumetric
rate than lime-precipitated sludge. However, most of this rate difference
is offset by the increased volume of sludqe due to the addition of soda ash
as a relatively dilute solution.
SEPARATION OF OILS AND GREASES FROM SPENT CHROME LIQUOR
A laboratory-scale, mass-balance evaluation of the filtration of alkali-
precipitated chromium hydroxide from spent chrome indicated that the contami-
nating fats, oils and greases are essentially non-filterable. Spent chrome
liquor containing 7,280 mg/1 of chromium and 8,820 mg/1 of Oil and Grease
was precipitated with calcium hydroxide or sodium carbonate to pH 9.0. The
sludges were filtered through Whatman No. 42 filter paper on a Btichner
funnel, and the filter cakes were dried and redissolved in hydrochloric
acid. As shown in Table 25, analyses of the filtrates and the redissolved
cakes revealed that less than one percent of the Oil and Grease was recov-
ered in the filtrates.
Therefore, it is apparent that to prevent Oil and Grease build-up to
intolerable levels in recycled chrome, they must be separated prior to pre-
cipitation and filtration and/or after acidification of the chromium hydrox-
ide filter cake. To search out feasible alternatives for Oil and Grease
separation several bench-scale and pilot-scale experiments were conducted.
TABLE 25. REAPPORTIONMENT OF OIL & GREASE DURING THE
FILTRATION OF CHROMIUM HYDROXIDE
Hydrated lime Soda ash
aliquot 100 ml 100 ml
Oil & Grease 882 mg 882 mg
volume after 121 ml 121 ml
precipitation
filtrate volume 106 ml 103 ml
dry cake weight 5.43 g 4.62 g
cake Oil & Grease- 16.3% 17.1%
dry basis
cake Oil & Grease 884 mg 791 mg
filtrate Oil & Grease 25.5 mg/1 31.0 mg/1
filtrate Oil & Grease 2.7 mg 3.2 mg
64
-------
Flotation
A 50 gallon "batch" of spent chrome containing 7,110 mg/1 of 011 and
Grease was settled for one hour. An aliquot of the underflow contained
only 181 mg/1 of Oil and Grease, indicating that 97.5% was concentrated
in the supernatant layer.
Similarly, reacidified chrome containing 6,820 mg/1 of Oil and Grease
yielded an underflow containing 990 mg/1, equivalent to an 85.5% transfer
into the supernatant, but after 72 hours.
Centrifugation
Centrifugation of a spent chrome liquor containing 12,060 mg/1 of Oil
and Grease in an International Model SVB laboratory centrifuge for 5 minutes
at 1500 rpm yielded an underflow which analyzed at 220 mg/1, indicating that
98.4% of the Oil and Grease was thus separated. Likewise, Centrifugation
for 5 minutes at 1500 rpm of a reacidified chrome containing 6,820 mg/1 of
Oil and Grease yielded an underflow containing 1,420 mg/1 with 79.3% being
transferred to the scum layer.
Extraction
An aliquot of spent chrome liquor which contained 12,060 mg/1 of Oil
and Grease was sequentially extracted at room temperature with two volumes
(25% and 10%) of chloroform. This two-stage, liquid-liquid extraction pro-
duced an aqueous phase containing 220 mg/1 of Oil and Grease. However,
extraction of a reacidified chrome with either chloroform or hexane proved
impractical since no discrete separation of the organic solvent phase and
the aqueous phase could be obtained.
Adsorption
A 20 ml aliquot of a spent chrome containing 12,060 mg/1 of Oil and
Grease was passed through a packed column containing 100 c.c. of 3 mm. dia.
glass beads. The effluent from the adsorption column analyzed 5,600 mg/1
indicating a 53.6% retention. Suspended 011s and Greases and fine fibrous
solids had a blinding effect on the column and made regeneration with organic
solvent impracticable. Likewise, reacidified chrome containing 6,820 mg/1 of
Oil and Grease was treated by application to the adsorption column resulting
in a 31.4% retention, with the underflow containing 4,690 mg/1 of 011 and
Grease.
Absorption
Filtration of spent chrome through a column containing 100 cc of blue
shavings proved Impossible. Once the blue shavings were wetted, the column
became completely impervious to spent chrome liquor.
65
-------
CLARIFIED EFFLUENT
SLURRY
RESERVOIR
SLUDGE
RECEPTACLE
Figure 18: HYDROCYCLONIC SEPARATOR,
66
-------
HYDROCYCLONIC SEPARATION OF CALCIUM SULFATE
FROM REACIOIFIED SPENT CHROME
The major disadvantage of using hydrated lime for precipitation in the
recovery of chromium from spent chrome tanning liquors has been the formation
of calcium sulfate. This sparingly soluble salt forms during the precipita-
tion of the chromium hydroxide and on acidification with HgSO^ to regenerate
the basic chrome sulfate.
A spent chrome liquor containing 6000 mg/1 of chromium, when precipi-
tated at pll 9.5 with hydrated lime followed by rea-cidification with concen-
trated sulfuric acid to pH 3.5, yielded a suspension of the rapidly settling
insoluble white salt. The suspension analyzed at 43,800 mg/1 of suspended
solids, of which 31,300 mg/1 was fixed suspended solids, primarily calcium
sulfate.
It appears, therefore, that sequential recycles through lime-
precipitation and acidification of settled sludqe with 112564 would result
in an unacceptable build-up of calcium sulfate.
To demonstrate the removal of precipitated calcium sulfate, 50 gallons
of calcium sulfate-chrome slurry prepared as above was charged to a hydro-
cyclonic separator (Bauer Clarifier No. 677-1)(22) depicted schematically
in Figure 18.
The hydrocyclonic separator utilizes a centrifugal force many times the
force of gravity to remove suspended solids from liquids. A slurry is
pumped under controlled pressures into the broad end of a fixed cone in a
tangential pattern. The cyclonic flow which develops causes particulate
matter to impinge on the wall of the cone and finally discharge through the
narrow end while the liquid follows a vortex back up the center of the cone
to be discharged through an overflow at the broad end. Thus, grit or other
solid matter which is substantially denser than the liquid carrier can be
removed quickly using equipment which occupies very little space.
The hydrocyclonic separator employed in this study operates at a feed
rate of 22 gpm and has overall dimensions (includinn pump and reservoir,
cyclone, and solids collector) of 31" x 31" x 66" high.
The results of three cyclone runs with varying orifices are shown in
Tables 26 and 27. The solids data are for grab samples collected at the
cyclone orifice and overflow, and from the main reservoir during operation
of the unit. Sludge volumes were determined by mass balance and are ex-
pressed as percent of feed volume.
With an orifice diameter of 3/16", and a feed suspended solids concen-
tration of 58,700 mg/1, an effluent containing only 9,770 mg/1 of suspended
solids was obtained, effecting an 83.4% reduction. The sludge was equiva-
lent to approximately 9% of the feed volume and contained 540,000 mg/1 of
suspended solids. Likewise, the fixed suspended solids concentration was
reduced by 94.5% from 39,100 mg/1 in the feed to 2,160 mg/1 in the effluent
while the sludge fixed suspended solids concentration was 413,000 mg/1.
67
-------
TABLE 2G. CYCLONE SEPARATION OF CALCIUM SULFATE FROM SPENT
CHROME LIQUOR AT VARYING ORIFICE DIAMETERS - PART I
oo
Orifice dia. ,
in
1/8
5/32
3/16
Influent
mg/1
43,000
57,800
58,700
SS, Effluent SS,
mg/1
10,800
9,790
9,770
Sludge SS,
mq/1
625,000
755,000
540,000
Sludqe vol., Removal
5.0 75.2
6.4 83.1
9.1 83.4
TABLE 27.
CYCLONE SEPARATION OF
CHROME LIQUOR AT VARY I
CALCIUM SULFATE FROM SPENT
MG ORIFICE DIAMETERS - PART II
Orifice dia. ,
in
1/8
5/32
3/16
Influent
fixed SS
nq/1
31 ,300
38,000
39,100
Effluent
, fixed SS,
rnn/1
4,260
2,300
2,160
Sludge
fixed SS,
mg/1
490,000
580,000
413,000
SI udge vol . , Removal ,
5.6 86.4
7.0 93.9
9.0 94.5
-------
The sludge had a specific gravity of 1.395 and contained 62 ml/100 ml of
settleable solids (1 hour). Further analysis of these three fractions indi-
cated that there was a slight disapportionment of chromium, with the efflu-
ent being slightly higher in chromium content than the recovered sludge.
The feed slurry contained 5,620 mg/1 of chromium while the effluent con-
tained 5,700 mg/1 and the s'ludoe 5,030 mg/1. Therefore, the solids thus
redistributed from the feed slurry to the sludge fraction are non-chromium
suspended solids.
The successful application of the cyclone for the separation of calcium
sulfate from spent chrome liquors suggested that the calcium sulfate which
is precipitated during the acidification of chromium hydroxide filter cake
could be separated in a similar manner.
REUSE OF CALCIUM SULFATE AS A FILTER AID
A bench-scale experiment indicated that the calcium sulfate which is
generated during the reacidification of lime-precipitated chromium hydroxide
filter cake can be reused as an effective filter aid for the separation of
Oil and Grease from subsequent reacidified chrome recovery solutions.
Approximately 500 g of wet calcium sulfate filter cake (60% solids) was
resuspended in the acidified, recovered chrome solution at a weight ratio of
one-to-one. Filtration of the slurry on a vacuum filter leaf at 20 inches
of mercury and at room temperature indicated that vacuum filtration is
possible at a slurry feed rate of 9.5 gal/ft2-hr yielding 35 Ib/ft2-hr of
dry cake solids (40% moisture) and 5.0 gal/ft2-hr of filtrate. The Oil and
Grease concentration was thus reduced by 90% from a feed concentration of
912 mg/1 to 92 mg/1 in the filtrate. In contrast to these findings, however,
direct filtration of reacidified chrome without "enrichment" with CaSO^ is
practically impossible.
69
-------
SECTION VII
CHROME RECOVERY SYSTEM DESIGN
CAPTURE OF SPENT CHROME
One-bath chrome tanning is the process in which hydrated chromic sulfate
complexes react with the biologically-occurring, fibrous protein substrate,
collagen, to form leather. One widely-used means of driving this chrome
fixation-reaction to practical completion is the offering of an excess of
chromium (III) sulfate. In modern practice, as much as 150% of the chrome
ultimately taken up by the hide may be offered. The excess chrome is dis-
charged to plant sewers when tanning is complete. Following the discharge
of these spent tanning liquors, relatively large volumes of wash waters are
used to remove surface impurities and to extract unwanted salts and 011s &
Greases from the tanned hides.
As demonstrated 1n Table 28, for a typical hide processor run approxi-
mately 47 pounds of excess chrome (as Cr) is discharged during the initial
800-1,000 gallon spent chrome drain. However, to capture the next 29 pounds
of chromium discharged, more than 4,000 gallons of combined drains and
washes would have to be collected. An additional four pounds of chromium
is "lost" to the floor drains during the unloading of the tanned hides. Thus,
as much as 80 pounds of chromium may be discharged in 5,100 gallons of waste
from a single hide processor. Even though these tanyard discharges are the
major contributors of chromium to the total tannery effluent, sequential,
24-hour wastewater surveys, conducted during the annual plant shutdown, demon-
strated that as much as one-third of the chrome comes from non-tanyard activi-
ties. For example, a mass-balance blue stock analysis (Section XVI) suggests
that approximately 14% of the chromium in the effluent is introduced through
blue stock wringing. Likewise, substantial loadings from piling, retanning,.
coloring, setting, pasting and even "dry-end" operations are suspected. Thus,
to maximize chrome capture, essentially all wastewaters generated after tan-
ning (approximately 720,000 gallons/day) would have to be collected for chrome
recovery. Clearly, for an urban tannery such a recovery scheme would not
optimize valuable space resources. Also, numerous non-chromium components
(such as vegetable tannins, buffing dust, 011s & Greases, etc.) would con-
tribute to the formation of a voluminous low-chromium sludge which would not
be suitable for chrome recycle. Thus, in the design of Pfister & Vogel's
chrome recovery system, collection, storage and treatment hardware were sized
for the recovery of chromium-rich tanyard wastes only.
The several alternative levels of chrome capture and recovery considered
are depicted 1n Table 29. Alternative A requires the collection of the most
70
-------
TABLE 28. DISTRIBUTION OF CHROMIUM IN HIDE PROCESSOR WASTES
Volume, gal
900
1,400
2,050
2,700
4,050
5,100
Cumulative
Chromium,
mg/1 Ib
6,260 47
5,060 59
3,860 66
3,150 71
2,250 76
1,890 80
TABLE 29. ALTERNATIVE LEVELS
Alternative Volume, gal /day
A 14,400
B 22,400
C 32,800
D 43,200
E 64,800
F 81,600
OF CHROME CAPTURE
Chromium
mg/1
6,260
5,060
3,860
3,150
2,250
1,890
Ib/day
752
944
1,060
1,140
1,220
1,280
71
-------
concentrated 900 gallons of spent chrome from each of 16 production runs
daily. An estimated 14,400 gallons of spent chrome liquor containing 6,260
mg/1 of chromium equivalent to 750 Ib/day would be processed.
Alternative B would require the collection and treatment of nearly
22,400 gallons of waste including 900 gallons of concentrated spent chrome
liquor and the first 500 gallons of wash water from each production run.
The chromium concentration would be diluted to 5,060 mg/1 while nearly 950 Ib
of chromium would be removed daily. Likewise, Alternatives C, D, and E call
for increased design loadings and corresponding reductions in average chro-
mium concentration. Since substantial quantities of water are required for
the practical unloading of tanned hides from hide processor equipment, the
final 1,050 gallons from each production run is currently being discharged
to floor drains and/or to common plant sewers. Therefore, Alternative F
remains hypothetical until a means of collecting this final waste fraction
1s determined.
At Pfister & Vogel, only 61% of the lower tannery production is done
in hide processors with pump out facilities. The remaining 39% of produc-
tion is conducted in process equipment utilizing gravity discharge of
wastes to common tannery sewers. Segregation of this latter portion of the
spent chrome liquors is not currently feasible. However, anticipating the
possible conversion of these production units to a pump-out design,
Alternatives A-F include the capacity to treat chrome-bearing wastes from
16 production runs per day.
PRECIPITATION
The principal objective in spent chrome recovery is the concentration
of chrome in relatively dilute process wastes to yield a sufficiently chrome-
rich by-product which can be recycled, while substantially reducing effluent
levels. As demonstrated during the chrome recovery feasibility studies
(Section VI), the most effective means of achieving this chromium separation/
concentration is by precipitation with alkali followed by sludqe thickeninn
and filtration.
One of the most important factors in the design of an efficient treat-
ment system for chrome recovery is the correct choice of alkali for precipi-
tation. While all of the alkalis investigated (soda ash, aqueous ammonia,
hydrated lime and caustic soda) demonstrate excellent chromium hydroxide
precipitation capability, hydrated lime was found to be superior.
Chromium hydroxide has a relatively low particle density and it is
highly hydrated in an aqueous system. Thickening is impractical within the
observed range of suspended solids concentrations without further chemical
treatment. Only hydrated lime-chromium hydroxide precipitate responds
favorably to the addition of coagulants and/or flocculants to yield large,
dense, rapidly settling floe particles. The formation of calcium sulfate
during the lime addition is believed to further improve floe density and
settling. The lime-precipitated sludge is readily filtered to yield a hiqh-
chromium filter cake suitable for land fill or subsequent acidification and
72
-------
reuse.
Soda ash proved to be the least effective precipitating agent during
the feasibility studies. The extremely slow settling rate of soda ash-
precipitated chromium hydroxide necessitates incorporation of excessive
thickener surface area for treating dilute wash waters. The soda ash-
precipitated chromium hydroxide sludge, however, is dewatered by vacuum
filtration as readily as lime-precipitated sludge. Further, the soda ash
sludge produces a substantially thicker filter cake leading to Improved cake
discharge during vacuum filtration.
Thus, while soda ash must be eliminated where treatment of the most
dilute chrome fractions is required, it remains a viable alternative for
chrome recovery from concentrated spent chrome liquors viz: Alternatives
A and B, where filtration of the unthickened sludge directly after precipi-
tation is favored.
One major disadvantage of the use of hydrated lime over the other
alkalis is the formation of large quantities of insoluble calcium sulfate
upon acidification with H2SOA. The suspended solids remaining after redis-
solvlng the lime-precipitatea chrome cake equal nearly 15% of the weight of
the reclaimed chrome solution. An additional unit operation must be employed
to reduce the calcium sulfate content to levels which are tolerable for
recycle. Removal of the calcium sulfate can be achieved by hydroclonic
separation and/or gravity settling. Ultimate disposal of the resultant
calcium sulfate-chrome sludge, however, remains a major drawback to the use
of Urne.
Where recycle of the chrome is not desired, the lime-precipitated
chrome cake can be burled in sanitary landfills. Use of an excess of lime
(i.e., more than the stoichiometric requirement) should Insure that the
trivalent chromium remains "fixed" as the insoluble hydroxide precluding
the possibility of leaching through the soil. Finally, on the basis of
chemical costs, hydrated lime at $1.25 per pound equivalent ($3.28/100 Ib)
is substantially cheaper to use as a precipitating aqent than is soda ash
which costs $5.33 per pound equivalent ($5.01/100 Ib).
Therefore, because lime is a superior precipitating agent yielding a
chromium hydroxide sludge which settles rapidly and filters readily, and
because of Its marked cost advantage over other alkalis, the treatment hard-
ware for this project was designed to recover chrome by the lime-
precipitation route. Fortuitously, this design also permits the optional
precipitation and filtration of the more concentrated chrome wastes by the
soda ash route, due to the similar filtration rates for the two types of
sludge.
COLLECTION
During normal production scheduling a maximum of two hide processors
(out of a total of eleven "on-line" productiofv units with pump-out facili-
ties) discharge chrome wastes within any two hour period. Therefore,
73.
-------
screening and collection facilities were sized to handle the simultaneous
discharge of wastes from two processing units. For screening, a 48" wide
Bauer Hydraselve with 0.040" screen surface openings, having a design load-
ing of 340 gpm was specified. The screened chrome-bearing wastes are dis-
charged by gravity to the 12 ft by 12 ft cylindrical fiberglass Spent Chrome
Collection Tank having a capacity of 10,200 gallons.
DESIGN
Sludge Thickening
The sludge thickening tank is amonn the major hardware Items required
for chrome recovery. Since space 1s a limiting factor in this Pfister &
Vogel demonstration project (as it is in most urban-located, high produc-
tivity tanneries), the comparative space requirements for each of the
alternative levels of spent chrome capture and recovery must be known.
Loadings for Alternatives A through F are shown in Table 30. Feed
chromium concentrations for the various alternatives vary from 6,260 mg/1 to
only 1,890 mg/1. Corresponding suspended solids loadinqs were extrapolated
from the suspended solids (lime-precipitated) vs. chromium curve (Figure 7)
generated during the feasibility studies. As shown, while suspended solids
concentrations are expected to vary from 47,500 mg/1 to 9,800 mg/1, the daily
solids loading would only vary from 5,700 Ib to 6,850 Ib.
Design surface areas were determined empirically from the graphical data
in Figure 10. Table 31 shows the extrapolated design volumetric loadings
which would vary from 400 gpd/ft2 at a feed suspended solids concentration
of 47,500 mg/1 to 1,800 gpd/ft2 at a concentration of 9,800 mg/1. Area
requirements calculated from design volumetric loadings and anticipated flow
vary from 36.0 ft' to 45.3 ft2, respectively. Corresponding decreases in
retention time and underflow concentration as extrapolated from Figures 11
and 12, respectively, are also indicated. Alternatively, surface area
requirements can be calculated from extrapolated values of sludge thickener
unit areas from Figure 13. Thus, as Table 32 shows, for Alternatives A
through E the unit area is a nearly constant 12.1 ft2/ton-day. For Alterna-
tive F, however, because of the-unhindered settling characteristics of the
low concentration sludge, the unit area is Increased to 13.3 ft2/ton-day.'
The surface area requirements thus calculated are in very close agreement
with the values determined by the design volumetric loading formula.
Filtration
The second major hardware item required for chrome recovery is a filter.
A rotary vacuum drum filter was chosen based on its demonstrated effective-
ness in dewatering chromium hydroxide sludges. This type of filter lends
itself to reliable, automated operation while the capital cost is substan-
tially lower than for a comparable pressure filter.
During the operation of the pilot vacuum filter, a sludge feed rate of
6.6 gal/ftz-hr was observed with a lime-precipitated chromium hydroxide
74
-------
TABLE
Alternative
A
B
C
0
E
F
30. SURFACE
Volume,
qal/day
14,400
22,400
32,800
43,200
64,800
81,600
AREA
FOR SLUDGE
Cr
mg/1
6
5
3
3
2
1
,260
,060
,860
,150
,250
,890
THICKENING
U/day
1
1
1
1
752
944
,060
,140
,220
,280
- PART I
Suspended
mg/i
47,500
35,500
24,900
19,000
12,360
9,800
solids,
Ib/day
5,700
6,630
6,800
6,850
6,650
6,660
TABLE 31. SURFACE AREA FOR SLUDGE THICKENING - PART II
Alternative
A
B
C
D
£
F
Suspended
solids,mq/l
47,500
35,500
24,900
19,000
12,300
9,800
Flow,
gal /day
14,400
22,400
32,800
43,200
64,800
81,600
Design
volumetric
loading,
qpd/ft2
400
550
800
1,050
1,500
1,800
Area,
ft2
36.0
40.8
41.0
41.2
43.2
45.3
tu,
min
66
49
35
27
18
14
cu,
mg/1
84,500
67,500
51,500
43,000
33,000
29,500
TABLE 32. SURFACE AREA FOR SLUDGE THICKENING - PART III
Alternative
A
B
C
D
E
F
Suspended
sol ids, tons/day
2.85
3.32
3.40
3.42
3.32
3.30
Design
unit area,
ft2/ton-day
12.1
12.1
12.1
12.1
12.1
13.3
Area, ft2
34.5
40.2
41.2
41.4
40.2
43.9
75
-------
containing 38,440 mg/1 of suspended solids. Table 20 (Section VI)
indicates that the predicted feed rate for a 30,100 mg/1 sludge (the minimum
sludge concentration anticipated is 29,500 mg/1 for Alternative F) at a drum
speed of 0.1 rpm is 7.19 gal/ft2-hr. Even though significant rate increases
are projected for higher solids concentrations, the more conservative 6.6
gal/ft2-hr pilot rate was used to calculate design filter areas for each of
the six alternative treatment levels because of our higher level of confi-
dence in the field pilot data.
Anticipated sludge volumes and concentrations for each of the six alter-
natives are shown in table 33. Likewise, design filter areas are shown along
with expected filter vields. Surface area requirements vary from a minimum
51 ft2 to handle 8,100 gpd at 84,500 mg/1 suspended solids to 171 ft*
required for 27,100 gpd at 29,500 mg/1.
While the calculated filter area requirements increase proportionately
with increases in sludge volume, the cost per square foot of filtering sur-
face for rotary vacuum drum filters decreases substantially within the
50-250 ft2 range. For example, a major filter supplier's estimated cost for
a 110 ft2 rotary vacuum drum filter (June 1973} equalled $184/ft2, while the
estimated cost per foot for a 227 ft2 filter was only two-thirds of that
value ($123/ft2). This cost effectiveness weighed heavily in the final
decision to design the chrome recovery facility for Alternative F levels.
However, the determining factor in the design for the treatment of all
81,600 gpd of tanyard chrome-bearing wastes was the observed space efficiency
for chromium hydroxide thickening.
TABLE 33. FILTER AREA FOR CHROMIUM HYDROXIDE
Alternative
A
3
C
0
E
F
Settled
sludge
volume,
gpd
8,100
11,800
15,900
19,100
24,200
27,100
Feed
mg/1
84,500
67,500
51,500
43,000
33,000
29,500
solids,
Ib/day
5,700
6,630
6,800
6,850
6,650
6,600
0
Area, ft/
51
75
100
121
153
171
Filter yield,
Ib/ft2-hr
4.67
3.68
2.83
2.36
1.81
1.62
The theoretical area required for chromium hydroxide thickening as shown
in Table 31 is 45.3 gpd/ft2. Since in full-scale operations settling always
deviates from the ideal, a design factor of 1.75 (the normal range of applied
factors is 1.0-2.0) has been applied. The area was increased an additional
25% to provide capacity for future production increases. Thus, the calcu-
lated area for thickening is:
45.3 ft2 x 1.75 x 1.25 = 103 ft2
76
-------
A twelve foot diameter, 90° conical bottom fiberglass tank, having a nominal
surface area of 113 ft^ and a 5 ft sidewall (overall depth 11 ft), was
chosen. After installation of the flocculation-distribution chamber, the
actual area available for thickening was reduced to 107 ft?.
As shown in Table 33, 171 ft2 of filter area is required to handle the
27,100 gallons of sludge to be generated daily. This area was increased by
25% to 214 ft?, anticipating eventual increases in production levels. A 6'
diameter x 12' wide rotary vacuum drum filter with string discharge and a
Nylon multi-filament filter cloth was selected. A vacuum of 22" is supplied
by a manufacturer-recommended 40 h.p. blower with a nameplate capacity of
600 cfm. Likewise, a self-priming centrifugal pump, rated at 70 gpm, was
chosen for filtrate removal. All wetted metal parts were ordered in 316
stainless steel because of the highly corrosive conditions.
Alkali Addition
Equalized spent chrome wastes can be pumped to the Alkali Addition Tank
or recirculated to the Spent Chrome Collection Tank with a centrifugal pump.
Any desired flow within the 0-80 gpm capacity range can be delivered for
precipitation, using an in-line flow meter for control by precise adjustment
of the recirculating valve.
An alkali make-up tank has been sized for one 8-hour shift operation of
the Chrome Recovery System. A dish-bottom fiberglass tank with an effective
capacity of 340 gallons will supply 835 Ib of Ca(OH)2 per shift as a 2.5 lb/
gallon slurry. The lime slurry is pumped to the Alkali Addition Tank for
chromium hydroxide precipitation. Feed rates varying from 0.4 to 5.0 gpin
using a progressing cavity positive displacement pump with a variable speed
drive are possible. After calibrating the drive setting for the output from
the pump, alkali can be delivered proportionally by adjusting the pump drive
to the setting required by any given waste acidity.
To insure maximum precipitation of the chromium hydroxide, the Alkali
Addition Tank was specified at 400 gallons capacity providing a minimum of
5 minutes retention. A 0.75 hp chemical mixer further insures optimum pre-
cipitation. The bottom-center outlet is piped upward to the settling tank
feed trough providing a head-leveling device. Thus, during steady state
operation, precipitated chrome is withdrawn from the bottom by gravity,
while the volume in the Alkali Addition Tank is maintained near the 400 gallon
mark.
Fl peculation
The precipitated chromium hydroxide is flocculated with a liquid poly-
electrolyte - hlalcolyte 677 at a 30 ppm dosage. Polymer solution must be
prepared daily in the 100 gallon Polymer make-up tank. A 0.75 hp chemical
mixer provides rapid solution of the polymer. A variable discharge 600 ml/
min capacity diaphragm pump delivers the high viscosity polymer solution to
the head of the settling tank feed trough in proportion to wastewater flow.
The inclined trough is baffled to effect a "flash mixing" of the polymer with
the chromium hydroxide as it is fed to the center of the sludge thickening
77
-------
SPENT CHROME
ALKALI
I
_ z
-lO v
2Ez
IJQ<
< a1"-
POLYELECTROLYTE
1 RAPID-MIX FEED TROUGH
SLUDGE
THICKENING
---TANK--
SLUDGE
TO
FILTER
-BAFFLE
CHAMBER
Figure 19: FLOCCULATION OF CHROMIUM
HYDROXIDE.
78
-------
tank. The chromium hydroxide-polyelectrolyte mixture flows into the sludge
thickener through the baffle chamber/distribution box, as depicted sche-
matically in Figure 19. A series of parallel plates alternates flow between
the center and the perimeter of the central chamber, as illustrated, to
effect a gentle mixing pattern conducive to good floe formation. As the
sludge leaves the central chamber it is redirected towards the surface by
an outer chamber and is ultimately released into the quiescent zone approxi-
mately one foot below the surface for settling.
Sludge Pumping
Anticipated operation of the chromium hydroxide sludge thickener will
generate sludge at widely varying concentrations. A progressing cavity posi-
tive displacement pump was therefore chosen for sludge drawoff because of
its nearly constant output per revolution over a wide range of total dis-
charge heads. Also, this type of pump minimizes break-up of the floe
particles, therefore improving filtration. A variable speed drive allows
operation of the sludge pump anywhere within a 4.0 to 45 gpm range.
Cake Acidification
Chromium hydroxide filter cake from the rotary vacuum drum filter is
discharged to a 1000 gallon capacity Cake Solution Tank containing a solu-
tion of concentrated sulfuric acid and dissolved filter cake. A 1.5 hp
chemical mixer accelerates resolution of the cake, and a compression load
cell under the tank allows accurate process control throughout the 0 to
10,000 Ib net weight range. During continuous operation of the vacuum
filter, chromium hydroxide cake is redissolved to yield 4,000 Ib batches
of recovered chrome. When the pre-set gross weight of 6000 Ibs is reached
in the Cake Solution Tank, the feed screw conveyor is shut off and the tank
is drained to the 2000 Ib level (under design loadings the screw conveyor
is sized to hold cake equivalent to one hour's filter operation). The
recovered chrome is pumped to a pair of hydrocyclonic separators rated at
22 gpm through-put each at a feed pressure of 50-60 psi. Calcium sulfate
sludge is collected in barrels for daily disposal, while the clarified
chrome is repumped to a 2,800 gallon capacity fiberglass storage tank. A
minimum of 2,000 Ibs of dissolved cake is retained in the Cake Solution Tank
as a diluent for subsequent feeds of concentrated sulfuric acid. A new
batch commences with the addition of the acid (400-500 Ib/batch) followed by
addition of fresh chromium hydroxide cake.
79
-------
CO
o
TJ
(5*
O
X
3J
O
s
m
DO
m
o
O
<
m
DO
0)
H
m
SPENT CHROME POLYELECTROLYTE
-LUJ ALKALI ACID
pH MONITOR-
AUTOMATIC
VALVE
SCREEN
SOLIDS
GREASE«-i
COLLECTION
TANK-
FLOW
METER
SPENT
CHROME PUMP
OVERFLOW^
ALKALI
ADDITION
TANK
SCREW
CONVEYER
r
SLUDGE
THICKENING
"TANK (VACUUM
I PUMP
SLUDGE VACUUM
PUMP FILTER
FILTRATE
CYCLONE
PUMP
•*
CYCLONE
CAKE
SOLUTION
TANK
CALCIUM
SULFATE
•REUSE
CHROME
STORAGE
TANK
SEWER
-------
SECTION VIII
CHROME RECOVERY SYSTEM—FULL SCALE OPERATION
The full-scale Chrome Recovery System, as illustrated in Figure 20,
incorporates collection, storage, precipitation, sludge thickening, filtra-
tion and cake dissolving facilities for all tanyard chrome-bearing wastes,
all within a 1000 ft2 plant area.
Chrome-bearing wastes are pumped from each processing unit through a
common collection "header" pipe to a gravity screen (0.040" openings). The
pH in the chrome header is monitored continuously to protect against a
catastrophic introduction of'sulfide bearing wastes into the Chrome Recovery
System. If the pH rises above the-alarm level, an automatic air solenoid-
operated valve closes and a horn .alerts operating personnel to the upset.
The screened spent chrome solution is discharged to a 10,200 gallon
capacity collection-equalization-flotation tank, while the relatively small
amounts of suspended solids are collected for regular disposal. In the
Spent Chrome Collection. Tank, Oils and Greases separate fron^solutlon by
gravity to form a surface-scum layer which can be skimmed (or siphoned) off
daily for disposal. The low Oil and Srease underflow from the collection
tank is metered to the Alkali Addition Tank and mixed with a proportioned
feed of alkali. The chromium hydroxide suspension overflows the Alkali
Addition Tank into a "rapid-mixing" trough feeding the conical-bottom Sludge
thickening Tank. Polyelectrolyte is added at the head of the mixing-trough.
The inlet to the thickener consists of a sub-surface baffle chamber which
induces floe formation and distributes the waste radially into the settling
zone. Flocculated chromium hydroxide, settles into the conical sludge thick-
ening zone, while clarified effluent overflows a weir at the tank surface.
A positive displacement pump conveys the thickened sludge to the slurry tank
of a rotary vacuum: drum filter-. Filtrate from the vacuum filter can be
sewered, or, if- necessary, may be recycled to the Alkali Addition Tank to
utilize any available alkali and improve chrome removal. The filter cake is
removed by a string discharge mechanism, dumped onto a screw conveyor and
transported to the cake solution tank where it is dissolved in concentrated
sulfuric acid. The recovered chrome solution is pumped under pressures of
50-60 psi to .a hydrocyclonic separator to remove suspended calcium sulfate.
The calcium sulfate sludge, thus generated, is discharged into a hopper for
subsequent treatment and/or disposal. The "overflow" from the cyclones is
pumped to the recovered chrome storage tank for reuse in the tanning process.
Full-scale, 8-hour continuous operation of the Chrome Recovery System
confirms the design data developed during the bench- and pilot-scale
investigations.
81
-------
At a capture level of approximately 500 Ibs of chromium per day, treat-
ment efficiencies In excess of 98%* removal were recorded. Recovered,
acidified chrome liquor containing 0.6 Ib Cr203/gal was produced at the rate
of 1,000 to 1,200 gallons per day. Conconritantly, 3,000 to 3,600 Ib of sus-
pended solids as well as 330 Ib to 585 Ib of 011 and Grease were also
removed.
Lime-precipitation of the most concentrated spent chrome liquors, dis-
charged during one 24-hour period, yielded approximately 10,000 gal of
chromium hydroxide suspension. Observed solids loadings to the sludge
thickening unit slightly exceeded the 164 Ib/ft^-hr design loading. Likewise,
volumetric loadings of 516 gpd/ft2 exceeded the design rate of 400 gpd/ft2
for concentrated chrome. The sludge, thus generated, contained 7% solids,
and was dewatered to yield a cake of 30-35% solIds. Incremental rates for
the filtration of lime-precipitated chromium hydroxide as high as 7.5 gal
sludge/ft2-hr were observed (design rate=6.6 gal/ft2-hr) with correspond-
ing filter yields in excess of 3.7 Ib dry solids/ft2-hr. Even though these
optimal operation rates for the Chrome Recovery System cannot be maintained
for the total 8 hour interval, average filter rates as high as 5.1 gal/ft2-
hr with a corresponding yield of 3.0 Ib dry solids/ft2-hr have been observed.
Further, removal of Oil and Grease as a 50% scum effected a 92-95% reduction
through the use of the underflow draw-off from the Spent Chrome Collection
operation. Less than one percent of the Oil and Grease was discharged to
the sewer as filtrate and settling tank overflow, while the remainder concen-
trated in the filter cake and was thus.recycled.
Solids which require datly disposal include approximately 2,000 Ib of
chrome contaminated calcium sulfate sludge, as well as about 150 gallons of
supernatant Oil and Grease "scum". Also, about 3 gallons of screenings are
collected which may be reduced to three'pounds of screened solids by a
second pass over the static screen.
CHROME RECOVERY BY LIME PRECIPITATION
In another 8-hour continuous chrome recovery run, 960.0 gallons of con-
centrated spent chrome liquor was collected from eleven hide processors over
a 24 hour period. The spent chrome, which contained 6,320 mg/1 of chromium
(506 Ib Cr), was precipitated by the addition of 1,050 Ib of hydrated lime
(as a 2.5 Ib Ca(OH)2/gal slurry) within 4.5 hours to generate 10,000 gal of
chromium hydroxide suspension. As shown in Table 34, the resultant sus-
pended solids concentration equalled 38,420 mg/1. Approximately 120% of the
stoichiometric dosage of lime was required to maintain pH within the 8.7 to
9.0 range. After the controlled addition of 30 ppm of liquid polymer
(Nalcolyte 677 added as a 1% solution), the chromium hydroxide sludge was
fed at a rate of 33.5 gpm to the sludge thickener which was pre-charged with
6,000 gal of supernatant from a prior run. The equivalent volumetric load-
ing was thus 516 gpd/ft2 (107 ft2 surface area), while the solids feed rate
averaged 166 Ib/ft2-day.
* Since "end-of-the-pipe" chromium levels of 1340 Ib Cr per day are typical
overall effluent reductions of approximately 37% are calculated.
82
-------
TABLE 34. CHROME RECOVERY SYSTEM PERFORMANCE - LIME PRECIPITATION - PART I
CO
CO
Volume,
gallons
Spent Chrome
Feed
Overflow
Sludge
Filtrate
Cake
Cyclone Feed
Recovered Chrome
CaS04
9
10
4
5
4
(9
1
1
145
,600
,000
,850 *
,150
,300 **
,960 Ib)
,155
,010
(2,020 Ib)
TSS,
mg/1
-
38
1
72
-
205
79
,420
,740
,060
82
,000
,200
Ib
-
3,200
70.3
3,090
39.4
-
1,970
666
TS,
mg/1 Ib
101
68
126
61
(30
300
(58
-
,480
,480
,460
,440
.25%)
-
,416
.8%)
8
2
5
1
3
2
1
-
,460
,760
,430
,810
,020
-
,530
,190
Cr,
mg/1
6,320
6,040
250
11,120
3.4
-
-
53,800
(2.22%)
Ib
506
503
10.1
478
0.12
-
-
452
44.9
* By difference
** Estimated at 83% of sludge volume, per feasibility studies.
-------
Throughout the course of the 4.5 hour thickening operation, severe
foaming was observed. A thick pillow of foam was also observed above the
Alkali Addition Tank and the sludge thickener feed trounh suggesting that
foaming was a consequence of the lime addition and mixing (turning off the
high speed chemical mixer, however, did not significantly reduce foaming).
This foaming adversely affected the separation process as 1,740 mg/1 of
suspended solids and 250 mg/1 of chromium were detected in the settling tank
overflow. Also, a sample of collapsed foam from the settling tank analyzed
at 9,570 mg/1 of chromium. As a result of this foaming, 2.2% of the sus-
pended solids and 2.0% of the chromium contained in the sludge feed was dis-
charged in the overflow. This problem of foaming was subsequently corrected
by re-design of the Alkali Addition Tank overflow-head leveling device and
the feed trough to decrease the velocity of the feed to the sludge thickener.
Withdrawal of freshly thickened sludge commenced two hours after the
chromium hydroxide feed began. Approximately 5,200 gal of sludge at a mean
suspended solids concentration of 72,060 mg/1 was drawn off at an average
rate of 15 gpm and filtered on the rotary vacuum drum filter in 5.67 hours.
Cake thickness ranged from 4.8mm to 6.4mm. At a drum speed of 0.115 rpm,
37.5% maximum submergence and under a cake form vacuum of 26" of mercury,
an average sludge feed rate of 4.1 gal/ft2-hr to the filter was maintained.
This corresponds to an average filter yield of 1,750 Ib of wet cake per hour
at approximately 30% solids-equivalent to 2.34 Ib dry solids/ft2-hr. During
the final hour of filter operation a maximum filter rate of 6.6 gal feed/
ft2-hr was observed with a corresponding filter yield of 2.86 Ib dry solids/
ft2-hr.
Approximately 4,300 gallons of filtrate containing only 82 mg/1 of sus-
pended solids and 3.4 mg/1 of total chromium was also produced.
Acidification of the filter cake required 1,500 Ib of sulfuric add
(66° Be/) and yielded 1,155 gallons of a recovered chrome-calcium sulfate
suspension containing 205,000 mg/1 of suspended solids (of which 153,000 mg/1
were fixed suspended solids). This slurry was fed at an average rate of
33.5 gpm to a pair of hydrocyclonlc separators with orifice diameters of
3/16" (Bauer Clarifier No. 677-1). Approximately 200 gallons of sludge was
generated and subsequently decanted to produce 145 gallons (0.72 yds) of
thickened calcium sulfate containing 58.8% solids (57% fixed solids) and
2.22% Cr. The "overflow" from the cyclones plus the decanted supernatant
from calcium sulfate sludge thickening equalled 1,010 gallons and contained
80,500 mg/1 of suspended solids for a 61% reduction. The recovered chrome
analyzed at 0.655 Ib CreOa/gal (0.448 Ib Cr/gal), with a pH of 2.9 and a
basicity of 39.4%.
As Table 34 illustrates, of the 506 Ib of chromium charged to the Chrome
Recovery System, 10.1 Ib was discharged as overflow, and another 0.12 Ib was
discharged as filtrate, for a net removal efficiency of 97.98%. A substan-
tial 8.86% of the chromium (44.9 Ib Cr) was transferred to the calcium sul-
fate sludge, which would ultimately require land disposal. Likewise, of the
3,200 Ib of suspended solids entering the sludge thickener, 70.3 Ib was
found in the overflow and 39.4 Ib in the filtrate, for a net removal of
96.58%.
84
-------
TABLE 35. CHROME RECOVERY SYSTEM PERFORMANCE - LIME PRECIPITATION - PART II
00
en
Volume,
gallons
Spent
Chrome
Feed
Overflow
Sludge
Filtrate
Cake
Cycl one
Feed
Recovered
Chrome
CaS04
9
10
4
6
5
(10
1
1
(1
,940
,300
,160*
,140
,100**
,800 Ib)
,185
,035
,850 Ib)
Suspended
solids,
mg/1
-
42,060
300
67,740
210
-
252,800
154,000
(58.2%)
Ib
-
3,620
10.4
3,460
8.9
-
2,500
1,330
1,080
Total
solids,
mg/1
-
105,200
72,180
125,680
73,380
(33-. 4%)
566,260
464,960
-
Ib
-
9,050
2,490
6,440
3,140
3,610
5,480
4,010
-
Total
fixed
solids
mg/1
-
100,160
71,108
115,980
72,260
(26.72%)
308,260
246,280
(44.0%)
8
2
5
3
2
2
2
. Oil & Grease Cr
Ib mg/1
7,460
,720 379
,460 39
,930 671
,070 18
,880 (0.31%)
,790
,130
814
Ib mg/1
618 6,000
32.5 5,820
1.35 10.3
34.3 9,340
0.76 2.10
33.7*
48,500
49,200
(2.65%)
Ib
497
500
0.36
478
0.09
-
480
426
49
* by difference
** estimated at 83% of sludge volume, per feasibility studies
-------
In another typical chrome recovery run 9,940 gallons of concentrated
chrome liquor from eleven hide processors, containing 497 pounds of total
chromium (6,000 mg/1), was processed within an 8-hour production shift. As
shown in.Table 35, 1,035 gallons of 25% basicity recovered chrome (at 0.60 Ib
CrgOi per gallon) and 1,850 Ib of calcium sulfate sludge were generated.
Elimination of foaming during lime-precipitation of the chromium hydroxide
greatly improved the chrome separation efficiency to 99.9%, yielding average
supernatant and filtrate chromium concentrations of only 10.3 mg/1 and 2.10
mg/1, respectively.
The spent chrome was precipitated with 1,200 Ib of hydrated lime (as a
4.0 Ib/gal slurry) and flocculated with 30 ppm of Nalcolyte 677 polyelectro-
lyte, to generate 10,300 gallons of chromium hydroxide suspension. The
slurry was fed to the sludge thickener at a mean suspended solids concentra-
tion of 42,060 mg/1 within a 6 hour interval. Accordingly, an average volu-
metric loading to the sludge thickening unit of 375 gpd/ft* was demonstrated
with a corresponding solids feed rate of 135 Ib dry solids/ft2-day.
An Oil and Grease reduction of 94.7% was effected through gravity sepa-
ration in the Spent .Chrome Collection Tank. While the raw waste analyzed at
7,460 mg/1 of Oil and Grease (618 Ib), the feed to the sludge thickening
unit contained only 379 mg/1. The Oil and Grease-laden scum was manually
skimmed from the Spent Chrome Collection Tank through an overflow, and col-
lected for eventual land disposal.
Of the Oil and Grease remaining after gravity separation, an estimated
33.7 Ib was concentrated in the filter cake, while the sludge thickener over-
flow and the filtrate contained only 1.35 Ib and 0.76 Ib, or 0.22% and 0.12%,
respectively.
An estimated 4,160 gallons of overflow were generated, containing 300
mg/1 of suspended solids and only 39 mg/1 and 10.3 mg/1 of Oil and Grease
and chromium, respectively.
Approximately 90 minutes after the initial precipitation of chromium
hydroxide, the withdrawal of thickened sludge for vacuum filtration com-
menced at an average rate of 19.3 gpm. Within 5.33 hours, 6,140 gallons of
thickened sludge containing 67,740 mg/1 of suspended solids was filtered at
an average feed rate of 5.1 gal/ft2-hr. The corresponding filter yield was
2,040 Ib wet cake/hour (33.4% solids) equivalent to 2.99 Ib dry solids/
ft^-hr. The maximum incremental feed rate to the filter was 7.5 qal/ft^-hr,
while a maximum incremental filter yield of 3,93 Ib dry solids/ft^-hr was
realized.
The filter cake, which averaged 33.4% solids, was dissolved in 1,350 Ib
of sulfuric acid to yield 1,185 gallons of chrome-calcium sulfate slurry,
containing 252,800 mg/1 of suspended solids. Hydrocyclonic separation of
this slurry effected a 39% reduction in suspended solids to yield 1,035
gallons of recovered chrome and 150 gallons of a 58.2% calcium sulfate
solids sludge.
86
-------
A mass-balance analysis (Table 35) of this chrome recovery run reveals
that while a total of only 0.45 Ib of chromium was discharged in the sludge
thickener overflow and filtrate, an additional 49 Ib or 10.2% of the chromium
contained in the cyclone feed was discharged with the calcium sulfate sludge,
requiring land disposal. Thus, the recovery efficiency for this run was
actually only 86%, with 426 Ib Cr in the recovered chrome and an estimated
19 Ib Cr was retained in the sludge thickener for recovery 1n the subsequent
run.
CHROME RECOVERY BY SODA ASH PRECIPITATION
To eliminate the calcium sulfate separation and disposal required for
chrome recovery by the lime precipitation process, soda ash was substituted.
The operating procedures which were eventually adopted differed from those
required for lime due to the extremely slow settling rate of soda ash-
precipitated chromium hydroxide. The most concentrated spent chrome liquors
from eleven hide processors could be precipitated and charged to the sludge
thickening unit within a single 8-hour shift. Only about 70% of the equiva-
lent quantity of chromium hydroxide filter cake, however, could be generated
within the same interval. After allowing precipitated sludge to thicken
overnight, approximately 11,000 gallons of fresh chromium hydroxide was
charged to the sludge thickener on the following day, while a proportional
quantity of day-old thickened sludge was simultaneously drawn off and fil-
tered within eight hours.
In one run, 9,100 gallons of spent chrome was precipitated with soda ash
and the freshly thickened sludge was simultaneously withdrawn for fil-tration
within the same 8-hour interval. Chromium hydroxide suspension containing
24,000 mg/1 of suspended solids was fed to the sludge thickening unit over a
7-hour period at an average rate of 24.2 gpm. The average volumetric feed
rate to the sludge thickener was 329 gpd/ft2, while a corresponding solids
feed rate of 66 Ib dry solids/ft^-day was observed.
The fact that a small amount of aged sludge from the previous chrome
recovery run remained 1n the sludge thickener is Illustrated by the observed
98,000-99,000 mg/1 sludge concentrations during the first hour of sludge
withdrawal. After two hours, however, sludge suspended solids leveled off
and remained nearly constant throughout the final six hours of sludge with-
drawal .
Filtration of 7,000 gallons of sludge, which averaged 35,000 mg/1 of
suspended solids, yielded 7,500 Ib of filter cake within eight hours. The
average rate of thickened sludge feed during 7.67 hours of vacuum filtration
was 4.0 gal/ft2- hr (912 gal/hr), while filter yield averaged 980 Ib wet cake
per hour, equivalent to 1.12 Ib dry solids/ft^-hr.
Acidification of the filter cake required 950 Ib of 66° Be'sulfuric acid
and yielded 850 gallons of recovered chrome liquor, which analyzed at
0.596 Ib Cr203/gal (0.408 Ib Cr/gal) and 33.7% basicity.
87
-------
In another soda ash-precipitation run, day-old thickened sludge was
withdrawn from the sludge thickener and filtered, while 9,700 gallons of
fresh spent chrome was simultaneously charged to the Chrome Recovery System,
increasing the yield of recovered chrome within an 8-hour production shift
to 1,200 gallons containing 742 Ib of 0^03 (508 Ib Cr).
The spent chrome which was collected from eleven hide processor runs
(four of which were completed during the 8-hour chrome recovery run) con-
tained 6,170 mg/1 of Cr and required the stoichiometric addition of 2,350 Ib
of NagCOs (as a 2 Ib/gal solution) to yield 10,850 gallons of chromium
hydroxide precipitate at a pH of 8.5-9.0. The chromium hydroxide slurry
was charged to the sludge thickener at the average rate of 22.5 gpm, equiva-
lent to a volumetric loading of 305 gpd/ft2. A solids feed rate of 62.5 Ib/
ft^-day was observed at a mean feed suspended solids concentration of
24,700 mg/1.
As shown in Table 35, the overflow generated equalled half of the feed
volume and contained 710 mg/1 of suspended solids, equivalent to 1.5% of
the feed solids. Chromium and Oil and Crease in the overflow were likewise
reduced to 169.5 mg/1 and 57 mg/1, respectively.
Approximately 5,400 gallons of sludne was withdrawn over the eight
hour run, averaging 49,000 ma/1 of suspended solids. Filtration of the
sludge yielded 10,600 Ib of cake which contained 4.85% Cr and 76.4% H?0. The
mean rate of filtration was 3.0 gallons of feed per ft2-hr while the filter
yield averaged 1.21 Ib dry solids/ft2-hr (1,160 Ib wet cake/hr) and cake
thickness ranged from 4.0mm to 8.0mm. The maximum incremental filter rate
observed was. 4.5 gal feed/ft^-hr and incremental filter yields exceeded
1,980 Ib '.vet cake/hr equivalent to 2.05 Ib dry solids/ft^-hr.
Reacidification of the cake required 1,350 Ib of HgSffy and yielded
1,200 gallons of recovered chrome liquor at a pH of 3.2 containing 0.618 Ib
Cr203/gal (0.42 Ib Cr/gal).
Approximately 4,300 gallons of filtrate were generated containing 302
mg/1 (10.8 Ib) of suspended solids. Thus, only 0.5% of the feed solids
charged to the sludge thickener was discharged as filtrate. Likewise, Oil
and Grease analyzed at 30 mg/1 and chromium at 40.5 mg/1.
As Table 36 illustrates, the 9,700 gallons of spent chrome collected
from eleven hide processors contained 498 Ib of chromium. The chromium thus
captured comprises approximately 37% of Pfister & Vonel's "end-of-the-pipe"
effluent chrome loading. A mass-balance analysis of our chrome recovery
operation using soda ash to precipitate the chrome shows that 98.1% recovery
of the isolated chromium has been achieved. The sludge thickener overflow
contained 1.6%, while the filtrate contained only 0.3% of the chromium input
to the Chrome Recovery System. The combined effluent from the settling tank
overflow and vacuum filter filtrate equalled 9,775 gallons at an average
chromium concentration of 118 mg/1. Likewise, a 98.9% reduction in Oils and
Greases sewered during this dynamic collection-recovery process has been
demonstrated. Approximately 92% of the Oils and Creases was separated as
scum in the Spent Chrome Collection Tank, while only 0.8% and 0.3% were
88
-------
TABLE 36. CHROME RECOVERY SYSTEM PERFORMANCE - SODA ASH PRECIPITATION
00
vo
Spent
Chrome
Feed
Overflow
Sludge
Fi 1 trate
Cake
Recovered
Chrome
Volume,
gallons
9,690
10,850
5,425
5,425
4,350
(10,590 Ib)
1,200
TSS,
mg/1
3,800
24,700
710
48,320
302.
(23.6%)
14,500
Ib
307
2,230
32
2,200
10.9
2,500
145
TS,
mg/1
103,190
107,050
81,520
120,520
85,480
(23.6%)
274,000
Ib
8,330
9,780
3,680
5,450
3,230
2,500
2,740
TFS
rag/T
81,610
100,140
78,000
107,580
82,750
(17.3%)
199,500
>
Ib
6,580
9,060
3,530
4,870
3,000
1,820
1,995
Cr,
m.q/1
6,168
5,495
179.5
10,620
40.5
(4.85%)
50,800
Ib
498
496
8.12
481
1.47
514
508
O&G,
mg/1
4,408
311
57
732
30
(0.32%)
.
Ib
357
28.1
2.58
33.1
1.09
34.0
„
-------
found 1n the thickener-overflow and vacuum filter filtrate, respectively.
The remaining Oil and Grease was retained in the filter cake and appeared
in the recovered chronne. Also, a 98.2% recovery of suspended solids as
filter cake from the chromium hydroxide precipitate has been demonstrated.
The combined sludge thickener overflow and vacuum filter filtrate contained
an average of 538 mg/1 of suspended solids.
CHROME RECOVERY OPERATING COSTS
The costs for recovering the chrome from approximately 10,000 gpd of the
most concentrated spent chrome liquors (containing 6,260 mg/1 of Cr) has been
calculated (Table 37). At January 1976 prices for chemicals, solids disposal,
electricity and labor, the cost of recovery by the soda-ash precipitation
process is $248.30 per day. The corresponding cost for lime-precipitation
is $155.50 per day. Thus, within the price range for commercially available
prepared chrome tanning solutions, the estimated net annual savings would
vary from $50,000-$126,000 for soda ash to $68,000-$135,000 for lime precipi-
tation.
TABLE 37. CHROME RECOVERY SYSTEM OPERATING COSTS -
SODA ASH VS LIME PRECIPITATION
Daily Cost,* $
Unit Cost Soda ashLime
L1me $3.28/100 Ib - 39.40
Soda ash $5.01/100 Ib 117.80
Sulfuric acid $2.58/100 Ib 34.80 34.80
Polyelectrolyte $1.12/lb - 2.90
CHEMICALS SUB-TOTAL $152.60 $ 76.10
Electricity 2.2<£/kwh 11.70 8.80
Solids disposal $3.50/yd3 - 2.60
Labor $8.00/hr 64.00 48.00
Sub-Total 228.30 135.50
Contingencies 20.00 20.00
TOTAL $248.30 $155.50
Spent chrome volume = 10,000 gpd.
Spent chrome concentration = 6,260 mg/1 as Cr.
90
-------
RECOVERY OF OILS AMD GREASES
Approximately 10,000 gallons of spent chrome was passed through the
48" wide static screen (0.040" openings) and collected in the Spent Chrome
Collection Tank. The Oil and Grease was allowed to separate overnight under
quiescent Conditions. After 16 hours a water hose was inserted midway down
the tank and clean water was carefully added to raise the level of the Oil
and Grease layer to that of the tank overflow pipe. Approximately 50
gallons of the surface scum (Sp.Gr.=0.945) was skimmed through the overflow
and collected. A two-step chloroform extraction of a portion of this scum
(pH 3.5) revealed an Oil and Grease content of 51.4%.
Laboratory tests to explore possible routes to purification and/or
recovery of the contained Oils and Greases were undertaken as follows:
An aliquot of the scum was allowed to separate under quiescent condi-
tions for two hours. A three-phase separation was observed in which the
lower aqueous layer equalled 17% of the original volume. The middle layer
which consisted of an oily suspension of tiny fibrous particles equalled 82%
of the original volume and the top layer which appeared as a brown semi-
translucent oil represented only 1% of the total volume.
Acidification to a pH of approximately 1 effected a two-phase separation
with a well defined interface within 2 hours. After 10 minutes boiling on a
hotplate and subsequent overnight settling, a three-phase system was estab-
lished in which the third phase consisted of a top layer of a brown oil.
Another 200 ml .aliquot was brought to a boil and allowed to cool and
settle overnight. The resultant three-phase separation gave approximately
20% of the oily substance.
Finally, a 7.5 liter portion of the scum was acidified to pH 0.7 with
a 5% sulfuric acid solution. Live steam was injected to bring the mixture
to a boil for 15 minutes. Overnight settling and cooling to room temperature
effected a separation into three phases. Decantation yielded two liters of
the brown oily substance. Likewise, another liter of the brown oily suspen-
sion which had composed the intermediate layer was segregated.
The brown oil which was separated had a specific gravity of 0.900 and
was not steam distillable at pll 1.0. While several materials were investi-
gated to decolorize the oil, including activated charcoal, distilled water,
saturated sodium chloride solution, Hyflo Supercel (a filter aid), Celatom
(a diatomaceous earth), Clay R and two anionic and one cationic polyelec-
trolyte solutions, none of these produced a significant decolorization of
the oil.
91
-------
SECTION IX
CHROME RECYCLE - MATCHED PAIRS LEATHER TESTS
Any optimum procedure to be used for the recycle of chrome in the manu-
facture of leather must be considered first for its "fit" with highly sub-
jective tanning methods and procedures employed by each tanner. At Pfister
& Vogel, without seriously disrupting existing processes or equipment utili-
zation, the following routes for chrome recycle could be postulated, and were
investigated with varying degrees of success.
REUSE OF SPENT CHROME AS PICKLE FLOAT
A matched-sides, leather-comparative test demonstrated the effect of
reusing spent chrome liquor as the pickle float in a subsequent hide proces-
sor run. Approximately 850 gallons of spent chrome (equivalent to 50%
float) containing 6,390 mg/1 of Cr was added to drained bated stock, replac-
ing the normal salt brine feed.
Analysis of the test pickle exhaust liquor revealed that 54% of the
chromium was "scrubbed" from the float during the pickle run. Even though
30% excess acid is offered in the recycled liquor, a pickle pH of 3.2 was
observed, rather than the "expected" 2.85 for the regular. Approximately
two-thirds of this spent pickle liquor (containing 14.3 Ib Cr) was sewered,
while the remaining 6.5 Ib Cr was carried over into the tanning run.
Though the amount of fresh chrome offered in tanning was reduced in the
test run by 5%, the chrome exhaust liquor contained 23% more chromium than
the regular.
This "one-time" recycle effected a 9% reduction in the total amount of
chromium sewered from 75.4 Ib Cr for the regular run to an average of 68.7
Ib Cr for the two consecutive runs required to demonstrate one reuse of the
spent chrome.
Chemical analysis of the finished leather showed only slight differ-
ences, with the test sides containing 6.5% more ^03 on a hide substance
basis.
Analysis of the blue stock after wringing, however, revealed a 14%
increase in the ^03 content of the test side grains and a 16% increase
for the test splits on a moisture free basis, suggesting that further sig-
nificant reductions in the amount of fresh chrome offered for tanning may
be possible. In addition, though there was no measurable difference in the
92
-------
fat content of the grain layers, the splits showed a 33% reduction in fats
on a moisture free basis.
At the final sort the test sides were decidedly better for break
(eight out of eight), slightly firmer, slightly poorer for flesh and com-
parable to the regular sides for smoothness, fat wrinkles, surface break,
hair cell and veins. The test sides were also consistently lighter in
weight than the regulars. Most notably, however, the physical strength of
the test sides was substantially poorer, as measured by the Satra grain
crack, Satra extension and slot tear. The Satra force was reduced in this
test from an average value of 63.8 kg for the controls to 54.6 kg for the
test sides, with a corresponding reduction in the extension-at-grain burst
from 9.49 mm to 9.04 mm. Likewise, a reduction in slot tear strength from
67 Ib to 57 Ib was observed.
REUSE OF RECOVERED CHROME BY BLENDING WITH STOCK CHROME
A matched sides leather test in hide processors has demonstrated the
leather quality effects of blended, recovered chrome-fresh chrome mixture
used as a primary tanning agent.
A soda ash-precipitated, filtered and reacidified spent chrome liquor
was used to replace 37% of the C^O-s used in a normal processing run. This
represents the maximum "theoretical recycle that can be calculated from the
overall material balance for chrome in our operations. The recovered chrome
was blended with fresh chrome and subsequently added in a single feed. Due
to the lower basicity of the recovered chrome the quantity of acid used in
the test run was correspondingly reduced. Evaluation of the normal process
control indicators showed that the "behavior" of this recycled chrome blend
was not significantly different than fresh chrome during the hide processor
chrome-tanning process. Likewise, chemical analysis of the finished
leather indicated no significant differences between the test and control
sides. However, chemical analysis of the full thickness blue stock had
shown a 7.35% increase in the C^Oo content of the test sides on a hide sub-
stance basis. Upon stratagraphic five-layer analyses, the middle blue-stock
layers showed the greatest difference, with the test sides containing 30%
more C^O-j on a hide substance basis, indicating increased chrome penetration
for the chrome recycle test run.
The final sort for the subjective parameters of leather quality indi-
cated that the test-sides were considerably softer than the regulars (six
pairs out of ten). They were equal for hair cell, finish break, veins,
flesh and break. Again, however, the physical testing for stitch tear and
grain burst indicated a substantial loss of strength in the test sides. The
Satra force decreased from an average 46 kg in the regular sides to 35 kg in
the test sides with a corresponding reduction in extension to grain burst
from 8.59 mm to 8.12 mm. Likewise, the stitch tear was reduced from 61 Ib
to 53 Ib.
93
-------
REUSE OF RECOVERED CHROME BY TWO-STAGE CHROME ADDITION
A matched sides leather test usinq recovered chrome (from soda ash-
precipitated chromium hydroxide) to replace 37% of the chrome offered in
our normal chrome tannage indicated no deleterious effects when the lower
basicity recycled chrome was fed separately before the regular fresh chrome.
The total amount of sulfuric acid used was decreased to adjust the acidity
of the combined tan liquor within normal limits. Also, the total chrome
used was decreased by 5% to compensate for the previously observed increased
take-up during tanning.
Chemically, the finished test sides did not differ significantly from
the regular sides. Likewise, the subjectively-evaluated leather qualities
were equivalent, except that six of nine test sides were noticeably softer
than their matching control sides, and four of the test sides were slightly
smoother.
In this test, the comparative physical strength of the test sides was
improved over our previous chrome recycle test run. The Satra grain burst
strengths were found to be equal, while the stitch tear, this time, was
better—with the test sides averaging 44 Ib vs 40 Ib for the controls.
A second matched sides-leather test using a 33% basic recovered chrome
liquor to replace 23% of our normal chrome was also conducted. This level
is especially significant as it corresponds to the recycle of the chrome
from the most concentrated spent chrome fraction {as a reacidified, soda ash-
precipitated and filtered chrome) from a previous processing run.
Again, the recovered chrome was fed separately, before the regular
fresh chrome, and the acid was decreased to adjust for the difference in
basicity. The chrome take-up efficiencies, as calculated from chrome ex-
haust liquor analyses, were equal.
Chemically, there was no significant difference between the finished
test sides and the controls. Analysis of full thickness blue stock after
wriglng, however, Indicated that the test sides contained approximately 10%
more Cr203 (moisture free) than the regulars.
At the final sort five of the nine test sides were improved with the
other four judged equal for fat wrinkles. The test sides were also slightly
better than the controls for smoothness, softness and flesh, while no differ-
ence was observed for break, veins, or v/eight. The physical strength of the
test sides averaged slightly less using the Satra grain burst strength as
the criterion (45 kg vs 50 kg). Likewise, the Satra extension was slightly
less at an average of 8.30 mm vs 8.52 mm for the controls. The stitch tear
strength of the test sides, however, equalled the controls.
REUSE AS MAKE-UP IN REDUCED CHROME PREPARATION
Basic chromic sulfate tanning agents are produced by the reduction of
sodium bichromate in acid solution. Most commonly, organic reducing agents
94
-------
such as molasses or corn syrup are used, although other reducing substances
including glycerine, alcohol, dextrin, flour or sulfur dioxide have also
been used {23}. Careful control of all aspects of the bichromate reduction
reaction including temperature, rate, reaction time and mixing must be main
tained to produce a reduced chrome liquor with uniform tanning activity.
Likewise, it is essential that the proportion of chemicals employed remains
constant from batch to batch, since even a slight change will significantly
affect the basicity and thus the activity of the finished chrome.
One hypothetical equation for the reduction of sodium bichromate by
glucose to yield a 33-1/3% basic chrome liquor (similar to the equation
expressed by Orthman)(24) is shown below.
4 Na2Cr207. 2!I20 + 12 !I2S04 + Celine + 58 II20 =
(Sodium Bichromate) (Sulfuric (Glucose) (Water)
Acid)
4 Na2S04 . 10 H20 + 6 C02 + 8 Qr (Oli)
(Sodium Sulfate) (Carbon Dioxide) (33-1/32 basic chromic
sulfate)
In practice, the proportions of reactants vary considerably from these
theoretical quantities. The "hydrated chrome" thus formed is actually com-
posed of a multiplicity of chromium complex salts whose molecular species
probably include the one for 33-1/3% basic chromic sulfate hydrate shown
above. The corresponding Theoretical formula for 43.5% basic chromic sulfate
hydrate is:
[cr(OH) (H20) "1
L 1.455 4.54_5]
0.7725
This mathematical model shows a decrease in hydroxylation vs the 33.3% basic
chrome species.
In a typical production-scale "chrome boil" 5,000 Ib of hydrated sodium
bichromate is reduced by corn syrup to produce a 48.5% basic chrome liquor.
After dilution of the sodium bichromate to 1,575 gallons, 4,300 Ib of 66° Be
sulfuric acid is added. Using the heat of dilution of the concentrated acid
to initiate the reaction, corn syrup equivalent to 1,070 Ib of sugar is
metered into the reaction medium at a sufficient rate to maintain a gentle
boil. After sugar addition is complete, a small amount of sodium bisulfite
is added to insure complete reduction to tri-valent chromium. Before any
batch of prepared chrome is released for subsequent use, it is tested and
must show a zero hexavalent chromium result. The approximate "chrome boil"
yield is 1,700 gallons of stock chrome liquor containinn 1.5 Ib Cr203 per
gallon.
Since our recovered chrome liquor analyses at approximately 0.6 Ib
Cr203 per gallon (0.4 Ib Cr/gal), either further concentration, or blending
with high concentration chrome liquor, or fortification with spray-dried
chrome sulfate is required before the material can be fed back into the
95
-------
"normal" chrome distribution-leather production system.
In practice, we employ the first of these techniques by utilizing the
energy of the sugar-sodium bichromate redox reaction in an augmented chrome
boil.
Recovered, acidified chrome sulfate from the alkali precipitation-
filtration process was used as a portion of the initial diluent in a typical
chrome boil. Because of the lower basicity of the recovered chrome the
amount of acid normally used can be reduced, taking a credit for the excess
acidity.
No major differences, in the conduct of the boil for either lime-
precipitated or soda ash-precipitated recovered chrome were observed.
Augmentation with recovered chrome containinq as much as 17% of the
0203 content of a "normal" chrome boil produced a high concentration chrome
liquor which analyzed within the desired (48-49%) basicity ranne. Dilution
of the chrome liquor to the normal stock concentration enabled subsequent
recycle of the chrome without modification of the tannim process.
96
-------
SECTION X
HYDROGEN SULFIDE CONTROL
TOXICITY
Hydrogen sulfide is a highly toxic colorless gas which, although
mal-odorous in low concentrations, rapidly becomes odorless at physiologi-
cally-active concentrations. It is both an irritant and an asphyxiant.
Exposure to concentrations as low as 20-150 ppm may cause eye irritation,
while slightly higher levels may cause respiratory irritation and lung
edema. A 30 minute exposure to 500 ppm of hydrogen sulfide may result in
dizziness, headache, staggering, loss of consciousness, diarrhea, dysuria,
bronchitis and bronchopneumonia. Finally, exposure to 800-1000 ppm can be
fatal in 30 minutes or less (25).
Hydrogen sulfide in solution exists in equilibrium with both sulfide
and sulfhydrate ions according to the following relationship:
HS" + H+
HS
MS" = S= + HH
Normally, below pH 12, the equilibrium concentration of sulfide ion is
extremely low, whereas relative concentrations of H2$ and HS~ are effectively
a function of pll. The extent of hydrogen sulfide occurrence in solution can
be calculated from the first ionization constant for the dissociation, as
follows:
log
HS"
] = pll - pK
H2S
While the value of pK varies slightly with the ionic strength of solution,
at 25°C, 7.0 (26) is a good approximation for exemplary calculations. Table
38 shows expected proportions of hydrogen sulfide at various values of pH
(27).
97
-------
TABLE 38. PROPORTIONS OF DISSOLVED SULFIDE PRESENT AS H2S
plf
10.0
9.0
8.0
7.0
6.0
5.0
4.0
Proportion
as H2S
0.001
0.010
0.091
0.50
0.91
0.99
0.999
Above pH' 9.0 the relative concentration of hydrogen sulfide is insignificant.
As the pH is reduced to 7.0, however, approximately half of the hydrosulfide
ion species is converted to hydrogen sulfide, and at pH's less than 7.0,
hydrogen sulfide predominates.
Hydrogen sulfide is moderately soluble in water (3,175 nig/1 @ 25° C).
It does exert a significant partial pressure, however, even_at very low solu-
tion concentrations. Only 1 mg/1 of hydrogen sulfide (as S~) in solution
(25° C), for example exerts a partial pressure of 0.0024- mm of mercury
(28), equivalent to 315 ppm of HoS by volume. The H£S concentration in the
air increases proportionally witn solution concentration.
Since the pi! of unhairing solutions generally exceed 11.5-12.00, the
dissociation equilibrium is shifted far to the right and the solution concen-
tration of hydrogen sulfide in these lime-sulfide wastes is inconsequential.
If the pH in these solutions is reduced, however, by chemical additions dur-
ing the process or due to the inadvertent mixing with acid wastes in plant
sewers, the relative concentration of hydrogen sulfide may increase drasti-
cally. The resultant release of hydrogen sulfide gas to the atmosphere could
be extremely hazardous!
IN-HOUSE CONTROL
Since no practicable alternative to the alkali-sulfide method of unhair-
ing currently exists, every tanner must maintain rigorous control over the
handling of all sulfide-containing raw materials, process liquors, and waste
streams. Required control measures include (but certainly are not limited
to):
Proper^Identification, Storage and Transfer of Hazardous Chemicals
In the recently proposed City of Milwaukee Code revision for "Labelinn
of Hazardous Materials Storage Systems," (modeled from a Pfister & Vonel
"in-house" system) all hazardous materials receiving pipelines, storage tanks
and transmission lines should be labeled with appropriate DOT labels,
generic names and directional arrows. Each receivinn pipeline connection
98
-------
must be padlocked with a different lock, and each key must he clearly iden-
tified and secured in a remote area with 24 hour supervision. Finally, all
connecting to plant receiving lines should be done under the control of duly
trained and competent plant personnel.
Training
All employees who use hazardous materials must be adequately trained in
safe handling procedures. This training, in addition, should include an
explanation of the consequences of careless use of the materials, a demon-
stration of the proper use of personal protective equipment, emergency egress
procedures, and the insistence that safe operating practices must not be
compromised under any circumstances.
Adequate Ventilation
During the "deliming" step of the beamhouse-tanyard process, the de-
signed reduction in pH may generate hydrogen sulfide from the reaction of
residual sulfides in the hides. Administrative and engineering controls must
be implemented to limit employee exposure within OSIIA requirements. Such
controls may require efficient exhausting of hydrogen sulfide from the
process area, dilution with make-up air, confining the reaction to a closed
system and/or finally, limiting the amount of time any person may work in
the contaminated area.
Process Drain Controls
Vlastewater "dumps" from process operations must be coordinated to pre-
clude the simultaneous discharge of "incompatible" unhairing wastes with
spent pickle or spent chrome liquors to open plant sewers or treatment
facilities.
Monitoring Work Areas
An effective and convenient tool for determining hydrogen sulfide gas
concentrations in the working environment is the Drager Hultigas Detector
which consists of a 'fixed-volume", bellows pump and alchemical packed glass
tube through which the air sample is drawn. In the Drager Tube 5/b hydrogen
sulfide discolors the lead-acetate in the tube; and the tube is calibrated
so that the length of the stain is proportional to the i^S concentration in
the air. The method is rapid and the equipment is portable. ' Thus, for a
quick approximation of hydrogen sulfide by grab sample analysis the tube-
type detector is invaluable.
When significant concentrations of hydrogen sulfide can be generated
without warning in confined working areas, a continuous, automatic monitor-
ing system should be employed. At the Pfister & Vogel Tanm'nq Company, a
Bacharach H2$ Continuous Has Monitoring System was installed in our "Sewer-
House", the last point-on-site before all wastes are screened and discharged
into the municipal sewage system. A low level alarm relay is set at 20 ppm
(acceptable ceiling concentration) and is connected to a flashing yellow
light. If the level exceeds 50 ppm ("Acceptable maximum peak above the
99
-------
ceiling concentration for an 8-hour shift"), a high level alarm relay sets
off a warning horn. Workers are instructed to evacuate the area within 10
minutes after a low-level alarm, and immediately for a high-level alarm.
Additionally, all persons working on the lower level (plus a "safety man"
on the ground level) are required to wear supplied-air type respirators,
regardless of the presence or absence of hydrogen-sulfide in the ambient
atmosphere.
SEWER HOUSE1 CONTROL MEASURES
As our conversion from a "hair-save" to "hair-burn" process effected
substantial increases in our sewage sulfide loadings, while dramatically
reducing the alkalinity (the pH of 24 hour-composited in proportion-to-flow
total tannery effluent samples were reduced from approximately 12.0 to 9.8)
and volume of cur effluent, it became increasingly evident that the genera-
tion of hydrogen sulfide gas in our sewers and in our Sewer House (the
building which houses our effluent screen and sewage pumps) could reach
serious levels. The probability of acidic waste streams combining in common
sewers with sulfide-bearing wastes increased dangerously. To protect per-
sonnel responsible for maintenance of the Sewer House from excessive expo-
sure to hydrogen sulfide, the following precautionary measures were insti-
tuted (in addition to the installation of the automatic continuous monitor-
ing system and the required use of respiratory protective equipment):
1. The levels of the sewage pump on-off float switches were
adjusted to increase the pump-on-frequency, thus reducing
the volume of wastewater retained in the sump, and mini-
mizing the potential for hydrogen sulfide formation.
2. The effective-volume dimensions of the sump were changed
to further reduce retention time within the Sewer House.
3. Ventilation was increased by installation of a blower to
supply heated make-up air to the lower level of the Sewer
House.
4. An Emergency Button integrated with the In-Plant Emergency
Alarm System was installed, and supervisory personnel were
instructed in the special precautions required of their
response to a Sewer House Emergency.
In addition to the measures instituted in the Sewer House, numerous
controls were required in the lower tannery. To reduce their retention,
within the work area, some of the sewers in the lower tannery were recon-
structed and/or rerouted. Although effluent volume-reduction was a pri-
mary objective of the grant program, we found it necessary, ultimately,
to put back almost 300,000 gpd of fresh v;ater as sewer flushes to provide
sufficient dilution to maintain a free-flowing condition.
To prevent the dangerous mixing of acid-containinn wastes with sulfide-
bearing waste fractions, an inter-connected system of "drain warning
100
-------
flashers" was installed to synchronize the discharge of incompatible waste
fractions. Process revisions require each operator to turn on the drain
warning flasher during three critical drains. If the flasher is already on
he may not proceed until he identifies the existing drain warning. If his
intended drain is compatible with the extant drain, he may proceed. If the
drain is not compatible, however, the operator must v/ait until the in-
process drain is completed and the Drain Warning Flasher is shut off.
Finally, a respiratory protective program has been established for
emergency protection against hydrogen sulfide. To comply with OSHA recom-
mendations, the following programs have been instituted:
1. Instruction (written and verbal) and training of
personnel in the proper use of respirators and their
limitations.
2. A regular schedule for inspection, repair, cleaning
and disinfection of respirators.
3. Storage of respirators in a convenient, clean and
sanitary location.
4. Surveillance of work area conditions and degree of
employee exposure.
5. Periodic review of respirator user's medical status.
Canister-type gas masks have been located in the lower tannery for emergency
egress in case of any hydrogen sulfide "accident." Also, two 30-minute dura-
tion, self-contained breathing apparatus have been located in our Power Plant
to insure the continued operation or orderly shutdown of plant energy facili-
ties during such an emergency. All regular power-house personnel and all
technical and engineering supervisors have been trained in their proper use.
HIDE PROCESSOR DRAIN MANIFOLD
A drain manifold has been incorporated at each hide processor to segre-
gate waste fractions for subsequent pretreatment, reuse or discharge to the
sewer. This manifold provides four pumping options:
1. Recirculate process solutions from the rear of the unit
to the front of the unit.
2. Discharge waste solutions directly to the sewer.
3. Divert sulfide-containinq wastes to the hair-burn header
for transfer to the Sulfide Oxidation System.
4. Divert spent chrome wastes to the Chrome Recovery System.
101
-------
The hair-burn drain line from each hide processor is connected to a
common header line leading to the Sulfide Oxidation System. Similarly, the
spent chrome drain lines are connected to a Spent Chrome Header leading to
the Chrome Recovery System. At present, selection of the proper drain mode
and valve configuration is an operator responsibility.
To prevent a catastrophic accident (which could result from the inadver-
tent mixing of process wastes within the hair-burn or spent chrome collect-
ing facilities) continuous automatic, in-line, pH monitors have been
installed (with Ultrasonic electrode cleaning devices employed to reduce
electrode fouling). If the pi I in the Spent Chrome Header goes above 4.5 a
preset relay closes an automatic air operated valve located just ahead of
the collection tank, and a warning alarm sounds. Likewise, if the p!l in the
llairburn Header drops below 10.0, an automatic valve closes and the alarm
sounds. The "closed valve-alarm" continues until the pll in the header pipe
is restored to the acceptable level, and the drain-collection system is
re-set.
102
-------
SECTION XI
SULFIDE OXIDATION IN UNMAIRING WASTES - FEASIBILITY STUDIES
LITERATURE REVIEW
Sodium sulflde and sodium sulfhydrate are widely used in the Tanning
Industry, to remove hair from the hides and skins prior to tanning. The
waste waters discharged from the unhairing operation, therefore, contain
substantial quantities of sulfide ion, sometimes exceeding 5,000 mg/1. When
these wastes are discharged to a sewerage system and mixed with acidic wastes
such that the pH drops below 9.0, hydrogen sulfide gas can be produced.
Bailey (29) reports that sulfide limitations of 0-20 ppm are generally
imposed by authorities to avoid the build-up of dangerous concentrations of
hydrogen sulfide in sewerage systems. A secondary concern for establishing
sulfide limitations -is the bacterial oxidation of sulfide to sulfuric acid
which can cause crown corrosion in concrete sewers.
The effect of sulfide on biological treatment usually becomes relevant
only for "on-site" treatment of tannery wastes. vanVlimmeren (30) reports
that while stream concentrations of 10 mg/1 or less can be toxic to fish, it
is erroneous to conclude that such low levels of sulfide are also toxic to
biological treatment system organisms. In fact, sulfide concentrations as
high as 50-100 mg/1 do not inhabit biological waste treatment operations.
The A. C. Lawrence Leather Company (31) demonstrated on a pilot scale
that 24-hour dosages of 100 ppm of sulfide caused no problems in the opera-
tion of an activated sludge unit. A two-hour shock loading of 400 ppm
failed to damage the activated sludge. According to Bailey (32), while
one-hour shock loadings of 250 mg/1 of sulfide showed negligible deteriora-
tion of an activated sludge system, the level that could be tolerated on a
continuous basis was much lower. Aulenbach and Heukelkian (33) concluded
that sulfide concentrations up to 25 mg/1 were tolerable in activated sludge
and up to 50 mg/1 could be handled for short intervals. Lawrence and
Rudolfs (34) reported a reduction in gas formation when anaerobic digester
feed sulfide concentrations were in the 150-200 mg/1 range.
Numerous treatment procedures have been proposed to reduce or eliminate
sulfides from unhairing wastes. Bailey (35) reports that in Poland, chemi-
cal treatment of tannery effluent by coagulation with ferrous sulfate
results in 100% reduction in sulfide. Further, the Institute in Lo'dZ' has
demonstrated 100% reduction in sulfide by activated sludge treatment.
103
-------
In South Africa (36), eight hours of paddling durino the liming process
with manganous sulfate as a catalyst (100 ppm Mn) reduced the sulfide level
from 431 to 200 ppm. Oxidation of spent Hme liquors using manganous sulfate
catalysis, nearly completely oxidized the sulfides within a reasonable time.
vanVlimmeren (37) reported 95% reduction of sulfide in situ in unhairing
vats or drums by addition of 200 g/m-* of manganous suTfate after unhairing
is completed, when followed by an additional three hours of paddling or
drumming. Separate paddle oxidation of spent unhairing liquors only
resulted in 90$ reduction after four hours of paddling. A 36% reduction in
alkalinity was also observed. Eye and Clement (38) investigated the labora-
tory scale oxidation of sulfide in lime-sulfide solutions using several
oxidizing agents. Ozonation was the most effective method but investiga-
tions were limited by lack of an effective contact device. Of the catalyzed
air oxidation systems studies manganous sulfate was found superior to tri-
valent chromium but inferior to potassium permanganate. The latter was
found to react as a primary oxidant as well as a catalyst for air oxidation
of sulfides. Optimum sulfide oxidation using the air-potassium permanganate
system occurred at a manganese-to-lnltial sulfide ratio of 0.15.
Chen and Morris (39) investigated a number of metal ions as catalysts
for the oxidation of sulfides. The order of effectiveness was Ni+2>Co+2 >
Hn+2>Cu-f2 in the mildly basic pH range. A two-fold increase in the concen-
tration of Ni+2 from 5xlO~5M to 10x10'% not only increased the rate of
oxygen uptake but also effected a change in the stoichiometry, reducing the
oxygen requirement from 0.8 moles 0? per mole of S= to 0.6 moles 02 per
mole Sa. Krebs (40) showed that Mn*+ was the most effective catalyst for
air oxidation of sulfides in more alkaline solutions. Bailey (41) indicated
that 1n Switzerland, where a reduction in sulfide in tanning effluent to
less than 0.1 mg/1 was required, liming effluents were treated by manganese
catalyzed air oxidation using floating turbine aeration. While the manga-
nese catalyst has been shown to concentrate in the foam layer, this problem
is less.severe with surface aeration than when diffused-air aeration is
employed." Black and Goodson (42) reported that sulfides present in natural
waters at concentrations of 1.90-5.65 mg/1 can be oxidized by«chlorine to
sulfur and sulfate end-products. The stoichiometry of the redox reaction is
highly dependent upon pH. Mole ratios of chlorine reacted to sulfide re-
acted varied from approximately 3.4 at pH 5 to approximately 1.8 at pH 8.
LAB-SCALE STUDIES
Lab-scale familiarization experiments using potassium permanganate
catalyst were conducted in a 1,000 ml pyrex beaker with a high speed labora-
tory stirrer as the aeration device.
A sample of unhairing waste was composited from the hair-burn drain and
subsequent washes from a typical hide processor run and passed through a 30
mesh screen to remove coarse suspended solids. To an aliquot of the
screened waste (2,120 mg/1 sulfide) a 4% solution of potassium permanganate
equivalent to a reaction mixture Mn++ concentration of 212 mg/1 (10% of the
initial sulfide) was added. "Once-through" oxidation resulted in an 80%
decrease in sulfides after one hour, as measured by the lodimetric Method
104
-------
outlined by Skoog and 1,'est (43). The progress of the reaction 1s shown in
Figure 21.
Severe foaming occurred during the oxidation run. The apparent failure
to achieve better than 80% reduction in sulfide content is believed due to
the non-specificity of the lodimetric Method used as the indicator of the
oxidation progress, rather than a characteristic of the oxidation process/
equipment. Therefore, in subsequent oxidation runs (lab, pilot and full-
scale) a titrimetric method employing a standard solution of potassium
ferricyanide as titrant was employed (Society of Leather Trades' Chemists
Method No. SLM 4/2).
Further bench-scale oxidation runs were conducted using screened,
undiluted unhairing waste which analyzed at 4,210 mg/1 S=. Potassium
permanganate solution equivalent to 20% of the initial sulflde (842 mg/1
Mn++) was added. After 60 minutes of aeration no sulfide could be detected
by the Society of Leather Trades' Chemists Method. The reaction mixture
changed from an Initial brownish green color to a light tan at the end of
the 60-minute aeration period. Approximately 30 minutes after turning off
the aerator an aliquot of the resuspended mixture analyzed at 40 mg/1 S=.
Within 75 minutes after discontinuing aeration a concentration of 56 mg/1 S=
was observed.
For the third run, a 10% catalyst (KMn04 solution) dosage was used and
75 minutes of aeration was required to reduce the sulfide from an initial
concentration of 4,210 mg/1 to zero. An observed increase in pH from 11.9
at 40 minutes to 12.25 at 75 minutes corresponded with the noted color change
from green-brown to light tan.
These bench scale "probe" experiments suggested that further Investiga-
tion of the various parameters for manganese catalyzed air oxidation of sul-
fides in unhairing wastes was warranted. The topics for further Investiga-
tion were:
1. Rate of oxidation
2. Types of aeration devices
3. Effect of screenable solids on sulfide oxidation
4. Effect of initial sulfide on rate
5. Foaming
6. The "reappearance" of sulfide
7. Color changes
8. pH change
9. Catalyst type and dosage
105
-------
25
10 20 30 40 50 60
TIME, min
Figure 21: BENCH-SCALE OXIDATION OF
SULFIDES IN UNHAIRING
WASTES.
106
-------
PRESCREENING HAIRBURN WASTES
Two commercially available wastewater screens were tested for the
pretreatment of unhairing wastes prior to manganese catalyzed air oxidation.
Neither the Bauer Hydraseive nor the Dorr Oliver 300° Pressure Screen per-
formed satisfactorily under the conditions encountered. The most serious
shortcomings were blinding of the Hydraseive and excessive sludge volume
from the Dorr Oliver Screen. Fortuitously, subsequent oxidation runs demon-
strated that prescreening did not significantly improve sulfide oxidation
rates.
The Bauer Hydraseive tested was an 18 inch wide gravity screen with
interchangeable screen surfaces having openings ranging from 0.010 inches
to 0.080 inches. Undiluted hairburn liquor was pumped over the screen at
the rate of 35-50 gpm (1.94-2.78 gal per inch of screen width). The screen
surfaces tested (i.e., 0.020", 0.030", and 0.040") all immediately blinded.
In another run, 3,350 gallons of unhairing waste including subsequent
washes from a typical hide processor run were collected and equalized. This
diluted waste was then pumped over the Hydraseive fitted with a 0.20" open-
ing screen at an average rate of 47 qpm. The feed to the screen contained
25,700 mg/1 of suspended solids and the screened waste analyzed at 24,400
mg/1. Approximately 100 gallons of sludge containing 9.5% total solids were
collected. Thus, a suspended solids reduction efficiency of only 5% was
observed.
Parallel feasibility studies were conducted using a Dorr Oliver 300°
pressure screen with slot openings of 300, 150 and 75 microns. During the
initial studies, 3,800 gallons of diluted unhairing wastes, containing
approximately 25,000 mg/1 of suspended solids, were pumped from a hide
processor to the collection vessel. Using a screen with 300 micron openings,
a sludge fraction equivalent to approximately 5% of the feed volume was
generated. Visual comparison of the sludge and effluent from the screen
indicated that there was no substantial solids disproportionate. Using
a screen surface with 75 micron openings, an unacceptable sludge volume of
greater than 50% of the feed was collected.
At a slot opening of 150 microns, however, a distinct separation was
noted. Sludge volumes were estimated at 10-20% of the feed.
Continued investigations were conducted using the 150 micron opening.
Unhairing waste Including spent unhairing liquor and subsequent washes were
collected and fed to the pressure screen at a rate of 125-150 gpm at 50 psi
(feed suspended solids = 24,600 mg/1 in 3,800 gal). Approximately 500
gallons of sludge were collected containing 65,600 mg/1 of suspended solids.
Thus, 35% of the suspended solids was concentrated into a sludge volume
equal to roughly 13% of the original volume.
Therefore, while a significant solids separation has been demonstrated,
the large volumes of sludge generated could create a more serious disposal
problem than the discharge of the untreated unhairing wastes directly to the
sewerage system.
107
-------
STOICHIOMETRY OF SULFIDE OXIDATION
The concentrations of sulfide species in water are defined by the
following chemical equilibria:
(H+) (MS-)
H2S = H+ + IIS; KI = (l\2 S)
MS" " H+ + S=; K2 = (H+) (S= )
(HS -)
While the first ionization constant, Ki , is well established at 7.02, the
second ionization constant, K2, can only be estimated to lie within the
12.35 - 15 range (44). Simple calculations show that within the pH 7.0 -
12.5 ranfje, the dominant sulfide species is the HS" ion.
Sulfide oxidation by elemental oxytien may consume or generate hydrogen
ions to produce several possible reaction products (45):
Thus, the oxygen requirement as predicted by these stoichiometric relation-
ships ranges from 0.5 moles to 2 moles oxygen per mole of sulfide.
. p
As demonstrated by Chen S Morris (4G) for the Mi catalyzed air oxida-
tion of sulfides, an increase in the catalyst concentration not only accel-
erates the initial rate of oxygen uptake but also changes the stoichiometry
of the reaction. Thus, when the catalyst concentration was doubled from
5x1 0~5 to 1x1 0~4 molar, sulfide converted to elemental sulfur doubled from
40% to 80%, the remainder being converted to thiosulfate. Likewise, the
oxygen requirement was reduced from 0.8 to 0.6 moles 02 per mole of sulfide
consumed.
From lab-scale studies using sulfide unhairing wastes, it is evident
that the oxygen requirements for manganese catalyzed-air oxidation_of sul-
fides in unhairing wastes lie within the 0.6 to 1.0 Ib 02 per Ib S= range.
To test the effect of sulfide concentration on this redox system, vary-
ing aliquots of hair-burn waste were adjusted to 150 ml and placed in stop-
pered quart containers along with 10% Mn as manganous sulfate solution.
Half of these samples v/ere analyzed for residual sulfides after 6 hours of
agitation on an automatic shaker. The other half were shaken intermittently
over a 24-hour period and then promptly analyzed for sulfides. By calculat-
ing the amount of oxygen initially available in the sealed containers (47),
(Table 39) the weight ratio of oxygen-to-sulfide reacted was determined.
llhen the initial concentration of sulfides present was 1,670 mg/1, the
108
-------
TABLE 39. OXYGEN UPTAKE FOR SULFIDE OXIDATION
o
10
Sample
1
2
3
4
5
6
Shake
time.hr
G
6
C
24
24
24
Waste
volume, ml
5.85
58.5
117
5.85
58.5
117
Hn++,mg
2.5
25
50
2.5
25
50
Initial
ST mg
25
250
500
25
250
500
Final
S= mg
0
35
226
0
33
153
AST mg
25
215
274
25
217
347
02/S=
-
1.01
0.30
-
1.00
0.63
Air offered = 0.795 liters
Oxygen offered - 218 mg 02*
Total solution volume = 150 ml
* One liter of air contains 273 mn Q£ @ STP, Fair Guyer & Okun, Vol. 2, Chap. 35, p. 24.
-------
Op/S~ ration was 1.0 both after 6 hours and after 24 hours, l.'hen the ini-
tial sulfide concentration was increased to 3,340 mg/1, this ratio decreased
to 0.80 Ib 02/1 b S= after 6 hours and 0.63 Ib 02/lb S= after 24 hours.
Thus, it is apparent that sulfide concentration substantially effects
the stoichiometry of this redox reaction.
SULFIDE OXIDATION BY PADDLE AERATION
On completion of the bench-scale studies of the oxidation of sulfides
in unhairing wastes, a pilot-scale oxidation system was employed to observe
the effect of several parameters on the sulfide oxidation process. Primary
emphasis during this phase of the program was placed upon the determination
of the effects of the physical form of the catalyst feed, catalyst dosage
and valence on the air oxidation of sulfldes. The effect of initial sulfide
concentration was investigated, as well as the phenomena of foaming, color
and pH change.
These pilot studies were conducted in a pair of tannery paddle vats
having surface dimensions of 9' x 14' with a capacity of approximately 4,000
gallons each.
The recirculating-drain pump servicing each hide processor was used to
transfer the wastes through specially installed wastewater collection lines
to the paddle vats.
Initial oxidation runs in the paddles were conducted using potassium
permanganate catalyst added as a single dry feed at the start of the run.
The unhairing wastes consisted of mixtures of concentrated hair-burn liquors
plus subsequent washes. Batch volumes of 3,100-3,300 gallons were used in
all but a few paddle-aeration runs involving sequential oxidations. For the
first five runs, an average time of 2.75 hours was required to reduce the
sulfide from an average initial level of 1,400 mg/1 to "zero". The mean
rate of disappearance of sulfide over the five runs was 14.2 Ib S=/hr. As
shown in Table 40, initial sulfide concentrations varied from 1,040 mq/1 to
1,750 mg/1 while the mean rate of oxidation and reaction time varied from
9.05 - 18.5 Ib S=/!ir and 1.5 - 3.5 hours, respectively.
During oxidation runs 1, 2 and 4, severe foaminn necessitated the use
of an antifoam (Dow Corning DB-31) to contain the reaction mixture within
the paddle vat (even though one foot of freeboard was provided). An anti-
foam dosage of 20-60 ppm applied as a 1:1 water emulsion and sprayed over
the surface of the foam, adequately controlled foaming for up to 30 minutes.
As the sulfide oxidation progressed, a gradual change in the color of
the mixture occurred. The initial dark brown color of the hairburn-
manganese catalyst mixture turned to a deep olive brown which eventually
lightened to a celery green and finally to a beiqe color. Appearance of
the beige color signaled the approach of the "zero" level.
After a "zero" sulfide reading was recorded, samples of the aerated
110
-------
mixtures were left undisturbed at room temperature for 24 hours. These
samples reverted to their earlier brownish-green colors and sulfide levels
ranging from 22-73 mg/1 were regenerated. Comparison of samples stored in
open containers vs air-tight containers indicated no significant difference
in sulfide. Thus, this reappearance phenomenon is probably independent of
the presence of an air-liquid interface.
TABLE 40. SULFIDE OXIDATION DY PADDLE DERATION
__
Run
1
2
3
4
5
Initial
mg/1
1,530
1,750
1,510
1,040
1,200
Sulfide
Ib
40.3
46. G
40.3
27.8
32.0
Time.min
150
180
210
90
210
Mean.lb S'/hr
16.3
15.5
11.5
18.5
9.05
Max. 15 minute-
incremental ,lb S=/hr
29.8
36.2
38.5
53.1
19.2
Mn++/S= = 0.15
Volume = 3,200 gal
The redox curve for Run Ho. 3 which is depicted in Figure 22 is typical
of results obtained for the five potassium permanganate catalyzed oxidation
runs. The outstanding characteristic of this family of curves is the
extremely high initial rates of sulfide disappearance. Thus, while the mean
rate of disappearance of sulfide in Run No. 3 was 11.5 Ib S=/hr, a maximum
incremental rate of 38.5 Ib S=/hr was observed durinn the initial 15 minutes
of the reaction. These initial rate maxima v/hich are consistent with the
observations of Eye and Clement (48) are probably due to direct chemical
oxidation of the sulfide by potassium permanganate—an extremely strong
oxidizing agent. As the potassium permanganate is reduced, the manganous
hydroxide floe which is formed provides the catalytic surface for the con-
tinued air oxidation of the sulfide at a slower rate. The mechanism for
this catalyzed air oxidation probably employs the same oxygenated manganous
species involved in the Winkl.er dissolved oxygen test (49, 50, 51).
When dry manganous sulfate replaced potassium permanganate as catalyst
a substantial difference in the rate curve was observed. In runs Ho. 6 and
No. 7 manganous sulfate monohydrate equivalent to an initial manganese-to-
sulfide ratio of 0.15 was added to the aeration mixture at time zero'.
The rate of oxidation which averaged 11.£ Ib S~/hr over the two runs
was slightly lower than that observed for potassium permanganate catalysis
111
-------
2000-
;r 150(5?-
D)
E
LU
Q
Li.
_l
D
CO
1000-
500-
30 60 90 120 150 180 210
TIME, min
Figure 22: OXIDATION OF SULFIDES BY PADDLE
AERATION (Run No. 3)
112
-------
(14.2 Ib S=/hr). The maximum incremental rate, however, which averaged
17.4 Ib S=/hr for the two runs was less than half that observed with
potassium permanganate. Contrary to results obtained using potassium
permanganate as catalyst, the rate maxima occurred during the third hour
of aeration after the sulfide had been reduced to approximately half the
initial levels.
Therefore, while manganous sulfate effects a substantial change in the
shape of the rate curve vs potassium permanganate, potassium permanganate
is only 25% more effective when used dry in a batch oxidation system.
Further oxidation runs were conducted in paddles using a solution of
manganous sulfate as catalyst. At a manganese-to-sulfide ratio of 0.15 with
all of the catalyst added to the reaction mixture at time zero, the average
rate of sulfide oxidation for two runs was 18.7 Ib S=/hr with maximum incre-
mental rates averaging 29.5 Ib S~/hr.
In three subsequent runs manganous sulfate solution was added in three
feeds at 30-minute intervals to bring the manganese concentration to 0.15
of the initial sulfide content. The observed mean rates averaged 18.1 Ib
S=/hr, while the maximum incremental rate averaged 26.2 Ib S=/hr.
Thus, introduction of the manganous sulfate catalyst in solution re-
sulted in a substantial improvement over dry catalyst feeds.
There does not appear to be a substantial difference between the use
of catalyst in a single solution feed vs addition in three feeds.
Additional paddle runs were conducted to determine the effect of manga-
nous sulfate dosage on sulfide oxidation. Catalyst was fed in solution in
a single feed at time zero. The data illustrated in Table 41 demonstrate
that reducing the catalyst-to-sulfide ratio .from 0.15 to 0.10 or less
adversely affects the kinetics when catalyst is added in a single feed.
However, no significant advantage was demonstrated by increasing the ratio
from 0.15 to 0.20. Thus, the optimum manganese-to-sulfide ratio.lies within
the 0.10-0.15 range when manganous sulfate solution is added in a single
feed in paddles.
Another experiment using the paddle aerator was conducted to determine
the catalytic activity of used manganese sludge in a subsequent sulfide
oxidation run. This experiment demonstrated that day-old recovered sludge
does exhibit catalytic activity when it is mixed with fresh hair-burn wastes
for subsequent aeration.
Approximately 3,200 gallons of hair-burn wastes having an initial sul-
fide concentration of 1,840 mg/1 were aerated using a catalyst dose of 20%
of the initial sulfides (catalyst added in solution). After the oxidation
was complete, the mixture was allowed to settle and the supernatant was
pumped off leaving-approximately one-third of the original volume as sludge.
The following day fresh hair-burn waste was added to the settled sludge at
a two-to-one ratio. The net sulfide ion concentration of the mixed liquor
was 1,540 mg/1. This mixture was aerated without any further addition of
113
-------
catalyst.
The first use of catalyst effected a reduction of the sulfide level
to zero in 2.75 hours. The second use of the catalyst in the recycled
sludge resulted in complete oxidation of the sulfides in 4 hours. These
data contrast to the approximately 9 hours required to reduce the sulfides
to zero by aeration without catalyst of any kind.
TABLE 41. EFFECT OF CATALYST DOSAGE Of! SULFIDE OXIDATION
BY PADDLE AERATION
Rate
Initial Sulfide Max. 15 minute-
MnTT/S~ mg/1
0.
0.
0.
0.
05
10
15
20
2
2
1
1
,080
,040
,780*
,840
Ib
53.
52.
46.
47.
7
7
4*
6
Time, mi n
285
195
150*
165
Mean,lb S=/hr incremental, Ib S=/hr
11.
16.
18.
17.
3
2
7*
3
16
23
29
30
.0
.2
.5*
.4
Volume = 3,200 gal
* average for two runs
FLOATING SURFACE AERATOR
Among the alternative aeration systems investigated for the oxidation
of sulfides 1n unhairing wastes was a floating surface aerator. A two
horsepower floating surface aerator manufactured by Aqua-Aerobic Systems,
Inc. (Aqua-Jet Mark 2) was installed in one of the paddle vats which had
been previously used for paddle aeration. This aerator had a manufacturer
rating of 3.9 Ib 02 transferred per nameplate-horsepower in clear water at
standard temperature and atmospheric pressure.
After removing the paddle wheel, the aerator was moored with four
chains suspended from support beams spanning the vat. Operation of this
pilot aeration system provided the information summarized in Table 42 which
was used to design full-scale facilities.
In four oxidation runs using this surface aeration system batch volumes
averaged 3,300 gallons and initial sulfide averaged 2,580 mg/1 (Runs 22, 25,
27 & 29). A manganese-to-sulfide ratio of 0.15 was employed. The manganous
sulfate solution was added in three feeds 30 minutes apart. An average of
5.3 hours of aeration was required to reach zero sulfide. A silicone anti-
foam was added as required to contain the foam within the 1.5 feet of free-
board provided. The mean rate of sulfide reduction over the four runs was
114
-------
TABLE 42. SULFIDE OXIDATION BY SURFACE AERATION - PILOT SYSTEM
Rate
Run
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Volume,
gal
3300
1650
3300
1650
3300
3300
3300
3300
3300
3300
3300
3300
3800
3800
3800
3800
Hn++/S=
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15+
0.15+
0.015+
0.10+
Initial
sulfide,
mg/1
1310
1280
620
2540
2640
4980
2620
2620
2580
2540
5470
2520
1320
930
1340
2740
Ib
36.0
17.6
17.3
35.1
72.8
136.5
56.1
71.6
70.8
69.5
121.8
69.5
41.8
29.5
42.4
86.3
Time,
min
150
90
120
345
240
900*
510**
450
390
390
870***
210
165
105
210
195
mean rate,
Ib S=/hr
14.4
11.8
8.6
6.1
18.2
9.1
6.6
9.55
10.9
10.7
8.4
19.8
15.1
16.8
12.1
26.7
max. incre-
mental rate,
Ib S=/hr
26.4
19.1
12.3
8.2
36.6
17.0
9.4
14.2
14.8
21.7
17.5
34.0
44.1
30.0
25.2
51.8
Aerator efficiency
mean eff . ,
Ib S~/hp-hr
7.2
5.9
4.3
3.0
9.1
4.5
3.3
4.8
5.4
5.3
4.2
9.9
7.5
3.4
6.0
13.4
max. incre-
mental eff.,
Ib S=/hp-hr
13.2
9.5
6.1
4.1
18.3
8.5
4.7
7.1
7.4
10.8
8.7
17.0
22.0
15.0
12.6
25.9
* impeller clogged; final sulfide - 662 mg/1 after 900 minutes.
** impeller clogged; final sulfide - 510 mg/1 after 510 minutes.
*** final sulfide - 1,010 mg/1 after 370 minutes.
-------
14.6 Ib/hr tor an aerator-oxidation efficiency of /.j ID ^ /np-nr. me maxi-
mum 30 minute-Incremental rate observed was 36.6 1b S=/hr for an incremental
aerator efficiency of 18.3 1b S=/hp-hr.
The effect of initial sulfide concentration on sulfide oxidation was
measured in subsequent runs maintaining the original batch volume of 3,300
gallons and the manganese-to-sulfide ratio of 0.15. Hhen the initial sulfide
concentration was reduced approximately 503 (Run Mo. 18), the aeration time
required to achieve a zero sulfide reading was reduced from 6 to 2.5 hours.
A mean sulfide reduction rate of 14.4 Ib/hr, equivalent to 7,2 Ib/hp-hr was
thus observed. A maximum 30 minute-incremental rate of 28.3 Ib/hr was
recorded.
Further reduction of the initial sulfide, however, to 620 mg/1 (Run !lo.
20) effected a substantial drop in the mean rate to 8.6 Ib S'/hr (4.3 lb/
hp-hr) and a drop in the maximum incremental rate to 12.3 lb S=/hr with two
hours of aeration required.
Hair-burn liquors, as discharged from the hair pulping operation before
being mixed with subsequent washes, contain substantially higher waste con-
centrations. Sulfides in excess of 5,000 mg/1 are common, while total solids
normally exceed 100,000 mg/1. A 3,300 nallon batch of concentrated hair-burn
liquor collected from four process runs was subjected to aeration in the sur-
face aerator system (Run No. 28). Manganous sulfate equivalent to 0.15 times
the Initial sulfide on a Mn+'VS3 basis was added in three solution feeds,
30 minutes apart. The sulfide was reduced from an initial 5,470 mg/1 to
1,010 mg/1 after 14.5 hours of aeration. Thus, the mean rate of oxidation
observed was 8.4 Ib S=/hr, equivalent to an aerator efficiency of 4.2 lb/
hp-hr. The maximum 30 minute-incremental rate recorded was 17.5 Ib S~/hr.
Thus, it is apparent that dilution of unhairing waste from the 5,470
mg/1 S* encountered in concentrated liquors to approximately 2,500 mg/1 by
including wash waters effects a substantial increase in oxidation rates.
Concentrations from 2,500 mg/1 down to 1,300 mg/1 clearly lie within the
optimum range. As further dilution lowers sulfide concentrations to 620 mn/1
the rate of oxidation again falls off.
Another experiment using this surface aeration system was designed to
determine the effect of basin geometry on the oxidation of sulfides in un-
hairing wastes.
Uhen the volume was reduced to half of the "normal" volume, doubling the
power-to-volume ratio from 0.6 hp/1,000 gal to 1.2 hp/1,000 gal (Run No. 21),
nearly 6 hours of aeration_were required to reduce the sulfide from an initial
2,540 mg/1 to zero (Mn^/S3 = 0.15). Thus, the recorded mean rate of oxida-
tion was reduced from 12.8 Ib S=/hr at 3,300 gallons to 6.1 Ib/hr at 1,650
gallons for an aerator-oxidation efficiency of 3.0 Ib/hp-hr. The maximum 30
minute-incremental rate observed was reduced to only 8.2 Ib S=/hr equivalent
to 4.1 Ib S=/hp-hr.
In another run, both the initial sulfide content and the batch volume
were reduced to approximately half of the normal levels (Run No. 19) while
116
-------
the 0.15 Mn"H'/S= ratio was maintained (mannanous sulfate in three solution
feeds 30 minutes apart). One and one-half hours of aeration were required
to reduce the sulfide from 1,280 mg/1 to zero.
Thus, at approximately half the normal batch volume (power/volume »
1.2 hp/1,000 gallons) and_half the normal sulfide level, the mean oxidation
rate dropped to 11.8 Ib S=/hr from the 14.4 Ib/hr observed when half the
normal sulfide level was used at a power-to-volume ratio of 0.6 hp/1,000
gallons.
While a substantial reduction in reaction time was observed, this
increase in the power-to-volume ratio sacrificed aerator efficiency.
Another experiment was conducted in the surface aeration system using
an "uncontaminated" unhairing liquor consisting of 150 pounds of hydrated
lime in 3,300 gallons of 2,580 mg Ss/l sodium sulfhydrate solution (Run No.
26). The temperature of the mixture was adjusted to a typical hairburn
waste discharge temperature of 86°F. Manganous sulfate solution was added
in three feeds 30 minutes apart at a total Mn++/S= ratio of 0.15. The mix-
ture required 6.5 hours of aeration to reduce the sulfide to zero. The mean
rate of sulfide removal was 10.9 Ib/hr (5.4 Ib/hp-hr) while a maximum 30
minute-incremental rate of 14.8 Ib/hr (7.4 Ib/hp-hr) was observed. Foaming,
throughout the run, was pronounced but did not increase in severity with
aeration nor was antifoam required. The color of the reaction mixture varied
with aeration time but did not follow the color trend observed during a nor-
mal hairburn v/aste sulfide oxidation. Instead, the initial reaction mixture
was a translucent pale yellow separating into a transparent yellow superna-
tant and a white sludge layer. As the oxidation progressed, the intensity
of the supernatant increased and the calcium hydroxide-manganese hydroxide
sludge layer turned dark brown. Near the end point of the reaction the
intensity of the supernatant decreased until finally when no sulfide could
be detected the supernatant was colorless. These results indicate that the
complex mixture of dissolved and suspended proteinaceous matter, inorganic
salts, greases, alkalis, etc. which accompany the sulfides and hydrated lime
present in unhairing wastes does not demonstrably inhibit the manganese
catalyzed, air oxidation process.
In one attempt to improve the rather low rate of oxidation (8.4 Ib S=/
hr) observed with concentrated hairburn liquors, concentrated v/astes were
mixed with previously oxidized unhairing waste, restrengthening the mixture
with manganous sulfate catalyst. In this mixture the sulfides were oxidized
at substantially higher rates than the concentrated waste alone.
A mixture of 3,300 gallons of hairburn liquor and wash water containing
69.5 Ib of sulfide at an initial concentration of 2,520 mg/1 was aerated
using 15% manganese (MnSO^HoO) as catalyst. At an average rate of 19.8 Ib
Sa/hr, the oxidation took 3.5 hours to complete. The mixture was then
drained to approximately 3,000 gallons and 850 gallons of concentrated hair-
burn liquor was added. The sulfide concentration of the mixed waste was
1,320 mg/1 (Run No. 30). After restrengthening with 15% manganese, the mix-
ture was aerated for 2.75 hours to a zero detectable sulfide level.
117
-------
This treatment procedure was repeated for four successive runs. The
sulfide concentrations after the addition of the concentrated liquor were
1,320 mg/1, 930 mg/1, 1,340 mg/1 and 1,160 mq/1, respectively (Runs 30, 31,
32 & 33). In the final run, 50 Ib of sulfide was added as a solution of
sodium sulfhydrate, increasing the concentration from 1,160 mg/1 to 2,740
mg/1. The catalyst dosages for the four runs were 15%, 15%, 1.5% and 10%,
respectively. Corresponding average rates of sulfide oxidation of 15.1,
16.8, 12.1 and 26.7 Ib S=/hr were observed.
Thus, as these data indicate, concentrated hairburn v/astes can be
treated at substantially higher rates when they are mixed with previously
aerated and oxidized wastes.
These oxidation rates sharply contrast to the aerator manufacturers
performance data. After correction for temperature, residual and saturation
D.O. under field conditions andoCand># according to the commonly accepted
formula for aerator sizing (52) (Figure 23) oxygen transfer rates on the
order of 4.0 Ib 02/hp-hr are predicted usinn the conservative assumptions
that«f and ft are unity and the 0£ residual is maintained at 0 mg/1. Even
with a minimum observed oxygen-to-sulfide ratio of 0.6 Ib 03 consumed per
1.0 Ib sulfide oxidized, the maximum sulfide oxidation efficiency which could
be predicted is 6.7 Ib S=/hp-hr.
Demonstrated incremental aeration efficiencies as hiqh as 26 Ib S=/hp-
hr and mean efficiencies as hiqh as 13.3 Ib S=/hp-hr sharply contrast this
predicted transfer rate for the manganese catalyzed air oxidation of sul-
fides in unhairing wastes.
Due to the oxygen scavenging nature of both the manganese hydroxide
and the sulfides contained in the waste, it is impossible to empirical!."
determine a D.O. saturation value and thus/^ can only be estimated. Due to
the high concentration of dissolved and suspended matter, it is reasonable
to expect, however, that/9 is less than one. The only remaining factor
which can be associated with this marked increase in oxygenation rates over
specified manufacturer rates for clean water operation is*', the ratio of
field 02 transfer rates to clean water transfer rates. This factor is
dependent upon both the stoichiometry of the sulfide oxidation reaction and
the oxygen requirement. The authors conclude, therefore, that aeration
equipment for the manganese catalyzed air oxidation of sulfides in unhairing
wastes may be accurately and conveniently specified in terms of sulfide oxi-
dized per horsepower per hour rather than the conventional Q£ transfer
specification.
118
-------
the 0.15 Mn^/S2 ratio was maintained (manganous sulfate in three solution
feeds 30 minutes apart). One and one-half hours of aeration were required
to reduce the sulfide from 1,280 mg/1 to zero.
Thus, at approximately half the normal batch volume (power/volume s
1.2 hp/1,000 gallons) and half the normal sulfide level, the mean oxidation
rate dropped to 11.8 Ib S=/hr from the 14.4 Ib/hr observed when half the
normal sulfide level was used at a power-to-volume ratio of 0.6 hp/1,000
gallons.
While a substantial reduction in reaction time was observed, this
increase in the power-to-volume ratio sacrificed aerator efficiency.
Another experiment was conducted in the surface aeration system using
an "uncontaminated" unhairing liquor consisting of 150 pounds of hydrated
lime in 3,300 gallons of 2,530 mg S=/l sodium sulfhydrate solution (Run No.
26). The temperature of the mixture was adjusted to a typical hairburn
waste discharge temperature of 86°F. Manganous sulfate solution was added
in three feeds 30 minutes apart at a total Mn^/S51 ratio of 0.15. The mix-
ture required 6.5 hours of aeration to reduce the sulfide to zero. The mean
rate of sulfide removal was 10.9 Ib/hr (5.4 Ib/hp-hr) while a maximum 30
minute-incremental rate of 14.8 Ib/hr (7.4 Ib/hp-hr) was observed. Foaminn,
throughout the run, was pronounced but did not increase in severity with
aeration nor was antifoam required. The color of the reaction mixture varlec
with aeration time but did not follow the color trend observed during a nor-
mal hairburn waste sulfide oxidation. Instead, the initial reaction mixture
was a translucent pale yellow separating into a transparent yellow superna-
tant and a white sludge layer. As the oxidation progressed, the intensity
of the supernatant increased and the calcium hydroxide-manganese hydroxide
sludge layer turned dark brown. Near the end point of the reaction the
intensity of the supernatant decreased until finally when no sulfide could
be detected the supernatant was colorless. These results indicate that the
complex mixture of dissolved and suspended proteinaceous matter, inorganic
salts, greases, alkalis, etc. which accompany the sulfides and hydrated lime
present in unhairing wastes does not demonstrably inhibit the manganese
catalyzed, air oxidation process.
In one attempt to improve the rather low rate of oxidation (8.4 Ib S=/
hr) observed with concentrated hairburn liquors, concentrated v/astes were
mixed with previously oxidized unhairing waste, restrengthening the mixture
with manganous sulfate catalyst. In this mixture the sulfides were oxidized
at substantially higher rates than the concentrated waste alone.
A mixture of 3,300 gallons of hairburn liquor and wash water containing
69.5 Ib of sulfide at an initial concentration of 2,520 mg/1 was aerated
using 15% manganese (MnSO^HoO) as catalyst. At an average rate of 19.8 Ib
S=/hr, the oxidation took 3.5 hours to complete. The mixture was then
drained to approximately 3,000 gallons and 850 gallons of concentrated hair-
burn liquor was added. The sulfide concentration of the mixed waste was
1,320 mg/1 (Run No. 30). After restrengthening with 15% manganese, the mix-
ture was aerated for 2.75 hours to a zero detectable sulfide level.
117
-------
This treatment procedure was repeated for four successive runs. The
sulfide concentrations after the addition of the concentrated liquor were
1,320 mg/1, 930 mg/1, 1,340 mg/l and 1,160 mrj/1, respectively (Runs 30, 31,
32 & 33). In the final run, 50 Ib of sulfide was added as a solution of
sodium sulfhydrate, increasing the concentration from 1,160 mg/1 to 2,740
mg/1. The catalyst dosages for the four runs were 15%, 15%, 1.5% and 10%,
respectively. Corresponding average rates of sulfide oxidation of 15.1,
16.8, 12.1 and 26.7 Ib S"/hr were observed.
Thus, as these data indicate, concentrated hairburn wastes can be
treated at substantially higher rates when they are mixed with previously
aerated and oxidized wastes.
These oxidation rates sharply contrast to the aerator manufacturers
performance data. After correction for temperature, residual and saturation
D.O. under field conditions andoCand^? according to the commonly accepted
formula for aerator sizing (52) (Figure 23) oxygen transfer rates on the
order of 4.0 Ib 02/hp-hr are predicted using the conservative assumptions
thatoC and & are unity and the 0£ residual is maintained at 0 mrj/1. Even
with a minimum observed oxygen-to-sulfide ratio of 0.6 Ib Og consumed per
1.0 Ib sulfide oxidized, the maximum sulfide oxidation efficiency which coul<
be predicted is 6.7 Ib S=/hp-hr.
Demonstrated incremental aeration efficiencies as high as 26 Ib S=/hp-
hr and mean efficiencies as high as 13.3 Ib S=/hp-hr sharply contrast this
predicted transfer rate for the manganese catalyzed air oxidation of sul-
fides in unhairing wastes.
Due to the oxygen scavenging nature of both the manganese hydroxide
and the sulfides contained in the waste, it is impossible to empirically
determine a D.O. saturation value and thus/^ can only be estimated. Due to
the high concentration of dissolved and suspended matter, it is reasonable
to expect, however, that/9 is less than one. The only remaining factor
which can be associated with this marked increase in oxygenation rates over
specified manufacturer rates for clean water operation is<<", the ratio of
field 02 transfer rates to clean water transfer rates. This factor is
dependent upon both the stoichiometry of the sulfide oxidation reaction and
the oxygen requirement. The authors conclude, therefore, that aeration
equipment for the manganese catalyzed air oxidation of sulfides in unhairing
wastes may be accurately and conveniently specified in terms of sulfide oxi-
dized per horsepower per hour rather than the conventional 03 transfer
specification.
118
-------
r 1 T-20
CtfTR [(Cpe x/) - Cflj x (1.024) o
FTR - ~~ ' " r" "~~
Where;
FTR = Field Transfer Rate, Ib/hp-hr
CU'TR= Clean Water Transfer Rate, Ib/hp-hr
CDC = Saturation Concentration of D£ at Design
Temp. & Altitude, mg/1
CR = Residual D.O., mg/1
Csc = Saturation D.O. at Standard Conditions, 9.17 mg/1
T = Temperature, C
0 = Sat. P.O. In '>jaste
Sat. D.O. 1n Clean Hater
•^ = Rate of Transfer of Q? into Haste
Rate of Transfer of $ into Clean Hater
1(7.5 mg/1 x 1.0) - 0 mg/1
FTR = (3.9 Ib/hp-hr) |(7.5 mg/1 x 1.0) - 0 mg/l| x 1.26x1.0
= 4.0 Ib/hp-hr
9.17 mg/1
Figure 23. Determination of fiejd transfer rate
119
-------
SECTION XII
SULFIDE OXIDATION SYSTEM DESIGN
CAPTURE OF SULFIDE
The sulfides introduced in tannery operations are concentrated in dis-
crete, beamhouse sub-process solutions. The segregation and collection of
the resultant sulfide-bearing wastes remain among the most difficult and
unresolved problems in the pretreatment of tannery effluents. Even when
modern manufacturing equipment which provides re-circulation and pump-out
mechanisms is employed, the major difficulty in collecting sulfide v/astes
is their widespread occurrence in low-concentration/hinh-volume wash waters,
etc. which, nevertheless, contribute substantially to sulfide loadings.
Approximately 40« of the sulfides generated from a single production run may
be contained in as little as 700-800 gallons of concentrated unhairing
liquor. The capture of 80% of the sulfides requires that more than 5,000
gallons be collected; and to achieve 99% capture, over 13,000 gallons per
unit-run must be collected. It has been demonstrated that processing steps
after the liming-unhairing operations including bating and pickling continue
to extract sulfides from the hides. Their combined contribution equals only
1-2% of the total sulfide discharge, but concentrations as high as 48 mg/1
and 10.2 mg/1 in the bate and pickle liquors have been respectively recorded.
Including these wastes in the sulfide treatment would increase volumes to
more than 20,000 gallons per production run.
Table 43 illustrates the relationship between volume collected and sul-
fide capture for a typical hide-processor run. It is apparent from Table 43
that while 90% of the total sulfide loading is concentrated in the first
6,300 gallons of effluent, only 1% of the sulfide is contained in the last
6,900 gallons. In fact, capturing just the last 4.5% of sulfide increases
the volume of effluent to be collected and treated from 8,600 gallons to a
total of 20,100 gallons.
Thus, when segregation of waste streams is possible, the volume of
sulfide-bearing streams collected for pretreatment will be a first-order
function of the effluent limitations set by the controlling regulatory agency
rather than cost-effective logic of Table 43 distributions.
At the Pfister & Vogel tannery only 61% of beamhouse-tanyard processing
is done in equipment utilizing pump-out systems. The remaining 39% of the
lower tannery effluent is discharged via gravity to common tannery sewers.
In normal tannery production, necessary process adjustments, "unscheduled"
process and/or mechanical upsets, and variability in hide weight, water flow,
120
-------
draining efficiency, etc. are commonplace. The time required for production
runs is sufficiently irregular as to preclude absolute control of discharges
for sulfide collection by rigorous scheduling, even if diversion of such
process effluents throuqh common sewers were practicable. Therefore, that
39% of the total sulfides which are gravity-drained are not considered
amenable to pretreatment by segregation and sulfide oxidation.
TABLE 43. CUMULATIVE CAPTURE OF SULFIDES
Production
drain sequence
Cumulative
waste-water
volume,gal
Cumulative sulfide
mg/l
Tb
% capture
hair-burn
>
re
>
f
ime
bate
pickle
730
1,800
3,300
4,000
5,100
6,300
8,600
9,400
12,000
13,200
19,700
20,100
6,490
3,780
2,570
2,240
1,870
1,730
1,350
1,260
998
910
615
604
39.4
56.6
70.7
74.6
79.4
90.7
96.4
98.7
99.8
100.0
100.9
101.0
39.0
56.0
70.0
73.9
78.6
89.9
95.5
97.8
93.8
99.0
99.9
100.0
Segregation of the 61% of sulfide-bearing wastes collected through the
hide processor pump-out system requires facilities to collect and treat
221,000 gallons per day containing approximately 1,100 Ib of sulfide. Alter-
native levels of sulfide-waste collection and treatment are illustrated in
Table 44 (for eleven of Pfister & Vogel's production units).
TABLE 44. ALTERNATIVE LEVELS OF SULFIOE COLLECTION AND TREATMENT
Treatment
alternative
Volume,
gal /day
Sulfide
Ib/day
% total loading
A
B
C
D
221,000
70,000
37,000
8,000
1,110
1,000
780
430
61
55
43
24
Sulfide capture requires errorless control of process discharges
through operator.recognition and execution of the appropriate drain
sequences. Absolute control against the mixinn of acidic wastes w11
line lime-suTfi'de waste streams must be maintained.' Thus, pickle and chrome
liquors must "be collected separately, pre-neutralized or sewered directly to
with alka-
121
-------
avoid possible generation of hydrogen sulfide within the collection facility.
Only alkaline streams can be admitted to the sulfide collection tank(s).
Process controls must be diligently executed to insure, for example, that
the pii of sulfide-bearing bate liquors do not drop below the optimal,
slightly alkaline levels.
Because of these restrictions and due to severe limitations on avail-
able space, alternative A was eliminated from further consideration.
Under normal scheduling conditions not more than two hair-burn dis-
charges are anticipated during any two-hour interval. The sulfide oxidation
system must be capable of collecting and treating unhairino wastes from two
hide processors in any two-hour interval. For a batch oxidation system,
therefore, a capacity equivalent to the volume of unhairing wastes from tv/o
hide processors is required. Also, storage capacity equivalent to the vol-
ume of one hide processor's unhairing waste should be provided for collection
of wastes which may be discharged out of phase during a batch sulfide oxida-
tion run. This additional storage capacity would equal 6,300, 3,300 and 730
gallons for treatment alternatives B, C and D, respectively.
SULFIDE OXIDATION PROCESS ALTERNATIVES
In evaluating designs for the sulfide oxidation system, the criteria
for aeration equipment and/or chemical treatment included energy efficiency,
chemical costs, and area and volume requirements. All of the aeration sys-
tems were evaluated assuming the use of manganous sulfate as a catalyst for
sulfide oxidation.
Among the proposals considered were:
1) Chemical Oxidation
2\ Diffused Air Aeration
3) Submerged Turbine Aeration
4) Paddle Vat Aeration
5) Surface Aeration
Chemical Oxidation
Uhile the oxidation of low concentrations of sulfides in water and
wastes has been demonstrated to be technologically feasible usinn such
strong oxidants as ozone, potassium permanganate, chlorine and hydrogen
peroxide, the cost of such chemical treatments is prohibitive when high sul-
fide levels are encountered.
Potassium Permanganate--
According to Stewart (53) potassium permannanate reacts with sulfide
in alkaline solution as shown below:
8f1nO" + 3S= * 4I!20 ^ 8Mn02 + 3SOJ + 30H"
122
-------
The potassium permanganate-to-sulfide ratio predicted for the stoichiometric
oxidation of sulfide"to sulfate is 13.2 15 KMn04/lb S=. At the September
1975 price of $0.48 per pound, potassium permanganate oxidation of sulfides
would cost $6.34 per pound of sulfide.
Hydrogen Peroxide--
Operating data for hydrogen peroxide control of hydronen sulfide at
Ft. Lauderdale, Florida, and Los Angeles, California (54), indicate that
hydrogen peroxide-to-sulfide ratios of 3:1 to 8.2:1 are required for elimi-
nation of sulfides in sewage mains. Usina $.33/lb hydrogen peroxide
(September 1975-tankcars) costs could vary from $.99 to $2.70 per pound of
sulfide treated.
Chlorine—
Theoretical considerations for the oxidation of sulfides using elemental
chlorine indicate that a stoichiometric conversion to sulfur requires only
2.22 Ib Cl2/lb S=, while a stoichiometric oxidation to sulfate requires
8.87 Ib Cl2/lb S= (55). Black & Goodson (56) showed that in dilute sulfide
solutions when chlorine was offered at a 10 to 1 ratio, 4.1 parts of chlorine
were consumed per part of sulfide at a pH of 8.95.
The Los Angeles County Sanitary Districts (57) estimated chlorine
requirements in laboratory tests to be 3-9 Ib Clo/lb S=. In practice, how-
ever, 12 mg/1 of chlorine reduced the sewage sulfide entering the treatment
facility from 1.05 mg/1 to 0.43 mg/1 for a 19:1 ratio of Cl2 offered-to-
sulfide oxidized.
Therefore, using elemental chlorine at 5.75 cents per pound (September
1975-tankcars) chemical costs for oxidation of sulfide are postulated to be
in excess of $1.09/lb S at the 19-to-l chlorine-to-sulfide ratio.
Ozone--
One possible reaction for the ozonation of sulfide in alkaline solution
is:
403 + S~ _^ S04 + 402
The dissolved oxygen thus generated would also be available for oxida-
tion of sulfides. However, the degree to which this secondary oxidation
proceeds is not known. Maximum ozone requirements, on the basis of
stoichiometry alone, are 6.0 Ib 03 per Ib S=.
According to Fair, Ocyer and Okum (58), approximately 11.4 to 13,6 kwh
of electricity are required to generate one pound of ozone from air. At
2.3<£/kwh energy would cost $0.26 to $0.31 per pound of ozone. On this basis,
ozonation costs of approximately $1.56 to $1.86 per pound of sulfide are
conservatively estimated. However, any sulfide oxidation by molecular 02
which may take place will significantly reduce these costs.
123
-------
With the chemical costs ranning from $0.99 Ib S= for hydrogen peroxide
to $6.34 Ib 5= for potassium permannanate, chemical oxidation of sulfides
in unhairing waste (at a daily cost of $1,100-$7,040) is clearly prohibitive.
It is conceivable, however, that such chemical aqents could be used in a
"polishing" step to provide a space efficient (but costly) alternative to
the aeration of larqe volumes of low concentration wastes.
Diffused Air Aeration
One method of achieving oxygen transfer into wastewaters is by air
diffusion. Compressed air is blown in through a network of air diffusion
nozzles strategically located near the bottom of the aeration tank. By
releasing the air as minute bubbles, an extremely large surface area is
created to maximize oxygen transfer at the air-liquid interface. Once the
oxygen requirement of the treatment process has been established, aeration
requirements can be determined using the equations illustrated in Figures
24 and 25 (59).
Treatment alternative B (Table 44) calls for the batchwise oxidation
of 180 pounds of sulfide contained in 12,600 gallons of waste (from two hide
processors) within a two-hour interval. After deducting 30 minutes for fill-
ing and draining the reaction vessel, the time available for aeration is
reduced to approximately 90 minutes, increasing the required design oxygen
transfer rate to 120 Ib Op/hr. The design temperature is 30°C and the eleva-
tion is +700 ft. For a "typical" air diffusion system, (the Rexnord Air
Lock Crown Diffuser)(60) the manufacturer's standard efficiency rating is
8.6% at a 13 ft submergence and 15 scfm air flow per diffuser. Thus, under
actual operating conditions, the transfer efficiency for each diffuser would
be 8.8%. At 15 scfm per diffuser approximately 0.255 Ib O^/min would be
offered (61). Each diffuser would transfer:
0.255 Ib 02/m1n x 60 min/hr x .088 = 1.35 Ih 02/hr
Therefore, 89 diffusers would be required to achieve the design transfer
rate of 120 Ib/hr. Since the manufacturer's recommended minimum surface
requirement is one square foot per diffuser, the minimum tank area would be
89 ft2.
Thus, a 12 ft diameter tank with a liquid depth of 15 ft (12,600 gallon
capacity) plus approximately three feet of freeboard could be used. Air
offered would equal 1,340 scfm at 5.85 psi and approximately CO hp of
blower capacity would be required (62).
Similarly, Treatment Alternative C would require approximately 50 hp of
aeration capacity to provide 1,050 scfm of equivalent to 94 Ib Oo/hr through
70 diffusion heads. The required tank diameter would be reduced to 10 ft.
Finally, Treatment Alternative D requires 30 hp of aeration capacity
to provide 54 Ib Op/hr (600 scfm) through 40 diffusion heads. The tank
diameter could be further reduced to eight feet.
124
-------
Ea = Es
(Alt) (ft) (Cm) -
Cm) - C]
m
["I
1
T-2C
.024
where Ea = transfer efficiency of the equipment in the
treatment system, %
E- - transfer efficiency of the equipment under
standard conditions, %
Alt= Altitude Correction = ratio of barometric pressure
at design altitude to pressure at sea level.
Cm = mean saturation value of oxygen in tan water at
' the aeration mid-depth at temperature T, mn/1. (Figure 25)
Csm= mean saturation value of oxyqen in tap water at
the aeration mid-depth at 20°C, mn/1. (Figure 25)
C-j = dissolved oxygen operatinn level, mo/1.
ft - ratio of oxynen saturation in waste to that in tap
water at the same temperature.
oC = ratio of oxygen transfer in waste to that in tap
water at the same temperature.
T = temperature, °C
Given: C-| = 0 mg/1 ;/9=l .0;* =1.0; T=30°C; Fs=8.6%; AH=0.975
Ea = 8.6% p-J/") 0-°) ^8-51 ^^ - ° ^/^"l I IT 30-201
*~0-.30 mn/1
= 8 6* |(-975) (1.0) (8.51 mg/1) - 0 nn/fl P IT 30-20]
* ' L TO-3 mg/1 J |_(1.0)||0.029) J
p a — o r\°/
L Q "* O*O/O
Ea = 8.6%
).3 mn/1
;.C05) (1.27)
H
Figure 24. Determination of fie/ld, transfer efficiency.
125
-------
fir * >]
Where Cg = 02 saturation concentration at a given temperature
in tap water, mg/1.
Cs (20°C) = 7.44 rng/1
Cs (30°C) = 9.12 mg/1
Cp = Csm at 20°C
Pb = absolute pressure at the depth of air release, psi
depth of air release = 13 ft
3 I
0^ = concentration of oxygen in air leaving aeration tank,
Ot =• 20.5 - (20.5 x 0.08C) « 13.7S
Csn, - 3.02 m/1 + . 10.3
8'51 "5
Figure 25. Determination of mean oxygen saturation value.
126
-------
Submerged Turbine Aeration
Another common method of aeration for waste treatment is submerged
turbine aeration. Like diffused air aeration, compressed air is blown into
the bottom of the treatment vessel for distribution throuqhout the reaction
medium. Unlike diffused air, however, this distribution is carried out by
the sparging effect of a turbine. Air is introduced through the hollow
turbine shaft into the bottom of the tank, and the turbine blades shear the
air into minute bubbles which are evenly distributed throughout the tank.
Given the oxygen transfer requirement for Treatment Alternative B of
120 Ib 02/hr and assuming that the ratio of field efficiencies to
manufacturer-standard efficiencies is 1.02 as computed for air diffusion,
aeration requirements for submerged turbine aeration can he calculated
from manufacturer-standard efficiencies reported for a submerged turbine.
The Eimco NS-8 submerged turbine aerators (C3) are nominally rated at 3.3 Ib
02 transferred per horsepower-hour at standard conditions. Field transfer
rates, therefore, of 3.4 Ib 02/hp-hr are predicted. The total power require-
ment estimated for the submerged turbine is 35.3 brake-horsepower. At a
drive efficiency of 82%, the actual nameplate horsepower requirement would
be 43.1. Therefore, a 50 hp turbine could be employed using an additional
20 hp to supply 282 scfm or air, for a total power requirement of 70 hp.
Hith a maximum manufacturer-recommended power-to-volume ratio of five horse-
power per thousand cubic feet, a tank capacity of 75,000 gallons would be
required. At a liquid depth of 15' the surface dimensions of such a tank
could be 30' diameter (or 26' square). For Treatment Alternatives C and D,
approximately 60 and 30 hp, respectively, would be required with correspond-
ing tank volumes of 60,000 (26' diameter) and 30,000 (18.5 ft diameter)
qallons.
Paddle Vat Aeration
One viable alternative for tanners makes use of abandoned (or new)
paddle vats for the aeration of sulfide-wastes. Pilot studies at Pfister &
Vogel Tanning Company demonstrated a mean sul fide oxidation rate of 17 Ib
S~/hr for 3,300 gallon batches in a 4,000 gal capacity paddle vat, using
manganous sulfate as catalyst.
For Treatment Alternative B, a pair of paddle vats must be available
for each 6,300 gallon process discharge.
At 34 Ib S=/hr (for two paddle vats) approximately 2.75 hours of aera-
tion would be required per 6,300 gallon batch. Allowing 0.5 hour to fill
and drain the two paddle vats, the total batch time would be increased to
3.25 hours for an effective rate of 27.7 Ib S=/hr per pair of paddle vats.
Therefore, to oxidize sulfides at the design loading of 90 pounds per hour,
eight paddle vats would be required (four pairs). Pfister & Vogel's paddles
were fitted with 15 hp motors. While the nameplate-horsepower thus pre-
dicted equals 120, actual power requirements may be substantially lower
since the paddle vats were originally designed for operation under the more
strenuous conditions of processing leather. The effective volumetric
requirement would be 26,000 gallons with a net surface area of approximately
1,000 ft2.
127
-------
For Treatment Alternative C, five paddle vats would be needed with a
total effective capacity of 16,000 gallons and 630 ft2 of surface area.
Likewise, the requirements for Treatment Alternative D include 3 paddle
vats with an effective total volume of 10,000 gallons having 380 ft2 of
surface area.
Thus, while paddle aeration can reduce capital expenditures when
abandoned or under-utilized equipment is available, power costs are relative-
ly high and a great deal of manufacturing floor space must be sacrificed.
Surface Aeration
Another type of aeration applicable to the manqanese-catalyzed air
oxidation of sulfides in unhairing wastes is surface aeration. In contrast
to air diffusion and air dispersion, surface aeration effects oxygen trans-
fer by pumping large volumes of treatment liquor radially into the atmos-
phere to form a high velocity spray pattern. The droplets thus formed dras-
tically increase the air-liquid interfacial area which is essential for
oxygen transfer. Also, impingement of the spray on the liquid surface
causes additional agitation, further aiding oxygenation as well as establish-
ing a dynamic mixing pattern.
The design aerator efficiency for Alternative R was previously estab-
lished at Pfister & Vonel in pilot studies utilizing a two horsepower
"Aqua-Jet" (C4) floating surface aerator with a manufacturer oxygen transfer
rating of 3.9 Ib 02/hp-hr. At a mean, initial sulfide concentration of
1,730 mg/1, an aerator-sulfide oxidation efficiency of 7.3 Ib S=/hp-hr has
been established for this aerator. Assuming that the sulfide oxidation
efficiency-to-oxygen transfer efficiency ratio of 1.87:1 remains constant
for all size "Aqua-Jet" aerators, the 20 hp "Aqua-Jet Mark 20" (oxygen trans-
fer efficiency = 3.2 Ib Og/bp-hr) would be capable of oxidizing sulfides at
the rate of 6.0 Ib/hp-hr. Allowing 30 minutes to fill and empty the reaction
vessel, 20 hp in aeration capacity would be required to achieve the 120 Ib
S=/hr design rate. Aeration tank capacity should equal 13,600 gallons,
incorporating the maximum surface area practicable within the limit of the
minimum manufacturer recommended liquid depth of four feet.
Such a tank would have a diameter of 23 feet with a total sidewall of
7 feet including approximately 3 feet of freeboard. This design fully
accommodates the 20 foot diameter impingement pattern exhibited by this
aerator. Where ample floor space is unavailable, however, a reduction in
design surface area is advisable as long as aerator performance is not
seriously impaired. An alternative reaction vessel design, therefore,-.:would
be a 15 foot diameter tank with a liquid depth of 9.5 feet (effective volume
12,600 gallons) plus an additional 3 feet of freeboard. Likewise, a square
surfaced tank could be employed having the advantane of more efficient space
utilization with an increased surface-to-volume ratio. Such a design,
however, would risk the development of static zones in the corners where
sulfide oxidation might be inhibited due to inadequate mixing.
Finally, it is possible to install two aeration vessels, each havinq
half of the design aeration capacity. The switch from a single 20 hp aerator
128
-------
to two 10 hp aerators has several practical advantages. First, in the case
of the Aqua-Jet aerators, the efficiency of the 10 hp aerator is slightly
higher (3.4 Ib Og/hp-hr vs 3.2 Ib 02/hp-hr) than that of the larger aerator,
so that a combined sulfide oxidation rate of 128 Ib S=/hr (vs 120 Ib/hr for
a 20 hp aerator) is anticipated. Second, the installation of parallel aera-
tion systems vastly improves operational and maintenance flexibility. Also,
such a "dual-reactor" system provides the capability of comparative testing
under controlled conditions.
At an aerator efficiency of 6.55 Ib S=/hp-hr (1.87 Ib S=/02 x 3.5 Ib
Oo/hp-hr for a 15 hp Aqua-Jet aerator), the requirements for Treatment
Alternative C would equal 14.4 hp at the design oxidation rate of 94 Ib S=/
hr. Thus, a 15 hp aerator would be chosen along with an aeration vessel
having a 6,000 gallon capacity such as a 15 ft diameter cylindrical tank
with a liquid level of 5 ft and an additional 3 ft of freeboard.
Finally, for Treatment Alternative D a sharply lower design efficiency
of 3.9 Ib S=/hp-hr must be used. Pilot studies indicated that for the 2 hp
aerator concentrated hair-burn liquors were oxidized at a rate of 4.2 Ib S=/
hp-hr for a sulfide oxidation efficiency-to-oxygen transfer efficiency ratio
of 1.08 Ib S=/lb 02« At the manufacturer-oxygen transfer rate for the Aqua-
Jet Mark 15 aerator of 3.5 Ib 02/hp-hr, a sulfide oxidation rate of 3.9 Ib
S=/hp-hr is predicted. The aeration capacity required, therefore, to achieve
the design sulfide oxidation rate of 54 Ib S=/hr is 13.9 hp. Thus, a 15 hp
aerator would be specified. A reaction vessel with an 8,000 gallon capacity
(15' diameter x 6'-l" deep plus 3' of freeboard) capable of containing a
full day's loading would be recommended. Anticipated operation of this
system would incorporate batchwise aeration over a 90 minute interval with a
subsequent partial draw-off of treated effluent to accommodate fresh wastes.
The data in Table 45 illustrates various chemical oxidation procedures
with operating costs of $0.99 to $0.34 per pound of sulfide to be 21 to 135
times as costly as the most expensive aeration alternative. Removing
1,000 Ib S"/day by the least expensive chemical oxidation alternative would
cost more than $250,000 annually. The operating costs projected for every
aeration alternative considered are drastically lower than for chemical
treatment.
Catalyst costs for each of the aeration alternatives were computed on
the basis of catalyst-to-sulfide ratios employed in the Paddle Vat_aeration
studies. The optimum was shown to lie within the 0.10-0..15 Mn++/S= range
for manganous sulfate. A ratio of 0.15 was therefore chosen as a conserva-
tive value for this feasibility study.
The pilot aeration studies show that "spent" catalyst retains nuch of
its activity and reuse of the "manganese hydroxide" sludne is expected to
greatly reduce catalyst requirements. It appears from Table 45 that catalyst
cost-to-power cost ratios would range from 1.7:1 to 13:1. In practice,
however, it is anticipated that these ratios could be substantially lower
with catalyst optimization.
It was assumed that the optimum catalyst dosage for the mannanese
129
-------
TACLE 45. OPERATING COSTS FOR VARIOUS SULFIDE OXIDATION ".ETHOOS
oo
o
Alternative
3
B
D
LJ
3
r>
D
C
c
C
c
c
D
D
D
D
D
Process
Air Diffusion
Air Dispersion
Paddle Aeration
Surface Aeration
Chemical
Oxidation
Air Diffusion
Air Dispersion
Paddle Aeration
Surface Aeration
Chemical
Oxidation
Air Diffusion
Air Dispersion
Paddle Aeration
Surface Aeration
Chemical
Oxidation
Volume, gal
per hide
processor
6,300
6,300
6,300
6,300
3,300
3,300
3,300
3,300
3,300
3,300
730
730
730
730
730
Operating cost (S/lb S~)
Tank
area, ft
113
707
1,010
177
79
531
630
177
50
278
370
177
—
Power,
kwh/lb S=
0.33
0.44
0.75
0.12
11.4 -*
13.6
0.40
0.43
0.60
0.10
11.4 -*
13.6
0.42
0.42
0.55
0.21
11.4 -*
13.6
Power
0.0037
0.0101
0.0173
0.0028
0.262 -
0.313
0.0092
0.0111
0.0134
0.0023
0.262 -
0.313
0.0097
0.0097
0.0126
0.0048
0.262 -
0.313
Chemical
0.0295
0.0295
0.0295
0.0295
0.99-6.34
0.0295
0.0295
0.0295
0.0295
0.99-6.34
0.0295
0.0295
0.0295
0.0295
0.99-6.34
Total**
0.0382
0.0395
0.0468
0.0323
0.99-6.34
0.0387
0.0406
0.0423
0.0318
0.99-6.34
0.0392
0.0392
0.0421
0.0343
0.99-6.34
* Ozone on'ly-
** excluding labor pumpinq and maintenance costs.
-------
catalyzed air oxidation of sulfides would bo the same for each of the
aeration alternatives. Power requirements and surface requirements were
weighted most heavily since it is impossible to predict the real differences
in catalyst requirement attributable to the various types of aeration. Fur-
ther, it is assumed that present trends in the cost and availability of
electric power will continue, so that the importance of power costs relative
to catalyst costs will continue to increase.
Surface aeration was chosen as the preferred system for aeration of
sulfide-bearinn wastes for all three treatment levels primarily because of
the much lower power requirements vs alternate aeration systems.
Even though area requirements were substantially nreater than those for
air diffusion, the 177 ft2 required for surface aeration was well within the
strictures of available plant area.
Further consideration was given to the desirability of a dual-reactor
capability in the sulfide oxidation system design. Two ten-horsepower
floating surface aerators and two 12 ft diameter reaction vessels were
incorporated in a modified design for Treatment Alternative C.
Ample aeration capacity was included so that at a future date, by in-
stalling a 6,600 gallon storage tank, the sulfide oxidation system could be
converted to meet the requirements of Treatment Alternative B.
Two 12 ft diameter by 11 ft deep fiberglass reinforced resin tanks were
installed, taking advantage of the extra head space offered by sinking them
into existing paddle pits and incorporating existing sewers. The resin used
in construction of the tanks was specified for use under both pi! extremes.
The aerators were designed in stainless steel to withstand the harsh
chemical environment as well as the turbulent hydraulics envisioned. A
three ft high stainless steel cylindrical motor shield with a one-way conden-
sate drain was specified to protect the aerator motors from inundation by
foam. Each aerator was moored using three guide posts installed vertically
at the corners of an equilateral triangle within the aeration tanks and
three slide rinos welded to the circumference of the aerator floats.
131
-------
SULFIDE CONTAINING WASTES
pH MONITOR-
AUTOMATIC VALVE-Q
BY-PASS
SULFIDE OXIDATION
HOLDING TANK
(Planned)
FILL-VALVE-^ $
CATALYST
FEED TANK
CATALYST
FEED TANK.
AERATOR
SULFIDE
OXIDATION
TANK
CATALYST
P"
CATALYST
x* FILL'
Q?~VALVE
AERATOR
SULFIDE
OXIDATION
TANK
SEWER
I
Figure 26: SULFIDE OXIDATION SYSTEM.
132
-------
SECTION XIII
SULFIDE OXIDATION SYSTEfl-
FULL-SCALE OPERATION
The full-scale Sulfide Oxidation System, as illustrated in Figure 26,
incorporates two ten-horsepower, floating surface aerators in separate
7,000 gallon capacity cylindrical reaction tanks. Each tank measures 12 ft
in diameter with an 11 ft sidewall. At the 7,000 gallon level, 2.75 ft of
freeboard is provided. A valve manifold at the end of the "hairburn header"
pipe allows the operator to direct sulfide wastes to either of the reaction
tanks or to bypass them to the sewer. A continuous in-line pH monitor is
located in the hairburn header with a low-level alarm relay set at pH 10.0.
If the pH in the header drops below the alarm level an automatic air-
operated valve closes and an alarm sounds. Since this automatic valve is
upstream of the two reaction tank fill-valves (but downstream from the by-
pass valve) acidic wastes cannot be admitted to the sulfide oxidation
system.
A 400 gallon fiberglass tank with a high speed chemical mixer has been
installed for catalyst make-up. A standard (1 Ib Mn++/flal) solution of
manganous sulfate is prepared daily and pumped to graduated, 30 gallon
capacity, gravity feed tanks for batch-wise addition to the aeration tanks
as required. Aerators are moored to guide posts within the aeration tanks
so that they can operate at liquid levels from 3 to 9 feet. Adequate space
has been provided overhead for the future installation of a 6,600 gallon
holding tank to increase treatment capabilities.
Full-scale operating data confirm pilot study results for the manganese
catalyzed air oxidation of sulfides in unhairing wastes and for the scale-
up of surface aeration equipment. For ten oxidation runs at catalyst-to-
sulfide dosages varying from 0.05 to 0.15 To Mn++/lb S=, a mean time of 2.75
hours was required to oxidize the sulfide in 6,400 gallons of unhairing
waste from 2,830 mg/1 to zero. The rate of oxidation over the ten runs
averaged 58.3 Ib 3~/hr for an average aerator efficiency of 5.83 Ib S~/hn-hr.
The time required for 98% reduction of the sulfide, however, is only
2.125 hours per sulfide oxidation run for an efficiency of G.89 Ib S=/hp-hr.
Thus, while the rate observed for oxidation to the zero sulfide level is
within 91% of the design efficiency of 6.4 Ib S=/hp-hr, the mean rate
observed for 98%. removal of sulfide actually exceeds the design rate. The
maximum incremental efficiency observed during full-scale operation was
15.2 Ib S=/hp-hr. However, during the "die-off phase" of the reaction where
the last two percent of the sulfide is oxidized the efficiency is reduced to
a mere 0.5 Ib S=/hp-hr.
133
-------
OPTIMUM CATALYST-TO-SULFIOC RATIO
To determine the effect of catalyst dosage durin" the full-scale oxida-
tion of sulfides from unhairing wastes in a surface aeration system, vary-
ing amounts of manganous sulfate monohydrate (Carus :!n23) were added in
three liquid feeds 30 minutes apart to approximately 6,000-7,000 gallon
batches of unhairing wastes. The unhairinq wastes for each aeration run
were composed of concentrated hair-burn liquor and approximately half of the
subsequent wash from two hide processor production runs. Initial sulfide
concentrations varied from 2,310 mg/1 to 3,600 mci/1 with the total initial
sulfide content of the batches varyinn from 117 Ib to 175 Ih. Initial
temperatures averaged approximately 04°F, while finals v/ere approximately
10° higher. Followinn each run the aeration tank was drained and hosed
with fresh water to remove all residual catalyst-siudne.
The data in Table 46 illustrate that a baseline sulfide oxidation run
without any catalyst required nine hours of aeration to reduce the initial
sulfide concentration from 2,720 mg/1 (148 Ib) to 65 mn/1 (3.5 Ib), for a
treatment efficiency of 97.6%. The mean oxidation rate '7as 16.1 Ib S=/hr.
The corresponding mean and maximum sulfide oxidation-aerator efficiencies
were 1.61 Ib S~/hp-hr and 2.55 Ib S~/hp-hr, respectively.
When manganous sulfate was added at a 0.025 Mn'H75= ratio in a subse-
quent run, the oxidation rate was substantially improved, with only 5.5 hours
of aeration required to reduce the sulfide from an initial concentration of
2,310 mg/1 (117 Ib) to a "zero detectable" level. The moan rate of 21.3 Ib
S=/hr v/as 32% greater than the no-catalyst rate, while the maximum 15-minute-
incremental rate soared to 87.2 Ib S=/hr, nearly 3.5 tines the "un-catalyzed"
maximum.
Likewise, by increasing the catalyst-to-initial sulfide ratio-to 0.0375
a marked Improvement in the mean oxidation rate to 31.3 Ib S=/hr was effected.
The time required to achieve a zero sulfide level was further decreased to
4.5 hours but a decrease in the maximum 15 minute-incremental rate to 70.7 Ib
S=/nr was also observed.
A further increase in catalyst levels demonstrated improved oxidation
rates in three separate runs at the 0.05 ttr\++/S= ratio. For these three runs
initial sulfides averaged 3,110 mg/1 (157 Ib) requiring slightly more than
2.75 hours of aeration to reach the zero sulfide level. The mean rate of
oxidation for the three runs was 55.5 Ib S~/hr, while the maximum 15 minute-
incremental rate recorded was 152 Ib S=/hr.
A sulfide oxidation run at the 0.10 Mn++/S ratio required 2 hours of
aeration to reduce the initial sulfide concentration of 2,660 mg/1 to zero
for a mean rate of 65.5 Ib G~/hr. The maximum incremental rate of 132 Ib
S=/hr was slightly lower than that observed at the 0.05 catalyst-to-sulfide
ratio, but higher than for the 0.0375 ratio.
Finally, five runs at the 0.15 Mn++/S= ratio demonstrated that this
level was clearly beyond the range of efficient catalyst utilization.
134
-------
TABLE 46-. EFFECT OF CATALYST DOSAGE ON SULFIOE OXIDATION
CO
01
Rate
4-4- =
Mn++/S
0
0.025
0.037
0.050
0.100
0.150
flo. of Volume,
trials
1
1
1
3
1
5
gal
6,520
6,100
6,800
6,070*
5,920
6,710*
Initial
Sulfide
mg/1
2,720
2,310
2,520
3, 110*
2,660
2,730*
Ib
148
117
143
157*
131
155*
Time
hr
9.0**
5.5
4.5
2.83*
2.0
2. 65*
Mean ,
Ib S=/hr
16.1
21.3
31.8
55.5*
65.5
58.5*
flaximun
incremental ,
Ib S"/hr
25.5
87.2
70.7
152
132
134
* Average Values
** Sulfide reduced to 65 ng/1 in 9 hours
-------
An average of approximately 2.G7 hours was required to oxidize sulfides
from a mean initial level of 2,730 ng/1 to zero for a mean rate of oxidation
over the five runs of 53.5 Ib S=/hr. Individual mean rates varied from 51.0
Ib G=/hr to 75.0 Ib S=/hr, while the maximum 15 minute-incremental rate ob-
served during the five runs was 134 Ib 5~/hr.
A mean rate of sulfide oxidation of 65.5 Ib/hr was observed at the
0.10 Mn++/S= ratio. The apparent increase over rates at the 0.05 and 0.15
levels, however, cannot be considered to be statistically significant since
only one run was conducted at this catalyst level. Further analysis of the
data (Two Sample Ranks Test: n-j=3, n2=5, sum of ranks of smaller group = 13)
indicates that the average of the mean rates for sulfide oxidation at the
0.05 and 0.15 Mn+7S= ratio do not differ significantly. Therefore, it
might be concluded that the optimum Mn++/S= ratio for the oxidation of sul-
fides in impairing wastes by surface aeration lies within the 0.05 to 0.15
range. Further investigation of these catalyst ratios is indicated, with
particular attention to cost-efficiency evaluations.
These full-scale sulfide oxidation runs confirmed the observations made
during previous pilot-scale oxidation runs. The color of the reaction mix-
ture followed the same pattern, with the initial dark brown color of the
unhairing waste-catalyst mixture gradually lightening to a celery green, and
finally turning to beige near the zero sulfide endpoint.
Foaminq also followed the trend observed in pilot runs. Hith approxi-
mately three feet of freeboard available, however, and due to the special
aerator motor shields incorporated into the aeration system design, antifoam
was not required to contain the mixture within the reaction vessel. A pro-
nounced increase in foam generation was generally observed after approxi-
mately half of the sulfide had been oxidized and following the observed
incremental rate maxima. Hear the endpoint of the oxidation the foam became
extremely stable and in a few cases completely concealed the aerator spray.
Consistent with the pilot studies observations, the phenomenon of
regeneration of sulfides following aeration was also observed in the full-
scale sulfide oxidation runs. In one run at the 0.15 catalyst-to-sulfide
ratio, aeration was halted as soon as the zero detectable sulfide level was
reached. After 24 hours, the aerator was turned on briefly to mix the con-
tents of the tank and a sample was withdrawn. This sample analyzed at 123
mg/1 of sulfide.
In another run the tank was sampled 3, 5 and 24 hours after the aerator
was turned off, but without mixing the contents. These samples taken at a
level halfway down the side of the tank analyzed at 18, 16 and 27 mg/1 of
sulfide, respectively. A sample of scum collected from the surface after
4 days.of settling contained 12 mg/1. At the same time, an interlayer sample
analyzed at 36 mg/1 of sulfide while the bottom sediments contained 288 mg/1,
suggesting that the regeneration phenomenon may be due to the formation of a
settable, metastable precipitate which later reverts to sulfide.
Tanners must be cautioned, therefore, that treatment of sulfide-bearing
wastes to a zero sulfide level does not insure the absence of sulfide from
136
-------
the final effluent. Rather, where appreciable storage or lagooning of
"de-sulfided" wastes prior to discharge is necessary, measurable quantities
of sulfide may reappear.
Another observation consistent with the pilot aeration studies indicates
a slight Increase 1n pH (approximately 0.2-0.4 pll units) which occurs during
the maximum rate intervals. This sudden p!l increase supports the hypothesis
that sulfide is being converted to elemental sulfur according to the equa-
tion:
2HS" + 02 +211+ „ 2 H20 +2S
The oxidation of sulfides to sulfate can be expressed as:
US' + 202 ^ SOj + H+
or
HS" + 202 + OH" —- SOj + H20
The consumption of hydroxyl ions precludes the occurrence of sulfate as the
principal oxidation product.
The sulfide oxidation curve for a typical full-scale aeration run which
is shown in Figure 27 very closely follows the shape of the sulfide oxida-
tion curve (Figure 22) generated for paddle aeration. This curve may be
characterized by an initial sharp decrease in sulfide concentration with
decreasing slope. At a point corresponding to approximately half the initial
sulfide concentration an inflection occurs followed by a phase of rapid in-
crease in slope.
The significance of the inflection point can be emphasized by double
differentiating the sulfide oxidation curve. The second differential corre-
sponds to a plot ofAS=/W vs T. This rate curve (Figure 28) emphatically
shows the rate maxima corresponding to 40.0 mg/l-min for this run (152 1b
S=/hr). With further aeration the rate of sulfide disappearance declines
and approaches zero.
CATALYST VARIATIONS
The effect of'potassium permanganate on the oxidation of sulfides.in
unhairing wastes by surface aeration was investigated in a full-scale aera-
tion run. A 6,700 gallon batch of unhairing waste containing 2,270 mg/1 of
sulfide was treated with three feeds of potassium permannanate (18.7 Ib
KMn04) 30 minutes apart to bring the Mn++/S= ratio to the 0.05 level. Ini-
tial and final temperatures of 85° and 92°F were respectively recorded.
After 2.75 hours.of aeration a zero sulfide reading was recorded for a mean
rate of sulfide oxidation of 45.5 Ib S=/hr. The maximum incremental rate
observed was 82.1 Ib S=/nr.
137
-------
3000-
2500-
2000-
O)
E
LU
Q 1500
LL
_J
D
CO
lOOOh
500-
60 SO 120
TIME, min
180
Figure 27: FULL-SCALE SULFIDE
OXIDATION CURVE.
138
-------
30 60 90 120 150 180
TIME, min
Figure 28: RATE CURVE FOR FULL-SCALE
SULFIDE OXIDATION.
139
-------
30
90 120
TIME, min
Figure 29: FULL-SCALE SULFIDE OXIDATION
RATE CURVE-POTASSIUM
PERMANGANATE CATALYST.
140
-------
The sulfide oxidation rate curve shown in Figure 29 closely resembles
the curves for full-scale aeration usino manganous sulfate as catalyst. The
shape of this curve, however, with the rate maximum appear!no between 75
minutes and 90 minutes, sharply contrasts the_earlier results for paddle
aeration using potassium permanganate (Mn'*"f/S"= 0.15), where the maximum
incremental rate was observed within the first 15 minutes of the reaction.
Since the cost of using potassium permanganate is approximately seven times
that of manganous sulfate at equivalent Mn+Vs3 ratios, and since no rate
improvement was demonstrated at equal Mn+VS3 ratios, it is concluded that
manganous sulfate catalysis is more cost effective than potassium permanga-
nate for the oxidation of sulfides in unhairing wastes.
One possible explanation for the sharp contrast between the rate curves
for the two aeration systems lies in the fact that in paddles, potassium
permanganate was employed at a 0.15 Mn"H7Sss ratio (in a single feed), three
times the level employed in the full-scale oxidation run. It is possible,
therefore, that in paddle aeration the activation energy for direct chemical
oxidation of sulfide by permanganate was immediately surpassed, causing the
early rate surge, while permanganate levels in the full-scale system were
insufficient.
Also, with a manganese sulfate catalyzed air oxidation rate of roughly
three times that encountered in paddle aeration, the relative contribution
of air oxidation vs permanganate oxidation to the overall oxidation rate is
much greater in the full-scale system. Thus, the paddle aeration results
closely follow a potassium permanganate-oxidation model while the full-scale
permanganate run is more closely described by a manganese catalyzed air oxi-
dation model.
Another variation in catalyst usage was investigated in which a total
Mn++/S= ratio of 0.15 was achieved by addition of manganous sulfate in eleven
liquid feeds, 15 minutes apart. The first feed, at time zero, was made at a
Mn+VS3 ratio of 0.05. At successive 15 minute intervals additional feeds
were made to Increase the Mn+VS" ratio in 0.01 increments to a final level
of 0.15. Two and one-half hours of aeration were required to reduce the
sulfide in approximately 6,500 gallons of waste from an initial concentra-
tion of 2,480 mg/1 to zero. The mean rate of oxidation was 53.7 Ib S=/hr,
while the maximum incremental rate observed was 112 Ib S~/hr, neither of
which shows an improvement over previous runs at the 0.05, 0.10 and 0.15
Mn+'Vs* ratios where catalyst was added in three liquid feeds 30 minutes
apart.
AGED UNHAIRING WASTES
To determine the effect of the aging of unhairing wastes on the manganese
catalyzed air oxidation of sulfides, approximately 7,000 gallons of unhairing
wastes were stored at ambient temperature for twelve days, and were subse-
quently treated by surface aeration. After aging, the waste analyzed at
2,930 mg/1 of sulfide and the temperature was 84"F. Manganous sulfate was
added in three liquid feeds 30 minutes apart to a Mn++/S~ ratio of 0.15.
After 9.5 hours of aeration a zero sulfide reading was established. The mean
141
-------
rate of oxidation was thus reduced to 37.6 Ib S=/hr, while the maximum incre-
mental rate was 66.0 1b S~/hr. This maximum occurred at a much lower sulfide
concentration than for fresh unhairing waste.
In a subsequent sulfide oxidation run with two month old unhairing waste,
six hours of aeration was required to reduce the sulfide level from 3,040
mg/1 to 145 mg/1. Thus, the mean rate of oxidation dropped further to 25.2
Ib S=/hr. The maximum incremental rate observed, however, increased markedly
over the previous run to 144 Ib S=/hr. Also, in contrast to the earlier run,
this maximum incremental rate was observed early In the reaction (15 minutes
to 30 minutes) reducing the sulfide concentration from 2,880 mo/1 to 2,190
mg/1 within a 15 minute interval. The pH of the reaction mixture Increased
from 11.48 to 11.89 over the 6 hour period, with most of the increase occur-
ring during the maximum rate interval.
Thus, 1t is obvious that the extended aging of unhairing wastes prior to
aeration seriously reduces the mean sulfide oxidation efficiency of this
manganese-catalyzed surface aeration system.
FOAMING
During these full-scale sulfide oxidations two runs were atypical with
respect to severity of foaming. In both runs during the second half of the
reaction, the foam layers were extremely stable measuring approximately 2.5
to 3 feet in thickness. The resultant reduction in mixinn efficiency caused
a stratification to occur within the reaction vessel.
To approximately 6,000 gallons of unhairinn wastes containing 2,580 mg/1
of sulfide, manganous sulfate catalyst was added in three liquid feeds at
30 minute intervals for a Mn++/S= ratio of 0.15. Only two hours of aeration
was required to reduce the sulfide level to 45 rng/1. Thus, a sulfide re-
moval efficiency of better than 98% was achieved at a mean rate of 61.0 Ib
S=/hr. At this point, foaming was so severe that the aerator spray was com-
pletely concealed and the foam layer was within 6 inches of the top of the
motor shield (three feet high). The reaction very closely followed the
pattern established by previous runs except that even with an additional
three hours of aeration the zero sulfide level could not be attained. After
a total of five hours of aeration, a grab sample taken approximately halfway
down the side of the reaction vessel analyzed at 4 mg/1 S=. Another sample,
however, taken from the bottom of the tank revealed a sludge layer which
contained 54 mg/1 of sulfide.
Another run using 5,500 gallons of unhairinn waste showed similar
results at the 0,05*!1n++/S= ratio. Again, unusually severe foaming was
observed during the declining-rate phase of the reaction. Only two hours
of aeration was required for a 982 reduction in the sulfide concentration
from an initial level of 2,590 mg/1. With another 2.5 hours of aeration,
however, the sulfide level could only be reduced to 13 mg/1.
1/hile the cause of this foaming remains undetermined, it is possible to
achieve partial control over foaming. In a subsequent oxidation run,
142
-------
silicone antifoam was added during the declining-rate phase of the reaction.
Approximately 7,000 gallons of unhairing waste containing 2,740 mg/1 of sul-
fide was aerated for 1.5 hours (Mn++/S~ = 0.15) to reduce the sulfide to
290 mo/1. A 1% solution of silicone antifoam was continuously added over
the next 1.5 hour interval. The antifoam partially restrained foaming so
that the aerator spray could just be observed. After a total of 3.0 hours
of aeration, a sample which was representative of the entire reaction vessel
contents contained no sulfide. The mean rate of oxidation was 51 Ib S=/hr
with a maximum incremental rate of 101.G Ib S=/hr.
Approximately 5.5 pounds of silicone antifoam was required to control the
foaming. At a price of $1.40 per pound for the antifoam, the cost to control
foaming would be $7.70 per sulfide oxidation run. Since this antifoam im-
proves sulfide removal efficiency for only the last two percent of sulfide
(approximately 3 pounds), the cost per pound of sulfide is approximately
$2.50. Thus, while such antifoams are capable of controlling foaming and
thus completing the oxidation, auxiliary mixing by air injection, fluid
recirculation into the bottom of the reaction vessel, or an aerator draft
tube should be investigated as more economically-acceptable ways of achiev-
ing the zero sulfide endpoint.
TEMPERATURE
To determine tho effect of temperature on the rate of sulfide oxidation,
aeration runs were conducted using unhairing wastes at 98°F and 72°F (the
normal temperature before sulfide oxidation is 02°-86°F).
For the high-temperature run, approximately 5,000 gallons of unhairing
wastes from two hide processor runs were collected in the 7,000 gallon
capacity reactor vessel and diluted with clean hot water to the 6,500 gallon
level. The temperature after dilution was 98°F and the sulfides analyzed at
2,640 mg/1. Manganous sulfate catalyst was added in three liquid feeds for
a 0.05 Mn+4"/S3 ratio. After 3.75 hours of aeration, a zero sulfide level
was achieved. Thus, a mean oxidation rate of 38.2 Ib Ss/hr was observed,
well below the 51.7 to 58.4 Ib S=/hr previously observed at this catalyst
level. The maximum incremental rate was 66.2 Ib S=/hr and the final tempera-
ture was 99°F.
In the low temperature run, 5,000 gallons of waste were diluted to 6,500
gallons with clean, cold water to give an initial sulfide concentration of
2,290 mg/1. With three hours of aeration required to reduce the sulfide con-
centration to zero, the mean rate of oxidation was 40.7 Ib S=/hr. The maxi-
mum incremental rate observed as 85.7 Ib S=/hr and the final reaction
temperature was recorded at 79°F.
In both runs two liters (approximately 4.5 pounds) of a silicone anti--
foam, Dow Corning DB-31, were added when foaming became severe. A 1% solu-
tion of the antifoam-was dribbled into the spray pattern to help control
foam and insure adequate mixing.
The lower-than-normal rate observed in the low temperature run is
143
-------
consistent with the basic rules of thermodynamics. The reduced rate at
elevated temperatures, however, is contrary to the expected results. The
reason for this lower rate nay, in part, be the reduced solubility of oxygen
at elevated temperatures. Also, as the limited one degree fahrenheit temp-
erature increase suggests, the heat of reaction (which normally produces a
temperature increase of approximately 10°F) was probably dissipated to the
ambient air as water vapor. Turbulence at the surface of the reaction
vessel increases the surface area for evaporation to take place. This steam
could generate convection currents carrying fresh air away from the liquid
surface. Thus, an "oxygen starvation" condition may exist at the elevated
temperature. It may be possible to correct this "oxygen starvation" by
installing a small fan to blow fresh cool air over the surface of the re-
action vessel.
Therefore, as these experiments indicate, the normal starting temperature
of 82-84°F is probably close to the optimum for the batch-wise manganese
catalyzed air oxidation of sulfides in unhairing wastes for this system.
AERATOR MOORING
One mechanical problem encountered during these oxidation runs was the
tipping of the floating surface aerator during aeration. This tipping
appears to be related to wastewater foaming, and is most pronounced during
the declining-rate phase of the sulfide oxidation reaction. Also, heavy
doses of antifoam added to reduce foaming rapidly improves the condition.
Although, after consultation with the manufacturer's technical representa-
tives, the exact cause of the aerator tipping remains unknown, it may be
related to an increase in buoyancy and a shifting of the center of gravity
relative to the liquid surface. Hhile it appears that this tipping had no
effect on the efficiency of sulfide oxidation, it is unacceptable from an
operations standpoint. Men the aerator tips approximately 30° from verti-
cal, the float "jams" within the aerator guide posts. Subsequent emptying
of the reaction vessel leaves the aerator precariously suspended, and upon
release serious damage could occur.
In one attempt to correct aerator tipping, a 400 pound (100 pounds had
no effect on tipping) counterweight was attached to the aerator motor hous-
ing. This counterweight which equalled almost half the installed weight of
the aerator, maintained the aerator in a nearly level position throughout
one oxidation run. Observation of the spray pattern revealed a below-
normal pumping rate. The 6,000 gallon-batch of unhairing waste required
4 hours of aeration to reduce the sulfide from an initial level of 3,970
mg/1 to 531 mg/1 for an average oxidation rate of 41.8 Ib 2=/hr (Mn /S"
= 0.05). After 8 hours of aeration, a sample of waste taken near the sur-
face contained 50 mg/1 of sulfide, while a sample from the bottom still con-
tained 3,330 mg/1, indicating that the mixing pattern was indeed very poor,
corroborating the low pumping rate observation.
In subsequent aeration runs, three guy lines were connected from the top
of the motor housing to the edge of the oxidation tank. By maintaining
equal tension on these lines, the aerator remained level throughout the
144
-------
oxidation run. Thus, another alternative to the present mooring system
would be to use a set of three or more spring-loaded or counterbalanced
mooring cables fixed to the sidewall of the reaction vessel. Such a mooring
system would provide horizontal stability while allowing the aerator to
float on the surface at any level within operating limits.
SEQUENTIAL OXIDATION
To test the hypothesis that the manganese floe-substrate in the sulfide
oxidation process retains its catalytic activity when reused in subsequent
sulfide oxidation runs, approximately 3,300 gallons of unhairing waste were
aerated using manganese sulfate catalyst at the 0.15 Mn++/S= ratio (added in
three liquid feeds 15 minutes apart). Aeration of this initial mixture for
1.75 hours reduced the sulfide from 2,620 mg/1 to zero for a mean oxidation
rate of 41.3 Ib S=/hr. While this first use of the catalyst gave a lower
than expected oxidation rate, reuse of the catalyst-floe in a subsequent run
showed that approximately 75% of the catalytic activity was retained. After
the first run had been completed, an additional 3,700 gallons of fresh un-
hairing waste was added to the reaction vessel to brinn the sulfide level
up to 1,400 mg/1. Without additional catalyst, 2.5 hours of aeration
effected a reduction to the zero sulfide level. The mean oxidation rate for
this reused catalyst run was 32.8 Ib S3/hr.
In another full-scale sulfide oxidation run, catalyst was added to a
5,700 gallon batch at the 0.05 Mn++/S* ratio. This first use of catalyst
effected a reduction in sulfide from 3,680 mg/1 to zero within 3 hours, for
a mean oxidation rate of 58.4 Ib S=/hr. The aerated waste was drained to
leave approximately 3,000 gallons in the reaction vessel. Fresh unhairing
waste was then added to increase the volume to 6,200 gallons and the sulfide
concentration to 1,780 mg/1. Additional manganese sulfate equivalent to a
Mn+Vs= ratio of 0.05 was added in three liquid feeds 30 minutes apart.
This fresh catalys.t-refortified reaction mixture required two hours of aera-
tion to achieve a zero sulfide level for a mean oxidation rate of 46.0 Ib
S=/hr. These data suggest, therefore, that while the catalyst floe does
indeed retain catalytic activity, reuse of this manganese sludge to augment
fresh catalyst, used at the optimum Mn'H'/S=! ratio, does not effect an
improvement in oxidation rates. Further investigations into reducing the
fresh catalyst feed below the optimum level while augmenting with used
catalyst up to the optimum, may further reduce catalyst consumption.
145
-------
SECTION XIV
RECOVERY OF CRUDE PROTEIN FROM OXIDIZED
UNHAIRING WASTE
In cattlehide unhairing, alkali and depilatory agents (usually in the
form of lime and sodium sulfide and/or sodium sulfhydratn) are used to solu-
bllize the keratins of the epidermis and hair. The end products of a "hair-
burn" process Include dissolved and colloidal protein as well as suspended
fibrous hair fragments, which are transferred to the process effluent. Both
dissolved and suspended inorganic salts are also present,
A typical spent unhairing liquor from a hide processor run contained
8,380 mg/1 of total Kjeldahl nitrogen, most of which is contributed by the
denatured hide proteins. Likewise, the total volatile solids concentration
of 51,000 mg/1 (of which 30,000 mg/1 are volatile suspended solids and
21,000 mg/1 are volatile dissolved solids) reflects the extremely high pro-
tein decomposition.
Happich e_t_ al (65) demonstrated that a recovery scheme employing removal
of suspended solT3s by gravity sedimentation, screening, centrifugation and/
or filtration; removal of soluble inorganic compounds by dialysis or ultra-
filtration; acidification with acetic acid to pH 5.0 and 3.3; washing and
drying; yields a protein fraction of 90 to 92 percent purity.
Approximately 30 percent of the COD (37% BOD) present in the untreated
hair-burn waste was separated in the solids fraction after a two-stage
centrifugation. Another 31,5 percent of the COD (34.9% BOD) was transferred
to the filtrate by ultrafiltration and 38.5% of the COD (28.1% BOD) was
actually removed with thi* protein.
Happich noted that vhe recovered protein contained the ten "essential"
amino acids and closely resembled the composition of native hair. He con-
cluded, however, that th« recovered hair protein was low in most of those
amino acids relative to whole egg protein and would require supplementation
for subsequent use as feod.
To determine the degree of effluent reduction possible through protein
recovery from manganese-catalyzed, air-oxidized unhairin^' wastes, a sample
of oxidized waste from an early Pfister & Vogel paddle aeration run was sent
to the U.S. Department of Agriculture, Eastern Regional [Research Center in
Philadelphia for protein recovery by a modified treatment scheme. Precipi-
tation by acidification of the sample, as received at the' U.S. Department
of Agriculture, to pH 4.2 yielded a tan color protein of 80% purity.
146
-------
According to Happich, 70% reduction of the COD of the effluent was achieved
and 65% of the protein in the effluent was recovered. The product could be
purified to 90% protein by re-precipitation. The amino acid composition did
not differ significantly from the protein recovered from unoxidlzed lime-
sulfide unhairing effluent.
Another bench-scale experiment was conducted at Pfister & Vogel to
further define the conditions for optimum precipitation of protein and sub-
sequent reduction of contaminant levels.
Concentrated hair-burn Hquor was air-oxidized using manganese sulfate
as the catalyst* at a Mn++/S~ ratio of 0.15. After oxidation^ the liquor
contained approximately 160,000 mg/1 of total solids of which 58,400 mg/1
was suspended solids.
The titration curve for acidification of (25 ml aliquot) hair-burn
Hquor with sulfuric acid is shown in Figure 30. The buffering effect of
the dissolved protein Is evident from the shape of the curve below pH 7.0.
Varying levels of concentrated HoSO*, estimated from Figure 30, were
added to eighteen 500 ml aliquots of the oxidized waste. The samples were
mixed using a Phipps and Bird laboratory stirrer for 5 minutes at 100 rpm
followed by 5 minutes of "slow" mixing at 50 rpm. After settling for 60
minutes, the supernatant liquids were decanted and analyzed for pH, total
solids and suspended solids. The pH's, which varied from 0.3 to 9.4, were
consistently higher than the expected values. This is most likely due to
the times elapsed in the precipitation runs contrasted to the 10-15 minutes
required for the original titration. While there is no apparent relation-
ship between supernatant pf! and suspended solids (supernatant suspended
solids concentrations were consistently within the 280-630 mg/1 range except
at pH 9.4 where a markedly increased level of 1,670 mg/1 was observed) the
optimum for reduction of total solids, as shown in Figure 31, clearly lies
within the 0.9 to 5.0 pll range.
The volumes of the resultant sludges varied from 300 ml to 475 ml as
shown in Figure 32, with the minimum lying within the 1.0 to 3.9 pH range.
Fortunately, this range of minimum sludoe volume is also within the optimum
range for precipitation of protein.
At pH 3.2, the supernatant analyzed at 100,200 mg/1 of total solids
(minimum observed for 18 runs) and 401 mg/1 of suspended solids. Thus,
approximately 75% of the total solids and 99.7% of the suspended solids
present in the original, air-oxidized waste sample were transferred to the
300 ml sludge fraction after precipitation and settling.
Undoubtedly, further sludge dewatering will be required to concentrate
these solids which occupied 60% of the original sample volume. Since a por-
tion of the sludge (which contained approximately 80% moisture) dewatered
readily by gravity on an 18-mesh screen, dewatering methods recommended for
further investigation include gravity screening and sand bed filtration.
Lab-scale attempts at vacuum filtration failed to generate a firm cake indi-
cating that conventional rotary vacuum drum fi/ltration is not a feasible
147
-------
14
o
12
10
0.
25ml aliquot
6 8 JO 12
H2SO4, me
14
16
Figure 30: TITRATION CURVE FOR
ACIDIFICATION OF OXIDIZED
HAIR-BURN LIQUOR.
148
-------
Il2000h
HOOOO
CO
V)
108000
106000
H-
^ 104000
CC
LLJ
Q.
D
CO
102000
100000
5 6
PH
8
O
10
Figure "31: EFFECT OF pH ON TOTAL SOLIDS
FOR ACIDIFICATION OF OXIDIZED
HAIR-BURN LIQUOR.
149
-------
250
10
Figure 32: EFFECT OF pH ON SLUDGE
VOLUME FOR ACIDIFICATION
OF OXIDIZED HAIR-BURN
LIQUOR.
150
-------
alternative.
Another production batch of concentrated unhairino waste was oxidized
by the nanganese catalyzed air oxidation method until a zero sulfide level
was reached. This aerated waste had a pli of 12.0 and contained 156,800 mo/1
of total solids, 06,500 mg/1 of COP and 7,6(6 mg/1 of Oil and Grease.
Varying amounts of 1:1 sulfuric acid v.'ere added to five, 500 r\] aliquots
of aerated hair-burn waste to reduce the pi! to within the 2.4-5.3 ration.
As shown in Table 47, after one hour of sett!inn, sludge volumes varied
from 275 ml at pi! 2.4 to 340 ml at p!! 5.3. The sharp decrease in sludge vol-
ume at pH 5.8 is indicative of a substantial decrease in the efficiency of
protein precipitation. This sludge reduction is consistent with the observed,
sharply higher total solids and COD levels in the supernatant at pH 5.S.
The lowest supernatant COD observed was 17,900 mg/1 at a p!! of 2.4 for a
79.3% reduction over the initial concentration of 36,500 mg/1. Thus, the
COD was reapportioned to yield 275 ml of sludge containing approximately 90»
of the original COD.
TABLE 47. TREATMENT OF UNHAIRING 'JASTES BY PROTEIN PRECIPITATION
Sample
1
2
3
4
5
PH
5.80
5.30
4.00
3.75
2.40
Sludge
volume,
ml
280
340
330
330
275
Supernatant
TS,
mg/1
113,350
107,910
105,610
104,130
103,870
COD,
mo/ 1
31,700
25,100
19,800
19,500
17,900
As indicated in SECTION XII, these concentrated hair-burn liquors com-
prise only a small fraction of the total wastewater flow (i.e., approximately
11,700 gal hair-burn liquor vs approximately 1.5 million gal total daily
flow), yet contribute substantially to pollutant loadings. These concen-
trated liquors contain an estimated 11% of the total BOD loading, 182 of the
total COD and 20% of the suspended solids. I/hen the capture volune is in-
creased from 730 gal to 3,300 gal per hide processor production run, the BOD,
COD and suspended solids as a percentage of total effluent loadings are in-
creased to 19%, 33%. and 30%, respectively.
Therefore, it appears from these results that precipitation of crude
protein from oxidized unhairing wastes can effect significant reductions in
151
-------
total plant loadings provided adequate methods of sludge dewatering and dis-
posal can be demonstrated. Therefore, future R & D efforts should be direct-
ed toward sludge handling methods and ultimate sludge utilization, in addi-
tion to optimization of this protein recovery pretreatment process through
operation of a pilot plant.
152
-------
SECTION XV
REMOVAL OF SOLIDS FROM SOAK WASTES
Our sampling and analyses of individual process waste fractions reveal
that up to 12,000 mg/1 of fixed suspended solids have been present in spent
soak liquors. Further investigation showed these fixed solids to be largely
sand. The soak liquors from four packs of conventionally cured, unfleshed
and untrimmed hides were found to contain more sand than any other hide-cure
types.
Equilibrated soak liquors from the soaking of various types of hides
were sampled and analyzed for acid insoluble ash according to the following
technique. An aliquot of the soak liquor (300-1,000 ml) was settled for
approximately five minutes and the supernatant was decanted and discarded.
The residue was transferred to an evaporating dish, evaporated to dryness
and ignited 1n a muffle furnace for 30 minutes at 550°C. After cooling, the
ash was resuspended in 6N hydrochloric acid, boiled to dissolve any metal
oxides or carbonates and then diluted with distilled water. The mixture was
filtered through ashless filter paper and washed with several allquots of
distilled water. The filter paper containing the residue was then folded
and transferred to a weighed evaporating dish and again ignited at 550°C for
30 minutes. The ash content, thus determined, was reported as sand from four
hide sources, as follows:
Lot Cure 1b sand/1000 Ib hide
A 3rine cured, unfleshed 1.44
B Brine cured, fleshed 0.28
C Conventional cured, fleshed 0.44
D Conventional cured, unfleshed 3.26
Average 1.34
Thus, the quantity of sand dislodged from cured hides from four separate
hide lots averaged 1.34 lb/1000 Ib hide, ranging from a minimum of 0.28 Tb/
1000 Ib hide for a lot of brine cured, fleshed and trimmed hides to 3.26 Ib/
1000 Ib hide for a lot of conventionally cured, unfleshed hides.
In the design of most conventional municipal wastewater treatment facil-
ities, grit chambers remove inert substances preventing wear of pumps and
machinery and reducing accumulation in treatment units. Most grit chambers
are installed to remove sand and other dense particles with a specific gravi-
ty of 2.65 or greater and with a minimum diameter of 2 x 10"^ cm (66).
153
-------
Likewise, collection systems are desiqned to maintain minimum transport
velocities of 2.0-3.0 fps (67) to maintain these particles in suspension.
A sieve analysis of a portion of sand collected from the soak liquors
of a typical lot of conventionally cured, unfleshed hides yielded the results
shown in Table 48. As these data indicate, nearly 80% of the particles
(weight basis) lie within the 0.074-0.297 mm diameter range. Also, only
slightly more than 53% of this sand would be expected to settle out in a
typical grit chamber.
Since wet sand slurry is approximately 40% water by volume (specific
gravity of the mixture is thus estimated to be 1.99) approximately 159 pounds
of sand would generate one cubic foot of wet grit. Thus, at the average sand
d-ischarge level from soaking, approximately 0.0065 ft3 of grit per 1000
pounds hide is generated. At a mean total tannery effluent flow of 6,590
gal/1000 Ib hide, this is equivalent to only one cubic foot of grit per
million gallons—less than one-quarter the amount of grit recovered from the
"average" municipal treatment facility per million gallons treated (68).
TABLE 48. SIEVE ANALYSIS OF SAND FROM SOAK LIQUOR
Mesh
Number
20
20-30
30-40
40-50
50-70
70-100
100-140
140-200
200-270
270
Millimeters
0.84
0.84 - 0.59
0.59 - 0.42
0.42 - 0.297
0.297-^0.210
0.210-0.149
0.149-0.105
0.105-0.074
0.074-0.053
0.053
% of total
weight
7.8
5.7
4.0
21.0
15.1
17.1
14.2.
10.4
3.6
1.1
Cumulative % of
total weight
7.8
13.5
17.5
38.5
53.6
70.7
34.9
95.3
9C.9
100.0
A probe into the separation of sand from these wastes has shown, at
least qualitatively, that hydrocyclonic separation is an effective and com-
pact means of treatment. In one trial run, unscreened soak wastes were fed
to a hydrocyclonic separator at approximately 20 opm under a pressure of 50
psi. With an orifice diameter of 1/8", approximately 5% of the influent was
recovered as sludge. Visual inspection of the clarified liquor revealed no
trace of settleable grit. Thus, where removal of nn't before discharge to
plant or municipal sewers is desirable, hydrocyclonic separation should be
further investigated as a space efficient (but power intensive) alternative
to settling.
Further investigation into the treatment of soak liquors was directed
at screening as a preparatory step required for recycle of this waste
154
-------
fraction. While a 300° pressure screen proved to be impractical because of
repeated clogging of the orifices (even at the largest orifice setting),
separation of hair and other coarse suspended solids by gravity screening
was highly effective. Removal of these solids prior to recycle is necessary
to prevent the clogging of process recirculating pumps and. the resulting
intolerable production delays.
Approximately 700 gallons of soak wastes from a single production run
were pumped over an 18" wide gravity screen with 0.040" openings at a mean
rate of 130 gpm.
At a throughput rate of 7.2 gal/min per inch of screen width all visible
traces of flesh and hair as well as a portion of the sand were removed.
Approximately eight gallons of sludge were recovered with no significant
blinding of the screen. After the sludge supernatant was decanted and
screened a second time, 39 pounds of wet solids were recovered.
Further work to optimize the solids removal/screen opening (and thus
feed rate) relationship is indeed'warranted where soak liquor pretreatment
or recycle is desired.
155
-------
SECTION XVI
GENERATION OF OIL & GREASE AND CHROME
DURING BLUE STOCK WRINGING
Substantial quantities of 011 and Grease and total chromium are ex-
tracted from the tanned sides during the "wringing process. Blue stock
leather analyses and weight -data before and after wringing Indicate that
this mechanical extraction of 011 and Grease is equivalent to approximately
10% of Pfister & Vogel's total tannery effluent loading. Likewise, total
chromium equivalent to approximately 13% of the total tannery effluent load-
ing is extracted from the blue stock during wringing.
Four packs of brine-cured hides from a single hide lot, which had been
fleshed and trimmed by the packer after curing, were processed to the blue
condition in a hide processor. After tanning, the sides were piled for 16
hours, weighed, wrung and then reweighed. Adjacent cuttings before and after
wringing were taken from the shoulder, kidney and belly of 12 sides (6 rights
and 6 lefts) selected at random. The relative position of the adjacent cut-
tings was alternated over the 12 sides. Two 2" x 1" pieces were taken from
each cutting, One set of nieces was immediately analyzed for moisture con-
tent while the second set was dried overnight at room temperature and subse-
quently ground and composited to yield six representative leather samples.
Analysis of the samples according to standard ALCA methods yielded the data
shown 1n Table 49.
The average fat content before wringing vias 1.37% while after wringing
fats decreased to 1.33% on an "as received" basis. Average chromic oxide
content increased however from T.32% ^03 before to 1.53% after wringing.
As Table 50 shows, the average weight of a-side of blue stock decreased
20.9% from 24.58 1b before to 19.44 Ib after wringing while the equivalent
weight of a cured side charged to the tanning process averaged 21.99 Ib.
The mass-balance analysis of blue stock wringing shown indicates that
wringing extracts approximately 3.55 pounds of fats and 0.813 Ib Cr (1.19 Ib
C^Og) per 1000 pounds of cured hide charged to the process. This is equiva
lent to 10.4% and 13.1% of our mean total tannery loadings of Oils and
Greases and total chromium respectively of 34.1 lb/1000 Ib and 6.20 Ib/
1000 Ib (based on actual raw-material mix).
156
-------
TABLE 49. EFFECT OF WRINGING ON BLUE STOCK ANALYSIS
%
%
-------
TABLE 50. CHROME & FATS DISCHARGED FROM BLUE STOCK U
Fats Cr?0-}
Blue, stock
before wrinninn/ Ib/sidc
after wringing, Ib/side
Discharged
Ib/side
lb/1000 Ib cured hide
0.33C7
0.2586
0.0781
3.55
0;3237
0.297C
0.0261
1.10
Cured weight of stock = 21.99 Ib/side
Blue stock weight before wringing = 24.58 Ib/side
Blue stock weight after wringing = 1D.44 Ib/side
158
-------
SECTION XVII
GRAVITY SEPARATION OF OIL & GREASE FROM BEAMHOUSE WASTES
Of the several wastewater pollutants which characterize tannery efflu-
ents, Oil and Grease is perhaps the most ubiquitous. It is generated (or
released) during practically every step of the "lower tannery" leather-
making processes, which contribute approximately 85% of the total tannery
Oil and Grease loadings. In fact, the Oil and Grease load in the waste
fractions from soaking, liming, reliming, bating and tanning each individ-
ually exceeds the promulgated EPA effluent standard (recently rescinded) for
the direct discharge of Oil and Grease. Soaking, for example, contributes
approximately 7.5 Ib of Oil and Grease per 1000 Ib of cured hide charged to
the process—ten times the EPA guideline for total tannery waste (0.75 Ib/
1000 Ib). Even pickling wastes, which alone contribute 0.44 Ib Oil X Grease/
1000 Ib hide, are discharged at a concentration more than 100 times the
allowed total effluent Oil and Grease concentration of 15 ng/1 (at the EPA
design flow of 6 gal/ Ib hide). Thus, to achieve substantial reductions in
total effluent Oil and Grease loadings every "lower tannery" wastewater
fraction requires treatment; and to meet the EPA guidelines, 97.8% of all
Oil and Grease must be removed (from 34 lb/1000 Ib to 0.75 lb/1000 Ib).
To this end, lab-scale experiments were conducted to determine the
treatability of segregated waste fractions by gravity separation or "simple
flotation." Samples of six different lower tannery process waste streams
were transferred to one liter graduated cylinders.
Each sample was thoroughly mixed and an aliquot was withdrawn to deter-
mine the initial Oil and Grease content of the raw waste. After one hour of
separation under quiescent conditions, another aliquot was withdrawn midway
between the supernatant and sludge layers and analyzed for Oil and Grease.
As Table 51 shows, the pickle and spent chrome samples initially con-
tained 2,467 mg/1 and 3,524 mg/1 of Oil and Grease respectively at corre-
sponding pils of 2.63 and 3.45. These Oil and Grease fractions which are pri-
marily insoluble fatty esters and fatty acids^float readily and concentrate
in the scum layers. The clarified interlayers analyzed at 272 mg/1 and 428
mg/1 respectively indicating a substantial redistribution of the Oil and
Grease component. Thus, under ideal conditions, reductions of 39.0% and
95.0# of the Oil and Grease in pickle liquors and spent chrome liquors by
"simple flotation" have been observed.
The soak liquor sample contained G,333 mg/1 of Oil and Grease at a pH
of 9.72. After settling, the interlayer contained 6,062 mg/1 indicating
that the Oil and Grease must be present in the, waste fraction as emulsions
159
-------
and/or as soluble soaps. Likewise, no substantial reduction was noted in
the slightly alkaline bate liquor sample which had an 8.3 pH.
A pronounced redistribution of Oil and Grease (probably as calcium
soaps) was observed in the hair-burn liquor. The combined scum and sludge
fractions, however, equalled 75% of the original volume, makino gravity
separation impracticable for the "removal" of Oil and Grease.
In the relime liquor sample also, a substantial volume of sludge was
generated while a reduction in Oil and Grease from 1,010 mg/1 to 134 mg/1
in the clarified interlayer was observed.
As these results suggest, the naturally-occurring Oil and Grease
(introduced by the hides) in several isolated "lower tannery" waste fractions
disproportionate between the sludge and scum layers and the clarified inter-
layers by gravity alone. Most notably, substantial reductions in Oil and
Grease are predicted for the "simple flotation" of spent pickle and spent
chrome liquors using surface skimming for scum removal. More advanced flo-
tation techniques such as dissolved air flotation, dispersed air flotation
and foam fractionation should be investigated as means of further improving
Oil and Grease separation from these tv/o waste fractions.
The results for lime and relime liquors suggest that substantial Oil
and Grease reductions are not practicable using simple flotation techniques,
but might only be made feasible in conjunction with the removal of major
quantities of suspended solids followed by sludge dewaterinn and final dis-
posal.
llhile bate liquors are not amenable to substantial reductions by simple
flotation, additional research needs to be carried out to determine whether
chemical additives such as polyelectrolytes and/or acid neutralization can
"free" the Oil and Grease for separation.
Similarly, acid addition to soak liquors reprecipitates stabilized or
dissolved Oil and Grease to induce separation. For that purpose, spent
pickle liquor was added to several soak liquor aliquots at pickle-to-soak
volume ratios varying from 1:1 to 2.25:1.
The pickle-soak mixtures were transferred to graduated cylinders and
settled under quiescent conditions. As Table 52 shows, the clarified inter-
layer after one hour analyzed at 256 mg/1, 414 mg/1 and 2,320 mg/1 for
pickle-to-soak ratios of 1:1, 1.35:1 and 2.25:1, respectively.
The maximum degree of redistribution of the Oil and Crease was thus ob-
served at the 1:1 ratio and at an initial pll of 7.9. The final pH was 8.5.
The addition of pickle liquor to the soak liquor followed by a short period
of gentle stirring produced a large, sturdy, flotable floe. After one hour
the floe particles developed into a scum layer equivalent to 390 ml/1, while
the volume of solids settling to the bottom was negligible. Thus, while a
94.3% reduction of Oil and Grease in the interlayer (from 4,515 mg/1 to
256 mg/1) was observed, the voluminous scum which generated would surely
require further dewatering. Dissolved air flotation of this material should
160
-------
TABLE 51.
Haste
fraction
Soak
Lime
Re lime
Bate
Pickle
Spent Chrome
Pickle/Soak
Ratio
1.0
1.35
2.25
EFFECT OF FLOTATION Of! THE OIL & GREASE CONTENT OF VARIOUS HASTE FRACTIONS
PU
9.72
12.23
12.40
3.30
2.63
3.45
TABLE 52.
PH
Before
flotation
7.9
7.1
6.0
Raw '..'aste
Settlab"le solids,
ml /I Oil 5 Crease, mcj/1 Oi
4 6,333
GOO 11,260
60 1,010
19 724
30 2,467
0.3 8,524
FLOTATION OF PICKLE/SOAK HASTEUATER MIXTURES
Oil & Grease, mq/1 Settlabl
After Before After Solids,
flotation flotation flotation ml/1
8.5 4,515 256 0.9
3.3 4,020 414 0.8
7.3 3,510 2,320
Clarified
inter! ayer
1 ?( Grease, nq/ 1
6,062
896
134
522
272
42G
e Flotable
solids,
ml/1
390
230
140
0
Vt 3
n>
1/1
n
(K
(0
Q.
%
3
i?
g
01
m
-a
-5
O
CT
Q>
o-
n>
ro
3
to
O
-h
T3
3
5.
-3
t/>
O
O
0
3
O
n>
Ol
1
-------
SECTION XVIII
RECOVERY OF LIME FROM SPENT LIME LIQUORS
A mass-balance analysis of Pfister ft Vogel's tannery operations indi-
cates that the sum of the waste water suspended solids loadings from the
individual process steps of soaking, lininn, bating, pickling and tanning
plus the loading from retanning, coloring, fatliquoring and subsequent
"upper tannery" operations equals only 60% of the suspended solids loadings
as determined from the analysis of 24-hour total effluent composite samples
(flow proportional) and total effluent flow data. These data demonstrate
that the combination of individual subprocess wastes in equalization/
neutralization basins or ponds will create a substantial Increase in the
suspended solids content. Host significantly, the presence of lime-ben ring
fractions will bring about the following solids-producing interactions: they
will react with spent chrome wastes to yield chromium hydroxide. The carbo-
nates from other processing steps combine with free calcium hydroxide to
yield sparingly soluble calcium carbonate. Likewise, the formation of
insoluble calcium soaps is expected. Finally, the divalent calcium ions act
as a coagulant to destabilize any colloidal solids present. Thus, the reduc-
tion of lime wastes fron tannery effluents by lime liquor recycle is expected
to reduce total effluent suspended solids loadings well beyond the levels
predicted by the mass balance contribution of the lime alone.
Preliminary investigations suggest that one possible recycle scheme is
the collection and concentration of spent lime liquors by flocculent settling
followed by refortification to produce a "stock" lime slurry for re-use. A
second alternative investigated was process modification to reduce relime
float volumes allowin^ direct reuse of a major fraction of the lime-
concentrate in subsequent unhairing or reliming runs.
Two matched sides leather tests in hide processors were conducted to
determine the more practical recycle "fit" for spent line liquors in our
leather-making process. In the first test, the relime float was drained
halfway through the relime, then followed by a sequential fresh water flood
and run. In the control runs, the relime float was retained such that after
the fresh water flood, the float volume was approximately twice the volume of
the test lime-concentrate drain. Analysis of the lime waste waters showed
that the first test drain contained alkalinity equivalent to 93 Ib Ca(0i!)2,
with the final drain after flooding containing an additional 85 Ib Ca(0!i)2.
The relime drain from the regular hide processor contained alkalinity equiva-
lent to 193 Ib of Ca(OH)2- Therefore, as the data shows, the intended con-
centration of lime by process modification was not forthcoming. In the test
hide processor run the initial drain contained only 52% of the combined
162
-------
alkalinity of the two test drains. Likewise, the initial test drain con-
tained only 48% of the lime in the control run drain, and at 18,000 nn/1
alkalinity (as €3(011)2) there was no significant difference in concentration
over the control.
In the first test, the finished leather showed the test sides to be con-
siderably smoother and better for fat wrinkles, finish break and hair cell
than the controls. However, they were substantially looser and somewhat
firmer than the control sides. There was no difference in thickness, in
woolincss of the flesh, or the veininess of the sides. Also, the physical
strength of the test sides as measured by Satra grain burst force, Satra
extension and two-hole stitch tear was equal to the control sides.
In a repeat of the test, the test sides remained slightly better for
smoothness, fat wrinkles, finish break and hair cell and this time were
equal to the controls for tightness, softness and veininess with an improve-
ment in the flesh noted.
In the second test, a substantial improvement in physical strength of
the test sides was observed. Satra grain burst force averaged 68 kg vs 61 kg
for the controls. The corresponding increase in extension was from 8.71 mm
to 8.90 mm. Likewise, stitch tear strength increased from 91 pounds to 100
pounds.
Since these results suggest that process float volume and drain modifi-
cations do not effect any concentration of contained alkalinity, alternative
measures including collection and thickening by sedimentation or hydrocy-
clonic separation were investigated.
In one probe experiment, 800 gallons of spent lime liquor containing
17,680 mg/1 of suspended solids and 17,800 mg/1 alkalinity as Ca(OH)2 were
collected from a single hide processor. This waste was pumped at a rate of
37.5 gpm to a pair of hydrocyclonic separators havinq orifice diameters of
3/16". The sludge which was generated at 4.0 gpm contained 52,750 mg/1 of
suspended solids and 40,000 mg/1 of lime.
Thus, approximately 25% of the contained alkalinity of the cyclone
supernatant and sludge was concentrated in the sludge fraction which equaled
10.6% of the original feed volume. The remaining 75% of the alkalinity,
however, was discharged in the supernatant which was equivalent to 89.4% of
the feed volume. Likewise, approximately 30% of the combined suspended
solids was separated in the sludge, and 70% remained in the supernatant.
Therefore, even within the constraints of the fixed orifice diameter and feed
pressure available for this experiment, a significant disproportionation of
alkalinity and solids between the cyclone supernatant and sludge has .been
effected. Further optimization of the operating conditions for hydrocyclonic
separation of lime from spent lime liquors should improve recovery efficien-
cies and, indeed, merits further investigation.
In another pilot-scale experiment 620 gallons of spent relime liquor
containing 18,100 mg/1 alkalinity as Ca(OH)2 was collected from a typical
production run. A 500 ml aliquot of the waste/was transferred to a graduated
163
-------
500
15 20
TIME, min
Figure 33: FLOCCULENT SETTLING OF LIME
IN RELIME LIQUOR.
164
-------
cylinder and treated with 100 ppm of an anionic polyelectrolyte (Rohm & Haas
Primafloc A-10). After one minute of rapid mixing followed by five minutes
slow nixinn the suspension was left to settle under quiescent conditions.
The settling curve-shown in Figure 33 demonstrates the rapid-settling
characteristics of the resultant flocculent mixture. After two hours an
aliquot of the supernatant was withdrawn and analyzed at 5,840 mg/1 Ca(OH)2
for a 67.7% reduction. A sludge volume of GO ml was recorded, and by mass-
balance calculations was estimated to contain 109,000 mcj/1 of Ca(0ll)2.
Subsequently, 100 ppm of Primafloc A-10 was added to the G20 gallon
batch, the mixture was agitated gently for 5 minutes, and then settled for
two hours. The first 50 gallons of thickened sludqe was drav/n off from the
bottom of the conical settling tank. This sludge analyzed at 101,000 mg/1
of Ca(OII)2 alkalinity equivalent to 42 pounds of lime. The supernatant con-
tained 5,900 mg/1 Ca(0!!)2 for a reduction of 67.4%. Hhile approximately
45% of the lime was actually recovered, mass-balance analysis indicates that
as much as 77% might be removed in continuous steady state operation.
These results demonstrate that a better than five-fold concentration of
alkalinity in spent relime liquors is possible, thus making recycle by forti-
fication with powdered hydrated lime and subsequent reuse in unhairing-liminn
production steps at least possible from a mass-balance standpoint. Of course
the most important consideration - the effect of such a recycle scheme on
leather quality remains to be determined.
165
-------
SECTION XIX
ACKNOWLEDGMENTS
This study was conducted by the Pfister & Vogel Tanning Company at
Milwaukee, Wisconsin under the direction of Dr. .lames !!. Constantin, Vice
President, Research and Development. George B. Stockman served as Project
Engineer.
Dr. J. David Eye, Professor of Environmental Health Engineering at the
University of Cincinnati served as consultant to Pfister ?< Vogel Tannina
Company during the project. Robert Fulton, Plant Engineer and Robert
Kusserow, Chief Engineer for Pfister & Vogel» assisted during the construc-
tion phase of the project.
This report was submitted in fulfillment of ^rant Mo. S-301037. The
support of II. L. Uanks, EPA Region VII, who served as Project Officer is
gratefully acknowledged.
166
-------
SECTION XX
REFERENCES
1. U.S. Environmental Protection Agency. Development Document for Proposed
Effluent Limitations Guidelines and Hew Source Performance Standards for
the Leather Tanning and Finishing Point Source Category. Washington, D.C.
EPA 44/1-73/015. November 1973. p.l.
2. llarnly, J.W. Journal of the American Leather Chemists Association,
46:169, 1951.
3. Benrud, ii.C. Journal of the American Leather Chemists Association,
46:171, 1951.
4. Miller, H. Report of the Symposium on Industrial Maste of the Tanning
Industry. Journal of the American Leather Chemists Association, Supple-
ment No. 15:25, 1970.
5. Davis, M.li. and J.r>. Scrongie. Investigation of Commercial Chrome-
Tanning Systems, Part III. Journal of the Society of Leather Technolo-
gists and Chemists, 57:53-58, 1973.
6. Davis, M.H. and J.fi. Scroggie. Investigation of Commercial Chrome-
Tanning Systems, Part II. Journal of the Society of Leather Technolo-
gists and Chemists, 57:35-33, 1973.
7. Petruschke, R. A Study of Chrome Recovery from Spent Chrome Tanninn
Liquors. M.S. Thesis, University of Cincinnati, 1959.
8. llauck, R.A. Report on Methods of Chromium Recovery and Reuse From Spent
Chrome Tan Liquor. Journal of the American Leather Chemists Association,
67:422, 1972.
9. Young, H.H. Effluent Treatment for a Small Tannery. Journal of the
American Leather Chemists Association, 68:308, 1973.
10. Ueber, W. Report of IULCS Effluent Commission. Journal of the Society
of Leather Technologists and Chemists, 57:63, 1973.
11. Das, B.M., J.K. De and S.N. Bose. Recovery of Chromium from Uaste
Chrome Liquors.' Bulletin of the Central Leather Research Institute,
.Madras, 1(5): 14, 1954.
167
-------
12. Abramovich, A. The Separation of Chromium from Tannery Effluents.
Izvest. Vysshikh Ucheb. Zavedenii, Technol. Legkoi Prom., No. 3:36041.
From C.A., Vol. 5C, No. 7068i.
13. Dianchi, A. French Patent No. 1,113,962, April 6, 1956. From C.A.,
Vol. 53, No. 9708g.
14. Hognes, T.R., !J.C. Johnson and A.A. Armstrong. Qualitative Analysis
and Chemical Equilibrium. Holt, Rinehard and Hilson, Inc., Chicago,
1966. p.565.
15. Sawyer, C.N. and P.L. McCarty. Chemistry for Sanitary Engineers.
McGraw-Hill Company, St. Louis, 1967. p.220.
16. Metcalf and Eddy, Inc. Hastewater Engineering: Collection, Treatment
Disposal. McGraw-Hill Book Company, New York, 1972. p.296.
17. Anon, h'aste '.'later Treatment with Air Flotation. Environmental Scieno
and Technology, 7(11):916, 1973.
18. Eckenfelder, VI.U. Jr. and D.J. O'Connor. Biological Haste Treatment.
Pergamon Press, IJew York, 1961. p. 182.
19. Bhattacharyya, D., J.A. Carlton and R.B. Grieves. Continuous Precipit
Flotation of Chromium (III) Hydroxide. American Institute of Chemical
Engineers Journal, 17:419, 1971.
20. Rich, L.G. Unit Operations of Sanitary Engineering. John Wiley and
Sons, Inc., Clemson, S.C., 1971. p.90.
21. ibid. p.167.
22. The Bauer Brothers Company. Bulletin G-422. Springfield, Ohio, 1969.
23. Qrthman, A.C. Tanning Processes. Hide and Leather Publishing Co., 194!
p.374.
24. ibid. p.376.
25. American Mutual Insurance Alliance. Handbook of Hazardous Materials.
Technical Guide No. 7. Chicago, IL, 1974. p.60.
26. U.S. Environmental Protection Aqency, Technology Transfer. Process
Design Manual For Sulfide Control in Sanitary Sewerage Systems. 1974,
p.2.2.
27. ibid. p.2.6.
28. ibid. p.2.3.
168
-------
29. Bailey, D.A. IULCS Tannery Wastes Commission, Minutes of the Fifth
Meeting of the Commission in Barcelona. Journal of the Society of
Leather Technologists and Chemists, 58(1):3, Jan.-Feb. 1974.
30. van Vlimmeren, P.J. Tannery Effluent. Journal of the American Leather
Chemists Association, 67:392-393, 1972.
31. A.C. Lawrence Leather Company. Activated Sludqe Treatment of Chrome
Tannery Hastes. Federal Water Pollution Control Administration, Depart-
ment of the Interior, Washington, D.C., Sept. 1969. p.135.
32. op. cit. Bailey, p.3.
33. ibid.
34. U.S. Environmental Protection Agency, Office of V'ater Program Operations,
Pretreatment of Pollutants Introduced into Publicly Owned Treatment
Works. Washington, D.C., October 1973. p.C-4.
35. Baily, D.A. Tannery Effluent Report to the Members of the Effluent
Commission of the International Union of Leather Chemists Societies.
Journal of the Society of Leather Trades Chemists, 50:52, 1972.
36. ibid. p.53.
37. op. cit. van Vlimmeren. pp. 397-399.
38. Eye, J.D. and D.P. Clement. Oxidation of Sulfides in Tannery Waste
Haters. Journal of the American Leather Chemists Association,
67:256-267, June 1972.
39. Chen, K.Y. and C.J. f.orris. Oxidation of Sulfide by 03: Catalysis and
Inhibition. Journal of the Sanitary Engineering Division Proceedings
of the American Society of Civil Engineers, 98(SA1):215-227, Feb. 1972.
40. Krebs, 11. A. Uber die 1,'irkung der Schwermetalle aus die Autoxydation der
Alkalisulfide Und des Schwefelwasser-stoffs. Biochim. Z., 204:343, 1929.
From: Chen, K.Y. and C.J. Horn's. Oxidation of Sulfide by 03:
Catalysis and Inhibition. Journal of the Sanitary Engineering Division
Proceedings of the American Society of Civil Engineers, 98(SA1):226,
Feb. 1972.
41. Bailey, D.A. IULCS Effluent Commission, Fourth Meeting of the Commis-
sion in Lyon. Journal of the Society of Leather Technologists and
Chemists, 57:64, 1973.
42. Black, A.P. and J.3. Goodson, Jr. The Oxidation of Sulfides by Chlorine
in Dilute Aqueous Solutions. Journal of the American Water '/orks
Association,-44:309-316, April 1952.
169
-------
43. Skoog, D.A. and D.H. West. Fundamentals of Analytical Chemistry, 2nd
Ed. Holt, Rinehart and Winston, Inc., Chicago, 1969. pp. 445-465.
44. Chen, K.Y. and J.C. Morris. Kinetics of Oxidation of Aqueous Sulfide
by Oo. Environmental Science and Technology, Easton, PA, 6:534.
June 1972,
45. ibid. p.530.
46. op. cit. Chen and Morris. Oxidation of Sulfide by 02. pp. 217-218.
47. Fair, G.M., J.C. Geyer and D.A. Okun. Mater and Wastewater Engineering,
Vol. 2. John Wiley and Sons, Inc., New York, N.Y., 1968. p.35.24.
48. op. cit. Eye and Clement, p.260.
49. ibid. p. 259.
50. op. cit. Sawyer and McCarty. pp. 383-393,
51. APHA, AWVIA and WPCF. Standard Methods for the Examination of Water and
!-!astewater, 13th Ed. American Public Health Association, Washington, D.C.
1971. pp. 474-488.
52. Aqua-Aerobic Systems, Inc. Engineering Manual. Roekford, II. p, 5.4.
53. Stewart, R.- Oxidation by Permanganate. Reprinted from: Oxidation in
Organic Chemistry, Part A, K.B. Wiberg, ed.'Academic Press, New York,
1965. p. 28.
54. op. cit. U.S.E.P.A. Process Design Manual for.Sulfide Control in
Sanitary Sewerage Systems, pp. 5.37-5.41.
55. Ibid. p. 5.31.
56. op. cit. Black and Goodson. p. 310.
57. op. cit. U.S.E.P.A. Process Design Manual for Sulfide Control in
Sanitary Sewerage Systems, pp. 5.31-5.32.
58. op. cit. Fair, fieyer and Okun, Vol. 2. p. 31.14.
59. Environmental Control Group. How Much Air? Rex Chainbelt Inc.,
Haukesha, Wisconsin. 7 pp.
60. Environmental Control Group. Air Lock Crown Diffuser. Rex Chainbelt
Inc. Bulletin AEP-398, 1970.
61. op, cit. Fair, Geyer and Okun. Vol. 2, p. 25.24.
62. Spencer Turbine Company. Spencer Turbo-Compressors. Bulletin 126J,
Hartford, p. 16.
170
-------
63. Envirotech Corporation, Eimco Processing Machinery Division. Eimco
NS-8. Salt Lake City.
64. op. cit. Aqua-Aerobic Systems, Inc. p. 1.10.
65. Happich, '.!.F., M.L. Happich, J.E. Cooper, S.ll. Feairheller, H.I!. Taylor,
0. Bailey, il.H. Jones, E.F. Mellon and J. .'laghski. Recovery of Proteins
from Lime-Sulfide Unhairinfl Effluents. Eastern Regional Research
Laboratory, Agriculture Research Service, U.S. Department of Agricul-
ture, Philadelphia, PA. Presented in part at the sixty-eighth Annual
Meeting of the American Leather Chemists Association, Pocono Manor, PA,
June 18-21, 1972.
66. op. cit. Fair, Geyer and Okun, Vol. 2. p. 25.26.
67. Fair, G.M., J.C. Geyer and D.A. Okun. 'later and Hastewater Engineerinq,
Vol. 1. John Wiley and Sons, Inc., New York, N.Y., 1966. p. 3.12.
68. op. cit. Fair, Geyer and Okun, Vol. 2, p. 25.29.
171
-------
TECHNICAL REPORT DATA
(/'lease rfnJ Insjsuctions on lite reverse before t\Mii>lclinx)
I. REPORT NO.
J. TITLE AND SUBTITLE
LEATHER TANNERY WASTE MANAGEMENT THROUGH PROCESS
CHANGE, REUSE AND PRETREATMENT
3. RECIPIENT'S ACCESSION-NO.
S. ntPQKT OAIF
.January 1977J.$.s.uiog .date
G. PI HI-ORMINl; ORGANISATION CODt
8. PfcRI-OHMING ORGANIZATION HH'OH) NO.
J. M. Constantin
JLJL. Stockman . __ ... ....... ___
9. PERI-ORMING ORGANI2A HOW NAMt AND ADDRESS
Pfister and Vogel Tanning Company
1531 North Water Street
Milwaukee, Wisconsin 53201
12. SPONSORING AGENCY NAME AND ADDRESS
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
10. PHOGHAM I I.LMENT NO.
1BB037
1 1. CONTRACT/GRANT NO.
S-801037
13. TYPE OF REPORT AND PERIOD COVERED
R&D April 197?-Marrh 1Q75
14. SPONSORING AGENCY CODE
EPA-ORD
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Reduction of tannery waste, i.e., trivalent chromium, sulfide and oil and grease
components has been accomplished by process change. Protein.recovery and hydro-
clonic separation of solids was shown to be possible in tannery processing in
reducing waste loading. All waste load reduction was accomplished without loss
of leather quality. Waste characterization through material balance was accomplished,
Chemical reactions and engineering design factors provide guidance for other plant
scale operation.
Kt'Y WORDS AND DOCUMENT ANALYSIS
DLSCNII'tOH!;
Leather tanning, Chromium recycle, Sulfide
reduction, Oil & Grease separation, Waste
characterization & loading
ll.lOf NTII:ll.»S/OPf.N KNDtO Tl: RMS C. COSATI Held/dump
Leather industry, Waste
reduction, Process modi-
fication, Engineering
design factors.
13B
8. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS flli/i Kepvrtj
UNCLASSIFIED
21. NO. OF PAGfS
J81
20. SECURITY CLASS (This
UNCLASSIFIFD
22. PRICE
EPA Form 2220-1 (9-73)
172
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Technical Information Staff
Cincinnati, Ohio 45268
OFFICIAL BUSINESS
PENALTY FOR PRIVATE USE. S3OO
AN EQUAL OPPORTUNITY EMPLOYER
POSTAGE AND FEES PAID
US ENVIRONMENTAL PROTECTION AGENCY
EPA-335
Special Fourth-Class Rate
Book
I
§ XX^T'V o
\
If your address is incorrect, please change on the above label;
tear off; and return to the above address.
If you do not desire to continue receiving this technical report
series, CHECK H£KE LU; tear off label, and return it to the
above address.
------- |