'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
------- |