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.)

1—B
_
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

-------