EPA-8 8 0/2-73-024
MARCH 1974
Environmental Protection Technology Series
Treatment and Recovery of
Fluoride Industrial Wastes
Office of Research and Development
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
RESEARCH REPORTING SERI2S
Research reports of the office of Research and
Monitoring, Environmental Protection Agency, have
been grouped into five series. These five broad
categories were established to facilitate further
development and application of environmental
technology. Elimination of traditional grouping
was consciously planned to foster technology
transfer and a maximum interface in related
fields. The five series are:
1, Environmental Health Effects Researen
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.
EPA REVIEW NOTICE
This report has been reviewed by the Office of Research and
Development, EPA, and approved for publication. Approval does
not signify that the contents necessarily reflect the views
and policies of the Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorse-
ment or recommendation for use.
-------
EPA-660/2- 73-024
March 1974
TREATMENT AND RECOVERY OF
FLUORIDE INDUSTRIAL WASTES
by
Christian J. Staebler, Jr.
Project S800680
Program Element 1BB036
Project Officer
John Ciancia
Edison Water Quality Research Laboratory
Edison, New Jersey 08817
Prepared for:
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D. C. 20460
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price $1.10
-------
ABSTRACT
This report presents the development and successful demonstration of laboratory
and pilot-scale fluoride treatment techniques for selected aerospace and metal-
working industry chemical processing solutions and rinse waters. It includes
laboratory-scale, lime treatment parameters for chemical processing solutions
such as temperature, retention time, pH, slurry concentration and fluoride influ-
ent and effluent levels, and ion-exchange treatment techniques to reduce the
fluoride concentration of rinse waters to levels less than three parts per million.
Pilot studies of centrifugal techniques to separate lime-precipitated sludges from
titanium chemical milling, titanium descaling and aluminum deoxidizing solutions
show that lime precipitation can give final effluents having fluoride concentrations
less than three parts per million. Aluminum conversion coating solutions, how-
ever, require secondary treatment with aluminum sulfate to give final effluents
having fluoride concentrations less than three parts per million.
Chemical and mechanical property tests show that it is potentially feasible to use
calcium fluoride sludge as a strength-maintaining additive for concrete. The re-
use of treated rinse waters, the economics of precipitation, and production plans
for chemical processing solutions and rinse waters are also presented.
This report was submitted in fulfillment of Project S800680 (12070 HGH) under the
(partial) sponsorship of the Office of Research and Development, Environmental
Protection Agency.
Key Words:
Fluoride Industrial Wastes
Lime Treatment
Ion Exchange
Calcium Fluoride Sludge
ii
-------
CONTENTS
Section Page
I CONCLUSIONS . . 1
H RECOMMENDATIONS 3
m INTRODUCTION 5
IV FLUORIDE PRECIPITATION AND CONCENTRATION
TECHNIQUES 6
Approach 6
Study Areas 6
Chemical Analysis of Solutions 6
Selection of Fluoride Precipitation Materials . . 9
Development of Laboratory Lime Precipitation
Techniques 9
Titanium Chemical Milling Solution 12
Titanium Descaling Solution 14
Ferrous Alloy Descaling Solution (5% H2SO4-
3 Oz/Gal NaNO0 - 3 Oz/Gal Actane 70) . . 18
o
Ferrous Alloy Descaling Solution (50% HNO,,-
5%HF) **. 18
Aluminum Deoxidizing Solution (2-6 Oz/Gal.
Amchem 7-17, 15%-20% by volume HNO,,
(42° B5)) 18
Aluminum Conversion Coating Solution
(Alodine 1200) 21
Selection and Evaluation of Ion-Exchange
Materials 21
Development of Isotherm Data Procedures. ... 25
Generation of Breakthrough Data 31
Amberlite IRA 400 Resin 31
Alcoa F-l Activated Alumina 35
Regeneration of Spent Amberlite IRA 400
Resin 38
Development of Two-Stage, Ion-Exchange
System for Titanium Chemical Milling Rinse
Water 38
iii
-------
CONTENTS (cont.)
Section Page
V PILOT CENTRIFUGE TREATMENT - TESTING AND
ANALYSIS 47
Approach 47
Study Areas . 47
Pilot Centrifuge Setup and Calibration 47
Treatment and Centrifugal Separation of Chemical
Process Solutions 50
Titanium Chemical Milling Solution 51
Titanium Descaling Solution 53
Aluminum Deoxidizing Solution. 54
Aluminum Conversion Coating Solution
(Alodine 1200) 54
Summary of Results 55
VI ECONOMICS OF CALCIUM FLUORIDE SLUDGE
RE-USE AND PRODUCTION SCALE-UP 58
Approach . 58
Study Areas 58
Evaluation of Calcium Fluoride Sludge as an
Additive to Concrete 59
Test Procedure. . 59
Test Results 63
Leachability of Sludge Concrete . . 63
Re-Use of Ion-Exchange Rinse Water Effluent . . 67
Lime Treatment Costs for Chemical Process
Solutions. 67
Production Scale-Up Plans for Treatment of
Process Solutions and Rinse Water . 68
Treatment of Process Solutions 68
Treatment of Rinse Waters ........ 71
Installation and Operation of Prototype
Production Lime Treatment System for Titanium
Chemical Milling Solution , . 71
Objective 71
f Treatment Procedure 71
Process Demonstration 73
Cost Studies for Prototype System 79
iv
-------
CONTENTS (cont.)
Section Page
VII BIBLIOGRAPHY 81
VIH APPENDIX 83
-------
FIGUBES
No. Page
1. Fluoride Electrode with Expanded-Scale pH Meter . . 10
2. Lime Treatment of 10% by Volume of 70%-Weight
Hydrofluoric Acid Titanium Chemical Milling
Solution 15
3. Lime Treatment of 40% Nitric Acid (42° Be) by
Volume/2% Hydrofluoric Acid (70% Weight) by Volume
Titanium Pickling and Descaling Solutions 17
4. Isotherm Plot for Amber lite IRA-400 Ion-Exchange
Resin 30
5. Isotherm Plot for Alcoa F-l Activated Alumina Ion-
Exchange Resin 33
6. Amberlite IRA-400 Ion-Exchange Resin Breakthrough
Curve . 36
7. Amberlite IRA-400 Ion-Exchange Resin Rinse Curve
for First-Run Spent Tower Following Regeneration. . 40
8. Amberlite IRA-400 Ion-Exchange Resin Regeneration
Curve for First-Run Spent Tower 42
9. Two-Stage, Ion-Exchange System for Titanium
Chemical Milling Rinse Water 43
10. Breakthrough Curve for Titanium Chemical Milling
Rinse Water 45
11. Sharpies/Fletcher Mark m Centrifuge 48
12. Pilot Centrifuge System 49
13. Sharpies Super-D-Canter Centrifuge 56
14. Calcium Fluoride Sludge from Lime Treatment of
Titanium Chemical Milling Solution 60
15. Mold for Casting Concrete-Calcium Fluoride Sludge
Flexural Strength Test Specimens and Cubical
Compressive Strength Test Specimens 62
16. Wiedemann-Baldwin 30B, 30,000-Pound Universal
Testing Machine 64
17. Concrete-Calcium Fluoride Sludge Compressive
Strength Specimen Being Tested in Wiedemann-
Baldwin 30B Universal Testing Machine 65
vi
-------
FIGURES (cont.)
No. Page
18. Test Rig for Leachability Test 66
19. Proposed Industrial Waste Treatment Facility for
Fluorides, Nitrates and Chromates 69
20. Proposed Ion-Exchange System for Fluoride
and Nitrate-Containing Rinse Waters . 72
21. Pilot-Line Lime Treatment System 74
22. Centrifugal Separation System 77
23. Ionic Activity Coefficient of Fluoride Ion as a
Function of Total Ionic Strength 85
vii
-------
TABLES
No.
1. Composition and Uses of Fluoride-Containing
Processing Solutions 7
2. Fluoride Content of Chemical Processing Solutions. . 8
3. Limes Used for Fluoride Precipitation 11
4. Laboratory Lime Treatment Results for Titanium
Chemical Milling Solution (10% By Volume
Hydrofluoric Acid) 13
5. Laboratory Lime Treatment Results for Titanium
Descaling Solution (40% Nitric Add plus 2%
Hydrofluoric Acid) 16
6. Laboratory Lime Treatment Results for Ferrous
Alloy Descaling Solution[Sulfuric Acid (2-10%) plus
Sodium Nitrate (1-5 Oz/Gal) plus Actane 70 (1-5 Oz/
Gal)] 19
7. Laboratory Lime Treatment Results for Ferrous
Alloy Descaling Solution (50% Nitric Acid plus 5%
Hydrofluoric Acid) 20
8. Laboratory Lime Treatment Results for Aluminum
Deoxidizer Solution (2-6 Oz/Gal Amchem 7-17
Deoxidizer plus 15-20% Nitric Add) 22
9. Laboratory Lime Treatment Results for Aluminum
Conversion Coating (Alodine 1200) 23
10. Isotherm Data for lonac P-6 Ion-Exchange Resin at
26°C 28
11. Isotherm Data for Amberlite IRA-400 Ion-Exchange
Resin at 26°C 29
12. Isotherm Data for Alcoa F-l Activated Alumina Ion-
Exchange Material at 26° C 32
13. Column Breakthrough Data for Amberlite IRA-400
Ion-Exchange Resin and Sodium Fluoride Solution
(119-124 Parts per Million of Fluoride Ion) at 26°C. . 34
14. column Breakthrough Data for Alcoa F-l Activated
Alumina and Sodium Fluoride Solution (130 Parts per
Million of Fluoride Ion) 37
15. Regeneration of Amberlite IRA-400 Ion-Exchange
Resin with 4% by Weight Sodium Hydroxide 39
viii
-------
TABLES (Cont.)
No. Page
16. Rinse Water Properties following Regeneration of
Amberlite IRA-400 Resin by 4% Sodium Hydroxide . . 41
17. Breakthrough Data for Amberlite IRA-400 Resin and
Alcoa F-l Activated Alumina Columns in Series for
Titanium Chemical Milling Rinse Water 44
18. Pilot Centrifuge Test Results for Chemical Process
Solutions 52
19. Flexural and Compressive Strength Test Results for
Regular and Sludge-Containing Concretes 61
20. Summary of Results of Centrifugal Separation Tests
with Treated Titanium Chemical Milling Solution . . 78
ix
-------
ACKNOWLEDGEMENTS
This grant with Grumman Aerospace Corporation, Bethpage, New York, was
conducted by the Advanced Materials and Processes Development Section,
Mr. Carl Micillo, Manager.
The program was directed by Mr. Christian J. Staebler, Jr., Advanced Ma-
terials and Processes Development Section. The principal laboratory inves-
tigator for the program was Mr. Richard Barbini, Chemical Engineering Sec-
tion of the Material and Processes Engineering Department. Mr. John Masek,
Advanced Materials and Processes Development Section, assisted in preparation
of the ion-exchange section. Mr. Guenter Baumann, Advanced Materials and
Processes Development Section, designed the prototype production installation.
The assistance provided by Mr. Donald Ingram and Mr. Kenneth R. Hancock of
the Minerals, Pigments and Metals Division of Pfizer Inc., Mr. William E. Bornak
of Rohm and Haas, Mr. William F. Torreyson, Jr., Alcoa, and Mr. James T.
Costigan, Sharpies Division, Pennwalt Corporation, is greatly appreciated.
The support of this project by the Office of Research and Development, Environ-
mental Protection Agency, and the help provided by Mr. John Ciancia, Chief,
Industrial Waste Technology Branch, Edison, New Jersey, the Grant Project
Officer, and Dr. Hugh Durham, Chief, Heavy Industrial Sources Branch, EPA
Grosse Isle Laboratory, Grosse Isle, Michigan, are acknowledged with sincere
thanks.
-------
SECTION I
CONCLUSIONS
Chemical granular quicklime (CaO) proved to be the most effective low-
cost material for fluoride treatment of acid-containing chemical pro-
cess solutions.
Titanium chemical milling solution (10% by volume hydrofluoric acid)
can be treated with chemical granular quicklime to give a final fluoride
effluent concentration less than three parts per million if a solids re-
tention time (without stirring) of at least 1.5 hours is used.
Titanium descaling solution (40% nitric acid plus 2% hydrofluoric acid)
can be treated with chemical granular quicklime to give a final fluoride
effluent concentration less than one part per million.
Aluminum deoxidizer solution (2-6 oz./gal. Amchem 7-17 deoxidizer
plus 15-20% nitric acid) can be treated with chemical granular quick-
lime to give a final fluoride effluent concentration less than one part
per million.
Aluminum conversion coating (Alodine 1200) can be treated to give a fin-
al fluoride effluent concentration less than three parts per million if pri-
mary fluoride treatment with chemical granular quicklime is followed by
secondary treatment with aluminum sulfate.
The low fluoride levels obtained on the above chemical process solutions
(less than the 7.9-ppm fluoride solubility for calcium fluoride reported
in the literature) are probably the result of excess calcium in the ef-
fluent, brought about by the large amount of lime required initially to
neutralize the acid before fluoride precipitation can take place.
Rohm and Haas Amberlite IRA-400 resin is an efficient ion-exchange
material for removal of fluorides from process solution rinse waters.
The fluoride concentration in such rinse waters was reduced from 60
parts per million to one part per million.
-------
Alcoa F-l activated alumina, when used in series with Amberlite IRA
400 resin, is an efficient material for maintaining pH control of rinse
waters as well as providing a polishing treatment for further reduction
of the fluoride content in the waste discharge.
Pilot studies on centrifugal separation and removal of precipitated cal-
cium fluoride sludge using a Sharpies Mark III centrifuge provided the
data necessary for specification of production equipment and showed
the viability of centrifugal separation.
Compression and flexural strength tests showed the feasibility of adding
calcium fluoride sludge to concrete to establish sludge re-use capability
while maintaining strength levels.
Continuous recirculation of water over concrete made with calcium
fluoride sludge showed that an insignificant amount of fluoride leached
out of the concrete. This should not present any potential fluoride
pollution problems as a result of runoff and infiltration of rainwater
into the ground.
An economic evaluation of the chemical costs for the process solutions
evaluated showed that fluoride removal would cost between 0.39 and
2.34 cents per gallon.
Based on rinse water ion-exchange effluent results for fluoride, pH and
conductivity, treated rinse water can be reused as rinse or process
solution make-up water, or disposed of in recharge basins.
Data generated in the studies to remove fluorides from process solutions
and rinse waters made it possible to prepare production scale-up plans.
Prototype lime treatment system studies showed that centrifugal separa-
tion can remove 95 to 97 percent of the solids generated in the treated 10
percent hydrofluoric acid, titanium chemical milling solution (initial
concentration of 177,500 ppm). Settling of the centrate provides a clear
effluent with 99.99 percent of the solids removed.
The estimated capital investment of a complete lime treatment system
which would include automated lime slaking equipment with a capacity
of 500 tons per year, a 20 to 30-gpm centrifuge and a 10,000-gallon
settling tank would be $60,000.
-------
SECTION n
RECOMMENDATIONS
This technology development program strengthened the conviction that further
studies are needed to provide the aircraft industry with a total approach to
pollution abatement. The following developments are planned under Phase II
of this program:
Although the data established the feasibility of using Rohm and Haas'
Amberlite IRA-400 resin in series with Alcoa F-l activated alumina
for fluoride removal from rinse water, the following additional in-
formation must be generated before a production-size installation can
be set up:
- Scale up system to pilot size (one cubic foot)
- Conduct more comprehensive treatment studies on rinse waters
generated (including re-use) at metal finishing aircraft facilities
using a strong base IRA-400 resin column followed by an activated
alumina column. The data obtained would include the fluoride and
nitrate removal capacity of the resin, fluoride removal capacity of
the activated alumina, effect of flow rate on the performance of the
activated alumina, and the fluoride and nitrate concentration in the
raw waste, effluent from the resin bed, and effluent from the acti-
vated alumina column.
- Determine if the strong base IRA-400 resin starts to hydrolyze upon
repeated cycling, which will be reflected in progressively lower
capacity of the resin.
- Determine if small quantities of heavy-metal cations have any effect
on resin upon repeated cycling and establish possible need for
cation exchange tower in a production installation.
Although it was established that chemical process solutions containing
fluoride concentrations as high as 100,000 ppm could be reduced to less
than three ppm using chemical granular quicklime, a problem still exists,
-------
however, in that some of these solutions also have nitrate concentrations
higher than 100,000 ppm. Present New York State standards require
that metal processing solutions contain less than 10 ppm of nitrate before
being discharged to ground waters. The centrifugal lime precipitation
technique will be used in conjunction with a mechanically aided evapor-
ator to provide a comprehensive waste treatment facility capable of
removing fluorides, nitrates and chromates. Parameters such as feed
rate and steam pressure, and the effect of solution analysis on effluent
purity will be established. Once pilot evaporator parameters are
established, the economics of nitrate sludge reuse, water recycling and
production scale-up will be determined.
The following information must be established before concrete containing
lime treatment sludge can be used for road construction:
- Maximum amount of sludge that can be added to maintain or increase
concrete strength.
- Short and long-term aging strength through structural testing
- Long-term leachability, slump, air content, and degree of
strength degradation due to environmental exposure.
-------
SECTION HI
INTRODUCTION
A steadily increasing number of applications for titanium, super-alloys, and
refractory metals in military and commercial products is making it neces-
sary to expand the use of fluoride-containing solutions for cleaning, scale re-
moval and other chemical operations. Mounting public concern about environ-
mental pollution and more stringent government regulations have put increased
pressure on metalworking industries and chemical firms that use fluorides, and
on those chemical firms that manufacture fluorides to use non-polluting fluoride
waste disposal procedures.
The increasing volume of fluoride-containing processing solutions and rinse
waters being generated, however, is making existing methods of removing these
solutions prohibitively expensive. Most aerospace firms use vendors to dispose
of waste solutions. Since vendors may sometimes use questionable treatment
and disposal methods, the firms which supplied the waste solutions can be held
responsible for vendor-caused pollution damage even though contracts with
"hold-harmless clauses" are if effect. Also, since vendors may not normally
regenerate fluoride solutions, the fluoride pollution problem may merely be
transferred from one location to another. Some waste disposal vendors re-
quire that fluoride solutions be completely segregated. This creates storage
and handling problems when both large and small processing tanks are being
used and when the vendor has tank trucks of only one size. Measures must
also be taken to insure the safety of personnel involved in pumping and han-
dling the corrosive and toxic fluoride solutions from user storage facilities to
vendor tank trucks.
This project was initiated, therefore, to develop a comprehensive fluoride
waste treatment technology for chemical processing solutions and rinse waters
used in the aerospace and metalworking industries. The fluoride treatment
approach was based on the use of lime precipitation for process solutions and
ion-exchange for rinse waters.
-------
SECTION IV
FLUORIDE PRECIPITATION AND CONCENTRATION TECHNIQUES
APPROACH
This phase of the program was directed toward the removal and recovery of
fluorides from spent, metal treatment acid solutions and rinse waters in a pri-
mary effort to prevent pollution of ground water by discharge of these untreated,
fluoride-containing wastes. The treatment approach was based on lime precip-
itation for process solutions and ion-exchange adsorption techniques for rinse
waters.
STUDY AREAS
The task of developing precipitation and ion-exchange adsorption techniques for
removal of fluorides consisted of the following technical efforts:
Chemical analysis of solutions
Selection of fluoride precipitation materials
Development of laboratory lime precipitation techniques
Selection and evaluation of ion-exchange materials
Development of isotherm data procedures
Generation of breakthrough data
Regeneration of Amberlite IRA-400 resin
Development of a two-stage, ion-exchange system for titanium
chemical milling rinse water
CHEMICAL ANALYSIS OF SOLUTIONS
Six fluoride-containing process solutions were chosen for study under this pro-
gram and analyzed for total fluoride content using the specific ion electrode
technique. The typical formulations, uses, and manufacturers for the process
solutions studied are shown in Table 1. The total fluoride content (in parts per
million) of the solutions is shown in Table 2.
-------
TABLE 1
COMPOSITION AND USES OF FLUORIDE-CONTAINING PROCESSING SOLUTIONS
Fluoride-Containing Pi
Solutions-Typical Formulations
Typical Uses
Manufacturer
Aluminum Deoxidizer Solution
2-6 oz./gal AMCHEM
7-17 deoxidizer (contains potassium
dichromate and fluoride salts)
10-20% volume nitric acid 42° Baume
Cleaning of aluminum parts prior to:
Conversion coating (Alodine)
Masking of parts for chemical milling
Spot welding
AMCHEM Products Inc., Ambler, Pa.
Titanium Descaling Solution
35-45% volume nitric acid 42° Baume
Actane #70 1.5-3% by weight
or
Hydrofluoric acid 1.5-3% by volume.
using 70% weight acid.
Cleaning of titanium parts for:
Welding
Removal of heat-treat scale
Painting
Chemical grade acids (sources open)
Actane #70 - Enthone Inc., West
Haven, Connecticut
Ferrous Alloy Descaling Solution
2-10% volume sulfuric acid.
66°Baume.
Sodium nitrate 1-5 oz./gallon.
Hydrofluoric acid 1-5% volume.
or
Actane #70 1-5% by weight
ron (Fe) 4% weight maximum.
Cleaning of ferrous alloys for:
Heat-treat scale removal
Plating
Acids and sodium nitrate (sources open
Actane #70-Enthone Inc., West
Haven, Connecticut
Ferrous Alloy Descaling Solution
50% volume nitric acid 42° Baume
Hydrofluoric acid 5% by volume,
using 70% weight acid
Cleaning of Ferrous alloys for:
Heat-treat scale removal
Plating
Chemical grade acids (sources open)
'itanium Chemical Milling Solution
Hydrofluoric acid 10% by volume
sing 70% acid.
Chemical milling and blanking of
Itanium parts to effect:
> Weight reduction.
> Integral stiffening.
» Blanked-out parts.
Acid - 70% commercial grade
(source open)
Aluminum Conversion Coating Solution
Alodine). Alodine contains chromic
acid and complex fluoride salts.
Used to produce a protective coating
on aluminum to:
Increase corrosion resistance
Increase paint adhesion
AMCHEM Products, Inc., Ambler, Pa.
-------
TABLE 2
FLUORIDE CONTENT OF CHEMICAL PROCESSING SOLUTIONS
Proem Solution
Total Fluoride Content, parts par million
Aluminum Deoxidizing
Titanium Descaling, Solution
Sulfuric Ferrous Alloy Descaling
Nitric Ferrous Alloy Descaling
Titanium Chemical Milling
Aluminum Conversion Coating
2,250
60,000
16,000
39,600
100.000
1,750
-------
The fluoride electrode (Figure 1), which is a selective ion electro-chemical
sensor designed for use with an expanded-scale pH meter, was selected as an
analytical tool because of its simplicity and reliability. Its operation is based
on a single-crystal sensing element. The sample solution contacts one face
of the crystal and an internal reference solution contacts the other. A standard
Calomel electrode is used as a reference electrode. The fluoride analysis
procedure used for all samples is shown in the Appendix.
SELECTION OF FLUORIDE PRECIPITATION MATERIALS
Before starting this program, an extensive literature search was conducted to
establish the optimum and most cost-effective fluoride-removal materials.
Based on this search, fluoride precipitation with lime to yield the relatively
insoluble calcium fluoride was selected as the most feasible and economical
approach for primary fluoride removal from processing solutions. The limes
shown in Table 3 were selected for the initial laboratory studies.
DEVELOPMENT OF LABORATORY LIME PRECIPITATION TECHNIQUES
The following parameters for lime precipitation of fluorides were established
for titanium descaling, titanium chemical milling, nitric ferrous descaling,
sulfuric ferrous descaling, aluminum deoxidizing and aluminum conversion
coating solutions:
Lime slurry concentration requirement
Slaking time (where required)
Starting and maximum reaction temperature
Process solution-lime slurry ratio
Solids retention time
Treated effluent pH
Treated effluent fluoride concentration
-------
Figure 1. Fluoride Electrode with Expanded-Scale pH Meter
-------
TABLE 3
LIMES USED FOR FLUORIDE PRECIPITATION
Commercial Name
Chemical Granular Quicklime
High-Calcium Hydrated Lime
Dolomitic Granular Quicklime
Quick Plastic Dolomitic
Hydrated Lime
High-Calcium Limestone
Dolomitic Limestone
Chemical
Composition
CaO
Ca(OH)
3
CaO-MgO
Ca (OH),-
Mg (OH),
CaCO,
CaCO,-
MgCO,
Cost Per Ton
(dollars)
$21.00
$20.00
$21.50
$28.00
$ 7.00
$ 7.00
Bulk Density
(lb./cu. ft)
68-72
20-30
70-75
30-35
169
80-90
11
-------
High-calcium hydrated lime [Ca(OH)2J was used as a standard to establish
the weight of lime required to achieve maximum fluoride removal. The alkali
equivalents of the other five limes tested were computed so that the same alka-
linity level could be used for subsequent tests. The initial fluoride precipitation
tests were conducted on. titanium chemical mining solution (10% hydrofluoric
add) and titanium descaling solution (40% nitric acid plus 2% hydrofluoric acid).
These two solutions were picked for initial tests, since the titanium chemical
milling solution is a weak add and the descaling solution is a strong acid. The
fluoride predpitation results would be indicative of expected results on the
other process solutions to be tested. Specific fluoride reduction results achiev-
ed for the titanium chemical milling solution and the titanium descaling solution
using the six limes tested were as follows.
Titanium Chemical Milling Solution
lime predpitation data for this test and all subsequent tests were established
by treating one-liter volumes of process solution with different lime slurry
concentrations. Process solutions and water used to prepare lime slurries
were at ambient temperature prior to slaking (where required) and treatment.
Lime slurries were mixed by adding the quantity of lime to a measured quantity
of water. A 15-ml sample for fluoride analysis was removed from the treated
process solution at each indicated retention time. This sample was centrifuged
for three minutes in an International Clinical centrifuge and then analyzed.
For the six limes tested, optimum fluoride reduction was achieved with titanium
chemical milling solution (10% by volume of hydrofluoric add) using either chem-
ical granular quicklime or high-calcium hydrated lime (Table 4). The initial
fluoride concentration was reduced from 100,000 to 6.45 parts per million.
The greatest fluoride reduction occurred at a pH of about 12. ft was found that
a minimum solids retention time of 60 minutes was necessary with either chem-
ical granular quicklime or high-calcium hydrated lime to achieve maximum
fluoride reduction. The heat of neutralization caused a temperature increase of
46°F-66 F, except where high-calcium limestone or dolomitic limestone was
used. The limestones produced temperature increases of only 6°F-14°F. The
temperature increase due to neutralization occurred within five minutes of lime
12
-------
TABLE 4
LABORATORY LIME TREATMENT RESULTS FOR
TITANIUM CHEMICAL MILLING SOLUTION
(10% BY VOLUME HYDROFLUORIC ACID)
Initial
Fluoride
Concentration
(ppm)
99.500
99,500
99,500
99,500
99,500
99,500
Lime
Used For
Treatment
Chemical
Granular
Quicklime
High-
Calcium
Hydrated
Lime
Dolomitic
Granular
Quicklime
Dolomitic
Hydrated
Lime
High-Calcium
Limestone
Dolomitic
Limestone
Slurry
Concentration
(grams/liter)
178
232
153
210
268
276
Volume Acid:
Volume Slurry
1:1
1:1
1:1
1:1
1:1
1:1
Temp.
Increase
During
Neutraliza-
tion (°F)
46
66
46
52
6
14
Solids
Retention
Time
(minutes)
5
45
60
1140
15
30
60
120
240
10
30
60
180
10
20
40
60
120
180
255
210
Effluent Temp.
Prior To
Solids
Separation
C°F)
154
125
105
70
110
98
86
80
78
125
117
100
80
115
104
96
90
86
84
-
.
Effluent
PH
12.2
12.2
12.4
11.7
12.2
12.2
10.5
12.0
12.0
2.9
4.6
7.3
9.7
6.0
7.0
8.1
8.1
8.1
8.0
2.5
1.2
Effluent
Fluoride
Concentration
(ppm)
6.75
6.45
6.22
9.35
9.22
6.74
153
10.4
4,900
740
122
66
202
154
141
133
108
99
18,600
30,800
-------
slurry-acid contact. The use of dolomitic limes and limestones was ruled out
because the lowest effluent fluoride concentration that could be obtained was
65 parts per million. This was probably due to the fact that the highest pH
that could be reached with the dolomitic limes was 3.65 while unreacted lime
was still in the beaker. The optimum reductions that were achieved in fluoride
content of titanium chemical milling solution with the six limes tested are shown i
in Figure 2.
Titanium Descaling Solution
For the six limes tested, optimum fluoride reduction was achieved on titanium
descaling solution (40% by volume HNO3(42°Be)/2% by volume HF of 70% weight)
using either chemical granular quicklime or high-calcium hydrated lime (Table
5). Chemical granular quicklime and high-calcium hydrated lime at slurry con-
centrations of 237 and 345 grams per liter reduced total fluoride concentration
from 60,000 ppm to 0.615 and 0.297 ppm, respectively. The waste acid-to-
slurry volume was 1:1. Maximum fluoride reduction occurred between a pH of
8 and a pH of 10. The low fluoride levels obtained (less than the 7.9 ppm fluo-
ride solubility for calcium fluoride) are probably caused by excess calcium in
the effluent resulting from the large amount of lime required to neutralize the
nitric acid. Low fluoride levels were also obtained on samples treated with
dolomitic granular quicklime (1.36 ppm F~) and high-calcium limestone (4.79
ppm F~). The fluoride concentration of samples treated with dolomitic lime,
however, rose to 22.6 ppm when solids retention times greater than 10 min-
utes were used. This was attributed to redissolving of magnesium fluoride.
High-calcium limestone may produce an effluent with a fluoride concentra-
tion lower than three parts per million, if it is added dry. Liberation of car-
bon dioxide during neutralization, however, would require control equipment.
The optimum reductions in fluoride content of titanium descaling solutions ob-
tained with the six limes tested are shown in Figure 3.
Since optimum fluoride reductions with both the titanium chemical milling solu-
tion (10% hydrofluoric add) and titanium descaling solution (40% nitric acid
plus 2% hydrofluoric add) were obtained using either chemical granular quick-
lime or high-caldum hydrated lime, these limes were chosen for all subse-
quent testing* The following results were obtained for ferrous descaling,
14
-------
100,000
10.000
1,000-
5
a.
a.
Z
o
p
^
tc
1-
z
u
z
o
u
Ul
o
£ 100
§
u.
. EFFLUENT
z
u.
10-
1
Ul
LARQUICKLII
Z
cc
o
o
5*
UJ
5
6.74
Ul
DRATED LIM
I
D
U
<
^7
^
o
i
100.000
18,600
99
65.6
ICKLIME
3
a
0
i-
^
0
o
a
Ul
S
HYDRATED L
u
l-
i
o
J
g
Ul
O
CO
Ul
5
-i
5
O
^
9
i
CJ
X
30,800
UJ
Z
o
i-
OT
UJ
5
U
i
3
i
-
0
K
a:
z
UJ
U
2
0
u
o
z
p
X
w
Figure 2. Lime Treatment of 10% by Volume of 70% Weight Hydrofluoric Acid
Titanium Chemical Milling Solution
15
-------
TABLE 5
LABORATORY LIME TREATMENT RESULTS FOR
TITANIUM DESCALING SOLUTION
(40% NITRIC ACID + 2% HYDROFLUORIC ACID)
Initial
Fluoride
Concentration
(ppm)
60.006
60,000
60,000
60,000
60,000
60,000
Lime
Used For
Treatment
Chemical
Granular
Quicklime
High
Calcium
Hydrated
Lime
Dolomitic
Granular
Quicklime
Dolomitic
Hydrated
Lima
High-Calcium .
Limestone
Dolomitic
Limestone
Slurry
Concentration
(grams/liter)
25?
345
225
312
425
439
Volume Acid:
Volume Slurry
m
1:1
1:1
1:1
1:1
1:1
Temp.
Increase
During
Neutraliza-
tion (°F)
8?
88
104
94
26
22
Solids
Retention
Time
(minutes)
n>
30
105
5
10
30
60
180
240
10
30
60
180
240
10
30
60
135
255
60
135
240
80
270
Effluent Temp.
Prior To
Solids
Separation
<°F)
T5B
130
92
150
136
112
90
78
78
132
115
100
80
80
132
106
86
76
74
72
72
72
80
78
Effluent
pH
4_g
7.9
10.2
7.6
7.8
7.9
7.8
8.0
7.8
5.9
6.2
6.8
7.2
6.4
6.8
7.2
7.5
7.2
7.7
4.5
4.6
5.0
2.4
3.0
Effluent
Fluoride
~ -.*
oncentration
(ppm)
Ao
1.04
0.62
3.56
0.65
0.30
1.44
0.71
0.40
1.4
5.9
13.8
12.4
22.6
33.7
18.9
66.0
24.4
34.4
5.55
5.23
4.79
831
330
-------
106-|
104-
103-
5
fc
Z
o
1
UJ -
0 I02.
O
u
Ul
Q
E
S
J
Ik
tu
"^ 10 "
u_ »»
u.
UJ
u
z
10 -
.,
.297
UJ
'DRATED L
f
D
U
<
^
0
1
.615
UJ
5
o
o
a.
.LGRANUL
4
U
Ul
5
1.36
UJ
3
u
ITICQU
DOLOMI
4.79
Ul
O
te
Ul
S
J
D
U
^
V
2
E
ROjOno
330
18.9
UJ
S
O
: HYDRATE
%_/
f
O
Q
Ul
O
tfl
UJ
S
*
0
H
g
3
g
Z
O
DC
1-
UJ
o
z
o
o
o
z
i
Figure 3. Lime Treatment of 40% Nitric Acid (42° Be) by Volume/2%
Hydrofluoric Acid (70% Weight) by Volume Titanium
Pickling and Descaling Solution
17
-------
aluminum deoxidizing and aluminum conversion coating solutions tested with
chemical granular quicklime and high-calcium hydrated lime.
Ferrous Alloy Descaling Solution (5% H2SO4 - 3 Oz/Gal. NaNOg - 3 Oz/Gal.
Actane 70)
The lowest fluoride level that could be achieved with this solution, which had
an initial fluoride concentration of 16,000' parts per million, using either chem-
ical granular quicklime or high-calcium hydrated lime, was 29 parts per mil-
lion (Table 6). The Actane 70 additive to the descaling solution contains am-
monium salts (NH4+) and about 76 percent by weight of fluoride. Ammonia
gas, which was liberated at a pH above 7.0, prevented the pH from going much
higher that 8.0. Since the calcium flouride formed is soluble in ammonium
salt solutions, an effluent fluoride content below 29 parts per million was not
produced. As shown in Table 6, more ammonia was driven off at higher re-
tention times while the pH increased and the fluoride level decreased. Even
at a retention time of five hours for chemical granular quicklime, however, an
effluent containing only 29.0 parts per million of fluoride could be obtained.
Ferrous Alloy Descaling Solution (50% HNOg - 5% HF)
This ferrous alloy descaling solution is very similar to the previously tested
titanium descaling solution. A reduction in fluoride concentration from 39,600
parts per million to less than one part per million was obtained using either
chemical granular quicklime or high-calcium hydrated lime (Table 7). A re-
tention time of at least 30 minutes was required to achieve maximum fluoride
reduction. The ratio of the volume of acid treated to the volume of lime slurry
used was changed from 1:10 to 1:1.6 in order to add the extra lime required to
neutralize the 50 percent nitric acid. The slurry concentration could not be
raised above 30 percent by weight, since it was difficult to keep it in suspen-
sion prior to acid treatment.
Aluminum Deoxidizing Solution (2-6 Oz/Gal. Amchem 7-17, 15%-20% by Volume
HN03 (42° Be))
Since aluminum deoxidizing solution contains 5500 parts per million of hexava-
lent chromium, the solution was treated with sodium metabisulphite to reduce
the chromium to the trivalent form prior to fluoride treatment. Treatment with
18
-------
TABLE 6
LABORATORY LIME TREATMENT RESULTS FOR
FERROUS ALLOY DESCALING SOLUTION [SULFURIC ACID (2-10%) + SODIUM
NITRATE (1-5 OZ/GAL) + ACCANE 70 (1-5 OZ/GAL) ]
Initial
Fluoride
Concentration
(ppm)
16,000
16,000
Lime
Used For
Treatment
Chemical
Granular
Quicklime
High -Calcium
Hydrated Lime
Slurry
Concentration
(gram/liter)
202
267
Volume Acid:
Volume Slurry
2:1
2:1
Temp.
Increase
During
Neutraliza-
tion <°F)
24
66
Solids
Retention
Time
(minutes)
5
15
30
60
150
180
240
300
5
15
30
105
180
240
Effluent Temp.
Prior To
Solids
Separation
<°F)
126
115
110
96
80
80
78
78
-
-
100
82
78
74
Effluent
pH
2.3
2.4
2.9
3.7
8.4
8.3
8.7
8.7
5.8
7.4
7.7
7.9
8.2
8.2
Effluent
Fluoride
Concentration
(ppm)
2,990
2,580
1,750
540
48
41
40
29
64.5
40.0
41.5
46.8
38.9
36.2
-------
TABLE 7
LABORATORY LIME TREATMENT RESULTS FOR
FERROUS ALLOY DESCALING SOLUTION (50% NITRIC ACID + 5% HYDROFLUORIC ACID)
Initial
Fluoride
-------
20 grams of dry Na^S^C^ per liter of waste produced a solution containing
less than 0.1 part per million of hexavalent chromium. Chemical granular
quicklime gave optimum results on aluminum deoxidizing solution (Table 8).
The initial fluoride concentration was reduced from 2250 parts per million
to less than one part per million using a 176 gram-per-liter slurry and a
waste-to-slurry ratio of 5:4. Although the solids retention time required was
three hours, the fluoride concentration could be reduced to less than three
parts per million in 15 minutes. The lowest fluoride concentration that could
be achieved with high-calcium hydrated lime at a slurry concentration of 232
grams per liter and a waste-to-slurry ratio of 5:4 was 3.45 parts per million.
The solids retention time was 30 minutes.
Aluminum Conversion Coating Solution (Alodine 1200)
The aluminum conversion coating solution, which contains 1000 parts per mil-
lion of hexavalent chromium, was treated with sodium metabisulphite to reduce
the chromium to the trivalent form prior to fluoride treatment. Treatment
with 4.16 grains per liter of dry sodium metabisulphite reduced the hexavalent
chromium concentration to less than 0.1 part per million. The chromate-re-
duced Alodine 1200 solution was treated with 25.6 grams of dry chemical gran-
ular quicklime per liter of waste. Fluoride concentration was reduced from
1750 to 19.4 parts per million with a 25-minute retention time (Table 9).
The lowest fluoride level that could be obtained using high-calcium hydrated
lime at a concentration of 17 grams (dry) per liter of waste was 23.2 parts per
million for a retention time of one hour. Since Alodine 1200 contains ammo-
nium bifluoride, ammonia gas was liberated during lime treatment until equi-
librium was established. The solubility of precipitated calcium fluoride in the
remaining ammonium salt prevented the fluoride concentration from being re-
duced below 19.4 parts per million. A secondary treatment will be required on
this solution to produce an effluent containing less than three parts per million
of fluoride.
SELECTION AND EVALUATION OF ION-EXCHANGE MATERIALS
Manufacturers of ion-exchange materials, sales representatives and literature
were consulted to determine those ion-exchange materials which could best be
utilized to remove fluorides from acid rinse waters. Based on this search, the
following three materials were selected for study:
21
-------
TABLE 8
LABORATORY LIME TREATMENT RESULTS FOR
ALUMINUM DEOXIDIZER SOLUTION (2-6 OZ/GAL. AMCHEM 7-17
DEOX3DIZER + 15-20% NITRIC ACID)
Initial
Fluoride
Concentration
(ppm)
2,250
2,250
Lime
Used For
Treatment
Chemical
Granular
Quicklime
High-Calcium
Hydrated
Lime
Slurry
Concentration
(gram/liter)
176
232
Volume Add:
Volume Slurry
6:4
5:4
Temp.
" Increase
During
Neutraliza-
tion (°F)
30
40
Solids
Retention
Time
(minutes)
5
15
30
60
180
260
15
30
60
120
255
Effluent Temp.
Prior To
Solids
Separation
(°F)
118
94
84
77
76
76
108
88
82
76
76
Effluent
PH
11.6
11.6
11.4
11.5
11.5
11.5
10.6
10.9
11.0
10.9
10.0
Effluent
Fluoride
Concentration
(ppm)
4.94
2.71
2.70
1.53
0.76'
1.23
5.95
3.45
4.09
3.47
3.07
to
to
-------
TABLE 9
LABORATORY LIME TREATMENT RESULTS FOR
ALUMINUM CONVERSION COATING (ALODINE 1200)
to
CO
Initial
Fluoride
Concentration
(ppm)
1,750
1,750
Lime
Used For
Treatment
Chemical
Granular
Quicklime
High-
Calcium
Hydrated
Lime
Dry Lime
Concentration
(grams/liter)
25
17
Temp.
Increase
During
Neutralize-
, « fO f*\
tion ( F)
9
4
Solids
Retention
Time
(minutes)
25
40
55
80
140
240
300
5
15
30
60
195
240
Effluent Temp.
Prior To
Solids
Separation
( F)
80
80
78
78
76
76
76
77
77
77
77
77
77
Effluent
PH
10.3
10.9
11.1
11.7
12.3
12.2
12.3
8.8
9.0
8.8
8.7
8.6
8.6
Effluent
Fluoride
Concentration
(ppm)
19.4
41.6
42.8
43.0
21.4
20.8
17.4
62.5
38.0
31.4
23.2
47.2
45.4
-------
lonac P-60 Resin, lonac Chemical Corporation
- Chemical Properties - A proprietary formulation which will remove
fluoride ions from water if specific conditions are met.
- Physical Properties
Bulk Density: 50-55 lbs/ft3
Mesh: -16 + 50
Moisture content as shipped: About 30%
- Cost - $20.00 per cubic foot
Amberlite IRA-400 Resin, Rohm and Haas
- Chemical Properties - A strongly basic ion-exchange material
consisting of an 8% cross-linked polystyrene matrix of the
trimethylbenzylammonium type
- Physical Properties
Bulk Density: 44 lb/ft3
Mesh: 20-50
Void Volume: 40-45%
Moisture Content (Drained): 55% by weight
- Cost - $61.40 per cubic foot
F-l Activated Alumina, Aluminum Corporation of America
- Chemical Properties - A porous form of aluminum oxide having a
high surface area
- Physical Properties
Bulk Density: 55 lb/ft3
Size: 1/4"to 8 mesh
Moisture Content (Drained): 44% by weight
- Cost - $9.30 per cubic foot
24
-------
These materials were evaluated according to the criteria set forth in the fourth
edition of Perry's Chemical Engineering Handbook by obtaining isotherm, break-
through, regeneration and rinse data.
DEVELOPMENT OF ISOTHERM DATA PROCEDURES
The fluoride rinse solutions under study can be considered as one of the sim-
plest forms of ion-exchange systems - namely a solvent or carrier and the
solute. Simple plots could readily be drawn of the solute concentrations in the
solid phase as a function of its concentration in the fluid phase. Since this
functional relationship holds true at one particular temperature, these plots
are known as isotherms.
Separation performance depends upon the curvature of a particular isotherm;
usually a convex-upward curve would signify favorable exchange and a concave-
upward curve would signify unfavorable conditions. However, these curves
should not be used solely as a basis for discarding a resin system; a concave-
upward curve does not mean a system won't work, but that more resin than
expected will have to be used to achieve the desired results.
The isotherm data was obtained by placing a known weight of resin in a 250-ml
plastic jar and then adding a 200-ml aliquot of a known concentration of sodium
fluoride solution. The bottle was shaken for two hours on a vibrating table and
then left standing for 24 hours. A sample of the liquid phase was drawn off and
analyzed for fluoride. Before being used for this test, the resins were regener-
ated as follows:
lonac P-60 Resin
2
- Regenerate with 5% sodium hydroxide, using a dosage of 4 Ibs/ft
Q
- Rinse with water at a rate of 1 gpm/ft for a total of 15 gallons
- Neutralize with 1% potassium bisulfate solution, using a dosage of
1 lb/ft3
Q
- Rinse with water at a rate of 1 gpm/ft for a total of 30 gallons
Amberlite IRA-400 Resin
q
- Regenerate with 4% sodium hydroxide, using a dosage of 6 Ibs/ft
25
-------
- Rinse with water at a rate of 0.25 gpm/1-bed volume and then at
3
a rate of 1.5 gpm/ft resin until a pH of 10 is reached
Alcoa F-l Activated Alumina
o
- Regenerate with 4% sulfuric acid using a dosage of 12 Ibs/ft
- Rinse with water until a pH of 4.7 is reached
The isotherm data was compared with plots on pages 16-5, 16-8, 16-9, and
16-10 of Perry's Chemical Engineering Handbook. The generalized plots show
the concentration in the fluid phase, c, as the abscissa. A plot of 't[fl versus
"c", which is convex-upward, is a favorable isotherm. Facile absorption
of material from the fluid phase onto the solid phase will occur. For sim-
plicity, the symbols defined below will be used throughout this report:
c = concentration in stream at equilibrium
CQ = initial concentration in stream (influent)
q = concentration in resin at equilibrium
Q = upper limit of concentration in resin
For this particular program, Q is the capacity of the resin for fluoride. This
is the total fluoride which the resin can absorb when fluoride breakthrough is
seen in the effluent. Then q/Q is the degree of exhaustion of the resin. When
q/Q = 1, the resin has absorbed as much of the fluoride as it can (vertical
saturation). At this point, c/C goes to one, since no more flouride is absorb-
ed. This is fluoride breakthrough and signals that the bed must be regenerated.
For the general ion-exchange of solution ion A with resin ion B, the exchange
reaction is:
A + RB i± RA + B,
assuming A and B are of the same valence (true for system under study).
Equilibrium is then given by:
K * -D = x = rr:
AS3 O » UT, A .
26
-------
where q is the concentration of A or B in the resin
c is the concentration of A or B in the solution
Y is the plot of q/Q
X is the plot of c/C
Q is the capacity of the resin for A
C is the influent concentration of A
o
The following results were obtained:
lonac P-60 Ion-Exchange Resia - The exchange data for the lonac P-60
resin was so meager that an isotherm plot was not warranted. By
comparing CQ (the total concentration of fluoride ions in solution)
with C (the equilibrium fluoride concentration), the low fluoride pickup
can be seen (Table 10). Since the fluoride pickup was so low, the
large volume of P-60 resin that would be required for fluoride removal
from the rinse waters to be tested would make its use impractical. As
a result, further evaluation of this resin was discontinued.
Amberlite IRA-400 Ion-Exchange Resin - Analysis of the data obtained
for the Amberlite IRA-400 resin (Table 11) shows that the degree of
pickup for this resin is much greater than that for the lonac P-60 resin.
The plot of equilibrium concentration of fluoride ions in solution (C)
versus the weight of fluoride ion removed over the total weight of
absorber (q) in Figure 4 was compared with that of Type V resin
(Figure 16-3, page 16-9, of the fourth edition of Perry's Chemical
Engineer's Handbook). If the equation is considered to be a Langmiur
type, then according to Table 16-5 this resin has unfavorable charac-
teristics at low concentrations and favorable at high concentrations.
27
-------
TABLE 10
ISOTHERM DATA FOR IONAC P-60 ION-EXCHANGE RESIN AT 26°C
Sample
No.
1
2
3
4
5
6
Sample
Slza
(ml)
200
200
200
200
200
200
(mg)?>
85
85
85
85
85
85
W
-------
TABLE 11
ISOTHERM DATA FOR AMBER LITE IRA-400 ION-EXCHANGE RESIN AT 26°C
Sample
NO.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Sample
Size
(mt)
200
200
200
200
150
200
200
200
200
175
200
200
200
200
200
Co
(mg/l)
119
119
119
119
119
135
135
135
135
135
126
126
119
119
119
W
(mg>
2676
2217
1787
1443
582
427
221
127
76
45
3471
6182
8924
16237
16727
w
(mg)
10.9
11.2
9.2
7.4
4.7
6.4
6.0
6.4
5.0
1.6
10.4
15.4
13.3
20.6
20.1
Q--B-
.00417
.00507
.00513
.00512
.00808
.01500
.02720
.05030
.06580
.04500
.00300
.00249
.00148
.00176
.00121
C
(mg/l)
64.3
63.0
73.0
82.0
87.7
103.0
105.0
103.0
110.0
126.0
74.0
48.8
52.4
15.8
18.4
Q
.0658
.0658
.0658
.0658
.0658
.0658
.0658
.0658
.0658
.0658
.0658
.0658
.0658
.0658
.0658
C
x-.-cr
.540
.529
.613
.689
.737
.762
.777
.762
.812
.932
.588
.388
.430
.131
.154
q
Y =
Q
.0636
.0775
.0780
.0780
.1230
.2290
.4140
.7680
1.0000
.6850
.0458
.0380
.0218
.0268
.0184
C - Total concentration of F~ ions in solution
W = Total weight of adsorber
w = Weight of F" ions removed
C = Equilibrium F~ ions removed
Q = Capacity of adsorber experimentally determined; for above,Q = .0658
-------
W
o
o
.08
.07-
.06-
OT .06
I
Q
o
UJ
§
8
.04-
.03 H
.02
.OI-
20 30 40 SO 60 70 80
CONCENTRATION IN FLUID PHASE . mg/l
90
too
110
120
Figure 4. Isotherm Plot for Amberlite IRA-400 Ion-Exchange Resin
-------
F-l Activated Alumina Ion-Exchange Material - The isotherm data ob-
tained for the activated alumina (Table 12) was plotted as C versus q
(Figure 5) and compared to applicable plots in Perry's Chemical
Engineer's Handbook. The curve for the activated alumina was simi-
lar to that for the IRA-400 resin. It will be shown subsequently in
this section that the exchange characteristics of activated alumina are
extremely dependent upon the rate of flow of the solution. The pH
regulating capability of the material, however, is excellent and will be
used for this purpose and as a fluoride polishing treatment in subsequent
evaluations.
GENERATION OF BREAKTHROUGH DATA
In simple fixed-bed operations, the ion being exchanged is removed continously
from the carrier fluid and accumulated in the solid phase. Such transfer pro-
ceeds until the concentration on the solid reaches a value corresponding to
equilibrium with the concentration in the feed stream. Although at this point
the fluid just leaving the solid layer at the top of the column reaches the feed
concentration, the column effluent remains practically free of this solute until
the last layer of the sorbent is nearly saturated. This change in effluent con-
centration with time is known as breakthrough or concentration history.
Amberlite IRA-400 Resin
Prior to the establishment of breakthrough data for the Amberlite IRA-400/120-
135 ppm sodium fluoride solution, the Amberlite resin was activated by flowing
4500 grams of 4% by weight NaOH through it. The resin was rinsed and the pH
of the effluent measured (10.5-11.5). The sodium fluoride solution was then
passed through the column at a rate of 6.7 gallons per hour. Breakthrough
occurred after about 215 liters of 120-ppm fluoride solution were passed through
the 3-5/16-inch diameter x 13-inch high Plexiglas column filled with 0. 0648
cubic feet of resin.
The breakthrough data was obtained as follows: After a known volume of solu-
tion was passed through the column (see Table 13), a 200-ml sample was taken
31
-------
TABLE 12
ISOTHERM DATA FOR ALCOA F-l ACTIVATED ALUMINA ION-EXCHANGE MATERIAL AT 26°C
Sample
No.
1
2
3
4
5
6
7
8
9
10
11
12
Simple
Size
(ml)
200
200
200
200
200
200
200
200
200
200
200
200
(m£l)
135
135
135
135
135
135
135
135
135
135
135
135
W
(ma)
1200
1004
801
601
305
214
130
67.2
43.0
20.9
14999
8376
w
-------
CO
CO
120
130
140
CONCENTRATION IN LIQUID PHASE (C),mg/l
Figure 5. Isotherm Plot for Alcoa F-l Activated Alumina Ion-Exchange Material
-------
TABLE 13. COLUMN BREAKTHROUGH DATA FOR AMBERLITE IRA-400
ION-EXCHANGE RESIN AND SODIUM FLOURIDE SOLUTION
(119-124 PARTS PER MILLION OF FLUORIDE ION) AT 26°C
Sample
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Fluoride
Input
(ppm)
119
119
119
119
119
119
118
117
126
124
124
124
124
124
124
Input
pH
6.4
6.4
6.4
6.2
6.2
Q.I
6.3
6.6
6.6
6.9
6.9
6.9
6.9
6.9
6.9
Input
Cofld
(ttmhotf
em)
645
645
546
545
545
545
515
535
530
600
500
500
500
500
500
Fluoride
Output
(ppm)
0.3
0.3
0.6
0.8
0.2
0.8
0.6
OS
1.8
4.0
16.0
9.9
18.3
89.5
119.0
Exit
pH
11.0
11.6
11.5
11.6
11.6
11.6
11.6
11.6
11.8
11.6
11.7
11.7
11.7
11.4
11.0
Exit
Cond.
Umhos/
cm)
380
1180
1200
1150
1100
1000
1120
1160
1140
1240
1280
1250
1250
1230
1200
Sample
Size
(ml)
2OO
200
200
200
200
100
200
200
200
200
200
200
200
200
200
Volume
Paswd
(ml)
400
1.000
2,000
3,600
4,800
6,500
13,700
25,900
39,920
98,250
163,885
168,085
181,035
219,920
229,255
09
Column = 35/16DX13H - ,065ft3
Flow Rate 6.7 Gallons/Hour
-------
and analyzed. This procedure was repeated until about 230 liters had been
processed. A plot of the data (Figure 6) showed that breakthrough (inflection
point) occurred at about 215 liters.
Alcoa F-l Activated Alumina
Activated alumina (0.07 cubic feet) was placed in the 3-5/16-inchKliameter x
13-inch-high Plexiglas column. The alumina was activated by passing 8850 grams
of 4% by weight HgSO. through it. The activated alumina was rinsed with one
liter of water at a rate of 7.10 gallons per hour and then rinsed with 15 liters
of water at a rate of 24.2 gallons per hour. The pH of the water at this point
was 3.6. A standard sodium fluoride solution containing 130 parts per million
of fluoride ion was passed through the column. After 13.6 liters had been
passed through the column, it was observed that the effluent fluoride concentra-
tion was very erratic. Two batches of solution were then passed backward through
the column to determine if better solution/alumina contact would improve
results. The same erratic results were again obtained. The sodium fluoride
was then continued through the column in a forward direction until 52.4 liters
had been passed. It was concluded that significant breakthrough data could not
be obtained, since the exchange operation was extremely flow rate-dependent
(see Table 14). It was observed, however, that the input pH and the exit pH
were very close in value. Since the ultimate objective of this experiment was
to keep a captive rinse (that is, to take process water, remove the fluoride and
then replace the water in the rinse tank), the pH of the water going back into
the tank would have to approximate the pH of the rinse before exchange. There-
fore, 10 liters of the effluent obtained from the IRA-400 study were passed through
the alumina F-l column. The influent had a pH of about 11 and the effluent
had a pH of about 5.5. The F-l activated alumina, in addition to its ion-
exchange properties, did serve as a pH regulating agent. It was decided to use
a two-stage, ion-exchange system for the chemical milling rinse studies. The
first column would contain Amberlite IRA-400 resin to remove the bulk of the
fluoride; the second column would contain Alcoa F-l activated alumina mainly
for pH control and for some limited ion-exchange.
35
-------
1.0-1
Co
C5
SODIUM FLUORIDE INFLUENT - 120 ppm
SERVICE FLOW RATE - 6.7 GALLONS/HOUR
BED-3 6/16-INCH DIAMETER. 13-INCH HIGH,OH~FORM
BREAKTHROUGH
100 150
VOLUME OF 120-PPM FLUORIDE ION, liters
200
..M
300
Figure 6. Amberlite TEA-400 Ion-Exchange Resin Breakthrough Curve
-------
TABLE 14. COLUMN BREAKTHROUGH DATA FOR ALCOA F-l ACTIVATED ALUMINA AND
SODIUM FLUORIDE SOLUTION (130 PARTS PER MILLION OF FLUORIDE ION)
W
Flow Mode
Forward
Forward
Forward
Forward
Backward
Backward
Forward
Forward
Forward
Forward
Fluoride
Input
(ppm)
130
130
130
130
130
130
130
130
130
130
Input
PH
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
Fluoride
Exit
(ppm)
115.0
9.3
77.4
108.0
80.6
102.0
99.0
15.3
16.2
46.0
Exit
pH
6.8
6.2
6.7
7.0
5.7
4.3
4.9
5.2
6.5
6.2
Sample
Size
(ml)
200
200
200
200
200
200
200
200
200
200
Total
Volume
Paned
(mO
4,200
8,400
13,600
18,800
24,000
28,200
32,400
45,000
48,200
52,400
Flow
Rate
(gph)
27.20
7.68
5.20
27.20
27.20
27.20
27.20
2.76
2.76
2.76
Column - 3 5/16 Inches in diameter by 13 inches high
-------
REGENERATION OF SPENT AMBERLITE IRA-400 RESIN
After passing 230 liters of 120 parts per million fluoride solution through the
Amberlite resin, the column had to be regenerated (breakthrough point passed).
Again, 4.5 liters of 4% by weight NaOH were passed at a rate of 2.5 gallons
per hour in a forward-flow direction. At various volume increments, samples
were drawn and the fluoride concentration and pH of the effluent were measured
(see Table 15 and Figure 7).
The bed was then rinsed with distilled water (2.5 gallons per hour for one bed-
volume and 3.65 gallons per hour for 6.2 bed-volumes). Fluoride concentra-
tion and pH of the effluent were measured (see Table 16 and Figure 8). At the
end of 7.2 bed volumes of rinse, the pH of the effluent was 10.6 and the fluoride
concentration was less than 0.02 part per million.
DEVELOPMENT OF TWO STAGE, ION-EXCHANGE SYSTEMS FOR TITANIUM
CHEMICAL MILLING RINSE WATER
A two-stage, ion-exchange system was designed and fabricated (Figure 9) to
establish fluoride removal capability on actual titanium chemical milling rinse
water. The columns for holding Amberlite IRA-400 ion-exchange resin and
Alcoa F-l activated alumina were constructed of two-inch-diameter by 36-inch-
high acrylic tubing. Polyvinyl chloride valves and piping were used. The
Amberlite IRA-400 resin used was the regenerated material from the 120-ppm
sodium fluoride breakthrough test. The Alcoa F-l activated alumina was virgin
material regenerated with sulfuric acid as previously described. The chemical
milling rinse water tested had a fluoride concentration range of 3 to 61 parts per
million (average value of 15.8 parts per million). The rinse water also contain-
ed nitrate up to 200 parts per million and trace quantities of aluminum, copper,
vanadium, tin and iron. The rinse water was flowed through the series of
columns at rates ranging from 2.4 to 20.2 gallons per hour. Samples of the effluent
effluent were taken at various intervals as described previously. A total 421.7
liters of rinse water was passed through the columns. Breakthrough occurred
after 375 liters (100 gallons) of rinse water were passed through (see Table 17
and Figure 10). The results of this test showed that the use of Amberlite IRA-
400 resin, in conjunction with Alcoa F-l activated alumina, is an efficient means
of removing fluoride from chemical milling rinse water while simultaneously
maintaining the pH of the solution.
38
-------
TABLE 15. REGENERATION OF AMBERLITE IRA-400 ION-EXCHANGE RESIN WITH
4% BY WEIGHT SODIUM HYDROXIDE
CO
to
Sample
1
2
3
4
Flow
Mode
Forward
Forward
Forward
Forward
Exit
Fluoride
(ppm)
5520
850
306
280
Exit
pH
12.60
12.90
13.65
13.60
Sample
Volume
(mil
4000
200
200
100
Volume
Passed
(ml)
4000
4200
4400
4500
Flow
Rate
<8ph)
2.5
2.5
2.5
2.5
Column - 3 5/16 inches in diameter by 13 inches high
-------
1
I
8
a
o
p
cc
111
o
I
111
o
cc
o
HI
V)
10J
5x 10°
LEGEND:
O RINSE FLOW RATE TO DISPLACE REGENERANT @ 2.50 GAL/HR, VOID VOLUME 735 ML.
DFINAL FLOW RATE @ 3.65 GAL/HR
BED - 35/16" DIAMETER BY 13" HIGH
10" 1.5x10"
RINSE EFFLUENT VOLUME, liters
2x 10
2.5 x
Figure 7. Amberlite IRA-400 Ion-Exchange Resin Rinse Curve for First-Run Spent Tower Following
Regeneration
-------
TABLE 16. RINSE WATER PROPERTIES FOLLOWING REGENERATION OF
AMBERLITE IRA-400 RESIN BY 4% SODIUM HYDROXIDE
Sample
1
2
3
4
5
6
Flow
Mode
Forward
Forward
Forward
Forward
Forward
Forward
Input
PH
5.65
5.65
5.65
5.65
5.65
5.65
Exit
pH
12.8
12.7
12.3
10.6
10.6
10.2
Exit
Fluoride
Concentration
(ppm)
19.7
21.5
18.2
<0.02
<0.02
<0.02
Sample
Volume
(ml)
200
200
200
200
200
200
Total
Volume
Pasted
(ml)
1,000
4,200
8,900
13,100*
17,300
21,500
Flow
Rate
(gph)
2.50
3.65 (After total of
1.84 liters
( passed)
3.65
3.65
3.65
3.65
Notes: (1) Input Sodium Fluoride Concentration -120 parts per million
(2) Column Dimension - 35/16 inches diameter by 13 inches high
7.2 Bed volumes
-------
S.S
6.0*
4.5.
O 4.0-
8
Ul
O
3.5-
3.0-
ui
Ul
O
m 2.0-
2
DC
1.5-
1.0
0.6-
0.25-
0
4% BY WEIGHT. NaOH @>6LBM NaOH/FT3 RESIN
REGENERATION FLOW RATE 2* GAL/HR @ .643 GAL/FT^/MIN
BED 3 6/16" DIAMETER 13" HIGH
4000
4200
REGENERATION WASTE EFFLUENT VOLUME, milliliters
4400
1
4500
Figure 8. Amberlite IRA-400 Ion-Exchange Resin Regeneration Curve for First-Run Spent Tower
-------
ROHM & HAAS
AMBERLITE IRA^OO
RESIN
ALCOA F-1
ACTIVATED
ALUMINA
ION-EXCHANGE
FEED PUMP
FLUORIDE-CONTAINING
RINSE WATER BEING
TREATED
Figure 9. Two-Stage, Ion-Exchange System for Titanium Chemical Milling
Rinse Water
43
-------
TABLE 17
BREAKTHROUGH DATA FOR AMBERLITE IRA-400 RESIN AND ALCOA F-l.
ACTIVATED ALUMINA COLUMNS IN SERIES FOR TITANIUM
CHEMICAL MILLING RINSE WATER
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Input
Fluoride
(ppffl)
12.1
835
8.85
835
12.1
12.1
61.1
61.1
3.08
3.08
17.8
173
173
10.6
10.6
10.6
12.3
129
123
12.6
Input
pH
23
3.3
3.3
3.3
2B
2£
2.6
2.6
23
23
2.7
2.7
2.7
2B
23
23
2.7
23
23
3.3
Input
Cond.
(fimhos/
cm)
1330
360
360
360
1150
1150
1700
1700
790
790
1500
1500
1500
1900
1900
1900
1100
1100
660
660
Exit
Flourkto
(ppm)
1.33
1.68
1.38
133
0.70
1.04
1.02
1.63
1.12
1.02
0.44
0.15
032
0.60
034
1.10
1.70
8.00
10.60
11.70
Exit
pH
3.1
4.3
4.3
4.3
43
4.5
4.4
4.7
43
43
4.6
43
5.0
5.1
5.7
5.3
6.2
53
33
4.7
Exit
Cond.
Cunhos/
cm)
570
180
140
180
112
280
77
62
112
50
69
58
57
55
74
57
108
117
195
180
SeMnpra
Size
(ml)
200
200
200
200
200
200
200
200
200
200
200
200
200
200
200
200
200
200
200
200
Total
Volume
Passed
(ml)
10,200
23,400
43.400
49.600
124.030
132330
153,850
173,230
201,430
212,130
224,330
236.530
248.730
260330
'273,130
290,580
373.130
385,130
409,530
421,780
Flow
Rate
(gph)
7.0
4.1
8.2
9.3
4.2
13.6
2.4
18.3
16.0
17.7
18.6
18.1
18.0
185
16.7
20.2
17.2
17.2
17.2
17.2
Column Dimensions: Amberlite IRA-400 (2 inch diameter by 32 inches high) and Alcoa F-1 Activated Alumina (2 inch
diameter by 31 inches high).
44
-------
C71
1.0 n
.9-
_ .8
0°
5
8
a
£
z
i-
.5
-I .4-
u.
u.
uj
Z
1 .3-
UJ
Q
.2-
o
o
cc
COLUMNSUSED:
2-INCH-DIA x 32-INCH HIGH PLEXIGLAS (AMBERLITE IRA 400)
2-INCH DIA x 31-INCH HIGH PLEXIGLASS (ALCOA F-1 ACTIVATED ALUMINA)
SERVICE FLOW RATE - 7.04 G.P.H.
BREAKTHROUGH
I
100
I
200
300
400
500
Figure 10.
VOLUME OF 15.8 PPM (AVG CONC) FLUORIDE SOLUTION, liters
Breakthrough Curve for Titanium Chemical Milling Rinse Water
-------
The laboratory test showed the feasibility of using the series ion-exchange
system for fluoride removal. Before a production system is set up, however,
it is recommended that a pilot system having a minimum scale-up factor of
eight be evaluated to establish more realistic operational parameters. The
amount of resin to be used should be based on 80 percent of laboratory data
to allow for inefficiency of the distribution networks. A pilot test would esta-
blish whether the functional groups of the strongly basic IRA-400 resin hydro-
lyze upon repeated cycling.
46
-------
SECTION V
PILOT CENTRIFUGE TREATMENT - TESTING AND ANALYSIS
APPROACH
This phase of the program was concerned with the development of centrifugal
techniques for concentrating and removing precipitated fluorides. Chemical
granular quicklime, which provided optimum fluoride removal in the Phase I
beaker tests, was used for all pilot tests. Pilot tests were conducted on titanium
chemical milling, titanium descaling, Amchem 7-17 deoxidizer and Alodine alumi-
num coating solutions. The ferrous alloy descaling solution was not evaluated on a
pilot basis because of its similarity to the titanium descaling solution.
STUDY AREAS
The task of developing pilot centrifuge treatment and removal techniques for
fluoride-containingprocess solutions consisted of the following technical efforts:
Pilot centrifuge setup and calibration
Treatment and centrifugal separation of chemical process solutions
PILOT CENTRIFUGE SETUP AND CALIBRATION
A Sharpies/Fletcher Mark in centrifuge (Figure 11) manufactured by the
Pennwalt Corporation was used for all laboratory pilot testing. This centrifuge,
which has a 14-inch-diameter by 6-inch-high basket and a speed capability of
3250 revolutions per minute, is designed for separation studies and performance
analyses to generate data for projected production equipment. The pilot cen-
trifuge system (Figure 12) consists of the following components:
55-gallon tank with Lightnin AG-100 air-powered mixer for making up
lime slurry
Vanton PY-60 centrifugal pump to transfer lime to treatment tank
Treatment tank with Holloway Super electric mixer and Jabsco P-6 cen-
trifugal feed pump
30-gallon polyvinyl chloride rectangular tank to receive treated effluent
(at rear of centrifuge)
47
-------
Figure 11. Sharpies/Fletcher Mark III Centrifuge
-------
SLUDGE-RECEIVING DRUM
SHARPLES/FLETCHER
MARK III CENTRIFUGE
CENTRIFUGE OUTPUT TO
30-GALLONPVCTANK
LIME SLURRY TANK
TREATMENT TANK WIT
ELECTRIC MIXER
JABSCO P-6
CENTRIFUGAL FEED PUMP
VANTON PY-60
IMEPUMP
Figure 12. Pilot Centrifuge System
-------
55-gallon drum to receive sludge skimmed from centrifuge bowl
The Sharpies/Fletcher Mark m centrifuge is completely variable in speed over
the 0-3250 rpm range. The Sharpies division of the Pennwalt Corporation rec-
ommended that testing be conducted at a force of 1300G and flow rates of up to
one gallon per minute. The centrifuge speed necessary to obtain the 1300G
force was calculated as follows:
Centrifugal force in multiples of the force of gravity
-5 2
= (1.42 x 10 ) (Bowl speed in rpm) (Bowl diameter in inches)
Bowl speed = 2550 rpm
Bowl diameter = 14 inches
The flow of treated solution to the centrifuge was regulated by using a bypass
valve on the exit side of the Jabsco P-6 centrifugal pump. By opening the
bypass valve, a portion of the treated solution was diverted back to the treat-
ment tank. Prior to an actual test, water was run through the system and the
time necessary to collect a measured amount of water was taken with a stopwatch.
TREATMENT AND CENTRIFUGAL SEPARATION OF CHEMICAL PROCESS
SOLUTIONS
Pilot-size treatment parameters were established for titanium chemical milling,
titanium descaling, Amchem 7-17 deoxidizer and alodine aluminum conversion
coating solutions. The following parameters for pilot-size treatment were
established:
Initial fluoride concentration
Solution-to-slurry ratio
Solids retention time
Temperature increase during treatment
Flow rate
Effluent pH
Effluent fluoride concentration
50
-------
Percent solids in effluent
Density of sludge
Percent solids in sludge
Chemical analysis of sludge
Chemical granular quicklime at a slurry concentration of 2.03 pounds of lime
per gallon of water was used for all pilot testing. The lime was slaked for 25
to 30 minutes; solution temperature rose 120°F above ambient temperature.
Titanium Chemical Milling Solution
Three pilot tests were conducted on spent titanium chemical milling solution
having a fluoride concentration of 93,000 parts per million and an acid concentra-
tion of 2.73 normal. Each test was conducted on 10 gallons of spent acid treat-
ed with 7.5 gallons of slaked chemical granular quicklime. Specific procedures
and results for the three tests conducted were as follows:
Test No. 1. Ten gallons of titanium chemical milling solution were
treated with 7.5 gallons of slaked chemical granular quicklime and
mixed with the Holloway Super mixer for one hour. Five gallons
of the treated solution were passed through the pilot centrifuge at a
rate of 0.6 gallon per minute. Mixing was continued for 6.5 hours,
after which an additional five gallons of treated solution were passed
through the centrifuge at a rate of 0.6 gallon per minute. The .re-
maining solution was passed through the centrifuge at a rate of 0.6
gallon per minute after 24 hours of mixing. Samples of the centrifuge
effluent taken at 1, 6.5-and 24-hour retention times were analyzed
and found to contain 8.75, 6.04 and 5.35 parts per million of fluoride,
respectively. The calcium flouride sludge was skimmed out of the
centrifuge bowl after each test run and physical characteristics were
determined. The results of this test are shown in Table 18.
Test No. 2. This test was conducted in exactly the same manner as test
No. 1 except for the mixing technique during solids retention. After treat-
ment with lime, the solution was allowed to stand for 1/2 hour. It
was then mixed for one hour. Five gallons of the solution were then
51
-------
TABLE 18. PILOT CENTRIFUGE TEST RESULTS
FOR CHEMICAL PROCESS SOLUTIONS
Solution
Titanium
Chemical
Milling
(Test No. 1)
Titanium
Chemical
Milling
(Test No. 2)
Titanium
Chemical
Milling
(Teat No. 3)
Titanium
Descaling
Solution
SN Acid
Aluminum
Deoxldlzer
Amchem
7-17
Aluminum
Convention
Coating
Atocttne
1200
Secondary
Treatment
On Above
Effluent
Initial
Concentration
of Pollutauta
(ppm)
93,000
Fluoride
2.73 N
Acid
93,000
Fluoride
2.73N
Acid
93.000
Fluoride
2.73N
Acid
60,000
Fluoride
119,000
Nitrate (N)
1466
Fluoride
126,000
Nitrate
2020
Fluoride
250
Nitrate
10.5
Fluoride
Ratio of
Volume
Wa»t« to
Volume
Slurry
4:3
4:3
4:3
1:1.1
14.57
J.islbs.
of Dry
CaOPer
10 Gal.
Waste
22.9 Grama
JM?A
Liters
Temperature
Rlae
During
Treatment
ft)
80
80
80
100
50
10
Solids
Retention
Time
(hours)
1
6.5
24
1.5
4.0
24
24
1
J
0.6
18
21
22
22
Temperature
Of Feed
To Centrifuge
<°F>
114
80
70
70
140
72
80
70
70
70
Centrifuge
"Q" Force
1300
1300
I
900
1300
1300
1300
1
Feed
Flow
Rate
(gpm)j
0.6
0.6
0.5
0.6
0.6
0.79
1
Volume of
Solids in
Feed, %
36.6
1
36.6
\
36.6
31.4
31.4
6.66
pHof
Effluent
11.0
11.4
11.75
11.4
11.6
11.9
12.5
10.2
11.6
9.6
9.8
11.9
2.7
2.7
Volume
of Solids In
Effluent,
%
0.67
1
0.67
1.0
0.67
0.67
0.33
Bulk
Density of
Sludge
(Ibs/ft3)
85.5
85.5
85.5
78.6
86.2
71.8
Sollda
In Sludge.
%
35
1
35
35
28
40
30.8
Volume
Of Sludge
For 10 Oat.
of Feed
(gal.)
2.2
1
2.2
I
3.7
3.1
3.1
0.665
Sieve
Size
(mesh)'
100%
-400
100%
-400
100%
-400
100%
-400
100%
-400
100%
-400
Effluent
Fluoride
Concentration
(ppm)
8.75
6.04
5.35
2.72
1.26
0.826
, 0.2
0.02
0.2
16.0
11.8
10.5
2.43
Chemical
Analysis
Sludge
(%)
T1O, -9.4
A1,T>,-1.1
Cif- .1)04
Sn - 0. 14
T10- - 9.4
A126 -1.1
Cu - . 004
Sn - 0. 14
TiO2 - 9- -I
A1203-1.1
Cu - .004
Sn - 0.14
T1O- -3.0
AUG. -0.6
C
-------
passed through the centrifuge as in Test No. 1. The solution was al-
lowed to stand for four hours, after which it was mixed for one minute;
an additional five gallons were then centrifuged. The remaining solu-
tion was allowed to stand for 24 hours, mixed for one minute and cen-
trifuged. Samples of the centrifuge effluent taken at 1.5-, 4- and 24-
hour retention times were analyzed and found to contain 2.72 1.26
and 0.826 parts per million of fluoride, respectively (see Table 18).
The significantly lower fluoride concentrations achieved in this test
as compared to those for Test No. 1 are attributed to the standing time
without mixing prior to separation. The more intense mixing in test
No. 1 apparently caused some redissolving of the calcium fluoride.
Test No. 3. This test was conducted to substantiate the low fluoride
concentrations achieved in Test No. 2 where the treated solution was
allowed to stand prior to separation. Ten gallons of the same spent
titanium chemical milling solution used in Test No's 1 and 2 were
r-
treated with 7.5 gallons of lime and allowed to stand without mixing
for 24 hours. The treated solution was then mixed for one minute
and passed through the centrifuge at a rate of 0.5 gallon per minute.
The centrifuge effluent was analyzed and found to contain only 0.2 part
per million of fluoride. This test again showed the need for solids
retention without mixing to obtain a low fluoride content in the ef-
fluent.
Titanium Descaling Solution
Ten gallons of titanium descaling solution having a fluoride concentration of 60,000
parts per million and an acid concentration of 6.0 normal were treated with 11
gallons of chemical granular quicklime slurry. After treatment, the solution
was allowed to stand for one hour. It was then mixed and passed through the cen-
trifuge at a rate of 0.6 gallon per minute. The calcium fluoride sludge was
skimmed out of the centrifuge bowl and physical characteristics were determined.
The centrifuge effluent was analyzed and found to contain only 0.021 part per
million of fluoride. Results are shown in Table 18.
53
-------
Aluminum Deoxidizing Solution (2-6 Oz/Gal. Amchem 7-17, 15-20% by volume
HNO_ (42°Be»
d
Ten gallons of aluminum deoxidizer were analyzed and found to contain 5500 parts
per million of hexavalent chromium and 1400 parts per million of fluoride. Prior
to fluoride treatment the deoxidizer solution was treated with 1.67 pounds of so-
dium metabisulphite which reduced the hexavalent chrome concentration to 0.055
part per million. The solution was then treated with 5.7 gallons of chemical gran-
ular quicklime slurry, allowed to stand for three hours, mixed and then passed
through the centrifuge at a rate of 0.6 gallon per minute. The sludge was skimmed
out of the centrifuge bowl and physical characteristics were determined. The cen-
trifuge effluent was analyzed and found to contain 0.16 part per million of fluoride.
Results are presented in Table 18.
Aluminum Conversion Coating Solution (Alodine 1200)
Ten gallons of Alodine 1200 solution obtained for treatment were analyzed and
found to contain 600 parts per million of hexavalent chromium and 2020 parts per
million of fluoride. Prior to fluoride treatment, the solution was treated with 157
grams of sodium metabisulfite which reduced the hexavalent chromium concentra-
tion to 0.05 part per million. The solution (10 gallons) was then treated with 2.13
pounds of dry chemical granular quicklime. The dry lime was used for treatment,
because the Alodine 1200 solution has a low acid concentration and only a small
amount of lime is necessary to treat the fluoride. The treated Alodine 1200 solu-
tion was mixed for ten minutes and then allowed to stand for 1/2 hour. About
three gallons of the treated solution were then passed through the centrifuge at a
rate of 0.79 gallon per/minute. The solution was allowed to stand for 18 hours
before being mixed. An additional three gallons of solution were run through the
centrifuge at a rate of 0.79 gallon per/minute. The remaining solution was allowed
to stand for 21 hours before it was run through the centrifuge at a rate of 0.79
gallon per minute. Samples of the centrifuge effluents taken at the 25-minute, 18-
hour and 21-hour retention times were analyzed for fluoride and found to contain
16.0, 11.6 and 10.5 parts per million of fluoride, respectively. The calcium fluor-
ide sludge was skimmed out of the bowl and physical characteristics determined.
Since an effluent having a fluoride concentration of three parts per million could
not be obtained with the Alodine 1200 solution using lime treatment, a secondary
treatment with aluminum sulfate to precipitate out aluminum fluoride was tried.
Four liters of the centrifuge effluent from the lime treatment test were treated
with 22.93 grams of aluminum sulfate.
54
-------
The treated solution was mixed and allowed to stand for 22 hours. A sample of
the treated solution was men removed and centrifuged for three minutes using
an International Clinical centrifuge. Analysis of the clear effluent showed
that the effluent had a fluoride concentration of 2.43 parts per million. Complete
test results are given in Table 18.
SUMMARY OF RESULTS
The results of using calcium oxide (chemical granular quicklime) to treat four
chemical processing solutions are summarized in Table 18. Test results show
that titanium chemical milling solution, titanium descaling solution and alumi-
num deoxidizer solution can be treated to reduce fluoride effluent levels to
less than one part per million using only primary treatment with lime. The
fluoride concentration of the aluminum conversion coating solution (Alodine
1200) was reduced to 10.5 parts per million after primary treatment with
lime. Secondary treatment of this solution with aluminum sulfate produced
an effluent with less than rnree parts per million of fluoride.
After completion of the pilot tests, discussions were held with representatives
of the Sharpies Division of the Pennwalt Corporation regarding the data gener-
ated. The Sharpies representatives indicated that for the solutions treated
with feed solids contents greater than 30 percent, a horizontal Super-D-Canter
type of centrifuge would be required (Figure 13). If a vertical basket type of
centrifuge were used for production treatment, the basket would fill up so rap-
idly that skimming and unloading times would be too short to get a reasonably
effective capacity. The calcium fluoride sludge should also be continuously
unloaded, because it is too heavy to permit effective skimming.
The Super-D-Canter type of centrifuge is built to take heavy sludge and se-
parate it while simultaneously delivering a clear effluent. Since the Super-D-
Canter centrifuge utilizes a force of over 2000 times gravity, a much clearer
effluent and smaller volume of drier sludge would be obtained compared to
that obtained with the pilot centrifuge.
55
-------
LIQUIDS
DISCHARGE
SOLIDS
DISCHARGE
SLURRY IN
Figure 13. Sharpies Super-D-Canter Centrifuge
56
-------
Based on test results obtained under this portion of the program, Grumman
Aerospace Corporation rented (at no cost to the program) a Sharpies P-600
Super-D-Canter centrifuge to run a production test on the 10 percent hydrofluoric
acid titanium chemical milling solution. This test is described in Section VI
of this report. The titanium chemical milling solution was selected for the
production lime treatment test because it is the most often dumped fluoride
solution used at Grumman and because it is the only one of the four solutions
pilot tested which contains only fluoride as a pollutant. The other solutions
that were pilot tested (titanium descaling, aluminum deoxidizer and A Iodine
1200) also contain nitrates at concentrations up to 60,000 parts per million in
their lime-treated effluents. In New York State nitrate concentration must be
reduced to less than 10 parts per million before the effluent can be discharged.
The technique for removing nitrate from these effluents is discussed in Section
VI of this report and will be fulfy developed under Phase n of this program.
57
-------
SECTION VI
ECONOMICS OF CALCIUM FLUORIDE SLUDGE RE-USE
AND PRODUCTION SCALE-UP
APPROACH
This phase of the program was concerned with establishing the feasibility of re-
using calcium fluoride sludge as an additive to concrete, developing a production
scale-up plan and generating cost data relative to fluoride treatment. Production
testing of lime treatment for 10 percent hydrofluoric acid titanium chemical
milling solution was also included in this phase of the program. This test,
which was not originally scheduled as part of the work for this EPA grant, is
being funded by Grumman Aerospace Corporation because of the significant
fluoride reduction results achieved on the titanium chemical milling solution
as described in Sections IV and V of this report.
STUDY AREAS
The task of determining the economics of calcium fluoride sludge re-use and
generating production scale-up data consisted of the following technical efforts:
Evaluation of the feasibility of using calcium fluoride sludge as a con-
crete additive
Evaluation of the feasibility of re-using ion-exchange rinse water
effluent
Determination of lime treatment costs for chemical process solutions
Production scale-up plans for treatment of fluoride and nitrate-con-
taining solution wastes
Installation and operation of a prototype production lime treatment system
for titanium chemical milling solution
58
-------
EVALUATION OF CALCIUM FLUORIDE SLUDGE AS AN ADDITIVE TO CONCRETE
This test was conducted to establish the feasibility of using calcium fluoride sludge
(Figure 14) from lime treatment as an additive for concrete. If the addition of
calcium fluoride sludge to concrete would increase or maintain the strength of
concrete, this technique would provide a means of disposing of large quantities of
sludge. At the present time, most lime treatment sludges are disposed of as land
fill. As land fill sites become filled and disposal regulations more stringent, how-
ever, this method will no longer be available. The problem is especially acute for
companies located near large cities where land is relatively scarce.
Test Procedure
Tests were conducted on concrete prepared using procedures of ASTM
Standards C109-70T and C305. The compositions of the concrete specimens tested
were as follows:
Mixture No. 1 (Standard Mixture)
- 250 grams of Type I Portland Cement
- 687.5 grams of sand, graded as per ASTM Standard C 109
Mixture No. 2 (Standard Mix Plus Extra Sludge)
- 250 grams of Type I Portland Cement
- 687.5 grams of sand, graded as per ASTM Standard C 109
- 757 grams of wet titanium descaling solution sludge from previous
pilot centrifuge tests - (100 grams as dry sludge)
Mixture No. 3
- 250 grams of Type I Portland Cement
- 623 grams of sand, graded as per ASTM Standard C 109
- 230 grams of wet titanium descaling solution sludge from previous
pilot centrifuge tests (64.5 grams as dry sludge)
The standard mixture as well as the mixtures containing sludge (Table 19) were
prepared using procedures outlined in ASTM Standard C305. Three 1.575 x
1.575 x 6.3-inch flexural specimens were cast from each mixture (Figure 15).
59
-------
0
o
Figure 14. Calcium Fluoride Sludge from Lime Treatment of Titanium Chemical Milling Solution
-------
TABLE 19 FLEXURAL AND COMPRESSIVE STRENGTH TEST RESULTS
FOR REGULAR AND SLUDGE-CONTAINING CONCRETES
Mixture
No.
1
2
3
Ratio of
Cement
To Solids
1:2, 75
1:3, 16
1:2.75
Percentage of
Cement
(%)
26.7
24.1
26.7
Percentage of
Sand
(%)
73.3
66.4
66.5
Percentage of
Sludge
(%>
0
9.5
6.8
Flexural
Test No.
1
2
3
1
2
3
1
2
3
Beam
Depth
(In)
1. 581
1.580
1.576
1.577
1.572
1.578
1.561
1.571
1.572
Beam
Width
(In)
1.535
1.551
1.543
1.565
1.568
1.561
1.538
1,526
1.537
Fall
Load
(pounds)
315
311
298
204
224
216
290
298
329
Flexural
Strength
(psl)
567
560
536
357
404
389
522
536
592
Average
Flexural
Strength
(pst)
554
383
550
Cotnpressive
Test No.
1-1
1-2
2-1
2-2
3-1
3-2
1-1
1-2
2-1
2-2
3-1
3-2
1-1
1-2
2-1
2-2
3-1
3-2
Sample
Height
(In)
1.582
1.582
1.579
1.578
1.577
1.577
1.579
1.579
1.572
1.571
1.560
1.580
1.564
1.568
1.573
1.575
1.576
1.570
Area
of Sample
(In*)
2. 319
2.333
2.334
2.369
2.316
2.321
2.503
2.490
2.493
2.488
2.492
2.488
2.402
2.420
2.421
2.462
2.457
2.390
Maximum
toad
(pounds)
5590
5825
5200
5160
5090
5740
4300
4220
4025
4000
4475
4450
6280
7280
7190
7200
6900
7280
Compresslve
Strength
(pel)
2410
2495
2230
2180
2195
24V5
1715
1695
1615
1610
1805
1790
2615
3010
2970
2925
2810
3045
Average
Compresslve
Strength
(pst)
2330
1705
2895
-------
Figure 15. Mold for Casting Concrete-Calcium Fluoride Sludge Flexural Strength Test Specimens
and Cubical Compressive Strength Test Specimens
-------
The specimens were cured for seven days at 90 percent relative humidity and
at a temperature of 23°C± 1.7°C per ASTM Standard C 109. The cured flexural
strength specimens were tested per ASTM Standard C 348 on a Wiedemann
Baldwin SOB, 30,000-pound universal testing machine (Figure 16). After testing,
the flexural strength specimens were cut into cubical compression strength
specimens using a diamond saw blade and tested per ASTM Standard C 116 on a
Wiedemann Baldwin 30B universal testing machine in (Figure 17).
Test Results
Flexural and compressive strength test results for the three mixtures evaluated
are shown in Table 19. The compressive strength of Mixture No. 3, which con-
tains 6.8 percent sludge (cement-to-solids ratio of 1:2.75), is higher than that
for standard concrete (Mixture No. 1). The flexural strength of Mixture No. 3
was about equivalent to that for standard concrete. The compressive and
flexural strength values for Mixture No. 2, which contains 9.5 percent sludge,
(cement-to-solids ratio of 1:3.16) are lower than those for standard concrete.
These results show that it is potentially feasible to use calcium fluoride sludge as
an additive to conprete.
LEACHABILITY OF SLUDGE CONCRETE
A leachability test was conducted to determine if concrete containing calcium
fluoride sludge would lose fluorides when continuously exposed to running water.
In this test, 45 ounces of sludge concrete Mixture No. 2 were subjected to con-
tinuous recirculation of two liters of tap water for a period of two weeks (Figure
18). The water was checked for fluoride content, pH and conductivity before and
after the test. Initial values were a pH of 5.9, a conductivity of 46 micromhos
per centimeter and a flouride content of 0.068 part per million. After two
weeks of continuous recirculation, the water had a pH of 11.1, a conductivity of
10,300 micromhos per centimeter and a fluoride level of 1.7 parts per million.
This test showed that an insignificant amount of fluorides is leached out of
sludge concrete. The rise in pH and conductivity is due mainly to other leach-
able materials in the concrete (e.g., lime). Based on the results of the flexural
and compressive strength tests as well as the results of the leachability test,
further development work to establish sludge concrete as a viable material
for roads will be performed under Phase n of this program.
63
-------
Figure 16. Wiedernann Baldwin 30B, 30, 000-Pound Universal Testing Machine
-------
Figure 17. Concrete-Calcium Fluoride Sludge Compression Strength Specimen
Being Tested in Wiedemann-Baldwin SOB Universal Testing Machine
65
-------
PLASTIC
BAG TO
PREVENT
EVAPORATION
FUNNEL TO
DISPERSE
LIQUID
SLUDGE
CONCRETE
4-LITER
BEAKER
TAP WATER
SUBMERSIBLE
PUMP
Figure 18. Test Rig for Leachability Test
66
-------
RE-USE OF ION-EXCHANGE RINSE WATER EFFLUENT
Re-use of the effluent liquids from ion-exchange columns as rinse water or
solution make-up water would eliminate the need for ground water disposal
of these liquids. Two important analytical characteristics that affect rinse
water re-use are pH (a measure of acidity) and specific conductance (a mea-
sure of a fluid's capacity to convey an electric current). Since .the specific
conductance is affected by the nature and concentration of the substances dissolved
in a fluid, it gives an indication of the degree of purify of that fluid.
The total flow of rinse water for the titanium chemical milling solution is
10,000 gallons for 16 hours use. Results obtained for this solution using the
series ion-exchange system described in Section IV showed that the solution
can be re-used. The pH, conductivity and fluoride concentration ranges for
titanium chemical milling rinse water before and after ion-exchange treatment
were as follows:
PARAMETER INFLUENT EFFLUENT
pH 2-4 3-5
Conductivity (micromhos/centimeter) 120-2000 60-600
Fluoride Concentration (ppm) 3-60 1-2
LIME TREATMENT COSTS FOR CHEMICAL PROCESS SOLUTIONS
The cost to treat waste titanium chemical milling, titanium descaling and
aluminum deoxidizing solutions with lime was calculated using the following
equation:
Ct=(V8/V0)(Dg)(Cl)(Vk)
where Ct = cost to lime treat waste solution, $
V_/V = ratio of lime slurry, to waste solution
8 C
Ds = lime slurry concentration, 2.03 Ibs/gal
GI = lime cost, $0.0105/lb
Vfc = total volume of waste solution to be treated, gal
The lime slurry/waste solution ratios used for titanium chemical mining,
titanium descaling and aluminum deoxidizing solutions were 3/4, 1.1/1.0 and
0.57/1.0, respectively. The lime treatment costs per 1,000 gallons
67
-------
of solution are as follows:
Titanium Chemical Milling Solution - $ 15.97
Titanium Descaling Solution - $ 23.41
Aluminum Deoxidizing Solution - $ 12.12
The cost to treat waste aluminum conversion coating solution (Alodine
1200) consists of both the primary treatment cost with lime and the sec-
ondary treatment cost with aluminum sulfate. These costs were calcu-
lated using the following equation:
Ct=(W)(C)(V)
where Ct = cost to treat waste solution, $
W = weight of treatment material, Ibs/gal of waste solution
(0.213 Ib of lime or 0.0478 Ib of aluminum sulfate)
C = cost of treatment material, $0.0105Ab for lime and
$0.035/lb for aluminum sulfate
V = total volume of waste solution to be treated, gal
Primary (lime) and secondary (aluminum sulfate) treatment costs per 1,000
gallons of waste aluminum conversion coating solution are $2.23 and $1.67,
respectively. Total treatment cost for this solution is $3.90 per 1,000 gallons.
The calculated costs pertain only to the treatment of waste chemical process
solutions to remove fluorides. The calcium fluoride sludge produced must
either be carted away to land-fill sites or be re-used in other applications (e.g.,
sludge concrete). Although the cost of dumping calcium fluoride sludge is rel-
atively low, the precise cost depends on the geographic location and availability
of the land-fill sites. Land disposal of calcium fluoride sludge, however, has
the potential for causing problems because of its relatively high solubility, par-
ticularly where the ground water is used for drinking water supplies.
PRODUCTION SCALE-UP PLANS FOB TREATMENT OF PROCESS SOLUTIONS
AND RINSE WATERS
Treatment of Process Solutions
A proposed layout of a production installation to treat fluoride and nitrate-con-
taining process solutions is shown in Figure 19. The system would utilize a
centrifuge for separation and removal of precipitated fluorides and a mechanically
68
-------
SODIUM
METABISULPHITE
FEEDER
\
\-
LIME
FEEDER
C J
MIXER
PUMP
WASTE
SOLUTION
RECEIVING
TANK
_cn
CHROMATE
RECEIVING
& TREATMENT
TANK
-Ol
WASTE SOLUTIONS:
HF
HF-HN03
HN03Na2S04
CKROMATE CONVERSION
CHROMATED OEOXIDIZER
HF-H2SO,
HN03
H3P04-HN03
HN03CH3OH
RINSE WATER REGENERATION
WASTES f ROM ION EXCHANGE
LINIE
SLAKING
& HOLDING
TANK
I
LIME
TREATMENT
TANK
1
, r (OVERFLOW)
SETTLING
TANK
EVAPORATOR
FEED
SLUDGE
DRUM
MECHANICALLY
AIDED EVAPORATOR
SOLIDS
PURIFIED
WATER
CaINO,
Figure 19. Proposed Industrial Waste Treatment Facility for Fluorides, Nitrates and Chromates
-------
aided evaporator for removal of nitrates from the centrifuge effluents. The
parameters for operation of the evaporator will be established under Phase n
of this program. Specific treatment procedures for the process solutions
evaluated under the pilot phase of this program would be as follows:
Titanium Chemical Milling Solution (10% HF)
Treat solution with lime slurry to precipitate fluorides as calcium
fluoride. Pass the precipitated solution through the centrifuge to
obtain solid calcium fluoride sludge and clear effluent.
Titanium Descaling Solution (40% Nitric Acid plus 2% Hydrofluoric
Acid)
Treat solution with lime slurry to precipitate fluorides as calcium
fluoride. Pass precipitated solution through the centrifuge to obtain
solid calcium fluoride sludge and nitrate-containing effluent. Pass
the nitrate-containing effluent through the mechanically aided eva-
porator to obtain calcium nitrate solids and clear effluent.
Aluminum Deoxidizing Solution (2-6 oz/gal Amchem 7-17 plus 15%-
20% by volume Nitric Acid)
Treat the solution with sodium metabisulphlte to reduce chromium
from the hexavalent to the trivalent form. Treat the chromium-
reduced solution with lime to precipitate fluorides as calcium and
chromium as hydroxide. Pass the precipitate solution through
the centrifuge to obtain calcium fluoride-chromium hydroxide
sludge and effluent containing nitrates. Pass the nitrate-contain-
ing effluent through the mechanically aided evaporator to obtain
calcium nitrate solids and clear effluent
Aluminum Conversion Coating Solution (Alodine 1200)
Treat the solution with sodium metabisulpbite to reduce chromium
from the hexavalent to the trivalent from. Treat the chromium-reduced
solution with lime to precipitate fluorides as calcium and chromium as
hydroxide. Treat the solution with aluminum sulfate to precipitate
aluminum fluoride and reduce the fluoride concentration to less than
three parts million. Pass the precipitated solution through the centri-
fuge to obtain calcium fluoride-chromium hydroxide-aluminum
70
-------
fluoride sludge and effluent containing nitrates. Pass the nitrate-
containing effuent through the mechanically aided evaporator to
obtain calcium nitrate solids and clear effluent.
Treatment of Rinse Waters
A proposed layout of a production installation to treat fluoride and nitrate-
containing rinse waters is shown in Figure 20. This captive rinse system would
utilize Rohm & Haas Amberlite IRA 400 resin for fluoride removal and Alcoa F-l
activated alumina for fluoride removal and pH adjustment. The sodium hydrox-
ide regeneration waste from the IRA 400 resin and the sulfuric acid regeneration
waste from the activated alumina would be passed back to the lime treatment tank
shown in Figure 19, and treated as follows:
Treat the regeneration wastes with lime slurry to precipitate fluorides
as calcium fluoride. Pass the precipitated solution through the centrifuge
to obtain solid calcium fluoride sludge and effluent containing nitrates.
Pass the nitrate-containing effluent through the mechanically aided
evaporator to obtain calcium nitrate solids and clear effluent.
INSTALLATION AND OPERATION OF PROTOTYPE PRODUCTION LIME
TREATMENT SYSTEM FOR TITANIUM CHEMICAL MILLING SOLUTION
Objective
The objective of this study was to demonstrate on a pilot-line basis the ef-
fectiveness of the lime treatment technique for fluoride waste disposal.
Specific goals were:
Treatment of spent titanium chemical milling solution with lime
to reduce fluoride concentration to three ppm or less
Separation of reaction solids (calcium fluoride and metal oxides)
from the treated solution using centrifugal/settling tank procedures
Treatment Procedure
The recommended procedure for treating titanium chemical milling solution
is as follows:
Lime Slaking. Prepare calcium hydroxide slurry by adding 2.03
pounds of chemical granular quicklime per gallon of water
71
-------
FLUORIDE AND
NITRATE CONT.
RINSE-WATER
-d
tc
BACKWASH
WASTE
TO LIME
TREATMENT
TANK
REGENERATION WASTE
TO LIME TREATMENT
TANK «
ROHM & HAAS
AMBERLITE
IRA 400 RESIN
FOR FLUORIDE
AND NITRATE
REMOVAL
SODIUM HYDROXIDE
REGENERANT
&
ALCOA
ACTIVATED
ALUMINA
FOR FLUORIDE
REMOVAL AND
pH ADJUSTMENT
SULFURIC ACID
REGENERANT
WATER
BACKWASH I
BACKWASH
WASTE TO
LtME TREATMENT
TANK
CLEAR
EFFLUENT
BACK TO
RINSE
TANK
REGENERATION
WASTE TO
LIME TREATMENT
TANK
3-WAY
VALVE
VALVE
PUMP
Figure 20. Proposed Ion-Exchange System for Fluoride and Nitrate-Containing Rinse Waters
-------
Fluoride Solution Treatment. Add three parts by volume of calcium
hydroxide slurry to four parts by volume of titanium chemical milling
solution. Mix the reactants thoroughly while adding the slurry.
pH Adjustment. The pH of the treated solution should be between 11.0
and 12.5 for optimum results. If necessary, adjust the pH by adding
calcium hydroxide slurry.
Solids Retention. After adding the calcium hydroxide slurry, allow
the treated solution to stand unagitated for at least 24 hours.
Centrifugal Separation. Centri fuge the treated solution to provide
effluent and calcium fluoride sludge.
Settling Tank Separation. Discharge the effluent into a holding/settling
tank and allow it to stand unagitated for at least 32 hours to permit
settling of the particles not separated by the centrifuge. At the end
of this period, the supernatant liquid is ready for discharge into leach-
ing ponds.
Sludge Removal. The solid material obtained from the centrifugal
separation operation does not require further treatment.
Process Demonstration
Based on the data generated during the course of this grant, the recommended
procedure described previously was evaluated using 1,000 gallons of spent ti-
tanium chemical milling solution.
Lime Slaking (Figure 21)
- 900 gallons of water were added to the lime storage tank (No. 779)
- 1,840 pounds of Pfizer chemical granular quicklime were added to
the three lime holding drums located above the lime tank
- The quicklime was slaked by rapidly adding it to the 900 gallons of
water in Lime Tank No. 779 and continuously agitating the slurry by
mechanical stirring. The slurry reached a maximum temperature
of 220°F about 20 minutes after addition of the quicklime.
73
-------
LIME
HOLDING
DRUMS
LIME STORAGE
TANK (N. 779)
ACID TREATMENT TANK :
(N. 595)
Figure 21. Pilot-Line Lime Treatment System
74
-------
- The lime slurry was continuously stirred over a three to four-day
period. Probe inspection showed the presence of clumps of un-
suspended lime in the corners of the tank due to inefficient agitation
in the square corners of the tank. The lime clumps were broken
up by directing compressed air against them. This brought about
suspension of the lime.
Fluoride Solution Treatment. 750 gallons of lime slurry were added
to 1,000 gallons of spent hydrofluoric acid (87, 700 ppm fluoride) in
the acid treatment tank (No. 595). The lime slurry was pumped into
the acid treatment tank at a rate of about 18 gallons per minute. The
temperature of the spent solution rose to 170°F.
pH Adjustment. The pH of the treated spent acid was 10.3. Mechan-
ical breakdown of the slurry pump prevented adjustment of the pH from
the recorded value of 10.3 to the recommended 11.0-12.5 range.
Solids Retention. Since the temperature of the treated solution (170°F)
was high enough to possibly damage the polyvinyl chloride (PVC) tank
liner, the treated solution was agitated for about 18 hours to dissipate
the heat of reaction and reduce the solution temperature to about 120°F.
This prevented damage to the PVC liner. The solution was allowed to
remain unagitated for four hours before the solution was analyzed to
determine the total fluoride concentration. The total fluoride concen-
tration was 3 ppm.
In previous laboratory studies using the recommended procedure,
fluoride concentrations as low as 0.2 ppm were obtained in the treated
solution. Departures from the recommended method (incomplete sus-
pension of the lime slurry, lack of pH adjustment and reduced solids
retention time) resulted in higher fluoride concentrations for the pilot-
line treatment process. Reduction of the fluoride concentration to the
goal value of 3 ppm in spite of the above process variations demon-
strated the suitability of the treatment process for production use.
Centrifugal Separation. The solids were separated from the treated
acid solution with a Sharpies P-660 Super-D-Canter centrifuge
75
-------
(Figure 22). This unit, which was rented from the Pennwalt Company
of Warminster, Pennsylvania, was used in conjunction with a variable-
speed feed pump that transferred the treated solution from the reaction
tank to the centrifuge feed tube. An auxiliary feed line was connected to
the centrifuge system to permit the addition of flocculants (chemical
agents that cause particle aggregation) to the treated solution. The
flocculating agents were transferred to the centrifuge from a 55-gallon
drum by a chemical feed pump. The separated solids (calcium fluoride
sludge) were collected in a drum located under the centrifuge. The
centrate was pumped to a holding tank.
Centrifugal separation was accomplished at a speed of 6,200 rpm. The
2,900-rpm pinion speed (rear drive) and No. 4 pond setting (position of
the centrate discharge plate for separation refinement) were in accor-
dance with the manufacturer's recommendations. Separations were
conducted at various feed rates with and without the addition of poly-
electrolyte flocculants. The two flocculants evaluated were Hercofloc
812.3 (cationic agent) manufactured by Hercules Incorporated, Wilming-
ton, Delaware, and Aquafloc 423 (anionic agent) manufactured by the
Dearborn Chemical Division of the Chemed Corporation, Lake Zurich,
Illinois.
The results of the separation tests are summarized in Table 20. In
Separations 1 and 2, the treated solution was passed through the cen-
trifuge at rates of 0.58 and 2.16 gpm, respectively. Analysis of the
centrifuge effluent showed that the effluent had a fluoride concentration
between 3 and 4 ppm. The amount of solids removed (based on an ini-
tial solids concentration of 177,500 ppm) was 95.04 percent. The
effluent from Test No. 2 was allowed to settle for 32 hours and was
again analyzed for suspended solids concentration. The analysis
showed that 99.99 percent of the solids was removed from the effluent.
During Separations 3 through 6, Hercofloc 812.3 flocculant was added
to the feed stream in an attempt to reduce the amount of suspended
solids. Comparison of the test results with those for Test No. 2 shows
that about four percent more solids were removed while the fluoride
concentration remained at 3 ppm. Although addition of the
76
-------
SHARPLES P-660 CENTRIFULE
CENTRIFUGE '
FEED
PUMP
Figure 22. Centrifugal Separation System
-------
TABLE 20
RESULTS FOR CONTRIFUGAL SEPARATION OF TREATED TITANIUM
CHEMICAL MILLING SOLUTION
Test
No.
1
2
3
4
5
6
7
8
9
Cone, of
Suspended
Solids
Initial
(ppm)
177,500
177,500
177,500
177,500
177,600
177,600
177,500
177,600
177,500
Centrifuge
Feed Rat*
(Oal/min.)
0.58
2.18
2.16
2.16
2.16
2.16
0.68
2.16
12
Polyelee-
trolytt
<%)
_
Hftrcofloc
81229
1/2%
Hereof! ex;
81239
1%
Hercofloc
81229
1/2%
rwrcofloc
81229
1%
Aquaflbc
42391%
Aquafloc
42301%
Aquafloc
42391%
Polyelec-
trolyte
Feed Rat*
(flal/min.)
1B
_
022
0.13
027
0.22
0.07
0.168
0.07
pHof
Effluent
104
104
104
104
104
104
104
104
104
Fluoride
Cone, of
Effluent
(ppm)
4
3
4*
3
3
3
12
9
19
Cone, of
Suspended
Solids in
Effluent
(ppm)
5280
8800
1330
1860
1210
1290
10
231
16
Solids
Removal
(%)
97.04
95.04
9925
98.95
9922
9927
99.99
99.87
99.99
Cone, of
Dissolved
Solids In
Effluent
(ppm)
464
-
-
-
814
460
278
Cone, of
Suspended
Solids
After
32 Mrs
(ppm)
M
18
-
-
-
-
-
Solid»
Removal
In Settled
Effluent
(%)
_
99.99
-
-
-
_
Sludge in
Solids
(%)
48
48
67
61
63>
61
-q
oo
Sharpies P-660 Centrifuge
Bowl Speed* 6200 rpm
Pinion Speed - 2900 rpm
Pond Setting No. 4
-------
polyelectrolyte flocculant improved the efficiency of the centrifugal
separation, the resultant effluent was cloudy and would not clear up
after standing. In Separations 7, 8 and 9, Aquafloc 423 polyelectrolyte
flocculant was evaluated. This flocculant gave high solids removal
values (99.87 to 99.99 percent). Analysis of the effluent, however,
showed that the fluoride concentration had increased to between 9 and
19 ppm. This anionic flocculant apparently reduced the concentration
of the free calcium ion, thereby permitting additional calcium fluoride
to dissolve and increase the fluoride ion concentration.
Sludge Analysis. Analysis of a sludge sample by optical emission
spectroscopy gave the following results:
Element Percent
Calcium > 10
Titanium >10
Aluminum 1
Vanadium 1
Tin 0.10
Copper 0.02
Iron 0.10
Silicon 0.05
The sludge consists mainly of calcium fluoride and metal oxides of
titanium, aluminum and vanadium. This type of material is routinely
sent to land fill sites for disposal.
Cost Studies for Prototype System
Equipment. The estimated equipment requirements for a production,
fluoride-treatment system for chemical processing solutions are as
follows:
- Lime Slaking System (500 tons per year). This would include a
camlock valve, piping, bin vent, bin level indicator, 50-ton capacity
storage silo, discharge valve, screw conveyor, slaking tank, sonic
79
-------
probes, slaking pump, flow meters, solenoids, three-way port
valve and storage tank. This system would cost about $20,000.
- Centrifugal Separation System. A system consisting of a Sharpies
Model P-3400 stainless steel centrifuge capable of handling 20 to
30 gallons per minute of treated chemical milling solution and a
10,000-gallon settling tank would cost about $40,000
Chemicals. The cost of chemical 'granular, quicklime ($21 per ton) to
treat 6,000 gallons of spent titanium chemical milling solution would
be $95.91. If a polyelectrolyte flocculant were used, the cost would
be about $90.00 for 6,000 gallons of spent titanium chemical milling
solution. This cost is based on the use of a one-percent solution of
flocculant at a feed rate of 0.1 gallon per minute.
80
-------
Section vn
BIBLIOGRAPHY
1. Beckman Technical Data Sheet No. 1735-A, "Fluoride Ion Electrode, "
BeckmanInstruments, Inc., Scientific Instruments Division, Fullerton,
California 92634, November 1968
2. Boruff, C.S., etal, "Adsorption of Fluoride from Salts by Alum Floe,"
Industrial and Engineering Chemistry, Volume 29, No. 10
3. De Moura, J., "Specific Ion Electrodes Used to Study Interactions of
Glucose -1- Phosphate and Mg++ and F~, " Arch. Biochem. Biophys,
University of Idaho, Moscow, Idaho, 134 (1), 258-9, 1969
4. Halbedel, H.S., "Acid Fluorides and Safety," Technical Sheet No. 768,
The Harshaw Chemical Company, 1971
5. Handbook of Physics and Chemistry, The Chemical Rubber Company, 45th
Ed, p. B-147
6. Kolthoff, I.M. and Sandell, E.B., Textbook of Quantitative Inorganic
Analysis. 3rd Ed., p. 58
7. Kuzyakin, E.B., et al, "Relationship of the Settling Rate of Waste Water
Precipitants to the Ionic Composition during their Defluorination with
Lime, Aluminum Sulfate, and Polyacrylamide."
8. Mac Ihtire, W.H., et al, "Removal of Fluorides from Natural Waters by
Calcium Phosphates," Industrial and Engineering Chemistry, Vol. 30,
No. 2.
9. Ockershausen, R.W., "Disposal of Hydrofluoric Acid Wastes," Industrial
Chemicals Division, Allied Chemical Corporation.
10. Vanderborgh, N.E., "Evaluation of the Lanthanum Fluoride Membrane
Electrode Response in Acidic Solutions, " Talanta, Vol. 15, p. 1009-1013,
1968
11. Warner, T.B., "Lanthanum Fluoride Electrode Response in Water and
1M Sodium Chloride," U. S. Government Research and Development
Report 69(17)-68, Naval Research Laboratory, Washington. D.C., 1969
81
-------
Report 69(l7)-68, Naval Research Laboratory, Washington, D. C., 1969
12. Zabban, W., and Jewett, H. W., "The Treatment of Fluoride Wastes,"
Water and Sewage Works, p. 415, November 1967
13. Instruction Manual, Fluoride Ion Activity, Electrode Model 94-09, Orion
Research Incorporated, Cambridge, Massachusetts
14. Chemical Engineer's Handbook, Perry's John H, Fourth Edition, 1963
15. Rohm and Haas Technical Bulletin 1E-73-63, Helpful Hints in Ion Ex-
change Technology
16. Chemical Lime Facts, Bulletin 214, National Lime Association, Wash-
ington, D.C 20005
82
-------
Section Vm
APPENDIX
FLUORIDE ANALYSIS PROCEDURE
Sample Make-Up Mixtures
Process Solutic
1 ml process solution
100 ml distilled water (exact volumes are necessary)
10 ml 0.1 M disodium E.D.T.A.
Bring pH to 10 with IN NaOH (exact volume added)
Rinse Solution
100 ml rinse solution (exact volume)
Bring pH to 10 with 1N NaOH (exact volume added)
Treated Process Solution (Two cases)
1) 10 ml of centrifuged sarnole free of solids if pH below 10
10ml distilled water (exact volumes are ncessary)
10 ml 0.1M ditodium E.D.T.A.
Bring pH to 10 with 1N or .5N NaOH
2) 10 ml of centrifuged sample relatively free of solids at pH above 10.
10 ml distilled water, (exact volumes are necessary)
Bring pH to 10 with 0.5N HCL
Approximately 100 ml of the following standards are necessary for analysis:
NaF concentration
(moles/liter)
10-'
10'1
10'3
10-*
10'5
Resulting Single
Fluoride Ion Activity
at 25°C. (moles/liter)
0.77x10"'
0.91 x 10T1
0.97 x 10"J
10'5
10'*
Expanded-Scale pH meter and Specific Ion Electrode. An expanded-scale pH meter similar to the Beckman
Expandomatic SS-2 is used in conjunction with the Orion Research Fluoride Electrode.
Analysis Procedure
1) Standardize the expanded scale of the meter for a range of ±200mv
2) Measure the millivolt potential of the six sodium fluoride standards employing the specific ion electrode
(Confd)
83
-------
APPENDIX (Cont.)
3) Prepare a process, rinte, or treated procwttample
4) Measure the millivolt potential of the sample; accept a reading which stabilizes for at least S minutes
5) Measure the Specific Conductance or "Conductivity" of the sample in micromho/cm
Calculations
Conductivity Ionic Strength
(Micromho/cm.) (Moles/Liter)
10 1.1x1 <19x 10s)
-------
00
01
1.0
0.9-
0.8-
ACTIVITY
COEFFICENT
(7)
0.7-
0.6-
O.S
4 6 8 10"
TOTAL IONIC STRENGTH (MOLES/LITER)
Figure 23. Ionic Activity Coefficient of Fluoride Ion as a Function of Total Ionic Strength
-------
SELECTED WATER
RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
|> Report No.
Accession
w
4. Title
TREATMENT AND RECOVERY OF FLUORIDE INDUSTRIAL WASTES
7. Author(s)
Staebter, Jr., Christian I.
Grumman Aerospace Corporation, Bethpage, New York }1714
; Report No*
SJ..''
!(). Project No,
S 800680
11. Contract/Grant No,
12070HGH
115. Supplementary Notes
Environmental Protection Agency report number*
EPA-660/2-73-024, March 1974.
16. Abstract
lids report presents the development and successful demonstration of laboratory and pilot-scale fluoride treatment techniques for selected aerospace
md metalworking industry chemical processing solutions and rinse waters. It includes laboratory-scale, lime treatment parameters for chemical
processing solutions such as temperature, retention time, pH, slurry concentration and fluoride influent and effluent levels, and ion-exchage
treatment to reduce the fluoride concentration of rinse waters to levels less than three parts per million.
Mot studies of centrifugal techniques to separate lime-precipitated bridges from titanium chemical milling, titanium de&caHng and aluminum
deoxidizing solutions show that lime precipitation can give final effluents having fluoride concentrations less than three parts per million. Aluminum
conversion coating solutions, however, require secondary treatment with aluminum sulfate to give final effluents having fluoride concentrations
less than three parts per million.
Chemical and mechanical property tests show that ft is potentially feasible to use calcium fluoride sludge as a strength-maintaining additive for
concrete. The reuse of treated rinse waters, the economics of precipitation, and production plans for chemical processing solutions and rinse
waters are also presented.
TMs report was submitted in fulfillment of Project S800680 (12070 HGH) under the (partial) sponsorship of the Office of Research and Development,
Environmental Protection Agency.
17a. Descriptors
*Iime Treatment, *Ion Exchange, 'Calcium, Fluoride Sludge
17b.
Fluoride Industrial Wastes
17c. COWRR t>:Jd & Group
18. Ava.^-iiity 19. Security Class.
(Report)
2A. Security Clas*
(Page)
Abstractor Christian J. StaeWer, Jr.
21. No', of
Pages
22. Price
Send To:
WATER RESOURCES SCIENTIFIC INFORMATION CENTER
U.S. DEPARTMENT OP THE INTERIOR
WASHINGTON. QjC. 20240
'^vtitutioji Grumman Aerospace Corporation
------- |