'A-670/2-74-044
June 1974
Environmental Protection Technology Series
                       AN ION-EXCHANGE PROCESS
                     FOR RECOVERY  OF CHROMATE
                  FROM PIGMENT MANUFACTURING
                                National Environmental Research Center
                                  Office of Research and Development
                                 U.S. Environmental Protection Agency
                                          Cincinnati, Ohio 45268

-------
                                      EPA-670/2-74-044
                                      June  1974
    AN  ION-EXCHANGE  PROCESS  FOR  RECOVERY  OF

      CHROMATE  FROM  PIGMENT  MANUFACTURING
                      By

              Donald J.  Robinson
              Harold E.  Weisberg
                Glenn I.  Chase
             Kenneth R.  Libby,  Jr.
                James L.  Capper
         Mineral  Pigments Corporation
          Beltsville, Maryland  20705
               Project  12020  ERM
            Program Element  1BB036
               Project Officers

      Richard B.  Tabakin  and  John  Ciancia
Industrial  Waste  Treatment Research  Laboratory
           Edison,  New Jersey  08817
    NATIONAL ENVIRONMENTAL RESEARCH CENTER
      OFFICE OF RESEARCH AND DEVELOPMENT
     U.S.  ENVIRONMENTAL PROTECTION AGENCY
            CINCINNATI, OHIO  45268

-------
               REVIEW NOTICE

     The National Environmental Research Center -
Cincinnati has reviewed this report and approved
its publication.  Approval does not signify that
the contents necessarily reflect the views and
policies of the U.S. Environmental Protection
Agency, nor does mention of trade names or com-
mercial products constitute endorsement or recom-
mendation for use.
                    ii

-------
                          FOREWORD


     Man and his environment must be protected from the adverse
effects of pesticides, radiation, noise and other forms of pol-
lution, and the unwise management of solid waste.  Efforts to
protect the environment require a focus that recognizes the
interplay between the components of our physical environment —
air, water, and land.  The National Environmental Research
Centers provide this multidisciplinary focus through programs
engaged in

           •   studies on the effects of environmental
               contaminants on man and biosphere, and

           •   a search for ways to prevent contamin-
               ation and to recycle valuable resources.


     The studies for this report were undertaken to demonstrate
the suitability of an ion-exchange system for the treatment of
concentrated chromium containing rinse waters from the manu-
facture of zinc-yellow pigment.  The system removes over 99%
of chromium thereby preventing contamination of the receiving
waters while at the same time permitting recycling of the
chromium, a valuable resource, to the succeeding batch of pig-
ment.  The system also permits recovery of the zinc contained
in the rinse water in the form of a saleable by-product, zinc
carbonate.  This new technology could have a major effect on
the industry's efforts to protect our Nation's water resources.
                              A. W. Breidenbach, Ph.D.
                              Director
                              National Environmental
                              Research Center, Cincinnati
                              in

-------
                               ABSTRACT
Strongly basic ion-exchange resins have been shown to exhibit a prefer-
ence for dichromate over many other anions in water solution.  Labora-
tory studies were conducted to show that this ion preference could be
used to remove chromate from waste waters which were discharged from a
zinc yellow pigment manufacturing plant.  It was also shown that the re-
covered chromate solution could be recycled into product manufacture
without sacrificing product quality.

From these laboratory studies, a full-scale ion-exchange treatment plant
was designed, constructed, and demonstrated.  The chromate composition
of the plant effluent is being reduced from 2700 ppm to one to two ppm.

The treatment system was designed to treat 60 gallons per minute of in-
fluent and to discharge an effluent which is within statutory limits
for pH and for heavy metal content.  The plant was designed to require
minimal manual supervision.  The steps in treatment and in resin regen-
eration are performed automatically and the control system is interlocked
to make it fail safe.  Operators are required only to make up regenera-
tion solutions, to clean pump strainers and filters, to answer to alarms
and occasionally to differentiate between turbidity and color as seen
by the colorimeter.

This report was submitted in fulfillment of Project Number 12020 ERM,
by Mineral Pigments Corporation under the (partial) sponsorship of the
Environmental Protection Agency.
                                  iv

-------
                               CONTENTS
Abstract                                                          iv



List of Figures                                                   vi



List of Tables                                                  vili



Acknowledgments                                                    x



Sections



I       Conclusions                                                1



II      Recommendations                                            2



III     Introduction                                               3



IV      Laboratory Evaluation                                      6



V       Scale-Up                                                  18



VI      Plant Design                                              19



VII     Plant Operation                                           42



VIII    Discussion                                                4 5



DC      References                                                5 3



X       Glossary of Terms                                         55



XI      Appendices                                                56



     A     Analytical Methods                                     56



     B     Details of the Laboratory  Study                         58



     C     Details of the Plant Operations                         70



     D     Useful Information for Designers and Operators          85

-------
                                FIGURES

No.

1    Typical Curve of Concentration of Bichromate
     in the Regenerant Versus Volume for One Pass                   8

2    Typical Curve of Concentration of Dichromate
     in the Regenerant Versus Volume When The Initial
     Volume is Recycled                                             8

3    Plot of Concentration Versus Volume for a
     Regeneration in Which Portions III and IV from
     The Previous Regeneration Were Re-used as
     Portions I and II                                              9

k    Curve Showing the Change in Exchange Capacity
     With Time                                                     11

5    Illustration of the Treatment System                          22

6    Illustration of the Comparison Between the
     System Influent and the Clear Effluent                        24

7    Illustration of the Regeneration System                       26

8    Illustration of the System of Interlocks                      29

9    The Arrangement of the Sock Filters                           31

10  Illustration of pH Loop #1 Which Controls the
     Influent to the Resin Columns                                 32

11  Illustration of pH Loop #2 Which Controls the
     Column Effluent                                               34

12  Illustration of the Level Probe in the Zinc
     Precipitation Tank                                            37

13  Illustration of the Temperature Control System                38

14  Illustration of the Pressure and Vacuum
     Controls on the Resin Columns                                 4Q

15  A Copy of a Typical Strip-Chart Recording for
     The System During Treatment                                   44

16  Illustration of the Resin Column With the Sand
     Support Bed                                                   49

                                   vi

-------
                          FIGURES (CONTINUED)

No.

1?   Illustration of the Resin Column After Removal
     Of the Support Bed and Inversion of the Bottom
     Diffuser                                                     5I

18   Plots of pH and Bichromate Concentration Versus
     The Volume of Regenerant Solution Passed                     59

19   Plots of pH and Bichromate Concentration Versus
     The Volume of Regenerant Solution Passed.
     Excess Alkali Was Added in the First 300 ml                  61

20   Plots of pH and Bichromate Concentration Versus
     The Volume of Regenerant Solution Passed.
     Excess Alkali was added to a Partially
     Exhausted Column                                             63

21   Plot of the Concentration of Potassium Chloride
     Solutions and Their Specific Gravities                       91
                              vti

-------
                                TABLES

Ifo.                                                               Page
tmm^f^                                                                ^™^^^^^™^^

1    1969 Effluent Specifications                                   5

2    Estimation of Bichromate Exchange Capacities
     For Pour Different Resin Samples                              12

3    Preliminary Data on the Effect of Influent pH                 13

k    Effect of Influent pH on the Capacity of the
     Ion-Exchange Resin to the Point Where Bleeding
     Occurs                                                        13

5    Effect of Influent pH on the Capacity of the
     Ion-Exchange Resin at Exhaustion                              14

6    Effect of Influent pH on the Concentration of
     Dichromate In the Recovered Regenerant Solution               14

7    The Influent Concentration, The Volume of
     Influent and the Quantity of Dichromate Required
     To Cause a 100 ml Resin Column to Bleed                       15

8    The Volume of Influent and the Quantity of
     Dichromate Required to Exhaust a 100 ml
     Resin Column                                                  16

9    The pH and the Dichromate Concentration of
     Samples of Regenerant Passed Through An
     Exhausted Resin Column to Study the Regeneration
     Process                                                       58

10   The pH and the Dichromate Concentration of
     Samples of Regenerant When Excess Alkali is
     Added Initially                                               60

11   The pH and the Dichromate Concentration of
     Samples of Regenerant When Excess Alkali is
     Added and the Column, Initially, is Only
     Partially Exhausted                                           62

12   The Sodium Dichromate Concentrations of Each
     of the Four Regenerant Portions Used to
     Regenerate the Rohm & Haas Resin                              54


                                  viii

-------
                           TABLES (CONTINUED)

No.

13   The Sodium Bichromate Concentration of Each Of
     The Pour Regenerant Portions Used to Regenerate
     The Dow Resin (Original Laboratory Sample)                    65

14   The Sodium Dichromate Concentration of Each of
     The Four Regenerant Portions Used to Regenerate
     the Dow Resin (New Plant Sample)                              66

15   Chromium Analyses for the Four Regenerant Portions
     Used to Regenerate the Dow Resin 3/12/70                      67

16   Chromium Analyses for the Regenerant Solutions
     Used to Regenerate the Dow Resin 3/13/70                      68

17   An Accounting for the Influent Dichromate in the
     Recovered Regenerant                                          69

18   Influent analyses from Nov*, 1971                             71

19   The weight of Sodium Dichromate Recovered from
     the Resin Columns, a Measure of Exchange Capacity             72

20   Typical Results of the Potassium and Dichromate
     Concentrations of Regenerant Solutions                        73

21   Comparison Between Influent and Recovered
     Dichromate in Plant Operation                                 74

22   A Comparison of the Dow and Rohm and Haas Resins              75

23   A Summary of the Recovery from the Ion-exchange
     System Over a Two-Year Period                                 7 6
                                    IX

-------
                            ACKNOWLEDGMENTS
We would like to acknowledge the kind assistance of Mr.  Sam Griggs,
formerly of the Charlottesville, Va. Office of E.P.A., our first project
officer; Mr. John Ciancia, Mr. Bernard Hornstein and Mr. Richard Taba-
kin, all of the Edison Water Quality Research Laboratory, Edison, N.  J.
who have worked closely with us in bringing the project  to a successful
completion.  We would like to thank Mr. Rodney L. Cast,  Dr. Jesse
Williams, Mr. Charles Hodge and Mr. David L. Seal for their careful
laboratory work, Mr. Donald W, Agee and Mr. John D. Meininger for their
assistance in wiring and instrument installation, and Mr. John Phillips
for his assistance in preparing many of the drawings.

We acknowledge the help of Mr. Charles T. Dickert and Mr. Robert Kunin
of Rohm and Haas Co. and Mr. James C. Hessler, Mr. Tony  Diblik and
Mr. William Ward of Nalco Chemical Co. all of whom made  many helpful
suggestions while the plant was in its design stages. Also Mr. Robert
Gerster of Jacoby-Tarbox who helped in the selection of  an appropriate
filter to detect the sodium dichromate color and for his determination
of the path length required in the colorimetry to give the required
1 ppm sensitivity.

Our pleasure in all of this is diminished by the fact that Dr. Harold
£. Weisberg and Mr. James L. Capper both of whom died before the pro-
ject's completion cannot share in this satisfaction.

-------
                              CONCLUSIONS
1.  Either of two strongly basic ion-exchange resins can be used suc-
    cessfully to reduce the chromate content of pigment plant effluent
    from 2JOO ppm to less than one ppm of total chromium expressed as
    Cr.

2.  The ion-exchange resins can be regenerated with an alkaline salt
    solution.  In this plant, the salt used, potassium chloride, is a
    raw material in manufacture of zinc yellow pigment.  It is "borrow-
    ed" long enough to accomplish the regeneration; then it is returned
    to the manufacturing unit, along with the recovered dichromate,
    for re-use*

3*  The recovered chromate solution can be used in the manufacture of
    some chromate pigments without degradation of the product quality.
    This, of course, is the key to the complete success of the plant
    operation.

4.  This plant has been designed so that steps in the treatment and in
    the resin regeneration cycles follow automatically.  These steps
    are quite dependably controlled by color, tank level, time, and pH.

5»  Operators are required only to make up solutions, clean filters,
    answer alarms, and occasionally differentiate between color and
    turbidity as seen by the colorimeter.

-------
                            RECOMMENDATIONS
This plant has been limited to treatment of waste waters recovered from
the manufacture of zinc potassium chromate.  The waste water contains
zinc and chromate ions and high concentrations of sodium chloride.  The
ion exchange resin, in the chloride form, is successful in removing
chromate in exchange for chloride.  Other chromate pigment wastes might
contain sodium sulfate, sodium nitrate, or sodium acetate.  Additional
laboratory work would be necessary to demonstrate whether the resin will
adsorb chromate preferentially in competition with high concentrations
of these ions and whether an appropriate use could be found for the re-
covered chromate solution.

Were we to start anew, we would consider having the bottom diffuser
openings through the tank bottom with the support screens external and
with each opening individually valved.  This would enable one to make
screen repairs without removing the resin from the ion-exchange column
and would make a uniform up-flow during backwash more easily attainable.
It does present the possibility of an inactive area on the tank floor
which would collect exhausted resin.  Some further study of the design
would be helpful.

More work with level sensing devices will be required.  We have eval-
uated the applicability of a number of conductance and capacitance
level probes and for some uses we have not overcome their limitations.
A study of sonic probes is in process.

The polyester screens used in the diffuser openings are quite succes-
ful in retaining the resin beads while permitting uniform flow.  Their
most serious limitation is that they fatigue with continued flexing and
they eventually rupture.  Monel cloth is currently being evaluated.
This system does require periodic maintenance and merits further study.

-------
                             INTRODUCTION
Chromate, a deep yellow-colored, ion, in dilute water solution, is a
necessary by-product of chromate pigment manufacturing.  The commer-
cial manufacture of chromium containing pigments can be represented by
chemical reactions such as equations (l) and (2).  Equation (l)

           + S  -» Cr203 + N^SO^                                (l)


           + 2MO + 2HY   -»  2MCrO^ + 2NaY + H20                 (2 )

represents an over-simplified version of the reduction of sodium di-
chromate by sulfur (carbon could also be used) to form the green pigment
chromium oxide and a water soluble salt, which in this case is sodium
sulfate.  Equation (2) represents the preparation of the yellow chromate
pigments.  M is a heavy metal such as lead, zinc, barium, strontium, or
calcium.  Y is an anion such as chloride, nitrate, or acetate.  Lead
and barium chromates are only very slightly water soluble, but stron-
tium, calcium, and zinc chromates exhibit an appreciable solubility in
water.  Because of this solubility, or because the reactions as depicted
by equations (l) and (2) above may not proceed quantitatively to the
right, the filtrates and wash liquors from the precipitation and recov-
ery of these pigments contain chromate ions together with the salts sod-
ium chloride, sodium sulfate, sodium nitrate, or sodium acetate.

Waste waters containing chromium in excess of 0.5 ppm may not be dis-
charged into surface streams.  The current practice is to collect these
filtrates and wash liquors and either dispose of them in deep wells-*- or
to acidify and reduce the resulting dichromate solution with sulfur
dioxide or sodium sulfite.  The chromium is then precipitated as a hyd-
rous oxide and is discarded in a land-fill area.  The overall reaction
of this latter method may be represented by equation (3).  This practice
                             (2+x)HgO  -» 2Cr(OH)3.xH20 + UNagSO^  (3)

is unsatisfactory, first because it is wasteful, secondly because the
hydrous chromium oxide is a light bulky precipitate which does not
settle readily.  The resulting supernatant solution is often turbid.
A third problem is that, in the presence of acetate or chloride ions,
coordination compounds are frequently formed which do not precipitate
to yield a water-white effluent.  Under these conditions the pigment
plant effluent is frequently green or pale blue-green and contains color
and chromium in excess of statutory limits.

-------
We have proposed and will describe here a water treatment plant which
will remove chromate from plant effluent liquors.  This plant utilizes
ion-exchange resins which exhibit a preference for dichromate over the
other anions present in the waste water.  The dichromate is adsorbed on
the ion-exchange resin and may be recovered from the resin and re-used
in the pigment manufacturing process.

The chromate ion exists as a monochromate.  On acidification it is con-
verted to chromic acid (monochromic acid.)  Two moles of chromic acid
can then lose a mole of water and become a dimer or dichromic acid.

          2H2Cr04 -

This process can continue, forming trichromic acid, tetrachromic acid,
etc.2  We will confine our discussion to chromate ions, existing in
neutral or alkaline solutions, and to dichromate ions, existing in acidic
solutions.  We will report all chromium analyses as sodium dichromate,
dihydrate (NagCrgOj^HgO), for ease of intercomparison, even though no
dichromate ion may be present in, for instance, an alkaline regenerant
solution or a neutral filtrate from the zinc yellow pigment manufactur-
ing operation.

The adsorption of dichromate on the resin is favored by low pH condi-
tions and is represented in equation (5)«  The recovery of the chromate
is accomplished at a high pH and may be represented as in equation (6).
                + 2ResiaCl  -> ( Resin )2Cr2Oy + 2C1               (5)

     (Resin )2Cr20T + 20H~ + 2Cl"  -» 2Resin Cl  + 2Cr04~ + HgO   (6)
The original work on this process as reported by Hesler3 and by Hesler
and Oberhofer*'? was generally limited to chromate concentrations of
30 to 100 ppm such as result from the discharge from cooling towers or
from boiler blow-down.  Our initial interest was to determine whether
this ion-exchange resin would withstand dichromate concentrations up to
2700 ppm and to determine whether the adsorption would be specific for
dichromate in the presence of much higher concentrations of other alkali
metal salts.

Other questions were immediately apparent.  At what concentration can
the recycled chromate be recovered?  Does the recovered solution, which
is of necessity high in chloride ion concentration, affect the color
or chemical properties of the pigment into which it is recycled?  Can
the ion-exchange resin be repeatedly regenerated in such a manner as to
always yield a water-white effluent, or will some further treatment
procedure, e.g., reduction-precipitation, be required?

When we discuss the chromium content of treated waste water we will ex-
press the concentration as ppm of chromium (Cr) and will understand
that this is total chromium in all oxidation states, and in water sol-
uble as well as insoluble forms.

-------
At the time of our grant application the effluent specifications were
as listed in Table 1.  Since that time there has been considerable dis-
cussion that the total chromium content of industrial effluents should

                 Table 1.  1969 EFFLUENT SPECIFICATIONS
Suspended solids
Turbidity
Color
B.O.D.
PH
Oil
Cr
kOO ppm max
300 ppm max
400 ppm max
100 ppm max
5.5 to 8.5
30 ppm max
1 ppm max
be under 0.5 ppm,total chromium and perhaps as low as 0.05 ppm of hexa-
valent chromium.

While the process to be described here has been shown to be satisfac-
tory to attain the 1.0 ppm chromium specification, some post treat-
ment would be required to attain the 0.05 ppm level.  An effluent of
0.05 ppm would require an extremely complete regeneration of the ion-
exchange resin.  Also, our control step to signal the need for second-
ary treatment (a second column in series) is predicated upon the appear-
ance of color in the effluent.  The appearance of color is equivalent
to a concentration far in excess of 0.05 ppm of hexavalent chromium.

In this report we will describe the laboratory studies which were under-
taken to answer the above questions regarding the resin performance and
regarding the properties of the recovered chromate solution.  We will
describe the process of scale-up from the bench unit to our final plant
design.  We will describe the design of the water treatment plant and
the details of pH, color, flow and level control.  We will describe our
experience in the first twelve months of plant operation and give in-
dications of the equipment costs and the prospects for amortizing this
expense.

-------
                         LABORATORY EVALUATION
HIGH BICHROMATE CONCENTRATIONS

The original work on dichromate removal by ion-exchange was generally
limited to chromate concentrations of 30 to 100 ppm.  Our first inter-
est was to discover whether this ion-exchange resin would withstand
concentrations up to 2700 ppm, and to discover whether the adsorbtion
would be specific for dichromate in the presence of much higher concen-
trations of other alkali metal salts.

We obtained a sample of Dowex "SBR" resin from Nalco Chemical Company?
and tested it on filtrates from our zinc yellow and our chromium oxide
processes.  The influent to the resin contained 2.7 grams of
NajiC^Oj^I^O/ liter.  The effluent from the resin was water-white
or pale yellow, depending on the rate at which the filtrate was allowed
to flow through the resin column.  In these preliminary or survey tests
we showed that chromate concentrations as high as 2700 ppm could be ad-
sorbed by the resin nearly quantitatively, as determined visually.

In the laboratory work which followed we at one time exposed the resin
to 11,000 ppm dichromate and later in plant operation the resin was
accidentally exposed for a brief period to 15,000 ppm dichromate with
no apparent damage.
RESIN SPECIFICITY

The initial survey was extended to include filtrates which contained
calcium and strontium as well as zinc as heavy metals, and nitrate and
acetate in addition to chloride and sulfate anions.  As long as the pH
was maintained in the 3 to k region, the heavy metals presented no prob-
lem and the resin showed an apparent specificity for dichromate over all
of the other anions tested.  The balance of this report will, however,
be limited to a discussion of our experience with zinc as the heavy metal
cation and chloride as the competing anion.
RESIN REGENERABILITY

We next attempted to study the regeneration process.  An exhausted col-
umn was flushed with water, then covered with the recommended"

-------
regenerant solution.  Small portions of the regenerant were removed every
fifteen minutes.  The pH of each sample was recorded and an aliquot por-
tion was analyzed for its chromate content iodometrically.  The analyt-
ical procedure for the chromium analysis and the details of this study
are reported in Appendices A and B respectively.  The maximum concentra-
tion of chromate in the recovered solution was not obtained until the
pH of the regenerant solution exceeded 8.5 after it was passed through
the dichromate-containing resin.  The pH of the recovered solution then
rose sharply from that pH, indicating that the dichromate on the column
had apparently been neutralized, or converted to chromate, and the con-
centration of chromate in the recovered solution then reached its maximum
value and dropped quite sharply.

Starting with freshly prepared regenerant solution and causing the re-
generant to flow through the column in a single pass one will obtain
a concentration relationship of the general shape illustrated in Fig-
ure 1.  Iff however, one recirculates the alkaline regenerant through
the resin until the acidity of the dichromate ion which has been ad-
sorbed on the resin has been neutralized one can displace the curve
toward the left axis as illustrated in Figure 2.

Further, now, if one re-uses volumes III and IV from one regeneration
as volumes I and II, respectively, of the next succeeding regeneration
one can obtain a relationship as shown in Figure 3*  In this manner
we succeeded in maximizing the concentration of the recovered chromate
solution for re-use and yet held the total regenerant volume and the
time required for satisfactory regeneration to a minimum.

By placing a second resin column in series with the first to collect
the dichromate which bleeds from the first column it is possible to load
the first column more completely.  As will be shown in a later section
in which two different ion-exchange resins are compared, different resin
samples gave markedly different exchange capacities (even fresh, unused
resin samples) ranging from 8.8 to 20 g of sodium dichromate per 100 ml
of the resin.  The consequence of this variation in exchange capacity
is to increase or decrease the time interval between successive regen-
erations depending upon which resin is in use during the treatment cycle.
RESIN LIFE

Our next endeavor was to make some estimate or projection of resin life.
Resin suppliers were pessimistic and were predicting as much as one-
third mortality of the resin each year if we exposed it to 2700 ppm
dichromate at a low pH.  As will be shown later, our experience is not
nearly that bad, and is closer to a 15# replacement over a two year per-
iod.

From the literature the Rohm & Haas "macroreticular" resin with a
"stronger backbone" appeared to offer the best chances for success.9

-------
             10 -
 CONCENTRATION
              8 .
              6 -
     100 ml
              2 -

              0
                 0
           100
200
300
kOQ
                              VOLUME OF REGENERANT
                                       ml
  FIGURE 1       TYPICAL CURVE OF  CONCENTRATION OF BICHROMATE IN
                 REGENERANT VERSUS VOLUME FOR ONE PASS«
             10  -
 CONCENTRATIO:
              6  -
g
     100 ml
              2

              0
                 0
           100
200
300
                              VOLUME OF REGENERANT
                                       ml
  FIGURE 2
TYPICAL CURVE OF CONCENTRATION OF BICHROMATE IN
REGENERANT VERSUS VOLUME WHEN THE INITIAL VOLUME
IS RECYCLED.
                       8

-------
0                100               200               300              1*00

                          VOLUME OF REGENERANT
                                   ml

FIGURE 3    PLOT OF CONCENTRATION VERSUS VOLUME FOR A REGENERATION IN
            WHICH PORTIONS III AND IV FROM THE PREVIOUS REGENERATION
            WERE RE-USED AS PORTIONS I AND II

-------
During eight months of testing the Nalco (Dow) resin was tested for
twenty-six cycles, and the Rohm & Haas resin for sixty-five cycles of
exhaustion and regeneration.  Figure U is a plot of grams of sodium
dichromate per "100 ml" of resin versus cycles or time.  The 100 ml is
in quotes because it was discovered after a period of time (26 to 37
cycles) when the volume was remeasured that we had only from 81.5 to
96.0 ml of resin remaining in each of four resin columns.  Part of this
discrepancy was from mechanical loss (i.e., carrying the resin beads out
of the column during backwash.  These beads were visible in the retained
backwash liquors.) and part of it is probably the result of a predicted
resin breakdown by contact with the relatively high concentrations of
chromium (VI) solution at low pH.

If the volume of resin lost is replaced (or if the capacity is calcu-
lated on the basis of the actual volume present) the resin retains essen-
tially its original capacity for chromium.

Throughout this report exchange capacities will be expressed in terms of
the volume of ion exchange resin present,, that is, as the weight, in
grams, of sodium dichromate which is adsorbed on 100 ml of resin, or,
in the plant operation, as pounds per cubic foot.  It should be pointed
out, however, that this volume is not a very precise parameter, first,
because the resin particles are small spheres and a given mass may occupy
a greater or lesser volume depending upon how closely packed the spheres
are, one to the other.  Also, the volume changes by as much as 20% de-
pending on whether the resin is completely regenerated (in the chloride
salt form) or completely exhausted (in the dichromate, or polychromate
salt form.)
THE ION-EXCHANGE RESIN

There are a number of suppliers of ion-exchange resins.  We began our
laboratory study with two and have, arbitrarily, limited our study to
a comparison of these two; Dow Chemical's Dowex 1X8 and Rohm and Haas1
Amberlite IRA-900C.  These are both strongly basic anion exchange resins.
They are polymers of styrene, cross-linked with divinyl benzene and
their functional groups are quarternary ammonium groups.9> 10

The Dowex 1X8 was reportedly the "SBR" resin used by Nalco Chemical
Company in their Patent 3,223,620 and which was shown by them to have a
specificity for dichromate over many other anions.

The Rohm and Haas Amberlite IRA-900C is a macroreticular (or sponge-like)
resin with larger pores and a stronger "backbone", which is advertised
to be less prone to mechanical breakdown.  Mechanical breakdown results
in an accumulation of fine particles which yield a high pressure drop
across the resin bed, and a loss of exchange capacity.
                                   10

-------
±  18-
tn
ui
ocr
,,  17-1
s  16^
O  14
 vC
i
oi
 •  13
O
   12-
   9-


   S -



   T
                                                 DOW (NEW PLANT  SAMPLE)
                                                                     DOW (LAB SAMPLE;
             ROHM & HAAS
•     i     i
10    It    14
18    20

CYCLES
                                                           i
                                                          Zt
 i
24
 i
26
i
28
 i
30
 i     i
34   36
38
      Figure h    CURVES SHOWING THE CHABGE IN EXCHAMGE CAPACITY WITH TIME

-------
From our initial laboratory tests it was apparent that there were advan-
tages and disadvantages to both of the resins.  The Dowex 1X8 had a
larger capacity, however, the beads were somewhat gellular and more prone
to mechanical damage.  The Amberlite IRA-900C, with its lower capacity,
would have to be regenerated more frequently in our plant operation,
(approximately every lU to 18 hours as compared with 20 to 2k hours for
the Dowex 1X8,) but it appeared to regenerate and rinse with a little
less care and time than the Dowex 1X8.  We elected to continue our com-
parison of the two resins into the plant operation by installing the Dow
resin in one of the columns and the Rohm and Haas resin in the other.
THE CAPACITY OP THE ION-EXCHANGE RESINS FOR DICHROMATE

In Table 2 we have recorded the data from our estimation of the capacity
of four different resin samples to adsorb sodium dichromate.  It can be
seen from this table that there is considerable variation in the exchange
capacity of the resin from sample to sample.  Also, the capacity of the
Dow resin is significantly greater than the capacity of the Rohm and
Haas resin.
Table 2    ESTIMATION OF DICHRCMATE EXCHANGE CAPACITIES FOR FOUR DIFFERENT
           RESIN SAMPLES
Resin Identity
Sample Source
Resin Volume (ml)
 Cycle

   1
   2
   3
   4
   5
Average
Capacity
  g/100 ml
Ibs/cu ft
     IRA-900C
Lab      Plant Stock
   81.5       96.0
         grams of
 8.84
 9.18
 9.86
 9.18
 9-18
 9-25

11.35
 7.08
              12.07
              12.56
              11.73
              11.90
              11.73
              12.00

              12.50
              7..80
                               Dowex 1X8
                            Lab       Plant Stock
                              89.0         93.0
                              adsorbed
13. &
12.92
12.58
12.92
12.92
13.06

14.67
 9.16
18.19
17.85
19.38
19. ou
19.21
18.73

20.14
12.57
EFFECT OF INFLUENT pH

The conclusions drawn from this work were;  l) that the lower the pH of
the influent, the greater was the capacity of the rssin for chromium,
2) the lower the influent pH the greater the permissible flow rate to
obtain satisfactory adsorption by the resin, 3) the lower the influent
pH the lower would be the chromate content (the color) of the total
averaged effluent.
                                   12

-------
Table 3     PRELIMINARY DATA ON THE EFFECT OF  INFLUENT pH

Influent          Volume            Flow rate      Color of Averaged
   pH          of influent                             Effluent
                    nil               ml/min          Taylor Scale

  1.10             1500               335               eoo 2oo <3oo
  3.10             1000               1U3               >3oo
  3-^0             1000               1^5               >300
  3.80             1000               1^3               >Uoo
Most of the laboratory work which immediately followed the above study
was done with an influent pH of 2.0 or 2.5.  Later, as we began to specify
materials of construction for the plant, this low pH of the influent pre-
sented problems.  Together with the high sodium chloride content, the
low pH ruled out the use of stainless steel.  Most of the available flow
meters, filters and multi-port valves were not available in materials
which would withstand the low pH, high cloride exposure as well as the
high pH of the alkaline regenerant solutions.  We extended this study
with more tests to evaluate the effect of the influent pH and obtained
the results as shown in Table 4-6.

From this latter series of tests we concluded that a satisfactory resin
performance could be obtained with an influent pH of 3-0 and only when
the pH of the influent approached U.O did the capacity at bleed (the point
where secondary treatment becomes necessary), the quantity of dichromate
required to exhaust a column and the concentration of the recovered re-
generant solutions begin to decrease significantly.  From these findings
we were then able to specify stainless steel as satisfactory for many of
the contact surfaces in the final plant design.


Table h    EFFECT OF INFLUENT pH ON THE CAPACITY OF THE ION-EXCHANGE
           RESINS TO THE POINT WHERE BLEEDING OCCURS

Resin Identity                     IRA-900C                 Dowex 1X8
Influent                               Average Capacity At Bleed
   pH                             grams Na2Cr207.2H20/100 ml of resin

   2.5                                9-72                    19.56
   3.0                                9.21                    19.65
   3.5                               10.09                    18.1*6
   k.O                                8.85                    18.28
                                   13

-------
Table 5     EFFECT OP INFLUENT pH ON THE CAPACITY OF THE ION-EXCHANGE
            RESINS AT EXHAUSTION

Resin Identity                      IRA-900C                Dowex 1X8
Influent                             Average Capacity At Exhaustion
   pH                              grams NaaCrgOj.aHaO/lOO ml of resin

   2.5                                12.50                   20.13
   3.0                                12.66                   21.57
   3.5                                12.66                   21.12
   k.o                                11.69                   19.92
Table 6     EFFECT OF INFLUENT pH ON THE CONCENTRATION OF BICHROMATE
            IN THE RECOVERED REGENERANT SOLUTIONS

Resin Identity                      IRA-900C                Dowex 1X8
                               Concentration of Combined Portions I and II
Influent                          Of The Recovered Regenerant Solutions
   pH                           grams of Na2Cr2Oy. 2^0/100 ml of solution
   2.5                                 5-85                    8.75
   3.0                                 6.08                    8.8U
   3.5                                 5.63                    8.71
   U.o                                 5.27                    8.21
 CONCENTRATION OF DICHROMATE IN THE INFLUENT

 Samples of the untreated zinc yellow plant effluent were acidulated to
 an  appropriate pH value and analyzed iodometrically for total chromium(Vl).
 The results of some of these analyses are reported in the second column
 of  Table 7.  Generally the concentrations, expressed as parts of sodium
 dichromate dihydrate per million, vary between 1600 and 2700 ppm.  The
 variations are due in part to variations in the properties of the zinc
 oxide which is used in the pigment manufacture.  Variations are also due
 to  failure of the product recovery filters to consistently function
 properly to remove all of the suspended pigment from the filtrate.  The
 reading of 11,000 ppm reported on line 3 of column 2 in the table was
 the result of such an abnormal loss.

 THE VOLUME OF INFLUENT AND THE QUANTITY OF DICHROMATE REQUIRED TO CAUSE
 A COLUMN TO BLEED

 In  our laboratory evaluation of the ion-exchange resins we recorded the
 volume and the dichromate concentration of the influent which was re-
 quired to cause a 100 ml resin column to "bleed" the yellow dichromate
 color.  During our earliest work little attention was paid to a faint
 yellow coloration in the effluent from the column.  Later, when we were
 insisting on a water-white, "dichromate-free" effluent, the first

                                   14

-------
visible coloration was recorded as the "bleed" volume.  This is the point
at which secondary treatment, or treatment through a second resin column
in series would be required.
Table 7
Sample
Number

   1
   2

   h
   5
   6
   7
   8
   9
  10
  11
  12
  13
  14
  15
  16
THE INFLUENT CONCENTRATION, THE VOLUME OF INFLUENT AND THE
QUANTITY OF BICHROMATE REQUIRED TO CAUSE A 100 ml RESIN
COLUMN TO BLEED
   ppm Na2Cr2Oj.2H20
    In The Influent
          2210
          2300
        11,000
          1600
          2000
          1900
          1900
          1600
          1600
          2700
          1500
          2000
          2100
          1900
          1700
          1900
 Volume
To Bleed

  6000
  5400
   900
  6000
  6000
  4800
  5400
  7800
  5600
  44oo
  7200
  6200
  6600
  6000
  6200
  6100
Weight Of
  13.26
  12.42
   9.90
   9.60
  12.00
   9.12
  10.26
  12.48
   8.96
  11.88
  10.80
  12. 40
  13.86
  11.40
  10.5^
  11-59
Average               1927*              5980              11.28
      * This average was determined omitting the 11,000 ppm reading.
The volume of influent of a corresponding concentration which is required
to cause a bleed is reported in column 3 of Table 7.  As might be pre-
dicted, when the concentration is high, the volume to bleed is low and
vice versa.  Because of the variation in influent concentration, one
could not expect, for instance, to run a fixed volume of influent through
the column and then initiate secondary treatment.  To do so would result
in other than a satisfactory or optimum performance.

The quantity of sodium dichromate in grams required to cause a 100 ml
resin column to bleed is reported in column k of Table 7.  This value
is somewhat more reproducible than the volume to bleed.

THE VOLUME OF INFLUENT AND THE QUANTITY OF DICHROMATE REQUIRED TO EX-
HAUST A RESIN COLUMN

Similar to the above study we observed the volume of influent and the
quantity of dichromate required to "exhaust" each resin column.  This
                                   15

-------
"exhaust" point was even more subjective because it was determined vis-
ually by comparing the color intensities of the influent and the efflu-
ent from the column.  Representative values are reported in Table 8.
Table 8
Sample
Number

   1
   2
   3
   4
   5
   6
   7
   8
   9
  10
  11
  12
  13
  14
  15
  16
THE VOLUME OF INFLUENT AND THE QUANTITY OF BICHROMATE
REQUIRED ,TO EXHAUST A 100 ml RESIN COLUMN
   ppm
    In The Influent

          2210
          2300
        11,000
          1600
          2000
          1900
          1900
          1600
          1600
          2700
          1500
          2000
          2100
          1900
          1700
          1900
  Volume
To Exhaust

   7200
   8500
   1400
   7900
   7400
   7500
   9200
 10,300
   7200
   6300
 10,700
   9400
   8200
   7900
   8300
   8200
Weight Of
  15.91
  19-55
  15.40
  12.64
  14.80
  14.25
  17.^8
  16.48
  11.52
  17-01
  16.05
  18.80
  17.22
  15.01
  14.11
  15.58
Average               1927*              8280              15.74
      * This average was determined omitting the 11,000 ppm reading.
CONCENTRATION OF DICHROMATE IN THE RECOVERED REGENERANT

After considerable study of the regeneration process we adopted a pro-
cedure in the laboratory of using four 100 ml portions of regenerant solu-
tion to regenerate a column which contained 100 ml of the ion-exchange
resin.  The first two portions of regenerant which contained the highest
concentrations of dichromate would be recycled to manufacturing.  Por-
tions three and four from one regeneration would become portions one and
two respectively for the next regeneration.  Each portion of regenerant
was analyzed for its chromium content.  Typical results of this work are
listed in Tables 12, 13, and 14 of Appendix B.

From 16 such cycles with the Rohm and Haas resin the average composition
of the first portion was 6.71 g of sodium dichromate per 100 ml and for
the second portion 4.14 g per 100 ml.  Since in the plant we will be com-
bining portions one and two, we can expect an average composition of
5.43 g per 100 ml from regeneration of the IRA-900C resin.
                                   16

-------
Similarly, from the original laboratory sample of the Dow resin we ob-
tained a dichromate concentration of 9-1^ g/100 ml for the first portion
of regenerant, 5.1*5 g/100 ml for the second portion and the two portions
combined can be expected to give a concentration of 7.30 g/100 ml.

A sample of Dow resin taken from the plant stock gave somewhat higher
results, namely 10.67 g/100 ml, 6.73 g/100 ml and an average composi-
tion of 8.70 g of sodium dichromate per 100 ml.  We will be comparing
these values with those obtained in our plant operating experience later
in the report.
AN ACCOUNTING FOR THE DICHROMATE INTRODUCED IN THE INFLUENT

Inasmuch as all of the dichromate which was introduced to a given column
was picked up either by that column or by the second column in series,
no dichromate should be lost from the laboratory system.  All dichromate
in the influent should be accountable either in the regenerant, in the
rinse water or in the back-wash water.  Again, since regenerant portions
three and four from a given regeneration become portions one and two for
the next regeneration cycle we must concern ourselves with the increase
in dichromate content in these re-used portions, not with their total
dichromate content.  In Appendix B we have given the details of this
study and in Table 17 we have accounted for the total dichromate content
in the influent, the final less the initial dichromate content of regen-
erant portions one and two, and the dichromate content of portions three
and four, the rinse water and the back-wash water for 58 such laboratory
cycles.  From this study we succeeded in accounting for 99»5# of the
influent dichromate.
                                   17

-------
                               SCALE-UP


Following the laboratory evaluation and the decision to proceed with
the plant design and construction, it was necessary to determine a
number of parameters for scale-up.

On completion of the laboratory determination of the resin capacity and
the rate of chromate loss in the pigment waste water, it was necessary
to decide how long it would take to regenerate, or how frequently we
wished to regenerate.  We elected to regenerate no more frequently than
every twelve hours, both to decrease the amount of laboratory analytic-
al work and to make the regeneration coincide with our manufacturing
schedule.  This determined the size of the resin columns, the size of
the regenerant tanks, and ultimately the sizes of the pipes, pumps,
flow meters, and filters.

Another requirement was to determine the rate at which concentrated hy-
drochloric acid would have to be added to the influent to maintain a pH
of 3*0 and the amount of 20$ (weight/volume) sodium carbonate solution
which would be required to continuously adjust the effluent pH to 9*0
to precipitate the remaining zinc ions so that they could be removed by
filtration before the waste water was finally discharged to a nearby
surface stream.

Other requirements were the length of the light path necessary to color-
imetrically detect one ppm of chromium present as the dichromate ion
and the size of the heat exchanger necessary to raise the regenerant
temperature to 50°C at a rate of 50 gal/min to the resin column*
                                   18

-------
                             PLANT DESIGN
GENERAL CRITERIA

There were a number of factors which influenced the design of the treat-
ment plant.  Among these were:

1.   The exchange of anions is an equilibrium process.  At any cross-
     section in the tower during treatment there is an equilibrium be-
     tween the dichromate and chloride concentrations adsorbed on the
     resin and the dichromate and chloride ion concentrations present
     in the water solution.  As the solution passes downward to a lower
     level in the column it contacts resin which contains a lesser con-
     centration of adsorbed dichromate and conditions are again favor-
     able for a further exchange.  Here the resin again adsorbs an addi-
     tional quantity of dichromate from the waste water and gives up to
     the solution an equivalent quantity of chloride ion.

2.   Exhaustion or saturation of the resin is accompanied by a twenty
     percent decrease in the resin volume and a corresponding increase
     in the particle density.  During backwash or when the resin is sus-
     pended in the regenerant solution these more dense, exhausted resin
     beads fall to the bottom of the column.  This is the main argument
     for up-flow regeneration.  In our experience, up-flow regeneration
     is only successful when the flow is completely uniform over the
     entire floor of the resin column.  If there are any areas not sub-
     ject to up-flow, these areas tend to collect the exhausted (most
     dense) resin, and the end result is an incomplete regeneration.
     An incomplete regeneration results in the inability to obtain a
     chromium-free water-white effluent on the next use of that column.


3.   Pressures of three to seven psig are required to drive a liquid
     through a four foot deep compact resin bed at reasonable flow rate.
     (l to 3 gal/cu ft/min32)  For this reason the resin columns and the
     column covers must be capable of withstanding these pressures and
     pumps must be capable of overcoming these back-pressures at the re-
     quired flow rates.

k.   The density of the regenerant solution is considerably greater
     than the density of water.  Another problem with up-f low regenera-
     tion occurs  when one attempts to displace the last portion of

                                   19

-------
     regenerant solution by pumping water upward from below.  The  less
     dense water tends to channel unevenly upward  through the resin
     bed and results in an undue dilution or an incomplete removal of
     the alkaline regenerant solution.   Removal by down-flow results in
     an undesirable re-exchange between the most thoroughly regenerated
     resin at the bottom of the tower and the chromate-containing  regen-
     erant which had previously passed  through the column during the
     latter stages of the up-flow regeneration.

5.   In order to obtain a dichromate-free, water-white  effluent, the
     regeneration of the resin must be  complete and uniform.  As in-
     dicated in paragraphs 1. through k. above, this  is an equilibrium
     process and the resin containing adsorbed dichromate is more  dense
     tending   to settle at the bottom  of the column.  Also it  is  neces-
     sary to back-wash the resin bed thoroughly to prevent channeling
     or a compaction of the resin beads with an accompanying pressure
     drop once the system is returned to the treatment  mode.  If,  follow-
     ing an incomplete regeneration, the resin bed is back-washed, the
     most exhausted resin beads are at  the bottom  of  the resin  column
     in contact with the column effluent.  Even if the  solution being
     treated is water-white above this  point, it will,  as it passes
     the incompletely regenerated resin beads at the  tank bottom,  re-
     dissolve dichromate from the resin and the chromium content of
     the column effluent will be above  statutory limits.

6.   To accomplish such a complete regeneration, flow through the  resin
     bed must be uniformly distributed.  Considerable attention to the
     design of both the top and bottom  diffuser assemblies was  necessary
     in order to prevent dead areas, bed inversions and channeling along
     the side walls of the resin columns.

7*   Regeneration proceeds more rapidly at elevated temperatures,  but
     the ion-exchange resin is more vulnerable to  oxidative attack at
     higher temperature^.  Since the regenerant storage tanks  were not
     insulated   rather than maintain the temperature of the regenerant
     solutions at 45° to 50°C continuously, we elected  to pass  the re-
     generant through a heat exchanger  on its way  to  the column to be
     regenerated.  Our data on resin life was obtained  utilizing a re-
     generation temperature of 45° to 50°C and we  have  concluded that
     this temperature is high enough to allow regeneration in a reason-
     able time and not so high as to place the resin  in Jeopardy be-
     cause of thermal or oxidative attack.

8.   The resin is also more vulnerable  to attack and  decomposition when
     it is in the free base (hydroxyl)  form than when it is in  the salt
     (chloride) form.  While the original plant design  had a provision
     for the gradual re-introduction of sodium hydroxide to the regener-
     ant as the base was utilized by the acidic dichromate; this was
     unnecessary.  Again, our resin life statistics which have  been de-
     veloped over two years were determined using  a regenerant  in  which

                                   20

-------
     the required quantity of the sodium hydroxide was added initially.
     This actually resulted in a further cost saying, inasmuch as any
     additional base added had to be re-neutralized with acid before
     the recovered chromate solution could be used in manufacturing
     the zinc yellow pigment.  The raw material required in zinc yellow
     manufacture is sodium or potassium dichromate, not sodium chrom-
     ate.

9.   Salt solutions can easily be made with a density of 1.14 (approxi-
     mately 21% weight by weight KCl), sufficient to float the resin
     beads.  When floated, the resin can be pumped with a diaphragm
     pump without any significant damage to the resin beads.  This tech-
     nique is utilized whenever it is necessary to replace the screens
     in the bottom diffuser, or to make repairs to the diffuser or to
     the tank coating. ^
TREATMENT FLOW

The treatment flow as depicted by equation (5) is illustrated in Fig-
ure 5.  Filtrate and wash liquors from a continuous vacuum filter are
recovered in a vacuum receiver.  They are pumped through a filter press
which removes most of the solid lost in the initial filtration process
and then they are collected in a holding tank.  In the holding tank hy-
drochloric acid is added continuously to adjust the pH to 3*0, the de-
sired treatment pH value.

The acidified waste water is then pumped through a sock-type filter to
remove any foreign matter which might accumulate in the ion-exchange
resin bed, and then through a flow meter to the resin columns.  Initial-
ly, passage through a single column is sufficient to remove all of the
dichromate.  As this first column adsorbs more and more dichromate a
low concentration will begin to bleed from the resin bed.  This bleed-
ing is picked up in what is called "Secondary Treat" by causing the
waste water to flow through a second column, in series, prior to dis-
charge.

All of the effluent from the resin columns passes through the colori-
meter and its color is continuously recorded on a strip chart.  From
the colorimeter the water solution passes to a small, well agitated
tank where sodium carbonate solution is added to pH 9-0.

When the pH of the ion-exchange column effluent is raised to 9.0 the
zinc ions precipitate as a mixed zinc hydroxide-zinc carbonate.  This
white solid is filtered from the effluent solution in plate-and-frame
filter presses.  The filter cake is washed with water to remove adher-
ing sodium chloride and then dried.  The dried powder contains 70 to
12% ZnO.  With the present shortage of zinc oxide the prospects are
good that we will develop a market for this by-product solid to a user
interested in its zinc content.  It does contain low level concentrations

                                   21

-------
  PIGMENT
  SUSPENSION
Ni
                           HYDROCHLORIC
                          ACID ADDITION
               FILTER
               PRESS
JT=P
                             HOLDING
                               TANK
                                #2
                              PUMP
      1  CONTINUOUS VACUUM FILTER
      2  VACUUM RECEIVER
      3  SOCK FILTER
      k  FLOW METER
                           RESIN
                          COLUMNS
                                               SODIUM
                                              CARBONATE
                                              ADDITION
                                                                               FILTER
                                                                               PRESS
                                     CH	HD
                                  COLORIMETER
                                                                      PUMP #8
                                                                                       .EFFLUENT
                                                           ZINC
                                                       RECOVERY
     FIGURE 5    ILLUSTRATION OF THE TREATMENT SYSTEM

-------
of iron and silica and occasionally chromium (ill) as contaminants.

As indicated in the introduction, with satisfactory regeneration of
the ion-exchange resin we can be assured of an effluent which contains
less than 1.0 ppm of chromium.  In the event it becomes necessary to
attain a specification of 0.05 ppm of hexavelent chromium some post
treatment will be required at this point.
TEE ELECTROMECHANICAL CONTROL OF THE TREATMENT SYSTEM

The various functions which are involved with the treatment of the di-
chromate-containing influent with the ion exchange resin and the sub-
sequent regeneration of the ion exchange resin are controlled through
one master electromechanical logic system.  For control purposes the
circuiting is divided into a treatment system and a regeneration system.
The two systems are electrically interlocked such that a column which
is being regenerated cannot be called upon for treatment or secondary
treatment and vice versa.  With the exception of a few common alarms,
the two systems operate essentially independently.

The entire system was devised so that the operator could control the
timing, and to some extent the sequence of events, manually or the sys-
tem could be called upon to follow the programmed sequence of events
automatically.  The system was further designed to include a number of
safety interlocks so that an operator could not, for Instance, attempt
to pump liquid to an already filled tank, or attempt to regenerate with-
out first making up the regenerant solutions by appropriate additions
of caustic soda and potassium chloride to the make-up water.

Treatment is initiated, first, by making a selection for manual or auto-
matic operation,and then by selecting the tank which is to be used for
treatment.  Depressing the appropriate "Treat" button on the control
panel; l) activates the colorimeter and the influent and effluent pH
control loops (which will be discussed more fully in later sections.)
2) starts the influent feed pump and the level actuated pump on the zinc
precipitation tank (Tank 911 in Figure 11) 3) positions the main multi-
port valves to direct the influent through the chosen resin column and
through the colorimeter to the zinc precipitation tank.  The operator is
called upon to adjust the flow to the column with a manually operated
valve.  He adjusts the set points on the colorimeter, cleans and main-
tains the influent sock filters and maintains the quantity of acid in the
storage tank for the influent pH adjustment and of sodium carbonate solu-
tion in the storage tank as required for the effluent pH adjustment and
zinc precipitation.

Secondary treatment, which again can be called for manually or automatic-
ally, is initiated when a yellow dichromate color begins to bleed from
the resin column which is being used for treatment.  If the second resin
column has been regenerated and is ready for use, on secondary treat, the
                                   23

-------
                       n
                                     CLEAR
                INFLUENT
EFFLUENT
Figure 6   ILLUSTRATION OF THE COMPARISON BETWEEN THE SYSTEM INFLUENT AND
          THE CLEAR EFFLUENT
                              24

-------
main multiportvalves will now be oriented so that the flow from the
column which is bleeding will be directed to the inlet of the fresh
column.  The effluent from the latter column is directed through the
colorimeter to the zinc precipitation tank for pH adjustment.  In the
event that the second resin column is not ready for use, the treatment
system will sound an alarm, display an appropriate light on the panel
alarm board and shut down.
REGENERATION FLOW

Regeneration flow as represented by equation (6) is illustrated in Fig-
ure (6).  When a resin column is saturated with dichromate, usually 2
to k hours after the initiation of secondary treat, the system proceeds
to regenerate this saturated, or exhausted, resin column.

Wash water is introduced in order to limit the amount of zinc ion which
might precipitate on the resin surface when the alkaline regenerant
solution is later added;  A visibly low concentration of chromate in the
rinse water has usually proved satisfactory.    That is, removal of
dichromate ion is also indicative that the accompanying zinc ion has
been removed.

The resin column is completely drained to avoid dilution of the regener-
ant solution.  The alkaline salt solution from the first regenerant hold-
ing tank (Tank 901 in Figure 7) is pumped through the bottom of the col-
umn to displace air from the bed and to begin the neutralization of the
acid dichromate which is adsorbed on the resin.  When contact is made
with the top level probe in the resin column, the flow is reversed.  The
regenerant is recycled through the bed for a brief period (3 to 5 min-
utes), then it is pumped to the chromate recovery tank (Tank 91^ in
Figure 7).  When the level in Tank 901 falls below the bottom level probe
in this tank, the valve beneath Tank 901 closes and the valve under
Tank 902 opens.

Four portions of regenerant are pumped in this manner through the resin
bed.  The first two which will be highest in chromate content are pumped
to Tank 9lU where they are re-acidified to the dichromate pH, analyzed
for chromium and potassium content and eventually used in the preparation
of a succeeding batch of zinc yellow pigment.  The second two regenerant
portions are used to refill Tanks 901 and 902 and will be used as the
primary regenerant of the next regeneration cycle.

Water is then flowed through the resin bed to remove adhering regener-
ant and this water is used to refill Tanks 903 and 90^.  Rinsing is
continued until the rinse pH drops below the value of 10.  This value
was obtained simply by observation.  The excess rinse water is returned
to the holding tank (Tank 2) where it will be recycled into treatment.

A dilute hydrochloric acid solution is then pumped through the resin
column to acidify the column and prepare the resin surface for the next

                                   25

-------
                            POTASSIUM CHLORIDE
                                    AND
                             SODIUM HYDROXIDE
                                          HYDROCHLORIC
                                              ACID
     TANK
      901
          TANK
           902
                                        I
                 TANK
                  903
TANK
 904
TANK
 924
                                              PUMP #7
-WATER
                                                                 •AP2
                HOLDING
                  TANK
                   #2
                               TANK
                                 91k
                                                              TANK
                                                               905
TANK     2
TANKS  901

TANK   905
TANK
TANK
AP2
PUMP  #6
PUMP  #7
924
                                          PUMP #6'

IS THE INFLUENT HOLDING TANK
THROUGH 90U ARE USED FOR THE ALKALINE REGENERANT
SOLUTION
IS ONE OF TWO ION-EXCHANGE RESIN COLUMNS
IS THE HOLDING TANK FOR THE RECOVERED CHROMATE
SOLUTION
IS USED FOR DILUTE ACID STORAGE
IS AN ACID PUMP
IS USED TO PUMP REGENERANT TO THE RESIN COLUMNS
IS USED TO PUMP REGENERANT AND RINSE SOLUTIONS
FROM THE RESIN COLUMN TO STORAGE OR RECYCLE
FIGURE 7    ILLUSTRATION OF THE REGENERATION SYSTEM
                                  26

-------
treatment cycle.  The hydrochloric acid solution of necessity by-passes
the stainless steel regenerant flow meter.

Potassium chloride and sodium hydroxide are added by the system opera-
tor to Tanks 903 and 90U to prepare these solutions for the next re-
generation cycle.


THE ELECTRCMECHAHICAL CONTROL OP THE REGENERATION SYSTEM

The regeneration system controls the sequence of all of the events which
must occur in order to recover the dichromate for re-use and to renew
the resin for a future treatment cycle.  It is somewhat more complex
than the treatment system.  Physically it is simply to: l) drain and wash
the spent resin, 2) pump two portions of regenerant from storage tanks
through the resin solumn to a holding tank for re-use, 3) pump two addi-
tional portions of regenerant from the make-up tanks through the resin
column to the intermediate storage tanks, U) wash the column free of re-
generant, 5) reacldify the column and re-fill it with water until it is
required for a future treatment cycle.  The events are physically con-
nected, i.e., the volume from one tank which is being emptied is used to
re-fill a second, and this is used to advantage in the sequencing scheme.

The heart of the regeneration control system consists of three sequencing
relays as illustrated in Figure 8.  One controls in which direction a re-
generant liquor shall be pumped through the resin column (i.e., upflow or
downflow).  The second relay controls from where the regenerant liquor is
being pumped and the third relay controls to where the regenerant liquor
will be pumped.  Each contact position on each of these sequencing relays
operates another relay which, in turn, activates the proper pumps, valve
actuators, level controls, timers, etc., which are necessary to perform
the particular function or combination of functions called for.  The se-
quencing relay is advanced to its next position by an electrical pulse
signal generated by a device which senses the completion of the previous
step.  For example, the second sequencing relay will first open the valve
beneath the first regenerant storage tank (Tank 901 in Figure ?) and al-
low the tank contents to flow to the pump for transfer to the resin col-
umn until a level probe at the bottom of Tank 901 loses contact with the
liquid.  This loss of contact will close a relay to indicate to the sys-
tem that Tank 901 is empty and advance the second sequencing relay to its
next position.  This next position will open the valve beneath Tank 902
for its contents to be used in regeneration.  When Tank 902 is emptied
Tank 903 will be tapped, etc.  During this time the flow from the resin
column is being directed to the holding tank for re-use (Tank 91^ in
Figure 7) by the proper positioning of the third sequencing relay.  When
Tank 91^ is full a level control advances the third sequencing relay to
direct the flow to re-fill Tank 901.  This type of process continues
throughout the regeneration until the sequencing relays have all returned
to their original positions, indicating the completion of regeneration.

The last position on the first sequencing relay turns the regeneration
system off.  In addition to using level probe relays to signal the

                                  27

-------
completion of a particular step in the sequence, adjustable timers are
also used to generate the relay advance pulse.  Relays connected to the
pH sensing system are used to signify when the resin has been satisfac-
torily rinsed to remove the alkaline regenerant and when it has been
properly acidulated at the end of regeneration.  Built within this frame-
work are all of the process function controls, alarms and interlocks re-
quired for satisfactory operation.
FILTRATION

The resin beds would act as filters to collect and accumulate any for-
eign matter present in suspension in the Influent.  To avoid this ac-
cumulation sock filters are Installed in the treatment and in the re-
generation pipelines ahead of the flow meters.  These filters are
arranged to back-wash themselves every 30 minutes, or the operator can
manually initiate back-wash in the event the flow drops below 30 gallons
per minute.  Filtered water from the parallel sock filter is down-flowed
through the blinded sock from the inside and, thus, releases most of
the accumulation.  This small quantity of water is directed to the
plant's reduction-precipitation process for waste disposal, where the
slight amount of foreign matter presents no disposal problem.  The ar-
rangement of the filters is shown in Figure 9-
 pH CONTROL

 There  are  three pH monitors and two pH control loops in the chromate
 recovery system.

 The first  loop is shown in Figure 10.  A sample stream from Tank 2 is
 directed through the pH flow-cell and returned to Tank 2.  The output
 signal from this pH amplifier is directed in series through high and
 low pH alarms, the controller of a precision chemical pump, and a sig-
 nal receiver which in this case is a pH recorder.

 The pH in  Tank 2 is monitored and controlled at the value 3.0.  If the
 pH in  Tank 2 falls below 3.0 the signal amplifier, with proportional
 control, causes the acid pump API to decrease its stroke and deliver a
 smaller quantity of acid to Tank 2.

 Pump U, the cartridge filters, and the flow meter are all fabricated
 in 316 stainless steel, and must be protected against low pH (and high
 chloride ion) exposure.  In the event the pH falls below 2.2, the low
 pH alarm will sound until answered by the system operator.

 Zinc yellow pigment, which would be in suspension in the filtrate and
 wash liquors as they enter Tank 2 at a pH of about 6.3, will dissolve
 completely at a pH of 3.0.  If the pH in Tank 2 is permitted to rise
 above  3.0, conditions for adsorption of dichromate by the ion-exchange
                                   28

-------
ro
                                               COLUMN A
                                               VALVE SELECTION
                                                  TREAT
       Figure 8    ILLUSTRATION OF THE SYSTEM  OF  INTERLOCKS  (SEE KEY ON PAGE 30).

-------
10
o
    The lines of Figure 8 represent "hot" wires to the control system.  The boxes  represent  relays.   In
    some Instances a current signal must be sustained once a given push-button is  released.  The  current
    is sustained through self-energizing relays which are not shown.

    If one pushes, for example, the RA push-button, this energizes the  relay REGENERATE A.   This  cannot
    be done if either the TREAT A or the REGENERATE B relay is already  actuated, because the normally-
    closed contacts of these latter relays will be open.
    Key
RA   Regenerate Column A
RB   Regenerate Column B
TA   Treat Through Column A
TB   Treat Through Column B
SEC  Secondary Treatment, two columns in series.
A/B  Column A Ahead of Column B in series.
B/A  Column B Ahead of Column A in series.

     Push-Button
     Normally Open Contact
     Normally Closed Contact
     Wire Cross-Over With No Contact
     Wire Cross-Over And Contact
     Actuates A Given Relay
    KEY TO FIGURE 8,  THE SYSTEM OF INTERLOCKS
A    Valves and Level Controls for Regener-
     ation of Column A.

B    Valves and Level Controls for Regener-
     ation of Column B.

1    Sequencing Relay #1, Controls The Direc-
     tion Of Flow Through The Resin Column.
     This Relay Also Stops Regeneration
     When The Last Required Event Is
     Satisfied

2    Sequencing Relay #2, Activates Pump 7
     (See Figure 7), Acid Pump #2, The
     Main Water Valve, The Heat Exchanger
     And All Of The Bottom Valves On The
     Regenerant Storage Tanks, (i.e., Con-
     trols from where the regenerant solu-
     tion is flowing.)

3    Sequencing Relay #3» Activates Pump 6
     and All Of The Top Valves On The
     Regenerant Storage Tanks,  (i.e., This
     controls to where the regenerant solu-
     tions will be pumped.)

-------
FILTE
LIQU
DISCONNECT A
\ ^
0\
A

SUPPORT
FILTER
SOCK
0=
f





\L /
\(
SOLUTION IN —
— t$E n


1
L

RED
ID
O
1
1


	 ki CTl

>



-£2^
r
F>r~^

-i \

\ f
V
— PRESSURE GAUGE
L
y— DISCONNECT
OHPRESSURE GAUGE
J
AUTOMATIC
3-WAY VALVE
BACKWASH
1 _n»ATW T.T101!!
|— AUTOMATIC
J~U DRAIN VALVE

Figure 9    THE ARRANGEMENT OF THE SOCK FILTERS
                             31

-------
U)
NJ
                 CHROMATE
                 IHFLUEHT

t




3-
* 4

.
^s


^
I

\
\



r












t.
nH ETjgfFRODE ASSEMBLY



nUtill F\)mr nCi. \

TO RESTN CrtT.IMTfS
RT


( '
\^>
•^
n
                                                           PUMP COMTROLLBR-
                      QfH>-
                                 '    1
                                     RECORDER
                                     -LOW pH ALARM

                                     HIGH pH ALARM

                                     pH AMPLIFIER
                                                                                  ACID
                                                                                  'ORAGE
                                                                                  TANK
                                                                                   909
         PICSURE 10    ILLUSTRATION OP pH LOOP #1 WHICH CONTROLS THE INFLUENT TO THE RESIN COLUMNS

-------
resin are less favorable. At a pH above k.O yellow pigment in suspen-
sion will accumulate between the resin beads in the column and cause
an increase in the pressure drop across the resin bed.  In the event
the pH rises above 3-5* the high pH alarm will sound, again to demand
operator attention.

A strip chart on the signal receiver records the influent pH contin-
uously (see Figure 15.)

The second pH loop is shown in Figure 11.  The discharge from the resin
columns is directed through the colorimeter and delivered into Tank
911.  The submerged pH electrode assembly is located at the bottom of
this well-agitated vessel.

A 10 to 50 miHiampere pH signal passes through a series loop which con-
sists of the amplifier, the high and low pH alarms, a controller and
the signal receiver (recorder).  The current signal from the controller
is converted to a 3 - 15 psig pneumatic signal by a current-to-pressure
converter and the pressure signal is used to adjust a pneumatically
actuated valve.  Sodium carbonate solution is pumped from Tank 908
through this pneumatically controlled valve to the top of Tank 911.  In
Tank 911 the influent pH, normally 3.0, is adjusted until it is slightly
alkaline.  At a pH of 9>0 the zinc ion in the column discharge is pre-
cipitated as zinc carbonate.

The liquid level in Tank 911 is sensed by appropriately placed level
probes.  These probes actuate the on-off or pump-up, pump-down sequence
of Pump 5*  The pH of the waste solution, is thus continuously adjusted
in Tank 911 and this tank is intermittently emptied as the solution is
pumped to the zinc carbonate recovery filters.

     The third pH measurement is made in a flow assembly in the discharge
line from Pump 6 (shown in Figure ?).  It is used to detect the point
when the alkalinity of the regenerant solution has been neutralized and
when the resin column has been re-acidified as required at the end of
the regeneration process, before zinc-containing influent can be intro-
duced to the resin column during the next treatment cycle.  A low pH
signal is used to start a timer which is used to determine the duration
of dilute acid addition  and recycling at the end of the regeneration
process.
COLOR CONTROL

The dichromate ion is so highly colored that it can be determined color-
imetrically15.  Solutions to contain 1 ppm, 2 ppm, and 5 ppm of hexa-
valent chromium were prepared by dissolving carefully dried and weighed
                                   33

-------
  RESIN
  COLUMN
  EFFLUENT
                                                          SODIUM
                                                          CARBONATE
                                                          STORAGE
                                                         TANK 911
                                                  |  TO ZINC RECOVERY
                                                  I  FILTERS
1   LIGHT SOURCE
2   FILTER
3   COLORIMETER LIGHT PATH
k   PHOTO CELL
5   RECORDER
6   pH AMPLIFIER
7   HIGH pH ALARM
 8   LOW pH ALARM
 9   CONTROLLER
10   pH ELECTRODE ASSEMBLY
11   PNEUMATICALLY ACTUATED
     NEEDLE VALVE
12   HIGH LEVEL PROBE
13   CURRENT-TO-PRESSURE
     CONVERTER
Ik   LOW LEVEL PROBE
FIGURE 11    ILLUSTRATION OF pH LOOP #2 WHICH CONSOLS THE
             COLUMN EFFLUENT
                              34

-------
C.P. potassium dichromate in distilled water.  These solutions were sub-
mitted to the colorimeter manufacturer and from these the appropriate
color filter and the colorimeter light path were determined.  A commer-
cial turbidimeter was equipped with this blue filter and arranged with
a forty-inch light path.  This arrangement is capable of sensing as
little as 1 to 2 ppm of chromium when the chromium is present as dichro-
mate.    Calibration of the colorimeter discharge by comparison with
similarly prepared standard solutions has been repeated periodically.

Figure 11 illustrates the arrangement of the colorimeter.  The entire
flow from either of the resin columns (Tanks 905 and 906 in Figure  5)
is directed through the colorimeter before the column effluent enters
Tank 911.

There are two control points on the colorimeter.  The first, usually
set at 30-kO% of full scale, is used to detect a bleeding of dichro-
mate from a resin column.  If this bleeding persists for longer than 30
seconds to one minute, the system automatically switches itself to
secondary treat, thus directing the solution under treatment to a second
resin column in series.

If the bleeding increases to sustained indicated levels above the second
set point, this actuates  a relay which sounds a "high color" alarm and
turns off the treatment system.  An operator is required to make an ap-
propriate correction to the system and start the treatment cycle over
again once this high color alarm has sounded.

Anything which interrupts the light path in the colorimeter is read by
the instrument as color.  Turbidity, air bubbles occluded, in the resin
column, oily films on the light path windows, abrupt temperature
changes which cause condensation on the exterior of the windows and
occasionally turbulent flow have each caused a "high color" alarm.
Momentary "high color" signals (up to 60 seconds on a 0-60 second
timer) alert the system operator without shutting the system down.
If, however, the condition persists until the timer times itself out,
the system shuts itself down automatically.  These non-color interfer-
ences would make it impractical to utilize such a color control system
with a 0.05 ppm chromium specification because the colorimeter signal
at this sensitivity would "paint" the strip chart of the recorder.

Provisions are made to flush out the colorimeter with fresh water to
permit start-up following a "high color" alarm.  This is also available
for cleaning both of the light path window,  the window at the light
source and the window at the photocell, both of which do require some
periodic attention to prevent signal attenuation.
                                   35

-------
FLOW MEASUREMENT

Flow through the chromate recovery system is measured by conventional
variable diameter flow meters.  The entire influent flow passes a treat-
ment flowmeter and the flow is controlled by a 3-inch manual gate valve.
There is a similar arrangement for regenerant flow measurement and con-
trol.  Wetted parts of the flowmeters are glass or 316 stainless steel.
For this latter reason, the dilute acid solution required in the last
step of resin regeneration bypasses the regenerant flowmeter.
LEVEL CONTROL

The automatic control of liquid levels in the chromate recovery system
is necessary for satisfactory treatment and regeneration by automated
control.  This has presented a problem.

As first designed, all controlled levels were controlled by conductance
probes installed in the walls of tanks.  Where usable, conductance
probes are the cheapest to buy and to maintain and the easiest to field
wire.  The solutions were "grounded" either by the wall of the tank, if
the tank was metal, or by a probe introduced for this purpose in the
non-metallic tanks.  At 300 volts potential between the "probe" and the
"ground", all liquids acted as current bridges sufficient to actuate re-
lays on the control panel.  However, settled solids, damp crystallized
salts and liquid films also served as sufficient current bridges.  In
some cases, placing the probe in a PVC sheath and extending this sheath
down to the desired level has worked.  The arrangement is illustrated
in Figure 12.

In the ion-exchange columns, particularly at the liquid-air interface
an oily film or scum accumulates.  This may be from lubricants used
in the processing equipment, pump seals and the like and it may be
due in part to the gradual break-down of the ion-exchange resin in con-
tact with the acidic dichromate.  This film fouls the conductance level
probes and gives a false contact signal.  In practice conductance probes
and probes in sheaths have not worked well inside of the resin columns.
Capacitance probes worked well and gave the "pump-up/pump-down" control
capability when tried in one of the resin columns, but these probes, too,
must be cleaned every two to three cycles.  The capacitance level probes
are thin and, of necessity, about six feet long and they can often suf-
fer from physical damage as they are removed and replaced for cleaning.

We are currently evaluating sonic probes, which are to be installed at
the desired liquid levels.  Since they depend on the filling of a ^"
gap with vibration-transmitting liquid or solid, a fouling film should
present no problem.

-------
     £
u>
                                                rt
                                                   P..
 COLORIMETER EFFLUENT
-SODIUM CARBONATE SOLUTION
-LEVEL CONTROL FOR PUMP 8
 AGITATOR  DRIVE


-UPPER LEVEL PROBE

-PVC PIPE  (NON-CONDUCTING)
                                                             •SUPPORT
-TANK 911

-LOWER LEVEL PROBE
 SS ROD (CONDUCTING)
-CUT-AWAY FOR CLEANING
                                                             •pH ELECTRODE ASSEMBLY
                                                             •TO ZINC RECOVERY FILTERS
                                                             -PUMPS
      Figure 12    ILLUSTRATION OF THE LEVEL PROBES IN THE ZINC PRECIPITATION TANK

-------
 TEMPERATURE CONTROL
 To maintain the regenerant temperature and the rinse water temperature
 at 50° C*, we elected to use a shell'and tube heat exchanger.  The
 shell in the exchanger is of steel and the tubes are of copper.

 During the regeneration steps which require heat, a main steam valve
 is automatically opened, and the  opening of this valve activates the temper-
 ature control system.  See Figure (13).  The temperature control loop
 consists of a thermal bulb, which is the temperature sensing element,
 with a direct fluid connection to the bonnet of the steam control valve.
 The control valve regulates the input steam flow to the exchanger.

 This loop would provide adequate control at a steady state condition,
 but since the time constant of the system is much shorter than the time
 constant of the controller, a forty-gallon surge tank was placed in
 the line.  This insures that the resin and the PVC pipe downstream
 from the heat exchanger are completely protected from thermal damage.

 A bi-metallic temperature switch, mounted at the surge tank outlet, is
 set to close the main steam valve if the outlet temperature rises
 above 60° C., the maximum recommended operating temperature for the
 ion-exchange resins.                                          REGENERANT
                    BI-METALLIC TEMPERATURE SWITCH-
                                SURGE TANK-
              MAIN STEAM VALVE

                       STEAM CONTROL VALVE
S-
                                                            £
-------
PRESSURE AMD VACUUM COKEROL

Shortly after the start-up of the system It became obvious that it would
be necessary to install a pressure control system on the resin columns.
The original resin columns were fabricated of fiber reinforced plastic
and had a design range of -0.5 to +15 psi gauge pressure.  We were exper-
iencing pressure and vacuum readings outside of these limits.  The pres-
sures resulted from an accumulation of fine particles in the resin and
the vacuum resulted whenever a bottom valve on a resin tank was opened
while the tank was filled with liquid and the top valves were closed.
This vacuum would have been sufficient to implode a partially filled tank
or to collapse the dome-shaped cover.

The pressure control device consists of a \ inch three-position electric-
ally actuated valve mounted on the top of each of the resin columns.
Position one on each valve is a closed position which isolates the column.
The second position is a vent to the atmosphere.  The vent pipe, however,
is directed into the influent holding tank (Tank 2 of Figure T) to con-
tain liquids from a possible blow-back when the pressure is relieved.
The third valve position makes a connection to a 12 psig air source.
This enables one to pressurize the system to facilitate drainage of the
column and to provide for an even down-flow of regenerant through the
column.

The three-position valves are controlled by an electromechanical logic
system which receives electrical signals from the process control relays
and also from the high pressure and low pressure switches which are
located above the respective resin columns as shown in Figure 1^.

The high and low pressure switches, which are the primary sensing devices,
are set at 15 psig and -0.5 psig (one inch of mercury), respectively.
During treatment or regeneration a high pressure signal will; l) cause
an audible alarm, 2) display a visual indication of this high pressure
problem on the alarm board, 3) turn the automatic pressure control valve
to the vent position, and £) cause an immediate shut-down of the treat-
ment or regeneration through the column.  During Secondary Treat there
is a 30 second delay before shut-down.  The high pressure in this situa-
tion may be due, not from a blockage, but from the additional pressure-
drop of the second resin column in series.  Similarly a low pressure sig-
nal will result in simultaneous alarm, light display, venting and immed-
iate shut-down under any operating conditions.

Under normal operating conditions the top valve is in the closed posi-
tion during treatment and secondary treatment.  It is in the pressurized
position during the steps of regeneration which involve flow out of the
bottom of the column.  The column is vented only during the water or
regenerant filling steps of regeneration as we displace air from the
resin bed.
                                   39

-------
                                                   VENT TO
                                                   TANK 2
    VACUUM SWITCH
    PRESSURE SWITCH



















II
                                                                180°  3-WAY VALVE
                                                                                  FILTER
                                                                            COMPRESSED
                                                                            AIR  SOURCE
         PRESSURE-VACUUM GAUGE
FIGURE lU    ILLUSTRATION OF THE PRESSURE AND VACUUM CONTROLS ON THE RESIN COLUMNS.

-------
Since the installation of the pressure control system we have replaced
the fiber reinforced plastic resin (FRP) columns with epoxy coated
carbon steel tanks which have much greater strength under pressure or
vacuum.  The tank covers are still the original FRP covers with nearly
the same design limitations.  In any case, we have found the pressure
limits which were originally instituted to be acceptable process operat-
ing limits.

-------
                           PLAKT OPERATION
There was a considerable period of trial and error associated with the
plant operation during its initial stages.   Problems were experienced
with:

     1.  Fine particles in the ion-exchange resin which plugged
         screen openings and created unexpectedly high pressures
         or low flows through the system*

     2.  Level probes.  These problems were primarily with the
         probe becoming coated and giving a false contact signal.

     3.  The colorimeter, which does not differentiate between
         color and any other light interference,  such as turbid-
         ity or air bubbles.

     U.  Failure to completely regenerate.   This  failure was
         caused by a variety of problems such as  uneven flow dis-
         tribution, improper drainage, too  low a  regenerant temper-
         ature, or a valve malfunction.

Each of these problems was solved, one by one, and the final plant de-
sign is such that a normal operation can be obtained by trained plant
personnel with only a minimum of supervision.

During a typical month, one million gallons of influent are treated at
an average flow rate of 55 to 60 gallons per minute.  The chromium con-
tent of the influent varies from 1,200 parts of JfegC^Oy.2HJ20 per  mil-
lion to 4,000 ppm.  The average is about 1,800 ppm.

In a typical month we would recover 15,000  pounds of sodium dichromate
for re-use in our pigment manufacturing operation.  We would recover
18,000 pounds from a possible 19,000 pound  (or 95%) of the potassium
chloride which was used for the regeneration of the ion-exchange resin.
In addition, we recover the zinc carbonate  which  precipitates from the
chromium recovery treatment effluent on neutralization.

-------
Figure 15 is a copy of the strip-chart recording.  The left-hand line,
which is green on our chart,  is a measure of the  influent pH on a scale
of 1-5 pH units.  It can be seen that the control loop #1 adequately
controls the pH of the influent in this  instance  between 2.2 and 2.5.

The center line (red on our chart) is a  measure of the effluent pH on
a scale of 5-9 pH units.  This pH loop was  set to control this pH be-
tween 6.6 and 7-3, and is normally quite close to 6.8.*

The right-hand tracing (blue  on our  chart)  is a measure of freedom from
the yellow (dichromate) color.  At the start (10:30 A.M.) flow and air
bubbles cause the reading to  go to zero  (maximum  color) and quickly
return.  The steps in this tracing are caused by  changing the "coarse"
setting on the colorimeter.   With the coarse setting on Ik the color
was satisfactory from noon until 7:^5 P.M.

At 7:^5 P.M. a timer (initiated automatically from the colorimeter sig-
nal) timed out and transferred the system to "secondary treat" or treat-
ment through the two resin tanks in  series.

It can be seen that the color promptly returned to the satisfactory
range (7 to 10).  Acid from the acid rinse  in Tank 906 was detected by
the effluent pH loop, corrected, and the pH system also returned to
normal.

     *As noted elsewhere in this publication, the  actual adjusted pH is
normally 9.0.  At  some later  point  in time, it was discovered that the
range of 6.6 to 7.3 did not consistently produce  quantitative precipita-
tion.

-------
                                                       Acidity from
                                                       Tank 906
                                                       Secondary Treat
                                                       instituted
                                                       Flew interrupted
                                                       as filter is
                                                       cleaned
                                                       Color tracing
                                                        0 = maximum
                                                             color
                                                      100 = minimum
                                                             color
                                                      (maximum light
                                                       transmitted)
                                                       Influent  pH
                                                       Scale  1 to  5
                                                       pH units
                                                      Effluent  pH
                                                      Scale  5 to 9
                                                      pH units
                                                       Colorimeter
                                                       Coarse  adjust-
                                                       ment  on 14
                                                       Start of Treat
                                                       Tank 905
Figure 15
A COPY OF A TYPICAL STRIP-CHART RECORDING FOR THE SYSTEM
DURING TREATMENT
                    44

-------
                              DISCUSSION
SUCCESS OF THE PROJECT

Prior to the installation of this  ion-exchange water treatment system,
our procedure for disposing of the chromium  in filtrates and wash liquors
from the zinc yellow manufacture was  as  illustrated in equation (3) of
the Introduction.  The solution containing the chromate was acidified
with sulfuric acid, then the resulting dichromate was reduced with sodium
sulfite or sodium bisulfite.  Following  this reduction, the pH was raised
to 8.5 to 9.0 with lime, and at this  pH  value essentially all of the
chromium and the zinc were precipitated.  This slurry of metal oxides
or hydroxides and calcium sulfate  was then concentrated and disposed of
in a land-fill area.  The installation of the ion-exchange treatment
system has been beneficial to our  overall plant operation from a number
of points of view.

First, we are recovering raw materials (chromium and zinc) in useable
form.  The chromium, recovered as  hexavalent sodium dichromate in
nearly an &f> water solution, is immediately  reuseable in the manufactur-
ing process from which it originates.  Its re-use has required no con-
cession to product quality.  The zinc, recovered as a mixed zinc hydrox-
ide-zinc carbonate, is presently being evaluated by users interested in
its zinc content.

Secondly, we have decreased significantly the quantity and particularly
the volume of waste material which it is necessary to deposit in our
land-fill area.  Both chromium hydroxide and zinc hydroxide are precipi-
tated as low solids, high volume products poorly suited for land-fill.
Typically this mixed hydroxide precipitate will contain only 0.85#
chromium.  Were we to discard all  of  the chromium Which we are currently
recovering annually we would, on this basis, be discarding approximately
7,600,000 Ibs of low solids sludge.

Thirdly, our plant operators have  expressed  an interest in the operation
of the treatment system and they exhibit a greater willingness to get
involved and assist us to maintain a  quality effluent within the re-
quired guidelines.  We have received  many helpful suggestions from the
plant operators which have in turn made  the  project more simple and more
successful.

Finally, it is possible for us to  attain discharge levels lower in both
chromium and zinc than were attainable prior to the installation.  This

-------
is true because the solid resulting from the conventional reduction-
precipitation procedure from chromium waste treatment was allowed to
settle in a lagoon.

At the time of treatment the removal of heavy metals was optimally com-
plete, a function of pH and the presence or absence of competing or in-
terfering ions.  In the lagoon area, however, the settled solid was later
subjected to contact with discharges from other operations in the plant.
While these discharges were within, for instance, the pH guidelines of
5.5 to 8.5, a water solution at pH 5.5 will redissolve a significant
quantity of zinc hydroxide from the lagoon and will actually be higher
in zinc content than when originally discharged from the plant process.
By removing these large quantities of both zinc hydroxide and chromium
hydroxide from the lagoon this re-solution process has been shown to be
diminished and both the zinc and chromium contents of the effluent are
lower than they were prior to the installation.
PROSPECTS FOR THE AMORTIZATION OP THIS CAPITAL EXPENDITURE

Sodium dichromate is a $.16 per pound raw material.   We estimate that
it costs approximately $.25 per pound (chemical cost) to dispose of it
in the above reduction-precipitation fashion.  »From Table 23 in Appen-
dix C it can be seen that we recovered approximately 338*000 Ibs of
sodium dichromate in 23 months of operation, which is equivalent to
176,000 Ibs per year.  At $.25 per pound, this represents a saving of
$44,000 per year.  In addition to this, we are precipitating and recover-
ing a saleable zinc carbonate which is estimated to sell between $25,000
and $45,000 on today's market.  One additional saving is the cost of the
by-product solids handling and disposal, which we estimate to be $14,000
per year.  The total potential annual saving  is, thus, in excess of
$82,000.

The labor associated with running this plant  is less than one man per
shift, but the actual allocation would depend  upon the plant's proxim-
ity to other units.  If we assume no other duties are assignable to
this individual, then the cost for the three-shift operation is esti-
mated at $38,000 per year, for a net saving of over $44,000 per year.
Thus to amortize the $125,000 expenditure will require less than three
years.
APPLICATION TO OTHER INDUSTRIES

This study has been limited to the investigation of the use of ion-
exchange resin in treating chromate-containing waste from a zinc yel-
low pigment plant.  The original Nalco study dealt primarily with efflu-
ents from cooling towers and water from boiler blow-down in which chro-
mates were used as corrosion inhibitors.
                                   46

-------
The secret to the utility of this technology is most certainly to find
a use for the recovered chromate-containing regenerant solution.  In
this pigment plant application the conditions are not very favorable
for exchange.  That is, in the pigment plant application ve experience
chromate contents in the range of 1800 ppm and chloride ion concentra-
tions as high as 21,000 ppm.  (The chloride ion is approximately 0.36M.)
If we were to assume that we would be faced with no less favorable con-
ditions for exchange in some other industry than in our present applica-
tion, it is reasonable that this exchange system would work to remove
dichromate ions from other solutions requiring treatment.

Again, under the conditions reported here, we are accomplishing approxi-
mately a forty-fold concentration of the dichromate.  If no use could
be found for the recovered solution which contains approximately 8#
potassium dichromate and &% potassium chloride, then this solution, one-
fortieth of its original volume, but still containing all of the original
dichromate, would still have to be'disposed of.  This is generally done
by reducing the dichromate to chromium (ill), precipitating it as a
hydroxide, filtering it or concentrating it in some fashion and finally
disposing of it in a land-fill area.  The same technique could probably
be employed on the original solution prior to the ion-exchange treatment
at a lower cost.

If a use can be found for the recovered chromate-containing solution,
then this technique can be used wherever there is a chromate, dichro-
mate, or chromic acid disposal problem.  This would include the manu-
facturing processes for most chromate pigments, zinc chromate, basic
zinc chromate, calcium chromate, barium chromate, strontium chromate,
and chromium oxide.

Outside of the pigment industry this system might have utility wherever
chromium is used as a corrosion inhibitor such as cooling towers and
boiler feed waters, and where it is used as a raw material in the manu-
facture of a chromium-containing product, such as electroplating and
unless there were interference from organic impurities, in tanning.
ATTAINIMG UNIFORM FLOW THROUGH THE RESIN BED

Doubtlessly our most troublesome problems were associated with our ef-
forts to obtain uniform flow through the resin bed, a prerequisite for
satisfactory regeneration.  In the laboratory we dealt with 100 ml to
300 ml of resin.  The resin was supported over glass wool and contained
in a glass addition funnel.  One could readily observe the flow, the
color changes, channeling, back-wash, etc.  As dichromate is adsorbed
on the resin the adsorption is accompanied by a color change (as well
as a change in the resin particle density).  The pale straw yellow color
of the freshly regenerated resin becomes a deep amber or orange-brown
color when the resin is saturated with dichromate.

-------
When, however, one places the resin in a five foot diameter resin column,
an epoxy-lined steel tank, one can no longer make these helpful visual
observations.  Our-first approaches were extremely naive in spite of much
helpful advise1^16.  We found liquids flowing through a resin bed to be
extremely lazy, always content to take the path of least resistance, to
channel along with what had preceded it.  This was particularly true
during upflow regeneration.  In spite of considerable attention to the
design and placement of the diffuser openings, we were repeatedly plagued
with the inability to obtain a "water-white" effluent.

An accepted practice in many ion-exchange installations1'*1" is to place
the resin on a porous sand support bed.  The presence of the coarse sand
is to assist the bottom diffuser in distributing the flow evenly across
the entire cross-section of the bed.  This assures that all of the resin
is "active" and does not serve simply as an inert filler in the column.
Because the sand is considerably more dense than the ion-exchange resin
there is a minimum of mixing of the two media and any mixing which may
occur on down-flow is essentially eliminated when the bed is back-washed.
Thus the sand-resin interface remains quite distinct.

Our original column design incorporated a sand support bed to a level
above the bottom diffuser openings.  The presence of this carefully
screened sand (from which all material smaller than 40 mesh had been
removed) acted to plug the screen openings and to yield an unexpectedly
high pressure drop.  This pressure drop was observed even before the
ion-exchange resin was added to the column.

The sand was removed to a level just below the bottom diffuser openings
and the ion-exchange resin was added to the column.  We operated for
nearly six months with the sand support bed in place as illustrated in
Figure 16.  During this time we were constantly troubled by the presence
of zinc yellow pigment which accumulated below the bottom diffuser.
The zinc yellow which was in solution in the column influent was re-
precipitated when the alkaline regenerant solution was added.  In an
insoluble form the zinc yellow was more difficult to remove from within
the sand support bed.  Its presence just below the bottom diffuser was
apparent when acidified influent was re-introduced to the column.  The
acid-soluble pigment would slowly redissolve, migrate upward in the
water solution and enter the column effluent as a hexavalent chromium
discharge.

Efforts to flush the sand support bed upward or downward through the
column's center bottom drain were only marginally successful and were
very time consuming.

After what was thought by the operators to be a careful and thorough
regeneration we would, upon occasion, sample the resin by driving an
empty pipe through the depth of the bed, cap off the upper end of the
pipe and withdraw it full of resin.  The resin could then be examined
both visually and by a laboratory-scale performance test to determine
completeness of regeneration and to locate levels of poor flow and poor
regeneration.

-------
INFLUENT
                                                          FREE SPACE FOR
                                                          BED EXPANSION
                                                          DURING BACK-WASH
REGENERANT
                                                          ION-EXCHANGE RESIN
                                                          BOTTOM DIFFUSER
                                                          REGENERANT OUT
                                                          GRADUATED SAND
                                                          AND PEA GRAVEL
                                                          SUPPORT BED
                                                          BOTTOM VALVE
      TANK 911
      Figure  16     ILLUSTRATION OF THE RESIN COLUMN WITH SAND AND
                    SUPPORT BED

-------
Most of these problems were finally resolved when we removed the sand
support bed and inverted the bottom diffuser, directing the openings
toward the dish-shaped tank bottom.  Down-flow from these openings now
swept along the tank bottom, stirring the resin and contacting any
potentially inert areas.  The flow then changed direction and flowed
upward toward the top diffuser.  This change is illustrated in Figure 17.

The problem of displacing the dense regenerant salt solution upward from
below by rinse water was only resolved by reversing the direction of
flow.  Normally one elects to regenerate in the opposite direction from
the direction of treatment flow.  This assures that the most satisfac-
torily regenerated resin beads will lie on the tank bottom in contact
with the effluent and thus assures the minimum dichromate content in the
effluent.  Because of the previously mentioned density change, however,
when the bed is back-washed the more dense (partially exhausted) resin
beads find their way to the tank bottom.

Best regeneration results were obtained by down-flow regeneration also.
If the chromium content of the regenerant solution toward the end of
the regeneration process is not permitted to be too high (not above 1.5$
sodium dichromate after use) one is generally assured of a complete re-
generation throughout the entire column, and there is no significant
dichromate concentration gradient throughout the resin bed.  If there
were a significant dichromate content remaining on the resin in the
column, it would be necessary to retain this higher dichromate concen-
tration a.% or neai; the top of the resin column; not at the bottom where
it would be in contact with the column effluent.  This is to say, only
when one has eliminated partially regenerated resin within the column
can one be assured that partially regenerated resin will not accumulate
at the bottom of the column following back-wash.
THE SYSTEM OF SAFETY INTERLOCKS, ALARMS AND WARNING LIGHTS

Considerable attention was devoted, in the design stage, to make the
system operable by relatively unskilled plant personnel.  We were inter-
ested in safety and thus, for instance, we wanted to make it impossible
to pump a concentrated alkali into a tank containing acids.  We were
interested in a quality effluent and we wanted to detect leaks and
avoid spills.  All of the effluent from the resin columns passes through
the colorimeter so we are not troubled with by-pass sampling of the
effluent stream.  We paid particular attention to the concentrated chrome
solution recovered from regeneration and made it impossible, for instance,
for an operator to attempt to pump on top of an already filled tank.

We attempted to protect the equipment.  In this system one cannot pump
a dilute hydrochloric acid solution through a stainless steel valve,
pump or flow meter.  The system is protected against pressure levels
and vacuum levels which are outside of the design specifications.  PVC
pipe and the ion-exchange resins are both protected from thermal damage.


                                    50

-------
                       -TANK WALL
                        BOLTED FLANGE
\_n
                                                               BOTTOM DIFPUSER
                                                                      Ik MESH SUPPORT SCREEN

                                                                      kO MESH SCREEN
                                                                      Ik MESH SUPPORT SCREEN

                                                                      FLOW SWEEPS THE TANK
                                                                      DISH BOTTOM
            Figure II    ILLUSTRATION OF THE RESIN COLUMN AFTER REMOVAL OF THE  SUPPORT BED
                         AND INVERSION OF THE BOTTOM DIFFUSER

-------
Finally, the operator is given assistance and some opportunity to antici-
pate problems.  He cannot successfully call for regeneration if he has
not re-made his regenerant solutions and emptied the storage tank which
holds the recovered chrome solution.  He is given warning when pH read-
ings are outside of the control specifications and when his supply of
sodium carbonate solution is being depleted.  The following is a list
of the safety interlocks, alarms and warning lights which were installed.

Interlocks
1.  The recovered chrome solution tank (Tank 91^ in Figure 7) is still
    full.  The tank contents must have its pH adjusted to the dichro-
    mate pH and be pumped back to the manufacturing process before the
    next regeneration can be initiated.
2.  The operator has failed to make appropriate additions of sodium
    hydroxide and potassium chloride to the regenerant holding tanks
    (Tanks 903 and 9C& in Figure 7) after these tanks were re-filled
    with the column rinse water.  The next regeneration cannot be
    started.
3.  One cannot pump the zinc-containing influent on top of an alkaline
    regenerant solution.  (One cannot treat through a particular column
    when that column is being regenerated, nor can one regenerate a
    column while it is still in use in treatment.)
k.  One cannot regenerate a resin column when it is being used during
    secondary treatment.  (During secondary treatment both of the
    columns are in use, so neither can be regenerated.)
5.  One cannot regenerate two columns at the same time.
6.  Except during secondary treatment, when the two columns are arranged
    in series, one cannot treat through two columns (in parallel) at the
    same time.
7.  One cannot operate under pressure levels or vacuum levels outside
    of the design limits.

Alarms

The operator receives an audible alarm (which also sounds in the foremen's
office) whenever:
1.  The influent pH is too high.
2.  The influent pH is too low.
3.  The column effluent is too high in color.
4.  The pH of the effluent from the zinc precipitation tank is too high.
5.  The pH of the effluent from the zinc precipitation tank is too low.
6.  Tank 9l4 (the recovered chrome solution tank) is still full.
7.  Tanks 903 and 9C& (the regenerant storage tanks for portions III and
    IV) have not been re-made.
8.  Tank 908 (the sodium carbonate solution storage tank) is empty.

Warning Lights

Appropriately labeled warning lights are displayed on the control panel
board to explain each of the above alarms, another to warn the operator
of an imprudently low reserve of sodium carbonate solution and another
to warn of a cautionary colorimeter response.

                                     52

-------
                              REFERENCES
 1.  "Deep Wells for Industrial Waste Injection  in the United States",
     U. S. Dept. of Interior, Publication WP-20-10  FWPCA, Cincinnati.
     Ohio, Nov. 1967.

 2.  Thorn, P. C. L. and E. R. Roberts,  "Ephraim, Inorganic Chemistry",
     Interscience Publications, N. Y., 6th Revised Ed. (1954) p 502 ff.

 3.  Hesler, J. C., "Industrial Water and Wastes/' 6, No. 3 (1961)
     P 75-79.

 4.  U. S. Patent 3,223,620, December 14, 1965.

 5.  Hesler, J. C. and A. W. Oberhofer, "Recovery and Reuse of Chromates
     in Cooling Tower Discharges by  Ion Exchange," 20th Annual Symposium,
     N. A. C. E., March 9-13,
 6.  Parker, Dr. C. L. , General Technologies Corp., private communica-
     tion with D. J. Robinson.

 7.  This system is operated under a licensing agreement with Nalco
     Chemical Company which includes a royalty based on the quantity of
     ion-exchange resin in use.

 8.  Sloan, L. and N. J. Nitti, "Operating Experiences With Ion Exchange
     Chromate Recovery System on Cooling Tower Slowdown," Nalco Chemical
     Company, 1964.

 9.  Rohm and Haas Company, Technical Bulletin IE- 112 -67, Ion Exchange
     Department, Phila., PA, 19105, Sept., 1967.

10.  "Dowex Ion Exchange Resins," The Dow Chemical Co., 1964.

11.  EPA Report 12090 ESG, "Zinc Precipitation and Recovery From Viscose
     Rayon Wastewater, "  EPA, Washington, D. C. 20460.

12.  Reference 9, page 3, Table 11.

13.  Dickert, C. T., Rohm and Haas Company, Phila. PA, private communica-
     tion with D. J. Robinson.

14.  For a further discussion see Appendix D.

                                   53

-------
                        REFERENCES (CONTINUED)

15.  Willard, H. H. and N. H. Furman, "Elementary Quantitative Analysis,"
     3rd edition, D. Van Nostrand Co. N. Y., (19^8) p k6Q.

16.  Diblik, T. and W. Ward, Nalco Chemical Co., private communication
     with H. E. Weisberg and D. J. Robinson.

IT-  Kunin, R., "Elements of Ion Exchange," Reinhold Publishing Corp.,
     N. Y.

18.  Marinsky, J. A., "Ion Exchange," Vol. 2, Marcel Dekker, Inc., N. Y.,
     1969.

19.  Belcher, R. and C. L. Wilson, "New Methods in Analytical Chemistry,"
     Reinhold Publishing Corp., N. Y., 1956, p 260.

20.  Reference IT, p 35, P 135.

-------
                           GLOSSARY OP TERMS
Backwash - The process of flowing water upward through the ion-exchange
resin bed to suspend the resin, remove resin fines and foreign matter,
to render the column ready for its next treatment cycle.

Bleed - The situation wherein a visible low level of chromium (2-15
ppm) exists in the effluent from a resin column on treat.

Exhaust - The situation when the effluent from a resin column begins
to approach the influent in visual appearance or in chromium content.
The saturation of the column with respect to chromium uptake.

Regeneration - The process of removing adsorbed chromium from the ion-
exchange resin and returning the resin to the chloride salt form.
Regeneration is required once a column is exhausted.

Secondary Treat - The process of connecting two resin columns together
in series so that the "bleed" from one column is collected in the
second column when the two columns are on treat.

Treat - The process of flowing an acidified influent through a regener-
ated resin column in order to adsorb dichromate from the influent
solution and thus remove this heavy metal contaminant from the effluent
stream.
                                     55

-------
                              APPENDIX A

                          ANALYTICAL METHODS
ANALYTICAL METHOD FOR POTASSIUM

Reagent s

IN acetic acid: 5.75 ml of glacial acetic acid is diluted to 100 ml
with distilled water.

Sodium tetraphenyl boron reagent:   Dissolve 0.15 g sodium tetraphenyl
boron in 50 ml of distilled water.

Wash solution:  Add a small amount of sodium tetraphenyl boron to 200 ml
of distilled water containing a drop of HC1.  Precipitate the reagent
with KC1 and filter.  Shake the freshly precipitated potassium tetra-
phenyl boron with 500 ml of distilled water, filter to remove the excess
solid and save the saturated wash  solution.

Method :

Pipette 5-00 ml of the recovered potassium chromate solution, transfer
to a 100 ml volumetric flask, and  dilute to the mark.  Pipette 5.00 ml
of this diluted solution and transfer to a 250 ml beaker.  Add 15 ml.
of IN acetic acid, 35 ml of distilled water, and, with stirring, 50 ml
of the sodium tetraphenyl boron reagent.  Filter immediately through a
tared Gooch crucible.  Dry in an oven above 100° C., but below 120° C.
for one hour or to constant weight.  Cool in a desiccator and weigh.
Calculation:

Wt. of precipitate x 0.2080
- v 0<2^ ml -
                              ,__     „._,      ,,_  ,
                            x 100 = g KC1  per 100 ml.

-------
ANALYTICAL METHOD FOR CHROMIUM

Reagents:

CP Potassium Iodide
6N Hydrochloric Acid; Dilute 500 ml of cone reagent grade HC1 to one
    liter with distilled water.
Standard 0.1 N Sodium Thiosulfate Solution:  Dissove 25 g of reagent
    grade sodium thlosulfate crystals NagSgOo^HgO and 2 g of reagent
    grade sodium carbonate in one liter of freshly boiled distilled
    water.  Store in a clean bottle, allow to cool and standardize
    with CP potassium dichromate.
Starch Indicator Solution: Paste 2 g of soluble starch in a small
    amount of water and add it to 100 ml of boiling distilled water.
    Add a small amount of mercuric iodide to prevent bacterial action.

Method:

Pipette 5-00 ml of the concentrated regenerant solution and transfer to
a 100 ml volumetric flask.  Dilute to the mark and mix thoroughly.  Pipette
from this flask a known aliquot to contain approximately 0.1 g of sodium
dichromate.  Dilute to 50 ml in a 250 ml iodine flask.  Add 2 g of the
potassium iodide and slowly, with swirling to avoid a localized excess,
add 8 ml of the 6N hydrochloric acid solution.  Titrate the iodine which
is released at once with the standard sodium thiosulfate solution.  Add
the starch indicator about 1 ml prior to the end point.  (The solution
will have turned to a pale yellow-green color, but not the blue-green
color of trivalent chromium. )  The end point is determined by the dis-
appearance of the deep blue color characteristic of starch in the pres-
ence of iodine.

Calculation:
 ml of NapSgO^ x Normality  x   0.04967  x  100 = g  NaoCrpO-r  .2HpO/100 ml
 ml of original sample  in the  aliquot              *  *  '
                                    57

-------
                              APPENDIX B

                    DETAILS OF THE LABORATORY STUDY

REGENERATION OF THE ION-EXCHANGE RESIN

One of the first studies of the elutriation or regeneration of the ion-
exchange resin was done by placing 300 ml of the resin in a tall,  narrow
column (one inch in diameter and 23-| inches deep).   The column was ex-
hausted by the introduction of an acidulated pigment waste solution
which contained sodium dichromate.  The column was  then rinsed with dis-
tilled water and covered with a regenerant solution which contained 10 g
KC1 and 2.5 g NaOH per 100 ml.  Every fifteen minutes a 10 ml portion
of the regenerant solution was removed from the bottom of the column.
The pH of each portion was measured and a 1.00 ml aliquot was titrated
iodometrically to determine its dichromate content.   A portion of  the
results of this study is recorded in Table 9 and the two curves, pH
versus the volume of regenerant passed through the  column and the  con-
centration of the recovered regenerant solution versus the volume  passed
are plotted together in Figure 18 from this data.


Table 9    THE pH AND THE DICHROMATE CONCENTRATION  OF SAMPLES OF THE
           REGENERANT SOLUTION PASSED THROUGH AN EXHAUSTED RESIN COLUMN
           TO STUDY THE REGENERATION PROCESS

Sample                              Titre*             Concentration
Number           pH               .1087 N                 grams of
                                                         per  100 ml

   5             3.0                0.00                    0.000
  10             6.0                0.00                    0.000
  15             7.3                0.61                     .329
  20             7.5                0.98                     .529
  25             7-5                1.12                     .605
  30             7.5           '    1.1*7                     .79^
  35             7.6                1.80                     .972
  ^0             7.6                2.10                    1.13
  ^5             7.9                3.80                    2.05
  50             8.1                7.19                    3.88
  55             8.3               10.07                    5.14
  60             9.0               11.90                    6.43
  65            12.2               11.00                    5.94

     * 1.00 ml aliquot  samples       53

-------
             6.0  -
CONCENTRATION
   GRAMS
    ^Oj
    PER
   100 ml
             5.0  -
U.o
VO
3.0
             2.0
              1.0
                                    loO
                                aio^   3^0
                                                                          CONCENTRATION
                                                                                           12
                                                                                           10
                                                                                            8
                                                                                                          pH
      FIGURE 18
                                     VOUUME OF REGENERANT
                                            ml
     PLOTS OF pH AND BICHROMATE CONCENTRATION VERSUS THE VOLUME OF REGENERANT SOLUTION
     WHICH WAS PASSED.

-------
It can be seen that the alkalinity of the regenerant is being consumed
by the acidic dichromate which is adsorbed on the resin.  When this di-
chrornate acidity has been neutralized the rate at which the hexavalent
chromium is being removed from the column increases, reaches a maxium
and then drops off again.  It should be noted that the abscissa is also
a time axis since the samples were taken at fifteen minute intervals.

In the next study the column was again exhausted by the introduction of
dichromate and then rinsed as before.  This time, a 300 ml portion of
regenerant, which contained 10 g KC1 and 8 g NaOH per 100 ml, was intro-
duced to hasten the neutralization of the column's acidity.  This 300 ml
portion was followed by regenerant which contained 10 g KC1 and 2.5 g
NaOH per 100 ml.  Again a portion of the results are recorded in Table 10
and are plotted in Figure 19.  It can be seen that the concentration
maximum has been displaced toward the left, or that we have successfully
removed a greater quantity of dichromate from the resin after only 320
ml has been removed, as compared with 600 ml during the previous test.
It should be pointed out that the concentration scales of Figures 18
and 19 are not the same.  The concentration of the recovered regenerant
is in part a function of the quantity of dichromate present on the col-
umn prior to the regeneration.
Table 10    THE pH AND THE DICHROMATE CONCENTRATION OF SAMPLES OF
            REGENERAOT WHEN EXCESS ALKALI IS ADDED INITIALLY

Sample                              Titre*             Concentration
Number           pH               .1015 N                 grams of
                                                          per 100 ml

   3             6.0                0.00                      .000
   5             7.9                ^-20                     2.12
  10             9.3               22.80                    11.5
  15             9-^               25.19                    12.7
  20             9-3               30.13                    15-2
  25             9-3               32.09                    16.2
  30             9-3               32.25                    16.3
  36            12.1               25.35                    12.8

     * 1.00 ml aliquot s
Table 11 and Figure 20 which were accumulated from data from the re-
generation of a partially exhausted column are included only to show
that the concentration curve when completed is essentially symmetrical
about the concentration maximum and the maximum occurs in a much nar-
rower region when the larger quantity of base is used in the regenerant,
This column was regenerated entirely with a solution which contained
10 g KC1 and 8 g NaOH per 100 ml.


                                    60

-------
            16 -
            14

CONCENTRATION
            12
   GRAMS
    PER
   100ml
             8
                         o
               160
240
320
                                                             400
480
560
                              VOLUME OF REGENERAHT PASSED THROUGH THE COLUMN
                                                    ml
      FIGURE 19
PLOTS OF pH AND DICHROMATE CONCENTRATION VERSUS THE VOLUME OF REGENERANT
SOLUTION PASSED.  EXCESS ALKALI WAS ADDED IN THE FIRST 300 ml

-------
Table 11    THE pH AND THE BICHROMATE CONCENTRATION OF SAMPLES OF
            REGENERANT WHEN EXCESS ALKALI IS ADDED AND WHEN THE
            COLUMN, INITIALLY, IS ONLY PARTIALLY EXHAUSTED

Sample                             Titre*                Concentration
Number           pH                  N**                    grams of
   5             4.1                o.oo                     .000
  10             6.0                0.81                     .076
  15             7.0                2.44                     .228
  20             7.1                4.11                     -385
  25             7.4               12.71                    1-19
  30             7.6               11.61                    1.09
  35             9.4               28.20                    2.71
  40            12.7               18.72                    1.80
  45            12.8                7=53                     -T23
  48            12.8                4.5                      -^32

     *  5.00 ml aliquot s
     ** For samples 1-31 the Normality was 0.09423
        For samples 32-48 the Normality was 0.09664


There is now an economic consideration.  The raw material which is re-
quired in the manufacture of the zinc yellow pigment is sodium or
potassium dichromate, not chromate, and any excess base present in the
recovered regenerant solution must be acidified before the accompanying
hexavalent chromium can be used.  That is to say, it is necessary to
neutralize the excess alkali in order to convert the chromate present
to dichromate.

As mentioned earlier in this report, an economy was realized by recycling
the first portion of regenerant through the column, a technique which
assures the maximum chromium content and the minimum quantity of excess
base in the recovered solution.
CONCENTRATION OF DICHROMATE IN THE RECOVERED REGENERANT

In Tables 12, 13 and 14 are recorded the concentrations of sodium dichro-
mate in the recovered regenerant solutions.  In each regeneration there
are four equal-volume portions.  Portions I and II, which have the highest
chromium content and the minimum of excess free base, are returned to
the manufacturing process for re-use.  Portions III and IV from one re-
generation become portions I and II respectively for the next regeneration.
Table 12 shows the data from regeneration of the Rohm and Haas resin.
Tables 13 and 14 are both developed from data from regeneration of the
Dow resin and illustrate a range of variation from sample to sample.  The
                                    62

-------
    3.5  -
    3.0  -
H
 8
8
CM
 ff
u
6  1.5  -
83
s
§
O
   1.0
   0.5  -
       0
  160       240      320
VOLUME OF REGENERAHT PASSED
            ml
400
Figure 20    PLOTS OF pH AND DICHRCMATE CONCENTRATION VERSUS THE
             VOLUME OF REGENERANT SOLUTION PASSED.  EXCESS ALKALI
             WAS ADDED TO A PARTIALLY EXHAUSTED COLUMN.
                                   63

-------
actual plant experience with the Dow resin (shown in Table 14 of Appendix B)
is closer to the new  sample.  Portions I and II are combined prior to
re-use, so the average analysis of these portions is what should be com-
pared with the reported plant results.


Table 12    THE SODIUM DICHROMATE CONCENTRATION OF EACH OF THE FOUR
            REGENERANT PORTIONS USED TO REGENERATE THE ROHM AND HAAS
            RESIN

Portion No.
  Cycle                  I           II          III         IV
   No.                      grams of NagCiO^O/lOO ml
    1                   6.49        4.33        2.16        1.27
    2                   6.61        3.80        1.98        1.08
    3                   6.73        4.22        2.22        1.06
    4                   6.44        4.o4        2.06        0.98
    5     ,              6.75        4.12        2.31        0.83
    6                   6.28        4.25        2.06        0.90
    7                   6.68        4.19        2.32        1.16
    8  '                 6.95        4.14        2.38        0.69
    9                   6.63        4.13        1.93        1.08
   10                   6.92        3.84        1.96        1.07
   11                   6.63        4.37        2.29        1-16
   12                   7.28        3.92        1.76        1.17
   13                   6.68        4.14        1.96        1.01
   14                   6.78      ,  4.04        1.92        0.85
   15                   6.99        3-90        2.26        1.05
   16                   6.58        4.87        2.82        1.31

Average                 6.71        4.14        2.15        1.04

Portions I and II combined          5.43

-------
Table 13    THE SODIUM DICHROMATE CONCENTRATION OF EACH OP THE FOUR
            REGENERANT PORTIONS USED TO REGENERATE THE DOW RESIN
            (ORIGINAL LABORATORY SAMPLE)

Portion No.              I           II          III         IV
  Cycle                     grams of Na2Cr20Y.2H20/100 ml
   No.

     1
     2

     k
     5
     6
     7
     8
     9
    10
    11
    12
    13
    1*
    15
    16
    17

 Average                 9-lk        5-*5        2-50        1.19

 Portions I and II combined          7.30
9.26
9-56
9-56
9.90
9.61
8.88
9.06
8.73
9.03
8.96
8.85
8.71
9-07
8.51
9.70
8.75
9.21
4.98
5.43
5.82
6.22
5.84
5.85
4.43
5.18
5.62
5-63
5.02
5.33
5.11
5.15
5-78
5.55
5.70
2.kk
2.40
2.74
2.60
2.34
2.81
2.14
2.58
2.84
2,39
2.50
2.25
2.40
2.30
2.35
2.75
2.73
1.10
1-36
1.19
1.24
1.15
1.05
1.15
1.09
1.15
1.19
1.15
1.00
1.15
1.16
1.25
1.34
1.52
                                     65

-------
Table 14    THE SODIUM BICHROMATE CONCENTRATION OP EACH OF THE FOUR
            REGENERANT PORTIONS USED TO REGENERATE THE DOW RESIN
            (NEW PLANT SAMPLE)

Portion No.              I           II          III         IV
  Cycle                      grains of NagCrgOY^^O/lOO ml
   No.
    1                  11.25        6.83        3-39        1-86
    2                  11.55        7.55        4.03        2.31
    3                  10.96        7.13        3-28        2.14
    1*                  10.60        7.39        4.23        2.14
    5                  10.38        7.02        3.58        1.94
    6                  10.65        7.06        2.99        2.23
    1                  10.74        6.84        3.19        1.75
    8                  10.53        7.2?        3.83        1.57
    9                  10.80        6.72        3.78        1.97
   10                  11.24        6.46        3.65        1-79
   11                  10.60        6.51        3-35        1.86
   12                  10.26        5.85        2.88        1.70
   13                  10.55        6.16        3-23        1-90
   14                  10.65        6.41        3-10        1.56
   15                  10.38        6.4o        3-75        1.88
   16                   9.51        6.19        4.06        2.45
   17                  10.66        6.65        3.65        2.30

Average                10.67        6.73        3.53        1.96

Portions I and II combined          8.70
DETAILS OF THE ACCOUNTING FOR DICHROMATE INTRODUCED IN THE INFLUENT

As indicated earlier in the report, we adopted as a laboratory proced-
ure for regeneration of the exhausted resin the use of four 100 ml portions
of regenerant solution to regenerate a 100 ml resin column.  The first
two portions of regenerant which contained the highest concentrations
of recovered dichromate were intended for recycle to manufacturing.  Por-
tions three and four from one regeneration would become portions one and
two respectively for the next regeneration.  Each portion of regenerant
was analyzed for its chromium content and the results for one day
(3/12/70) were recorded as in Table 15.
                                      66

-------
Table 15    CHROMIUM ANALYSES FOR THE FOUR REGENERANT PORTIONS USED TO
            REGENERATE THE DOW RESIN 3/12/70
Sample
Number
  B4
  B5
  B6
  BT
The calculation of Table 15 is done as follows:

Titre x Normality x Milliequivalent Wt.  x Total Volume
               Aliquot

Aliquot

1.00
1.00
1.00
1.00
Titre
.1089 N
NagSg^
11.91
8.51
5.30
2.91
Weight of
Sodium Dichromate
in 100 ml
6.44 g
4.60
2.87
1.57
                                      g NagCrgOf^HgO in the sample

Substituting the values for sample B6 of Table 15 we obtain:
   5.30 x .1089 x .0^96?  x 100 = 2.87 g
           j»* OO

On the next day (3/13/70) the following were recorded:

l) The influent concentration:
   A 10 ml aliquot of the influent required 20.41 ml of .1089 N sodium
   thiosulfate solution in the iodometric tit rat ion.

   20.41 x .1089 x .OU967  = 0.01104 g m2Cr207.2E20/mlL
          1.O* OO

2) Total dichromate in the influent:
   The 100 ml Dow resin column was exhausted by 1400 ml of the above
   influent solution, so the total dichromate in the influent was:

       1400 ml x 0.01104 g/ml = 15.46 g

3) Dichromate recovered:
   Regenerant portions B6 and B7 from the previous regeneration cycle
   were reused and portions B8 and B9 which were freshly prepared were
   then used to regenerate the exhausted resin column.  The dichromate
   analyses for these solutions, the rinse and back-wash are recorded
   in Table 16.
                                    67

-------
Table 16
Sample
Number
  B6
CHROMIUM ANALYSES FOR THE REGENERANT SOLUTIONS USED TO
REGENERATE THE DOW RESIN 3/13/70
  Titre
.1089 N
 Aliquot
 1.00 ml     13.31* ml
             10.50
              5.09
              3-25
              1.66
              1.00
    Total Diqhromate Contained
B7
B8
B9
Rinse
Backwash
1.00
1.00
1.00
1.00
50.00
 Total
Volume
              100.0 ml
              100.0
              100.0
              100.0
              100.0
              500.0
  Weight of
Na2Cr2°T*2H20
  Recovered
U)  Net Dichromate recovered:
    Regenerant portion B6 contained 2.8? g and B7 contained 1.57 g
    of dichromate from the previous regeneration which total k.kk g.
    If we subtract this amount from the total dichromate from Table 16
    we have a net value of 13.92 g.  This amounts to 90$ of the influent
    dichromate.  (i.e., 13.92 is 90# of 15.1*6))

On a given day, primarily because the "exhaust" point was determined
visually, we could account for as little as 90$ of the dichromate, or
as much as 108$.  Over fifty-eight cycles, however, it can be seen from
Table 17 that we accounted for 99-5$ of the dichromate which was intro-
duced in the influent as being recovered either in the regenerant or in
the rinse and back-wash waters.
                                    68

-------
Table IT    AN ACCOUNTING FOR THE INFIAJENT BICHROMATE IN THE
            RECOVERED REGENERANT
                        (grams of
      Rohm & Haas Resin
Dichr ornate
in influent

    9.18
   11.73
    8.90
   11.39
    8.90
   11.39
    8.90
   11. 7k
    9-43
   11.39
    8.72
   11.39
    9.07
   12. 46
    8.90
   11.03
    9.61
    8.18
   10.72
    7-95
   11.1*0
 7
12
      99
      11
    7.61
   12.^5
    9.16
   11.93
    8.83
                Bichromate3
                 recovered

                    9-53
                   12.^7
                    9-27
                   11.97
                    9-27
                   11.78
                    8.91
                   11.56
                    9-40
                   11.34
                    9.21
                   11.55
                    9.64
                   13.06
                    9.02
                   11.51
                   10.38
                    8.1*7
                   10.71
                    8.51
11.18
 7.
12,
 7.
  .73
  • 59
  .02
11.30
 8.71*
11.94
 8.79
                                                Bow Resin
Bichromate
in influent

   19.21
   12,92
   11*.06
   11.55
   16.02
   12.82
   16.91
   13.12
   17.62
   13.27
   18.15
   13-27
   18.15
   ll*.22
   18.15
   13.17
   12.U6
   18.15
   12.62
   16.78
   12.69
   18.33
   13.11
   17.99
   12.80
   16.57
   13.32
   18.76
Dichromatea
 recovered

   19-15
   13-53
   14.10
   12.70
   16.00
   12.95
   15.62
   12.92
   17.39
   13-1*0
   17.04
   13.40
   16.99
   13.96
   17.77
   13.26
   11.85
   18.35
   12.28
   16.78
   12.49
   17.73
   12.26
   17.11
   12 = 52
   16.79
   13.15
   18.12
Total dichromate in influent

Total recovered  dichromate

Percentage recovery
                                      736.59 g

                                      733-28

                                       99.55*
a: Total  (net) recovery in four regenerant portions plus the rinse
   and backwash  solutions.
                                     69

-------
                              APPENDIX C
                    DETAILS OP THE PLANT OPERATIONS
INFLUENT ANALYSES

In Table 18 we have recorded one month's influent data,  Again we can ob-
serve variations from 965 ppm to 3286 ppm,  with the average figure 1952
ppm quite close to the figure of 192? ppm reported in our laboratory work
in Table 18.  As in the laboratory study we had one instance where color
was introduced in error into the waste recovery system.  This time analy-
ses of 8U?2 and 15,376 ppm were recorded.  These concentrations for a
brief period presented no problem to the recovery system.


RESIN EXCHANGE CAPACITY

One means of estimating the exchange capacity of the ion-exchange resins
is to determine the quantity of sodium dichromate recovered from the ex-
hausted column.  This data is reported in Table  19.
THE CONCENTRATION OF THE RECOVERED REGENERANT SOLUTION

The regenerant solution which is originally composed of 8% KC1 and 2.5$
NaOH, contains potassium dichromate, excess potassium chloride and sod-
ium hydroxide after it has been used to regenerate an ion-exchange resin
column.  In our efforts to re-use this solution we are concerned with a
number of considerations: l) The potassium content.  Potassium, as
either potassium chloride or potassium chloride, is a required raw mater-
ial in the zinc yellow manufacture. 2) The dichromate content.  Sodium
dichromate or potassium dichromate can be used in the zinc yellow manu-
facture.  3) Excess base.  The required raw material is sodium dichro-
mate, not sodium chromate, and any excess base must be acidulated to
the dichromate pH with hydrochloric acid prior to our use of the re-
covered regenerant solution. 4) The concentration.  In the manufactur-
ing process there is a limit to the quantity of water which may be
added.  The more concentrated the recovered regenerant solution, the
simpler it is to recycle it to manufacturing without the incorporation
of excess water.  5) The relative concentrations of potassium and dichro-
mate.  Generally the potassium content is the limiting factor as to
                                   70

-------
Table 18    INFUJEKT ANALYSES FROM NOVEMBER,  1971

Date                own                      Date
11-1                 1763                      u-11            1899
                     1930                                       2231
                     2478                                       2241
                     I608                      11-12            1631
H-2                 1906                      n-i5            1U|6
                     2647              .                         i?79
H-3                 1930                      11-16            2089
                     2352
                     1864
11-4                 2524
                     1194                      11-17            0965
11-5                 1718                                       1598
                     1447                      11-19            1971
                     I960                                       1945
                     2020                      11-22            1945
11-8                 1919                                       2246
                     1812                                       2382
                     1839                                       2382
                     2023                      11-23            2382
                     2138                                      *8472
11-9                 1918                                     *15376
                     2080                                       1357
                     2005                      11-24            3286
                     1749                                       2358
                     1622                                       2487
11-10                1764                      11-26            2059
                     1749                                       2200
                     1824                      11-29            1688
                     1583                                       2517
                     1667
                                              Average           1952 ppm*

     * Excluding 8472 and 15,376 readings resulting from an error
       in pumping zinc chromate to the holding tank.
                                   71

-------
Table 19    THE WEIGHT OF SODIUM DICHROMATE RECOVERED FROM THE RESIN
            COLUMNS, A MEASURE OF EXCHANGE CAPACITY
Cycle                Pounds Of                     Pounds Of
Number             Na2Cr20y.2H20                 NagCrgO
                   Recovered From                Recovered From
                      IRA-900C                      Dowex 1X8

  1                      U80                            965
  2                      602                           1021
  3                      6jk                           1142
  k                      86k                            753
  5                      509                            925
  6                      892                            705
  7                      633                            9W»
  8                      768                           1005
  9                      820                           1101
 10                      717                            924
 11                      6U2                            939
 12                      766                            999
 13                      719                            982
 14                      738                           1081
 15                      770                           1079
 16                      7^4                            999
 17                      650                            905
 18                      802                            923
 19                      620                            954
 20                      76^                           1053
 21                      798	

Average Capacity
  lbs/100 cu ft          710                            970
                                   72

-------
the  quantity of regenerant solution which can be recycled to a given
batch.  No more than  hOOO Ibs of potassium (expressed  as  potassium
chloride) may be recycled per batch.

As we explained in our description of the laboratory work, we have
attempted to minimize the dilution of the regenerant (by  rinse water),
minimize the loss  of  potassium (via the necessary rinsing and back-
washing of the resin  column, and maximize the dichromate  content to
facilitate the recycle and prevent the accumulation of dilute solu-
tions in storage.

In Table 20 are listed typical results of the potassium and dichromate
concentrations of  regenerant solutions recovered following plant use.
 Table  20    TYPICAL RESULTS OF THE CONCENTRATION OF THE RECOVERED
             REGENERANT SOLUTIONS

                  Potassium Chloride            Sodium Dichromate
                             Concentrations in g/100 ml
 Resin           Dowex 1X8      IRA-900C      Dowex          IRA-900C

 Cycle
  No.
   1             6.960          6.173         7.265          k.khO
   2             7.879          6.921         7.492          5.334
   3             7.053          8.095*        8.443          5-564
   4             7.817          8.108         7.874          6.275
   5             7-460          8.721         8.380          6.604
   6             9.136*         8.651         8.565          6.614
   7             7.180          6.9^3         8.600          6.176
   8             8.391          7.660         3-252**        5-688

 Ave             7.74           7.66          8.08           5-69

     * Values of greater than 8% KC1 result either when the column
        rinse used to refill the regenerant tanks contains a signifi-
        cant amount of potassium, or when solutions of greater than
        Q% are synthesized to suspend the resin to facilitate removal.
        See Appendix D.

     ** This low value was the result of a re-regeneration of a column
        because it was bleeding the yellow dichromate color initially.
        The column was not fully exhausted prior to regeneration.
        The low value was not included in the calculation  of the
        average concentration.
                                    73

-------
COMPARISON BETWEEN INFLUENT BICHROMATE AND THE BICHROMATE RECOVERED IN
THE REGENERANT SOLUTIONS

We were interested in comparing our recovery in the plant with the re-
sults previously obtained in the laboratory.  The plant is not equipped
with an integrating meter, or any continuous influent monitoring device.
To determine total influent to the exchange system we sampled the influ-
ent solution approximately once per shift and analyzed each sample
iodometrically for its chromate content.  We recorded the flow rate to
each column and the plant operators attempted to maintain this flow
rate, or they recorded those periods when the flow to the system was
diminished or stopped.  We then accumulated influent flow in gallons
and, by assuming that the sample analysis was representative of the
entire interval between samples, we were able to estimate the influent
chromium content as shown in Table 21.
Table 21    COMPARISON BETWEEN INFLUENT AND RECOVERED DICHROMATE IN
            PLANT OPERATION

Date           Estimated      Concentration Of      Pounds of Sodium
July, 1971     Gallons        The Influent          Dichromate In The
               Treated             ppm                 Influent

1/1             U2,700             2000                   712
7/2             52,725             2000                   876
7/3             61,600             2000                  1027
7/7             25,200             3063                   639
7/7             U2,000             3060                  1072
7/8             55,200             2560                  1179
7/9             37,200             2306                   717
7/10            5^,300             2317                  10^9
7/12            71,000             2195                  1300
7/13            30,900             2195                   56U
7/1^            85,200             2120                  1506
7/15            81,500             271+3                  1862
7/19            20,000             3369                   562
7/21            19,950             3636                   605
7/26            29,500             2032           '        500
7/27            25,800             1998                   430

Totals         73^,775 gallons                         lU,600 Ibs

Total Na2Cr2Oy.2H20 in the influent                    lU,600 Ibs
Total Na2Cr20j.2H20 recovered                          13, ^83 Ibs
     (From Table 23  p 7 7 )

Difference (Recycled during rinse following             1,117 Ibs
regeneration and filter rinsing.)

                                   74

-------
 Table 22   A COMPARISON OF THE DOW AND ROHM AND HAAS RESINS

                                         Rohm & Haas
                                          Amberlite
                                           IRA-900C

 Average Volume of Influent
 (Gallons/100 cubic feet of resin)          44,000            59 QOO

 Average Weight of Dichromate Recovered
 (Pounds NagC^Oj^H^O/lOO cubic feet)
 Average Concentration of the Recovered
 Solution (Grains Na2Cr20T.2H20/100 ml. )     5.69^)           8.0&W
 NOTES:
      (l)   Based on our laboratory work, we predicted between 708
            and ?80 pounds.  (See Table 2)

      (2)   Laboratory results were widespread from two different resin
            samples.  One lead us to predict 916 pounds, the other 1,257
            pounds.  (See Table 2)

      (3)   This is slightly higher than our laboratory result of 5.43.
            (See Table 12)

      (U)   Again, laboratory results varied from 7«30 to 8.70.
            (See Tables 13 and 14)
 A SUMMARY OF TWO YEARS OF RECOVERY

 The recovered regenerant solution is routinely analyzed for its potassium
 content by the tetraphenyl boron procedure reported  in Appendix A and
 for its chromium content by the iodometric procedure reported in Appen-
 dix  A.   These potassium and chromium analyses are required before the
 recovered regenerant  solution can be returned  to the processing tanks
 for use in the manufacture of the zinc yellow  or zinc potassium chromate
 pigment.   Thus it was routine to record the amounts  of potassium chlor-
 ide and sodium dichromate recovered  from the ion-exchange unit.  These
 data are  recorded in  Table 23.

Table 23  is sub-divided on a monthly basis to  coincide with our plant
operating information.   It shows the number of cycles each column has
used, the quant it ites of potassium chloride and sodium dichromate which
were recovered from each cycle,  the  average quantities of these chem-
icals recovered from  each of the resin columns and some trends, with


                                   75

-------
time as a variables (i. e., an indication of resin durability or resin
life.)  This study accounts for 362 operating cycles, divided approxi-
mately equally between the two resins.  It reports the recovery of
319,801 Ibs. of potassium chloride and 337,88? Ibs. of sodium dichro-
mate over approximately a two year study period.
Table 23    A SUMMARY OF THE RECOVERY FROM THE ION-EXCHANGE SYSTEM OVER
            A TWO-YEAR PERIOD

From Start-Up Through April, 1971
Cycle
 No.
KC1 Recovered Pounds
             Dowex 1X8
              IRA-900C
   1
   2

   U
   5
   6
   7
   8
Sub-Tot
Ave
Total 16 eye

May, 1971

   1
   2

   U
   t;
   6
   7
   8
   9
  10
Sub-Tot
Ave
Total 17 eye

   1
   2
   3
   U
Sub-Tot
Ave
Total 7 eye
816
1027
1099
1000
976
1216
1076
Or-*
Opj
1008
15,177
720
11*1*1*
1001
735
776
790
958


918
ll*,368
1359
935
938
1155
1*W
1097
575
600
675
85!*
1025
810
790
1785
711k
889
Ibs
772
693
985
830
783
675
801*
695
833
79^
Ibs
775
632
833

22T5
7^7
       Recovered

Dowex 1X8      IRA-900C
                                  215
                                  820
                                  868
                                  971
                                  219
                                  61*0
                                  786
                                  55
                 200
                 1*32
                 607
                 724
                 1*58
                 300
                 803
                                  61*0
                                       9181* Ibs
                 508
       6627 Ibs
802
85!*
936
785
798
81*9
835



5^59
837
12,031
1086
1079
999
5o59
1017
570
691
738
61*5
257
578
689
61*7
66U
693
ol72
617
Ibs
670
585
722
1977
659
            Ibs

-------
Table 23    (CONTINUED)
Cycle
No.
July,  1971
KC1 Recovered Pounds
              Dowex 1X8
               IRA-900C
1
2
3
4
5
6
7
8
Sub-Tot
Ave
Total 16
November
1
2
3
4
5
6
7
8
9
10
11
Sub-Tot
Ave
Total 19
December
1
2
3
4
5
6
7
8
Sub -Tot
Ave
Total 15

884
1003
880
1113
957
1125
996
1077
B035
1004
eye 14,943
, 1971
986
750
1045
934
987
992
952
948
937
938
889
10735B
941
eye 17,736
, 1971
1030
1004
948
939
1154
730*
898
1022
7725
966
eye 14,220

776
983
1044
1125
982
1100
898
(352)2
6908
987
Ibs

745
1000**
962
932
960
894
1073
812



737B
922
Ibs

988
970
901
946
877
878
935

6495
928
Ibs
77
       Recovered

Dowex 1X8      IRA-900C
923
954
1053
1121
1076
1054
1193 ,
(417)1
7375
1053
13,483
558
688
718
783
744
849
799
5692
712
Ibs
                                                  910
                                                 1131
                                                 1131
                                                 1182
                                                 1194
                                                  638*
                                                  761*
                                                 124?
                                                  674*
                                                  564*
                                                 1057
                                               10,489
                                                  954
                                                  883
                                                  800**
                                                  947
                                                 1051
                                                 1024
                                                 1045
                                                 1086
                                                  987
                                                 7B23
                                                  979
                                        19,312 Ibs
                                    538           942
                                                 1168
                                                  939
                                                 1051
                                                  975
                                                  864*
                                                  964
                                                 _ _
                                                 690!
                                   1022           986
                                        15,077 Ibs
                                                  1037
                                                  1137
                                                  1124
                                                   764*
                                                  1130
                                                  1030

-------
Table 23   (CONTINUED)
Cycle
 No.
KC1 Recovered Pounds
            Dowex 1X8
               IRA-900C
January, 1972
1
2
3
4
5
6
7
8
9
10
Sub -Tot
Ave
Totals 19 eye
February, 1972
1
2
3
4
5
6
7
8
Sub -Tot
Ave
Total 16 eye
March, 1972
1
2
3
4
5
6
7
8
9
Sub -Tot
Ave
Total 17 eye
967
989
797
972
990
1066
1014
987
976
1010
9755
977
18,098

957
1097
984
1123
904
955
450
1104
7574"
947
14,829

961
1024
864
1013
1020
1280
8o4
773
904
8643
960
15,712
973
903
877
887
838
966
912
938
1036

B330
926
Ibs

954
893
792
998
737
995
852
1034
7255
907
Ibs

886
1055
971
1003
886
755
625
888

70^9
884
Ibs
      Recovered

Dowex 1X8      IRA-900C
                                   1028
                                   1010
                                   1041
                                   1103
                                   1102
                                   1129
                                   1010
                                   1115
                                   1010
                                   1113
                                                               924
                                                               865
                                                               880
                                                               895
                                                               837
                                                               727
                                                               758
                                                               874
                                                              1006
                                                1066           865
                                                     18,427 Ibs
1027
832
996
1050
1078
1145
391
1498
8017
1002
15,724
882
872
873
991
1074
990
892
1000
1139
3713
968
15,900
960
910
949
887
884
965
1153
999
7707
963
Ibs
1039
943
819
916
875
900
775
920

7l8T
898
Ibs
                                  78

-------
Table 23  (CONTINUED)

             KC1 Recovered  Pounds
Cycle
 No.
            Dowex  1X8
April, 1972
1
2
3
k
5
6
7
Sub -Tot
Ave
Total 14 eye
928
1100
1126
1012
820
980
963
3929
990

 May,  1972
                      13,307 Ibs
1
2
3
4
5
6
7
Sub-Tot
Ave
Total 13 eye
908
974
896
900
900
863

5541
907

                             IRA-900C
                                 975
                                1044
                                 884
                                 801
                                1029
                                 849
                                 796
                                5378"
                                 911
                      12,336 Ibs
 June,  1972

    1            1025
    2             781
    3             650
    4             957
    5             673
    6             837
    7             916
    8             945
    9             882
 Sub-Tot.         7666
 Ave              852
 Total 18 eye
                                1131
                                 666
                                 812
                                 652
                                 535
                                 901
                                 952
                                 900

                                7468"
                                 830
                             Ibs
Recovered
Dowex-lX8
1124
859
767
1207
997
966
1098
7055
1007

1080
1040
1057
800
1176
966

Bli9
1020

1117
1077
1037
1207
719
848
870
872
901
B54~8"
961

IRA-900C
1047
1030
1297
548
1000
860
876
6658
951
13,706 Ibs
924
803
896
1056
869
935
1220
6703
958
12,822 Ibs
535
986
936
617
676
974
1109
1036
920
7789
865
16,437 Ibs
                                    79

-------
Table 23   (CONTINUED)
Cycle
 No.
KC1 Recovered Pounds
            Dowex 1X8
July, 1972

   1
   2

   4
   5
   6
Sub-Tot
Ave
Total 11 eye

August, 1972

   1
   2
   3
   1*
   5
   6
   7
   8
   9
  10
  11
Sub-Tot
Ave
Total 21 eye
    895
    923
    867
    912
    918
   .513
   5*025
    838
September, 1972
1
2
3
U
5
6
7
8
9
10
11
Sub-Tot
Ave
Totals 21 eye
913
952
904
939
952
956
938
795
897
874

9120
912

               IRA-900C
 789
 879
 907
 984
 463
"5022
 8o4
        9050 Ibs
   1341
    952
    876
   1196
   1091
    985
    662
    814
    856
    965
   1003
 10,741
    976
 973
 637
 785
 938
 94o
1020
 795
 879
 812
 828

8607
 861
        19,3^8 Ibs
                     18,901* Ibs
                                990
                                843
                                900
                                917
                                854
                                985
                                850
                                806
                                861*
                                831
                                944
                               9785
                                889
                      Recovered

             Dowex-lX8       IRA-900C
1027
986
807
892
954
721
5357
898

197
1163
935
902
818
907
667
898
785
7^8
881
8901
809

924
1040
874
924
889
766
799
788
832
864

8700
870

872
820
908
938
726

1*254"
853
9,651 Ibs
684
820
841
901
893
873
991
1019
896
812

8730
873
17,631 Ibs
803
951
800
993
927
1037
899
880
944
900
815
9949
904
18, 649 Ibs
                                    80

-------
Table 23    (CONTINUED)
Cycle
 No.
KC1 Recovered Pounds
             Dowex 1X8
                IRA-900C
October,  1972
1
2
3
4
5
6
7
8
9
10
Sub -Tot
Ave
Total 20 eye
November, 1972
1
2
3
4
Sub-Tot
Ave
Total 7 eye
December, 1972
1
2
3
4
5
6
7
8
Sub -Tot
Ave
Total 12 eye
866
918
856
881
964
84l
900
891
883
732
8732
873


802
775
807

2355
795


861
893
903
1079
944
1526

_934
7140
892

833
701
1061
830
766
877
800
834
745
1070
11,310
1131
17,249 Ibs

812
800
770
764
3146
787
5,530 Ibs

759
1294
699
1007



3119
780
10,259 Ibs
         Recovered

Dowex 1X8      IRA-900C
                                                  983
                                                  130?
                                                  1278
                                                  1206
                                                  1421
                                                  1315
                                                  800
                                                  932
                                                  1211
                                                  857
                                                 8517
                                                  852
                                                   772
                                                   930
                                                   966
                                                   987
                                                   948
                                                  1085
                                                  1000
                                                   852
                                                   842
                                                   602
                                                      20,294 Ibs
                                                   898
                                                 1216
                                                 1204
                                                 1037
                                                 1152
                                                      7,859
                                                  1135
                                                  1211
                                                   945
                                                  1111
                                                  4~5o2
                                                  1100
                                                  988
                                                  654
                                                  1223
                                                  1457
                                                  803
                                                  1801
                                                  1008
                                                   989
                                                  1112
                                                   904
                                                  1047          ,	
                                                  B6T5          £013
                                                  1077           1003
                                                       12,626  Ibs
                                     81

-------
Table 23   (CONTINUED)
Cycle
 No.
KC1 Recovered Pounds
            Dowex 1X8
              IRA-900C
January, 1973
1
2
3
4
5
6
7
8
9
Sub-Tot
Ave
Total 17 eye
February, 1973
1
2
3
4
5
6
7
8 '
9
Sub-Tot
Ave
Total 16 eye
March, 1973
1
2
3
4
5
6
7
8
9
10
Sub-Tot
Ave
Total 20 eye
967
527
1045
854
816
821
797
785

3512
827
13,691

888
823
727
750
466
812
889
805
1126
7281
809
12,647

861
761
821
784
747
911
730
776
750
759
7900
790
15,361
976
823
745
906
779
682
750
771
647
7079
787
Ibs

816
759
663
771
749
808
796


5362"
766
Ibs

715
823
587
717
795
615
818
785
811
695
7431
746
Ibs
         Recovered

Dowex 1X8      IRA-900C
                                                1245
                                                 642
                                                132?
                                                1175
                                                1110
                                                1072
                                                1167
                                  1255
                                  1248
                                  845
                                  1030
                                   90
                                  800
                                  1127
                                  945
                                  1140
                                 B4~S6
                                  942           940
                                      15,061 Ibs
                                                              1131
                                                              1187
                                                              847
                                                              988
                                                              750
                                                              623
                                                              1055
                                  82

-------
Table 23    (CONTINUED)
Cycle
 No.
  KC1 Recovered Pounds

Dowex 1X8       IRA-900C
April,  1973
1
2
3
4
5
6
?
8
9
10
Sub-Tot
Ave
Total 19 eye
May, 1973
1
2
3
4
5
6
Sub-Tot
Ave
Total 11 eye
937
827
948
1055
836
746
1137
954
912

B352
928


927
761
1095
444
1063
377
"5567
778

754
736
746
773
868
536
996
945
743
947
BoTO
8o4
16,396 Ibs

835
794
958
849
766

4204
840
8,869 Ibs
                       Na2Cr2°7*2H2°
                        Recovered
                                             Dowex 1X8
                              IRA-900C
1340
980
761
895
1055
580
1456
580
811

B4~58"
940
17,996
899
453
1526
587
1436
587
5^58
915
874
991
994
999
847
670
1238
1056
926
9^3
9538
954
Ibs
721
1055
1246
1042
878

4942
988
                                                      10,430 Ibs
 SUMMARY
                                                    Pounds of
 April, 1971
 May
 June
 July
 November*
 December
 January, 1972
 February
 March
 April
 May

 Sub-Total
      Cycles

        16
        17
         7
        16
        19
        15
        19
        16
        17
        14
        13

       169
Pounds of KSC1
Recovered

   15,177
   14,368
    6,627
   14,9^3
   17,736
   14,220
   18,098
   14,829
   15,712
   13,307
   12,336

  157,353
       83
Recovered

  9,184
 12,031
  6,046
 13,^83
 19,312
 15,077
 18,427
 15,724
 15,900
 13,700
 12,822

151,706

-------
Table  23  (CONTINUED)

SUMMARY
June, 19T2
July
August
September
October
November
December
January, 1973
February
March
April
May

Sub-Total

Total
Cycles

  18
  11
  21
  21
  20
   7
  12
  17
  16
  20
  19
  11

 193

 362
Pounds of KC1
Recovered

   15,13*
    9,050
   19,3*8
   18,90*
   17,2*9
    5,530
   10,259
   13,691
   12,6*7
   15,361
   16,396
    8,869

  162,4*8

  319,801
                                                   Pounds of
Recovered

 16,*37
  9,651
 17,631
 18,6*9
 20,29*
  7,859
 12,626
 18,995
 15,061
 20,552
 17,996
 10,U30

186,181

337,887
     * The system was down for three months to accomplish extensive
       repairs and revisions to the resin columns.
                                   84

-------
                              APPENDIX D


            USEFUL INFORMATION FOR DESIGNERS AND OPERATORS

TYPICAL OPERATING PARAMETERS
                                   Lab Scale
                                                       Plant Scale
The Influent
Chrome Cone.
Chloride Cone. (NaCl)'

Density @25°C
Viscosity @25°C
Volume
                            or
                                    .18^0 ppm
                                 21.0 g/1
                                  0.36 mol/1
                                    1.015 g/cc
                                      5.5 cps
                               from 50,000 to 80,000 gal/day
                                      2100 to 3600 gal/hr
                                            35 to 60 gpm
                                                       .0153 lb/gal
                                                       .175  lb/gal
                              or
                              or
The Recovered Regenerant Solution (Combined Portions I  and II)
Chrome Cone. (NagCrgOj.2HpO)
     =8.70 g/100 ml        .726 lb/gal
     Dowex 1X8
     Amberlite IRA-900C
Potassium Cone. (KCl)
Density @30°C

The Original Regenerant Solution
Volume
Composition
     NaOH
     KCl
Density @U3°C
Viscosity @Uo°C

The Ion-Exchange Resin
Exchange Capacity
     Dowex 1X8

     Amberlite IRA-900C
Volume Normally Used
Treatment Flow Rate
Backwash Flow Rate
Regenerant Flow Rate
Rinse Volume
                                  7.30
                                  7.87
                                 1.08 to 1.10 g/cc
                                  U x 100 ml

                                  k g/100 ml
                                  8 g/100 ml
                                        1.0764 g/cc
                                        1.0796 g/cc
                                         10 cps
                              Ik.7 to 20.1 g/100 ml

                              11.k to 12.5
                                  100 ml
                                  15 to 20 ml/min
                                  20 to kO
                                   3 to 6
                                    1000 ml
                                                       .609
                                                       .656
                                                       k K 750 gal

                                                       250 lbs/750 gal
                                                       500 lbs/750 gal
                                                       916 to 1257
                                                        lbs/100 cu ft
                                                       708 to 780
                                                       100 cu ft
                                                       40 to 60 gpm
                                                       25-50
                                                       25 to 50
                                                       to 7500 gal
                                   85

-------
                                  Lab Scale            Plant Scale

Hydrochloric Acid Requirement
     To adjust influent from pH 6.3 to 3.0    1000 - 1200 Ibs/day
     To acidify the resin column following
     regeneration                                to 300 Ibs/day

Sodium Carbonate Requirement
     To precipitate zinc carbonate by
     adjusting pH from 3.0 to 9.0                    to 800 Ibs/day

CONVERSION FACTORS

     To Convert From                  To             Multiply By
        g/100 ml                  Ibs/cu ft            0.62^3
        g/100 ml                  Ibs/gal              0. 083^5

CALCULATION OF EXCHANGE COEFFICIENTS FOR THE ION-EXCHANGE RESINS

In the Discussion section we touched briefly on the applicability of
this system to other industries.  Doubtlessly, the surest way to as-
certain this applicability is by careful laboratory experiment aion.  It
is often helpful to be able to justify such a research program initially.
One approach is by the use of meaningful calculations.

The exchange of dichromate for chloride ion on the ion-exchange resin
is an equilibrium process, and, as indicated by Kunin  , it should be
possible to calculate a selectivity constant or an exchange coefficient.
With the aid of an exchange coefficient it would appear that one could
then estimate such values as: What is the maximum permissable level of
chloride ion in the influent for satisfactory exchange to occur?  How
much of the adsorbed dichromate must be removed from the resin during
regeneration in order to assure a water-white effluent?  If one has an
intermediate regenerant solution which contains x moles of dichromate
per liter, should this be used for a further regeneration, or should it
be recycled to manufacturing?

We will describe here our calculation of the exchange coefficients and
give some examples for their use.

From the equation representing adsorption of dichromate on a resin in
the chloride form, i.e., equation (5)
          ~
             + 2 Resin Cl  ^=^   (Resin )2 Cr20   + 2Cl"         (5)

an exchange coefficient for the treatment step can be written as equa-
tion (T), where the  [   J   represents the concentration expressed in
moles per liter.
                                   86

-------
            Sol'n
            fj      |C1- Resin|2                              (7)
     a.  For the Rohm  and Haas IRA-900C resin during treatment with an
         influent pH between 2.2 and 2.5,  as  indicated in Table  2 an
         average loading at  exhaustion is  11  g/100 ml resin.  This is
         equivalent to 0.367 mole/liter.

         Thus  [cr20T=Resin]      « 0.36? mole/liter

     to.  This implies  that 0.36? x 2 or 0.13k mole of Cl" is released
         from the resin to solution during treatment.  If the volume to
         exhaust the column  is 6300 ml., then . J3k/6. 3 liters = 0.117
         mole NaCl/liter is  the increase in the NaCl concentration.

         The concentration of NaCl coming  to  the resin column is approxi-
         mately 0.36 mole/liter, so the concentration of NaCl leaving
         the resin tower is  therefore approximately O.U77 mole/liter
         and
           JC1~ Sol'n] 2   =     [Vrf] 2 -   .2275 (moles/liter)
                                                                   2
     c.  According to the data from Rohm & Haas^ the minimum capacity
         of the  IRA-900C resin is  1 mole/liter.  From our data it would
         appear  that a  reasonable  figure would be 1.115 mole/liter.  If,
         during  treatment,  0.73^ mole  has been exchanged for Cr20j, this
         leaves  0.381 mole  of chloride ion remaining on the resin and
                Ici'Resin]2    =    |p.38l| 2    =    0.1^5 (moles/liter)2

     d.  The dichromate content leaking from the column toward the end
         of the exhaustion  cycle was determined to be 0.000712 mole/
         liter, and
                           ""1                               P
                fcr207=Solfn]     =    0.000712 (moles/liter)
Therefore

                                    . 808.7
           f.?67)(.2275)  .
           (.000732 )(•1^5)
2.

      i.  For the Dowex 1X8 resin a typical loading is  ikg/WO ml of
a.
    resin or
                       Resin]    =    O.VTO mole/liter

                                    87

-------
     b.  This implies 0.470 x 2 = 0.940 mole of Cl" were released from
         the resin to solution during treatment.  Since 7840 ml of
         influent vere required to exhaust the column,

                .940/7.84  =  0. 1199 mole/liter.

         Again, since the NaCl content of the influent is 0.36 mole/
         liter, the concentration of NaCl leaving the resin column is
         0.480 mole/liter and

                 |cr Sol'n] 2  =   (.480)2  =   (.230) (mole /liter)2

     c.  From the Dow data sheet on Dowex 1X8, the capacity of the
         resin is approximately 1.4 moles/liter.  If during treatment
         0.940 mole  of chloride is exchanged for dichromate, then
         1.400 - 0.940 = 0.460 mole  of chloride remain per liter and

                [ci" Resin]2  =   (0.46o)2  =  .2116 (mole /liter)2

     d.  As in Id.

                 Icr207= Sol'nl   =   0.000712

         KDOW  =  (.230K.470)            = 718
                  (7-12xlO-^)(.21l6)

Sample Calculations:

1.   Using KR&H = 809 calculate   jcr20y~ Sol'nj  at the point of leaking
     from the column.

     From our operating data we found 4000 ml. of influent which con-
     tained an average of 7-50g sodium dichromate was sufficient to
     ""bleed" beyond a water-white effluent.

         7-50g Sodium dichromate/100 ml or

         0.2517 mole/ liter =   Cr20j= Resin]

Again, if 0.25 mole of sodium dichromate is exchanged, 0.50 mole of Cl"
is removed from the resin to the effluent.  0.50 mole/4 liters = 0.125 mole
mole NaCl/liter added to the original 0.36 mole/liter = 0.485 mole NaCl/
liter =  [cr Sol'nj
If the original cr on the resin is 1.115 mole/liter, and we have re-
moved 0.50 mole, there is 0.165 mole remaining
         and   {ci~ resin]   =   0.615  mole/liter
                                    88

-------
Therefore

          809
          [Cr207=  Sol!n]  =  1.92 x 10'1* moles/liter


          809=   (°-3T2)g  (Q.U5)
             ^          =
     or, rearranging

          [Cr207=  sol'o}   .   .00167 mole/liter

     Thus, the  effluent concentration at  "exhaustion" would have de-
     creased from 1,840 g sodium dichr ornate/liter to 0.498 g/liter.

     This is equivalent to 1.00 x 10"3g.  Cr/liter or 1.00 ppm in good
     agreement  with  our observation that  1 ppm is colorless and 2-5 ppm
     are required to detect the yellow color visually in a 4 oz. Jar
     sample .

2.   Using K=809,  calculate the maximum allowable Cl" to exchange CrgOf*
     if the influent concentration is 2400 ppm, and regeneration removes
     Cr20j= to  a  level of 0.004 mole sodium dichromate/liter of resin.
     2400 ppm is  equivalent to  0.00805 mole of sodium dichromate per
     liter.

          809      _  Cl" sol'n.  2  (.004)
                      (. 00805 )( 1.142)^
                                                             /
          (Cl-)2      ( 809 X. 00805) (1.304)
          (Cl-)2   « 2123

          (cr)    - 46

     Thus chloride concentrations of 0.36 to 0.50 mole/liter should never
     present a problem if the regeneration is satisfactory.

3.   If 1840 ppm influent is being treated, what is the effluent concen-
     tration from this column at "exhaustion" if we assume a maximum load-
     ing of 0.45 mole sodium dichromate per liter of resin?

          1.840g. NapCr90T.2H 0/liter    = 
-------
would have been treated by the one liter of Ion exchange resin
at the time of exhaustion.

If 0.90 mole of Cl~ has been released from the resin into 73-0
liters, the average increase in the chloride content of the resin
effluent is


           , 0.0.23 mole Cl'/Hter


0.36 + 0.0123 = 0.3T2 mole/liter « Cl" sol'n.).   Again,  if the
original Cl" content of the resin is 1.115 mole/liter,  and 0.90
mole has been removed in the exchange, then   Cl~ resin   =0.215
moles/liter.
                             90

-------
We have mentioned from time to time the technique of suspending the ion-
exchange resin  in a salt  solution so  that  it  can be pumped with minimum
attrition or damage to the resin beads.  Figure 21 was made up from data
taken from the  Handbook of Chemistry  and Physics, ^5th Edition, and shows
the relationship between  the  percentage composition of a potassium
chloride solution and its density.  Also shown in the figure are the
densities of the two ion-exchange resins under study here.

If the density  of the salt  solution  is greater than the density of the
ion-exchange resin, then  the  resin will float in the solution.  Generally
the resin is handled most easily if  it is  suspended in a solution in
which it will settle, but only very  slowly, because of a very slight
difference  in the densities of the resin and  the suspending salt solution.
Agitation from  the  action of  the diaphragm pump is usually sufficient to
maintain the resin  in suspension.
            KC1
Dowex

Rohm & Haas
                        1.05       1.10         1-15
                               s.g. @ 20° C
      Fieure 21    PLOT OF THE CONCENTRATION OF POTASSIUM CHI0RIDE
      Figure tL                  THEIR SPECIFIC GRAVITIES
                                     91

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
1. REPORT NO.
      EPA-670/2-74-044
 2.
3. RECIPIENT'S ACCESSIO(*NO.
 . TITLE AND SUBTITLE
    AN  ION-EXCHANGE PROCESS FOR  RECOVERY OF CHROMATE
          FROM PIGMENT MANUFACTURING
                               5. REPORT DATE
                                June 1974;  Issuing Date
                               6, PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
 Donald J. Robinson,  Harold
 Kenneth  R. Libby,  Jr., and
                                                            8. PERFORMING ORGANIZATION REPORT NO
E. Weisberg, Glenn  I.  Chase,
James L. Capper
9. PERFORMING OHG -\NIZATION NAME AND ADDRESS
  Mineral  Pigments Corporation
  7011  Muirkirk Road
  Beltsville, Maryland  20705
                                                            10. PROGRAM ELEMENT NO.
                                1BB036:ROAP 21 AZQ:Task 03
                               11. CONTRACT/GRANT NO.
                                                                   12020 ERM
 12. SPONSORING AGENCY NAME AND ADDRESS
  National  Environmental Research  Center
  Office of Research and Development
  U.S.  Environmental Protection Agency
  Cincinnati, Ohio 45268
                               13. TYPE OF REPORT AND PERIOD COVERED
                                   Final Report	•	
                               14. SPONSORING AGENCY CODE
 15. SUPPLEMENTARY NOTES
 16. ABSTRACT
       Strongly basic ion-exchange resins have been shown to  exhibit a preference for
  dichromate over many other  anions in water solution.  Laboratory studies were con-
  ducted to show that this  ion  preference could be used to remove chromate from waste
  waters which were discharged  from a zinc yellow pigment manufacturing plant.  It was
  also shown that the recovered chromate solution could be recycled into product manu-
  facture without sacrificing product quality.

       From these laboratory  studies, a full-scale ion-exchange  treatment plant was
  designed, constructed, and  demonstrated.  The chromate composition of the plant
  effluent is being reduced from 2700 ppm to one to two ppm.

       The treatment system was designed to treat 60 gallons  per minute of influent
  and to discharge an effluent  which is within statutory limits  for pH and for heavy
  metal  content.  The plant was designed to require minimal manual  supervision.  The
  steps  in treatment and in resin regeneration are performed  automatically and the
  control system is interlocked to make it fail safe.  Operators are required only to
  make up regeneration solutions, to clean pump strainers and filters, to answer to
  alarms and occasionally to  differentiate between turbidity  and color as seen by
  by  colorimeter.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                 b.lDENTIFIERS/OPEN ENDED TERMS
                                                                         c.  COSATI Field/Group
 industrial  waste treatment
 *Ion exchangers
 *Pigments
 *Materials recovery
  Waste  treatment
  Adsorption
  Chromium
                  *Wastewater treatment
                   Z i nc
                  *Chemical manufacturing
                   wastes
                     13B
 3. DISTRIBUTION STATEMENT
       Release to public
                 19. SECURITY CLASS (ThisReport)'
                   Unclassified
                                                                          21. NO. OF PAGES
                                                                                102
                 20, SECURITY CLASS (Thispage)
                   Unclassified
                                                                         22. PRICE
EPA Form 2220-1 (9-73)
                92  **•*• GOVERNMENT PRINTING OFFICE: 197*-757-5»/5325 Region No. s-ll

-------