EPA-600/2-77-049
February 1977 Environmental Protection Technology Series
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are
1. Environmental Health Effects Research
2. Environmental Protection Technology
3 Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Developmenl
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-77-049
February 1977
TREATMENT OF METAL FINISHING WASTES
BY SULFIDE PRECIPITATION
by
Richard M. Schlauch
and
Arthur C. Epstein
The Permutit Company
Permutit Research and Development Center
Princeton, New Jersey 08540
For
Metal Finishers' Foundation
Upper Montclair, New Jersey 07043
Grant No. R802924
Project Officer
John Ciancia
Industrial Pollution Control Division
Industrial Environmental Research Laboratory
Cincinnati, Ohio 45268
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
U.S. Environmental Protection Agency
Region 5, Library (PL-12J)
77 West Jackson Boulevard, 12th
Oiicago. II 60604-3590
-------
DISCLAIMER
This report has been reviewed by the Industrial
Environmental Research Laboratory - Cincinnati, U. S.
Environmental Protection Agency, and approved for
publication. Approval does not signify that the contents
necessarily reflect the views and policies of the U. S.
Environmental Protection Agency, nor does mention of
trade names or commercial products constitute endorsement
or recommendation for use.
Agency
11
-------
FOREWORD
When energy and material resources are extracted, processed, con-
verted, and used, the related pollutional impacts on our environment
and even on our health often require that new and increasingly more
efficient pollution control methods be used. The Industrial Environ-
mental Research Laboratory - Cincinnati (lERL-Ci) assists in developing
and demonstrating new and improved methodologies that will meet these
needs both efficiently and economically.
The project was undertaken to demonstrate that ferrous sulfide
treatment is a more effective process than conventional hydroxide
treatment for heavy metal removal. In particular, the studies were
aimed at treating metal finishing wastewaters that are difficult to
treat with present-day standards by conventional treatment methods.
The information contained in this report will be of value to EPA and
to the industry itself. Within EPA's R&D program, the information will
be used as part of a continuing program to develop and evaluate advanced
precipitation methods for removal of heavy metals from industrial waste
discharges. Besides its direct application to metal finishing wastes,
this technology may find application in the control of heavy metals from
effluents generated by a host of other industries. For further informa-
tion concerning this subject, the Industrial Pollution Control Division
should be contacted.
David G. Stephan
Director
Industrial Environmental Research Laboratory
Cincinnati
111
-------
ABSTRACT
This project involved precipitating heavy metals normally
present in metal finishing wastewaters by a novel process which
employs ferrous sulfide addition (Sulfex), as well as by
conventional treatment using calcium hydroxide for comparison
purposes. These studies consisted of laboratory jar tests and
bench scale tests to determine the chemical and physical require-
ments for the precipitation of the heavy metals and the subse-
quent dewatering of the resulting sludges. Following the
laboratory tests, pilot plant tests were made to confirm the
validity of the laboratory test results and provide realistic
operating data.
As a result, it was demonstrated that Sulfex is a technically
viable process that is superior to conventional hydroxide precip-
itation for removal of copper, cadmium, nickel, and zinc from a
given influent. And, when operated in the pH 8-9.0 range, the
Sulfex process will remove total chromium to a concentration
which is less than or equal to that from a conventional hydroxide
precipitation process. Hexavalent chromium can be removed by
Sulfex in a one-step operation. The effluent quality from either
process is dependent on the type and concentration of complexing
agents present in the influent.
This report was submitted in fulfillment of Grant No.
R802924- by Metal Finishers' Foundation under the sponsorship
of the U. S. Environmental Protection Agency. This report
covers the period April 1, 1974 to April 24, 1975 and work
was completed as of May 2, 1975.
IV
-------
CONTENTS
Page
Forewo rd i i i
Abstract iv
List of Figures vi
List of Tables vii
List of Abbreviations and Symbols viii
Acknowledgements x
I Introduction 1
II Conclusions 9
III Recommendations 11
IV Laboratory Te.sts 12
V Pilot Plant Tests 26
VI Sludge Dewatering Tests 59
VII Cost Estimates 62
VIII References 75
-------
LIST OF FIGURES
Number Page
1 FeS dosage, sludge blanket concentration and
mixing time vs. complexed copper removal 17
2 Ca(OH)2 dosage, sludge blanket concentration
and pH vs. complexed copper removal 19
3 Demonstration test pilot plant flow plan 27
4 Pilot plant package precipitator with
chevron settlers 28
5 Mixing zone solids concentration vs. residual
copper value for complexed Cu influent 34
6 Effluent pH value vs. effluent copper and
iron concentration for complexed Cu influent.. 39
7 Precipitator operating data for removal of
complexed copper 43
8 Precipitator operating results for removal of
complexed copper 44
9 Precipitator operating data for removal of
complexed Cu, Cd, Cr, Ni , and Zn 53
10 Precipitator operating results for removal of
complexed Cu, Cd, Cr, Ni , and Zn 54
11 Forty gpm Sulfex plant for combined removal of
Cr1"6, Zn, Cu, Cd, Ni , and Fe 74
VI
-------
LIST OF TABLES
Number Page
1 Common Complexing Agents Used in the Metal
Finishing Industry 2
2 Solubility of Sulfides 6
3 Simulated Plating Waste Compositions 14
14 Sulfex Jar Test Results of Metal Removal With
and Without Complexing Agents Present 20, 21
5 Jar Test Results Showing Influence of Complexing
Agent Concentration on Metal Removal Comparing
Hydroxide vs. Sulfex 23, 24
6 Pilot Plant Effluent Residuals for Copper and
Iron at Different Operating Parameter Levels ... 33
7 Copper Removal in Various Stages of the Pre-
cipitator 36
8 Sludge Composition from Various Precipitator
Zones 41
9 Effluent Metal Residuals from Precipitator Tests
with Cu, Cd, Cr(III), Ni, and Zn in Influent ... 51
10 Chemical Costs of Hydroxide vs. Sulfex for
Various Copper Concentrations 63
11 Chemical Costs of Hydroxide vs. Sulfex on Wastes
Containing Mixtures of Common Heavy Metals 64
VII
-------
LIST OF ABBREVIATIONS AND SYMBOLS
ABBREVIATIONS
DCG -- Permutit Dual Cell Gravity sludge dewatering unit
ft/sec -- feet per second
gpm -- gallons per minute
gpm/sqft -- gallons per minute per square foot
I.D. -- inside diameter
JTU -- Jackson turbidity units
MFF -- Metal Finishers' Foundation (of National Association)
mg/1 -- milligrams per liter
MRP -- Permutit Multi Roller Press sludge drying unit
ppm -- parts per million
rpm -- revolutions per minute
X(FeS dose) -- times theoretical FeS requirement
SYMBOLS
Ag2S
Bi2S3
Ca
Ca(OH)2
Cd
CdS
CoS
Cr
Cr(OH)s
Cu
CuS
EDTA
Fe
Fe(OH)2
Fe(OH)3
FeS
FeS04
silver sulfide
bismuth sulfide
calcium
hydrated lime
cadmium
cadmium sulfide
cobalt sulfide
chromium
trivalent chrome
hexavalent chrome
chromate ion
chromic hydroxide
copper
cupric sulfide
ethylene diamine tetraacetic acid
iron
ferrous iron
ferric iron
ferrous hydroxide
ferric hydroxide
ferrous sulfide
ferrous sulfate
VII 1
-------
SYMBOLS (continued)
HgS -- mercuric sulficie
H2S -- hydrogen sulfide gas
F^SO^ -- sulfuric acid
K -- potassium
MnS -- manganous sulfide
Na -- sodium
NaHS -- sodium hydrosulf ide
-- sodium sulfide
-- sodium metabisulf ite
NHo -- ammona gas
Ni -- nickel
NiS -- nickelous sulfide
PbS -- lead sulfide
S° -- sulfur (elemental)
S~ -- sulfide ion
Zn -- zinc
ZnS -- zinc sulfide
IX
-------
ACKNOWLEDGMENTS
The authors gratefully acknowledge the cooperation and
assistance of the Metal Finishers' Foundation, the sponsor
of this project. Permutit performed this work under Metal
Finishers' Foundation Subcontract for Research No. 74-1
entitled "Treatment of Metal Finishing Wastes by Sulfide
Precipitation". The financial assistance provided by the
sponsor was made possible under Grant No. R802924-01 from the
Environmental Protection Agency. The advice and cooperation
of Mr. E. Durkin, Vice President and Chairman of the Pollution
Control Committee of MFF, Mr. P. Kovatis, Secretary of MFF,
and Mr. J. Ciancia, former Chief of Industrial Pollution
Control Branch, EPA, Edison, New Jersey are gratefully appreciated,
Acknowledgment is made to Dr. J. Anderson and Dr. C. Weiss,
the co-inventors of the Sulfex process. Their varied contribu-
tions were necessary to the undertaking of this project. The
proficient direction of Mr. W. A. Beach, the Program Manager,
is also recognized. His guidance throughout the project is
sincerely appreciated. Mr. C. Wroblewski , Engineering
Technician, provided expertise in assembling and operating
the pilot plant that greatly assisted in obtaining a smooth
running project. His skillful control of the plant and
attentive data and sample collection helped in achieving
reliable evaluation of results. The help of Mr. R. Shuler,
Mr. R. Freeman, and Mr. A. LaTourette in assembling and
operating the pilot plant is also appreciated,
Mr. W. Schwoyer and Mr. S. Blair were responsible for the
laboratory evaluation tests for sludge dewatering. Their
recommendations and assistance in running the pilot plant
dewatering equipment were also of great importance to the
project.
We wish to express gratitude to Mrs. C. Glover and her
analytical staff, Mr. T. Savonick and Mr. R. Irons, who
performed all the analytical services for this project. This
work involved time-consuming analytical procedures on hundreds
of samples. Their cooperation and tolerance deserve special
appreciation.
The advice of Dr. L. Luttinger and Mr. W. Malkin, the
Project Consultants, was helpful in establishing the project
objectives and evaluating the results. Additional consultation
and reviewing of the final report by Mr. K, Quazza, Mr. M. Scott,
and Mr. A. Mindler is also gratefully acknowledged.
-------
SECTION I
INTRODUCTION
A. Hydroxide Precipitation
/ The function of waste treatment in metal finishing
facilities is to reduce the metals, dirt, grease and oil
to a level allowable for discharge to a receiving body of
water. The contents of the waste dictate whether one step
or more steps are needed. If cyanide and/or hexavalent
chromium are present in the waste, they must be pretreated
individually, and then recombined with the main waste stream
for metal removal.
The common process for metal removal is the precipitation
of the metallic hydroxides. The wastewater is treated with
either lime or caustic to cause precipitation of the metal
hydroxides. Since heavy metal hydroxides are gelatinous and
difficult to flocculate, a coagulant or flocculant aid is
usually used.
There are two basic objections to the hydroxide precipita-
tion process. First, the solubilities of certain heavy metal
hydroxides are sufficient as to allow excessive amounts of
metal to remain in solution at normal pH discharge limits.
An example of this is cadmium hydroxide, Cd(OH)2, with a
solubility ranging between 0.1 to 1.0 mg/1 as Cd + +. Compare
these numbers with many statutes5'5 which prohibit discharges
containing no more than 0.01 mg/1 of cadmium. Second, in the
presence of certain complexing agents, the precipitation of
the metal ion is incomplete. The prevalence of these complexinj
agents in metal bearing waste treatment systems is often over-
looked or unrecognized. This becomes increasingly true as
industry turns, more frequently, to proprietary chemicals for
their metal finishing formulations.
* Examples: 1. Montana State Department of Health and
Environmental Sciences MAC 16-2.4
(10)-S14480 Water Quality Standards
2. Definition Regulation Missouri Clean
Water Commission May 1974-.
-------
Some of the metal complexes are so stable that hydroxide
precipitation is thermodynamically impossible to initiate even
at a pH well above that normally permissible for discharge.
In addition, the complexes are usually found in concentrations
that are one or more orders of magnitude greater than the
concentration of the metals. The mass action principle,
therefore, enhances the stability of the complexes. A list
of the common complexing agents, together with the operations
that utilize them, is given in Table 1.
Table 1. COMMON COMPLEXING AGENTS USED IN THE
METAL FINISHING INDUSTRY
Complexing Agerrt Process
Phosphates Alkaline Cleaners
E.D.T.A. Electroless Plating Baths
Cyanides Plating Baths
Tartrates Electro and Electroless Baths
Ammonia Metal Etchants
When in solution, complexing agents such as these prevent
the complete precipitation of heavy metal hydroxides by
competing with the hydroxyl ion for possession of the heavy
metal, e.g. ,
Zn(NH3)4++ + 20H~ = Zn(OH)2 + 4NH3 (a } (1)
Equation (1) indicates that solutions which contain dissolved
ammonia tend to drive the reaction to the left, thereby
preventing removal of zinc as the hydroxide. Calculations
show that for a solution containing 100 ppm of dissolved NH3,
at a pH of 8.0, nearly 3.0 ppm of Zn++ will remain unprecipitatd.
All complexing agents will solubilize certain heavy rnetals in
a fashion similar to that given in the above example,.
Another example^ of a complexing agent encountered in
electroplating is the tartrate ion which reacts with certain
heavy metals to form metal-tartrate ion complexes, e.g.,
REFERENCES
1. Fales, H. A. and F. Kenny. Inorganic Quantitative Analyses.
New ed. Appleton Century Co., New York-London, p. 233, 1939
-------
1. In acidic solution
COOH
I
CHOH
CHOH
I
COOH
Tartaric
Acid
coo-
I
CHOH
CHOH
coo-
Tartrate
Ion
2H-
(2)
2. In alkaline solution
coo-
CHOH
i
CHOH
coo-
Tartrate
Ion
20H-
coo-
CHO-
I
CHO-
COO
Complex-forming
Tartrate Ion
(3)
3. In presence of metal ion
COO"
CHO~
i
CHO-
I
COQ-
Complex-form ing
Tartrate Ion
Cu
Cupri-tartrate
Ion Complex
increases its complex forming
value. In addition, with the
The tartrate ion, therefore,
character with increasing pH
increasing use of proprietary formulations containing various
brighteners, wetting agents, etc. complexing agents of unknown
composition can be introduced into plating wastewaters without
the operator's knowledge.
-------
B. Sulfide Precipitation
Realizing the need for a metal removal process effective in the
presence of complexing agents, Permutit initiated a program to develop
such a process. The effort was guided by the following requirements:
1. The process must be workable over a broad spectrum of waste
compositions.
2. It should not require any specialized equipment or skills
foreign to the customer.
/A precipitation technique that satisfies these requirements has
been developed. The technique involves an exchange of ions between
the sulfide of an added heavy metal and the ligand of the pollutant
ion(s). Hence, the name "Sulfex". Sulfide precipitation, by forming
extremely insoluble species, overcomes both the problem of limited
hydroxide insolubility and the problem of complexing agents."^
The question that immediately arises is, "Why hasn't seflfide
precipitation been universally adopted as the treatment method of
choice?" Two basic reasons follow: Sulfide systems have earned
reputations for generating noxious hydrogen sulfide gas and forming
colloidal and otherwise difficulty settleable precipitates. The
"Sulfex" process overcomes these objections and provides discharges
of such purity that the user can generally satisfy the most stringent
clean water codes.
How, then, does the "Sulfex" process differ from pH controlled
addition of H2S or Na2S? In simplest terms, the "Sulfex" process uses
a source of sulfide ion, a sparingly soluble but non-toxic heavy metal
sulfide, such as FeS. Its reaction with Cu ions in a solution con-
taining EDTA as, for example, the rinses following an electroless
copper plating bath takes place as follows:
CuEDTA
-2
FeS(s)
= CuS,
FeEDTA
-2
(5)
In contrast, the reaction using a soluble source of sulfide ions
such as NaS or NaHS occurs as follows:
CuEDJA"
S'2 = CuS(s)
+ EDTA=
(6)
-------
Both processes are effective in breaking up the complex
and precipitating copper sulfide. However, because FeS has a
very low water solubility, its tendency to react with water,
and thereby generate H^S gas, is low as com'pared with an
equivalent amount of a very soluble sulfide, i.e. the reaction
FeS(s) + 2H20 = Fe++ + 2 OH- + H2S ( } (7)
does not proceed as completely as does
Na+ + HS~ + H20 = OH' + H2S , , + Na+ (8)
^ o '
Calculations show that for solutions containing 20 ppm
of sulfide, added in one case as NaHS and in the other case
as FeS, the vapor over the solution contains, in the former
case, 0.8 ppm of H~S but in the latter case only about 0.003
ppm. This last number is very close to the minimum value
detectable by smell, viz. 0.002 ppm as reported in the Eighth
Edition of the Merck Index.
However, it is more important to recognize that, in the
case of reaction (7), the amount of FeS added to the solution
has no effect on the amount of H2S in the air above it. This
is true because once you have saturated a solution with an
insoluble substance, like FeS, no more can dissolve. However,
in the case of reaction (8), any increase in the amount of
sulfide that would occur with poorly controlled addition of a
soluble sulfide leads to an increase in the concentration of
H2S in the air. These are extremely important concepts in the
design of the chemical metering systems and from the point of
view of operating comfort and safety. We can sum up by saying
that the "Sulfex" system is, therefore, self-regulating with
respect to the sulfide content of the solution. No more
sulfide dissolves than is required to precipitate the toxic
ions. Systems using soluble sulfides as precipitating agents
do not have this advantage.
In theory, any metal sulfide more soluble than the
pollutant sulfide can be used as a precipitant providing
the equilibrium is such as to give residual pollutant levels
in accord with existing legal specifications. This point
can be better understood from examination of Table 2 which
is a listing of the common metal sulfides arranged in order
of decreasing solubility.
-------
In order to maximize the driving force for the precipitation,
one naturally chooses as a precipitant a sulfide having the
highest possible equilibrium sulfide ion concentration. If,
for example, one had a waste containing cadmium, copper and
mercury, one could choose as a precipitant any sulfide from
PbS or MnS. However, one would not, for example, select ZnS
as a precipitant for Ni"1"1" because their nearly equivalent
equilibrium sulfide ion concentrations would not provide
sufficient driving force for the precipitation of NiS.
In
chosen
practice, ferrous sulfide
is the precipitant normally
for obvious reasons. It has a comparatively high
solubility, it can be prepared inexpensively; and iron is,
generally, non-toxic.
Table 2. SOLUBILITY OF SULFIDES
Metal Sulfide
MnS
FeS
ZnS
NiS
SnS
CoS
PbS
CdS
Ag2S
Bi2S3
CuS
HgS
Ksp (18 to 25QQ
Sulfide Concentration
(Moles/1)
1.
3.
1.
1.
1.
3.
3.
3.
1.
1.
8.
i~i
4
7
2
4
0
0
4
6
6
0
5
0
X
X
X
X
X
X
X
X
X
X
X
X
10
10
10
10
10
10
10
10
10
10
10
10
-15*
-19
-23
-24
-25
-26
-28
-29
-49
5t
ft
f.
ft ft
ft
ft
ft
ft
-97**
-45
-49
ft
ft
3.
6.
3.
1.
3.
1.
1.
6.
3.
4.
9.
4.
74
1
46
18
16
73
84
0
4
8
2
5
X
X
X
X
X
X
X
X
X
X
X
X
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
10-
8
10
12
12
13
13
14
15
17
20
23
25
REFERENCES
Handbook of Chemistry and Physics. 50th Ed., Robert C.
Weast, Ed. The Chemical Rubber Company 1969, p. B252
** 3. Handbook of Analytical Chemistry.
McGraw-Hill, Inc. 1963, pp. 1-15,
Louis Meites, Ed.
1-19.
-------
The program used to demonstrate the superiority of ferrous
sulfide, as compared to lime, for removal of complexed and
uncornplexed heavy metals from solution was as follows:
A. Phase I - Precipitation and Filtration Studies
1. Section 1 - Laboratory Tests (jar tests)
2. Section 2 - Assembly and piping of the pilot plant
3. Section 3 - Precipitator tests
4. Section 4 - Optimization tests
5. Section 5 - Analytical work
B. Phase II - DCG (Sludge Dewatering) Tests
1. Section 1 - Laboratory tests
2. Section 2 - Full scale DCG tests
3. Section 3 - MRP tests
4. Section 4 - Analytical work
C. Phase III - Final Report
This plan was based on testing only one simulated wastewater
an electroless copper plating waste containing 20 mg/1 of
copper complexed by
Permutit
scope of
1
2
EDTA and
in conjunction with
work to include:
Rochelle salt-
However
the MFF and EPA increased the
An influent containing 50-80 mg/1 of dissolved
iron in addition to the 20 mg/1 of complexed copper
An influent containing only 5 mg/1 of complexed
copper.
An influent containing 60-90 mg/1 of complexed
copper.
An influent containing five metals (copper,
cadmium, chrome, nickel, and zinc) complexed
with EDTA and Rochelle salt. The concentration
of each metal was 4 mg/1.
An influent containing five metals (copper,
cadmium, chrome, nickel, and zinc) complexed
only with Rochelle salt.
Large scale tests of hydroxide precipitation
using lime and influent 5.
Rochelle salt is sodium potassium tartrate.
-------
For reasons of effectiveness, availability, and economy,
the ferrous sulfide is not added as ferrous sulfide but
instead is freshly precipitated by reacting an iron salt with
a soluble sulfide such as sodium sulfide, sodium hydrosulfide,
or hydrogen sulfide. These soluble sulfides are used effec-
tively in the process, to form FeS, if a source of alkali such
as lime or sodium hydroxide is simultaneously added to main-
tain the pH at a value higher than 7.0 in order to prevent
evolution of hydrogen sulfide gas.
Another advantage of the Sulfex process is its abaility
to remove hexavalent chromium in one step as opposed to the
typical two-step process used with hydroxide precipitation.
cro;
s°,
Fe(OH)3(s) + Cr(OH)3(s)+ 20H' (9)
-------
SECTION II
CONCLUSIONS
/ As a result of the laboratory jar test studies and pilot plant
demonstration tests, it is concluded that:
A. The Sulfex process is a technically viable process.
B. When treating the same influent, the Sulfex process obtains
lower residuals of copper, cadmium, nickel, and zinc than
can be obtained with the hydroxide process.
C. At the same pH value, the Sulfex process removes chromium to
a concentration which is less than or equal to that from a
hydroxide precipitation process.
/
D. The reduction of hexavalent chrome to trivalent chrome can be
performed by Sulfex in one step at slightly alkaline pH values
(i.e. pH 8.0 to 9.0). For this reason, Sulfex does not require
segregation of chrome containing wastes from other heavy metal
containing wastes.
E. Satisfactory effluent quality is usually obtained with the Sulfex
Process within the 8.5 to 9.0 pH range which is within the 6.0 to
9.5 pH range permitted by the EPA for discharge. The hydroxide
process is often not satisfactory in this pH range and sometimes
not effective even if raised above pH 10 and readjusted within
permitted discharge limit.
F. The effluent quality from either process is dependent on the
type of complexing agents present in the effluent. Examples of
the effluent quality obtained by the Sulfex process with 4 mg/1
of each metal in the effluent are:
1. For metals complexed with high concentrations of EDTA and
Rochelle salt.
copper - 0.2 mg/1
zinc and cadmium - 1.0 mg/1
nickel and chrome - 3.0 mg/1
-------
2. For metals complexed with high concentrations of
Rochelle salt only.
cadmium - <0.01 mg/1
copper - 0.01 mg/1
zinc - 0.03 mg/1
nickel - 0 . H mg/1
total chrome - 3 mg/1
3. For metals with no complexing agents
cadmium - <0.01 mg/1
copper - 0.01 mg/1
zinc - 0.01 mg/1
nickel - <0.05 mg/1
total chrome - <0.05 mg/1
G. Using the Sulfex process, the removal of a particular
heavy metal is more effective when it is in a solution con-
taining other heavy metals than when it is the only metal in
solution.
H. The Sulfex process can be applied in Precipitators
(and similar devices) at surface rates up to 2.0 gpm/sq.ft.
when tube settlers are used.
I. The required dosage of ferrous sulfide reactant is
dependent upon the type of waste being treated. It should
normally vary from about 1.5 times theoretical requirement
for wastes with no complexing agents to 3 or more times
theoretical for wastes containing complexing agents.
J. The concentration of settleable ferrous sulfide solids
in the mixing zone, the pH of the process, and use of certain
polyelectrolytes are important to obtaining satisfactory
results in the Sulfex process.
K. To treat many common plating wastes (without strong
complexing agents) containing an average of about 20 mg/1
total heavy metals, the operating cost of Sulfex compares
favorably with that of conventional hydroxide precipitation.
L. It may be more economically desirable to pretreat
wastes containing high concentrations of dissolved heavy
metals (i.e., a total heavy metal concentration greater than
50 mg/1) by hydroxide before polishing with Sulfex,,
M. Conventional dewatering methods can dewater the sludge
from the Sulfex process to a bladeable solid suitable for
disposal.
10
-------
SECTION III
RECOMMENDATIONS
It has been demonstrated in these tests that Sulfex treats
various heavy metal influent compositions to lower effluent
metal concentrations than conventional hydroxide precipitation
processes. The operational costs for Sulfex appear to be
reasonably practical when treating most plating waste compositions
containing up to a total of about 50 mg/1 of dissolved heavy
metals. However, further test work should be performed on
actual plating wastes both in the laboratory and on-site in the
field. Such tests are necessary to evaluate the effectiveness
of Sulfex under conditions that cannot be simulated. In addition,
the economics of Sulfex is best evaluated under the actual
wastewater conditions existing at the plating operation.
11
-------
SECTION IV
LABORATORY TESTS
A. Composition of Plating Wastes
There were several different compositions of simulated
plating wastes used during the course of these studies. The
initial work used a waste which would be extremely difficult
to treat by conventional hydroxide precipitation. The
composition of the waste is shown in Column I of Table 3.
The simulated waste solution was prepared by diluting a
published1^ electroless copper plating solution formula 444-. 4
times with tap water. This dilution was used in order to
obtain a waste solution containing 20 mg/1 copper ion. This
copper concentration is assumed to be representative of the
concentration of all the heavy metal ions, except iron, found
in waste effluents from plating processes.
This waste is assumed to be typical of the waste from
many electroless copper plating facilities. Although it is
not typical of the waste from the majority of metal finishing
plants, it was selected because of the high concentrations of
a strong complexing agent (EDTA) and a moderately strong
complexing agent (Rochelle salt). Such wastes are not
satisfactorily treated by conventional hydroxide precipitation
methods. The satisfactory treatment of this waste by Sulfex
would, therefore, show superiority of Sulfex over the hydroxide
method.
The fact that it is particularly difficult- to remove
dissolved copper from this waste by hydroxide precipitation
is indicated by its already high pH (11.2) without copper
precipitate being formed. The pH of the concentrated plating
solution before dilution was 13.7 and 13.8, and no precipitation
of copper was evident even after several months of standing.
This is further evidence that this solution of strongly
complexed copper is very resistant to complete copper removal
by hydroxide precipitation processes.
REFERENCES
Metal Finishing Guidebook and Directory for 1970.
Nathaniel Hall, ed., Palmer H. Langdon, pub. 1970
p. 473.
12
-------
Because of a substantial cost increase for tartrates, the
composition of the waste used for the precipitator tests was
modified by the reduction in concentration of the Rochelle
salt to one-third of its original concentration (see Column II,
Table 3). Since the concentration of the strong complexing
agent was not changed, the results were not expected to be
significantly different from those obtained in jar tests.
During the program, the EPA and MFF requested we perform
tests with other metals which were of interest. These metals
were copper, cadmium, chromium, nickel, and zinc. It was
pointed out that the majority of the waste streams they were
interested in would not contain EDTA or Rochelle salt. In
fact, there would only be low concentrations (if any) of
complexing agents. However, in order to study the effects of
these complexing agents on the heavy metals, either both, one,
or neither of them were included in jar tests with simulated
five metal wastes. These formulations are shown in Columns
III, IV, and V, respectively, of Table 3. The effect of each
of these complexing agents on the five heavy metals was then
observed.
B. Parameters Studied in the Jar Tests
1. Description of Jar Tests
The jar tests were performed on 4 liter and 0.5 liter
samples of the simulated plating wastes. When the chemical
reactants were added (i.e. FeS, Ca(OH)2, etc.), the sample
was mixed rapidly (100 rpm) for periods of 30 to 60 seconds,
This step simulates the fast mix operation of a coagulation
system. The FeS was added to the jar tests samples by
either of two different methods: (1) separate solutions
of ferrous sulfate (FeSOi4 7H20) and sodium sulfide
(Na2S-9H20) were added directly to the rapidly mixing
heavy metal waste solution or (2) the ferrous sulfate
solution was slowly added to the rapidly mixing sodium
sulfide solution, forming a concentrated slurry of FeS
that was then added to the rapidly mixing heavy metal
waste solution. These tests confirmed previous studies
which showed the first method was unacceptable because it
promoted colloidal heavy metal sulfide suspensions, whereas
the second method promotes larger FeS particles and clearer
supernatants.
13
-------
Table 3. SIMULATED PLATING WASTE COMPOSITIONS
Chemical
(1) Rochelle salts (as I
(2) EDTA as (Na4EDTA)
(3) Sodium hydroxide (as NaOH)
(4) Soda Ash (as Na2C03)
(5) Copper Sulfate (as CuSO4-5H20)
Copper (as Cu++)
(6) Cadmium Chloride (as
Cadmium (as Cd++)
(7) Chromic Acid (as
+4.5 mg/1 H2SO4
(to reduce Cr+6 to Cr+3)
Chrome (as Cr+++)
Nickel (as Ni++)
(9) Zinc Chloride (as ZnCl2)
Zinc (as Zn++)
(10) pH
(11) Suspended Solids
I II III
ppm ppm ppm
a Tartrates) 333 127 126.9
45 45 45
aOH) 112.5 10 19.9
67.5 67.5
04-5H20) 78.5 78.5 15.75
20 20 4.0
6% CdCl2-23jH20) - - 10.17
4.0
7.7
mg/1 Na2S205)
3) - -
4.0
Ni(NO3) -6H20) - - 19.3
4.0
2) - - 8.6
4.0
11.2 8.8-9.0 7-7.5
less
than
1.0
IV
ppm
126.9
-
19.9
-
15.75
4.0
10.17
4.0
7.7
4.0
19.8
4.0
8.6
4.0
6.8
_.
V
ppm
-
-
-
-
15.75
4.0
10.17
4.0
7.7
4.0
19.8
4.0
8.6
4.0
-6.0
_
14
-------
The rapid mix period was followed by slow speed mixing
(25 to 30 rpm) lasting from 10 to 6G minutes. The samples
were then allowed to settle without mixing for periods of
1 to 30 minutes. The amount of timr allowed for settling
depended on whether precipitate settling rates were being
measured or the sample was only being evaluated for
dissolved heavy metal reduction after settling and filtering
To determine the effect of sludge blanket contact,
previously settled solids formed in the larger volume jar
tests were added to the mixing stage of the smaller volume
jar tests. To measure the total concentration of dissolved
heavy metals, iron, calcium, etc. remaining in a sample
after settling, the supernatant was immediately filtered
through a O.U5 micron Millipore filter to remove colloidal
constituents before the chemical analyses were made. Before
the analyses, the samples were acidified to a pK less than
2.0 in order to prevent post-precipitation of heavy metals.
All analyses were made by atomic absorption spectrophoto-
metry.
2. Jar Tests_on Waste Containing Complexed Copper
The objective of these jar tests was to determine the
effects of various operating parameters on precipitating,
coagulating, and settling the copper from solution. These
operating parameters would then be optimized during
Precipitator tests. It was assumed the effect of these
operating parameters on copper removal would be similar for
the removal of several other heavy metals. The target for
residual copper ion in a filtered effluent was less than
or equal to 0.1 mg/1 as Cu++. This is at least ten times
lower than is usually obtained from the hydroxide process
(1.0-1.5 mg/1). The parameters studied were:
a. Ferrous sulfide (FeS) dosage
b. FeS sludge blanket concentration
c. FeS sludge blanket solids contact time
with the liquid waste (i.e. mixing time)
d. Jse of coagulant aids for promoting faster
clarification when FeS is used as the reactant
(1) Type of coagulant aids required
(2) Concentration of coagulant aids required
(3) Addition time of coagulant aids
e. Use of lime, Ca(OH)2, as the reactant at various
pH values and sludge blanket concentrations
15
-------
3. Results of Jar Tests on Complexed Copper Waste
The jar tests show that increasing FeS dosage, increasing
mixing time (i.e. liquid-solids contact time), and increasing
sludge blanket concentration each have individual positive
influences on copper removal. Combining all of these
"positive" parameters is beneficial, but the most important
single parameter appeared to be the sludge blanket solids
concentration. The sludge blanket forms as FeS, CuS., and
other possible precipitates, build up and concentrate in the
mixing and settling zones of the Precipitator.
a. The jar tests indicate that for each equivalent of
complexed copper between two and three equivalents of FeS
give excellent copper removal when a concentrated sludge
blanket is maintained. The pH during these jar tests was
maintained in the 7.0 to 8.0 pH range. Figure 1 shows the
effect of increased FeS dosage on copper removal as a function
of sludge blanket concentration. The best results are
obtained when the sludge blanket is the most concentrated
(5000 mg/1 suspended solids). Copper removal is improved as
the FeS dosage is increased in the presence of a sludge
blanket. It is also seen that with no sludge blanket, the
removal of copper is improved with increasing dosages of
FeS.
b. Figure 1 demonstrates that a sludge blanket concen-
tration of several thousand mg/1 suspended solids (as FeS)
should be maintained to obtain essentially complete copper
removal.
c. Figure 1 shows that increasing the mixing time
from 10 minutes to 60 minutes with
gives only a modest improvement in
with an adequate FeS concentration
with a sludge blanket sufficiently
liquid-solids contact period of 45
a dilute sludge blanket
copper removal. However,
continuously supplied and
concentrated with FeS, a
to 60 minutes is sufficient
to reduce the residual copper to the desired level.
d. Results of the jar tests indicate that a coagulant
aid is important to maintaining a concentrated sludge
blanket and a clear effluent. The presence of metal
complexing agents apparently prevented the common inorganic
coagulants such as aluminum sulfate and ferric sulfate from
flocculating well in the simulated waste sample. It is
found that 1 to 5 mg/1 of a cationic natural gum derivative
polymer is very effective in maintaining a concentrated
sludge blanket. Some synthetic cationic polyelectrolytes
also displayed effectiveness in coagulating the colloidal
16
-------
20
10
£ I
0.
Q.
0 05
o w-°
0 I
LU W>
QC.
- 0.05
0.01
0.005
0.001
0
Sulfex process jar tests with complexed Cu influent
pH values maintained between 7 and 8 during tests
1 1
FeS DOSAGE
30 MINUTES
MIXING
NO SLUDGE
BLANKET
EFFLUENT COPPER
IOMIN. MIX.-
30MIN. MIX.-
60MIN. MIX.-
500 mg/l
SLUDGE
BLANKET
45MIN.MIX.
5000 mg/l
SLUDGE BLANKET
X=THEORETICAL EQUIVALENT CONCENTRATION OF FeS
REQUIRED TO PRECIPITATE 20 mg/l Cu++ =27.7 mg/l FeS
I I I I i i i I i i
2X 3X
FeS DOSAGE
4X
5X
6X
Figure I. Influence of FeS dosage, sludge blanket concentration
and mixing time on copper removal.
17
-------
metal sulfide particles. However, during extended mixing
periods, the sludge appears to be more resistant to
mechanical breakdown into colloidal particles when
coagulated with the natural gum polymer than with the
synthetic polyelectrolyte. The cationic polyelectrolytes
are most beneficial when the polymer solution is added to
the mixing stage a few minutes after the addition of the
other chemicals.
A series of jar tests were performed to evaluate the
copper removal capability of the hydroxide process when
treating a strongly complexed copper waste. The hydroxide
was added as lime. The results of these tests are shown
in Figure 2. For these tests, the simulated copper plating
waste at pH 11.2 was first acidified to pH 4.5 for 5
minutes in an attempt to destroy the copper complexes.
Then, lime, in various dosages, was added to the copper
plating waste which was mixed, settled, filtered and analyzed
by the same procedure as used in the Sulfex tests. Figure 2
shows the pH of the solution must be raised to 12 in order
to reduce the copper residual to 4 mg/1. Figure 2 also
shows that lime in massive doses is required to achieve a
pH of 12.
4. Jar Tests on Wastes Containing Copper, Cadmium,
Chromium, Nickel, and Zinc
In addition to copper, the removal of cadmium., chromium,
nickel, and zinc was studied. The influent concentration
of each metal was 4 mg/1. In one series of Sulfex jar tests,
the five metals were in a mixed solution. In another
series, the five metals were studied as individual solutions.
Parameters which were studied included the presence versus
the absence of complexing agents, the influence of pH, and
the ferrous sulfide dosage. The results of both series of
jar tests (i.e. individual versus mixed ions) are shown in
Table 4. The conclusions from these Sulfex jar tests are:
a. The removal of a particular metal is usually
better when it is in a solution containing other heavy
metals then when it is the only metal in solution.
b. The presence of high concentrations of Rochelle
salt and EDTA give higher residuals for nickel and chrome
than for copper, cadmium, and zinc.
c. Increasing the pH reduces the residual metal
concentration for all five metals whether complexing
agents are present or not. The iron which is introduced
into the solution as ferrous sulfide also has its
effluent concentration reduced by increasing the pH.
18
-------
hydroxide process jar tests with complexed Cu influent
>
IE
QL
12
10
8
20
18
16
-------
X
o
o
rH
X
0
rH
-
H
H
H
*
O
^
Q
«.
C3
U
fa
o
,_q
f~Q
O
1
w
EH
D
i
EH
W
tf
I j
X
H
CO
EH
W
CM
CO
EH
W
O
O
H
X
W
I-H1
s
o
U
EH
D
O
ffi
EH
H
IS
P
'Z
ffi
EH
H
C
CO
Q
H
X
in
X
ro
X
CM
X
in
r
^__^
4J
c ^
^ 0) CD
6 * -P
(1) (D --^ G X
rH M CD 0)
Q T! ^ "?
cd 3 C PS
EH tJ1 0)
Q) tji CT>
tf <; c
H
w cn x
t)
CD CD
X C
H O
ii r~i
(D CD
X C
H O
S H
O
CD CD
X C
H 0
S H
T3
CD CD
X C
H O
S H
1
H CD
TS C
C 0
H H
CD CD
X C
H 0
S H
T3
CD CD
X C
H O
S H
O
0) CD
X G
rH O
S H
H CD
t) C
C 0
H H
-f-
4.
-f.
1
+
+
1
1
1
^J* i-H
O O CN
* *
0 0 rH
rH rH
O O O
t
O O CN
O CN CTi
i
O O CN
CN CN CN rH rH CTl
OHO OOO CNHrH
OOO OOO OOO
V V V V
in
in en o m H i i
HHO cocnr^ ^cn^
OOV OOO CNCNO
-P
i i
fd
00 H CD
H en ^r
-------
H
H
W
CO
u
O
CO
EH
W
3
U
!Z
H
X
w
Oi
u
CO
EH
§
U
CO
CQ
Q
§
a
w
x
w
a
D
CO
XI
(0
N
Q
Q) "w
e * -p
3 C <
&1 Q)
0) Cr> Cn
« «; c
H
co Cn X
0) C 0)
fa -H rH
X ft
H 0) g
ft) rH O
o a u
H 6
4-> O -P
0) U 3
M O
O £ X!
0> 4-> 4->
X + I
X
o
0
rH
X
o
rH
X
in
X
ro
X
CM
X
m
rH
X
t3
a) in
X c
H 0
a H
m tn
X C
H 0
S H
T3
0) CO
X G
H O
S H
rrj
0) tn
X C
H O
a H
H tn
a G
G 0
H H
ro in
ro O o
romin r-\ -3? oomm
roOO rHOO CMOO
( t*l
ooo ooo ooo
V V
rH in <3- rH in in
iD^or- oomcN mr^o
roroo OOO rHOO
V
m
ID ro O
* »
CM rH O
V
CM in m
ooo
t
ooo
V
oo >*o m
ooo
ooo
V
in rH m
ooo
it
ooo
V
^ i i m
^ o o
ooo
V
o
rH
21
co
G
N
cr>
e
co
Cn
4J
rH
a
u
o
Cn
6
r-
CM
O
G
LO
1)
T)
D
rH
U
C
H
tn
-p
tn
H
X
0)
rH
a.
o
u
-------
d. One to one and a half equivalents of sulfide per
equivalent of heavy metal is sufficient to reduce the
residual concentration of each heavy metal to the desired
level when no complexing agents are present. When
complexing agents (EDTA and Rochelle salt) are present,
higher sulfide dosages are necessary to achieve the
desired removal of copper, cadmium, and zinc. Complexed
nickel and chrome were not reduced to the desired levels
with economically acceptable dosages of sulfide. However,
removal of complexed nickel may be improved by operating
at pH values greater than pH 10.
Jar tests were also performed to determine the influence
of Rochelle salt and EDTA concentrations on the precipita-
tion of copper, cadmium, chromium, copper, and zinc by both
sulfide and hydroxide precipitation. In these tests,
solutions containing all five metals, each at a concentra-
tion of 4 mg/1, were used. To these solutions either
Rochelle salt or EDTA was added at its standard concentra-
tion (127 mg/1 for Rochelle salt, 45 mg/1 for Na^EDTA).
Also, to determine the effect of the presence of both
Rochelle salt and EDTA, the two complexing agents were
combined in the five metal solutions as fractions of their
standard concentration. The results of these tests are
shown in Table 5. These results indicate that treatment
with the Sulfex process reduces all the heavy metal
concentrations to substantially lower values than treatment
with the hydroxide precipitation process except for
Chrome III which is reduced to about the same concentration
by both processes. For most heavy metals, the presence of
EDTA has a more deleterious effect on the removal of the
heavy metals than the presence of Rochelle salt (i.e.
tartrate ion). However, the presence of Rochelle salt
does interfere with the removal of Chrome III more
severely than does EDTA. The results of the jar tests
also indicate that Chrome III removal, even in the Sulfex
process, is as chrome hydroxide.
Jar tests were also made on an actual metal finishing
waste which contained 32 mg/1 of total chromium of which
25 mg/1 was dissolved hexavalent chromium. This sample
also contained 5.9 mg/1 of dissolved zinc and traces of
copper, iron, and manganese. The presence of strong
complexing agents was assumed to be low or absent. Sulfex
was tested to determine its effectiveness in removing the
hexavalent chromium without the conventional preliminary
step of adding a reducing agent at low pH (such as sodium
22
-------
EH
2
EH
ff
co
w
l_3
H
ffi
U
O
CM
O
w
CJ
2
w
D
CM
2
H
O
2
H
S
O
CO
CO
EH
!H
D
CO
ffi
EH
CO
W
EH
OH
(^
1-3
w
s
EH
W
H
CM
fn
O
2
O
H
EH
EH
H
ft
H
U
§
ft
2
O
2
O
H
EH
CN
EH
2
W
U
2
O
U
|C{J
EH
O
W
Q
£n
53
§
EH
X
W
CM
D
CO
C3
W
Q
H
X
o
Q
>H
as
o
2
H
On
(dj
s
o
u
1
CO
! "I
^
EH
S
>H
rtj
W
as
in
CO
1
00
oo
as
ft
rl
n3 -P
-P fd
0)
g W
jj
4S W
O d>
rd 4J
Q)
X
M-( 0}
Q 14-1
i-l
0)
fo
f-i
"\
0^
g
r§
i i
X.
en
e
H
2
r-J
'^\
en
e
(H
o
rH 3 rH
\ co! \
tT1
e -
o
^J* '
c^
w
fS ffi
H ft
IT)
-P -P
c id
0
u w
-p
11 C/5
o 01
rl 4-1
-P
3 0)
i ( n3
O -r!
W X
O
to 4S M
Cn
e
TJ
u
, 1
v^
Cn
e
O Til'
J 3
H (j as
5
U
Cn
g
-K
co
OJ
Cw
*
X
as
o
*
to
P
as
o
K
CO
01
CM
X
O
K
CO
(U
CM
O
K
CO
0)
ft.
3S
0
X
CO
o;
ft,
as
/-N
O
LT)
co
OQ
o
co
o
sT
o
r--
o
o
o
o
o
CM
O
O
CN
O
co
1-1
c
o
o
C-J
o
CN
C
O
o
V
O
V
CO
rH
o
v
c.
o
v
00
o
V
CO
-p
0)
B
X
0)
w
Q)
-P
O
H
O
(1)
^
ft
0)
T!
H
X
O
^
T3
N
C
H
0)
X
0)
i-i
ft U
e c
o o
O CJ
X
0)
s
i-H 0)
\i-H
Cn H
g OJ
f,
O
s
w
ml
P O (N 0
CO CJ rH (2
O C
O O
U
T3 C
-P O
CO O
0)
u
o
o;
ol
-------
^-^
>l
£>
EH
K?
<
CO
W
r^
Hi
W
K
U
O
«
fa
0
W
o
2;
w
D
1-3
fa
S
H
O
2
H
s
0
K
co
co
EH
A
D
co
w
S
EH
CO
W
EH
PS
<
1-2
a:
K
W
H
fa
fa
O
&
o
H
EH
<
EH
H
ft
H
O
W
S
ft
2;
o
2;
o
H
EH
<
K
EH
&
W
o
^
0
u
<
EH
Q
W
Q
S
rij
^_J
C
o
o
«*
g
P*J
§
EH
<<
H
K
EH
X
W
fa
g
CO
^3
W
Q
H
X!
§
Q
><
K
O
S
H
OS
<
ft
S
o
u
1
co
hH1
g
PL]
23
in
,
CO
1
n
CO
K
ft
H
tfl 4-1
4J (0
0)
g
£ CO
U CU
tO 4-1
0)
X
4-1 OJ
O M-t
rH
H 3
\ CO
Cn
o
^J1
o>
cn
C ffi
H ft
(0
4-1 4-1
C tO
O
o cn
4-1
C CO
O 01
H 4J
4-)
D 01
rH -O
0 -H
cn x
o
£ M
u -a
/rt S*,
' tu t^f
in H a
(D
rH
f\
rM
Ifl
EH
0)
fa
rH
\
CT
g
C
(SI
, ;
\
Cn
H
Z
rH
\
.r
^i
u
rH
\
t?
T3
U
rH
\
Cn
g
D
U
i (
\
g
*
co
01
fa
*
K
K
O
*
CO
m
fa
ffi
o
K
W
0)
fa
EC
o
*
co
CD
fa
o
K
co
0)
fa
ffi
o
X
CO
0)
fa
K
O
X
Q)
rH .
ft U
g C
0 O
0 0
co
o
CO
<£
o
in
rH
d
Oi
^
o
IT)
o
CN
rH
O
O
CM
(N
ro
O
O
in
CN
in
o
o
v
CN
O
o
o
ro
m
O
IT)
O
O
V
LO
O
O
V
O
V
ft
O
u
JP
MH
O
X <\
0)
0
U
wl
24
o
u
fal
O
-I
\
LO
o
o
v
10
o
CM
o
oo
I-l
o
in
o
o
v
in
o
o
v
CN
^J<
CD
o
v
O
O
0)
rH
OH O
g C
0 O
U U
4-1
C
0>
s
fO
Q)
X
Q)
CO
cn
4J
C
O)
cn
CD
JH
a
co
-------
metabisulfIte) to convert the hexavalent chrome to
trivalent chrome^. The results of these tests show that
between 50 and 100 mg/1 of ferrous sulfide (about the
same dosage as that required for removal of 20 mg/1 of
complex copper) is sufficient to reduce the effluent
total chrome level to less than 0.05 mg/1 (as Cr) and
the effluent zinc level to about 0.05 mg/1 (as Zn).
Therefore, a one step Sulfex process appears to be a very
effective method of removing hexavalent chrome as well as
other heavy metals from solution.
REFERENCES
Ward, A. E
Treatment.
Handbook, 3rd
New York, Van
pp. 359-360.
Chapter 11, Wastewater Control and
In: Electroplating Engineering
ed., Graham, A. K. (ed.), New York
Nostrand Reinhold Company, 1971,
25
-------
SECTION V
PILOT PLANT STUDIES
A. Description of Pilot Plant
The Sulfex pilot plant was assembled and piped in The
Permutit Research and Development Center which is located
in Princeton, New Jersey. A flow plan of the pilot plant
is shown in Figure 3. Figure 4- shows cross-sectional views
of the Precipitator used in the pilot plant system.
Supply to the Precipitator was a mixture of well water and
a prepared solution containing the desired metals. Well water
was pumped through a 2 inch inlet line to the upper part of the
mixing zone of the Precipitator. The complexed copper plating
solution and additional metals were fed into the inlet line,
by means of chemical metering pumps, approximately 25 feet
upstream of the Precipitator. The flow rates of the well water
and metal solutions were controlled to provide the desired
metal concentrations in the influent to the Precipitator.
Preparation and feed of treatment chemicals was accomplished
in the following manner. Freshly precipitated ferrous sulfide
(FeS) slurry was introduced into the influent line just prior
to entering the Precipitator mixing zone. The ferrous sulfide
slurry was produced by reacting sodium hydrosulfide and ferrous
sulfate in a 10 gallon head tank under rapid stirring,. A 1.5
gpm dilution water stream entered the bottom of the slurry tank
where it was detained and mixed for about 7 minutes. Sodium
hydrosulfide (as a M-5% solution of NaHS) was fed, by means of
a small chemical metering pump, to the dilution water line
before entering the head tank. Ferrous sulfate (FeS04) solution
and lime (Ca(OH)2) slurry were metered into the sodium hydro-
sulfide solution in the FeS head tank.
The influent, dosed with fresh FeS slurry, passed downward
through the mixing zone where it was mixed with accumulated,
previously precipitated FeS particles. At the bottom of the
mixing zone, the stream changed direction and flowed upward
through the port area into the settling zone. The cross-
sectional area perpendicular to the upward flow continuously
increased so that the upward flow velocity continuously
decreased and allowed the precipitated solids to settle out
of the upward moving liquid. At the top of the settling zone,
the liquid surface was covered with chevron settlers which
settled out most of the fine or light precipitates that did not
26
-------
LU
IS
Figure 3. Sulfex pilot plant flow plan.
27
-------
o
Figure 4. 5'-8" wide * 5-6" long xll'-0"high package Precipitalor
with chevron settlers.
28
-------
settle in the lower part of the settling zone. A very concen-
trated fluidized sludge blanket accumulated in the settling
zone to achieve additional solids-liquid contact time. The
clarified liquid then rose out of the chevron settlers
establishing a clear liquid zone at the top of the Precipitator.
The clarified liquid flowed, by gravity, through an effluent
collection pipe into the effluent line. For these tests, most
of the Precipitator effluent was discharged directly to waste.
A small portion (0.5 gpm) of the Precipitator effluent was
introduced into a 4.75 inch I.D. filter column containing 12
inches of fine sand as a bottom media layer and 18 inches of
fine anthracite as a top media layer. The effluent from the
filter was collected in a tank to obtain composite effluent
samples.
Both the sludge accumulated in the concentrator at the
bottom of the Precipitator, and the sludge accumulated in the
top of the sludge blanket (just below the bottom of the
chevron settlers) was discharged periodically through blowoff
lines into a sludge holding tank; therefore, it was not
necessary to continuously operate the dewatering equipment.
When sufficient sludge was accumulated in the sludge
holding tank, the sludge was pumped via a diaphragm sludge
pump to a head tank ahead of the DCG (Dual Cell Gravity)
filter. Here, a polyelectrolyte solution was metered into
and mixed with the sludge before entering the DCG so that the
sludge was conditioned into large firm particles. After this
conditioning stage, the sludge was gravity fed to the DCG
where it was changed from a fluidized sludge to a moist cake
by a gravity dewatering method. In this unit, the loose
liquid was separated from the solids by a moving screen.
The loose liquid, called the filtrate, was discharged to waste.
The solids collected on the screen, called the cake, was
discharged to the MRP unit.
The MRP or Multi Roller Press enabled further dewatering
of the DCG cake discharge. This unit pressed more water out
of the sludge cake and changed it from a moist cake-like solid
to a dryer bladeable solid. The liquid forced out of the sludge
cake by the MRP was drained, by gravity, out of the bottom of
the unit to waste. The pressed solid then dropped from the
end of the MRP into a 50 gallon drum in which it was collected
for disposal.
29
-------
Provisions were made for a portion of the sludge blanket,
in the lower settling zone of the Precipitator, to be recycled
back into the FeS head tank in order to increase the concen-
tration of solids in the FeS feed. If desired, sludge could
also be recycled from the concentrator at the bottom of the
Precipitator or from the mixing zone. However, the lower
settling zone was used since this sludge was fresher than
either the sludge in the upper settling zone or that in the
concentrator. Although the sludge in the mixing zone was
fresher than that in the lower settling zone, the sludge
concentration was much higher in the lower settling zone. The
lower settling zone, therefore, appeared to be the best point
source of a high concentration of relatively fresh sludge.
This pilot plant was also used for the Precipitator tests
of the hydroxide process. For these tests, no ferrous sulfate
or sodium hydrosulfide were fed and the lime, as Ca(OH)?, was
introduced directly into the inlet line instead of into'the
head tank (FeS slurry tank). All other details shown in
Figure 3 remained the same.
B. Sulfex Precipitator Tests w_ith 20 mg/1 Complexed Cu in
Influent
The objective of the initial Precipitator tests was to
optimize those parameters which were found to be important
in the jar tests. Since the jar tests in which the effects of
operating parameters were studied used the complexed copper
waste, these Precipitator tests were also performed using the
complexed copper waste (see Table 3, Column I). Other
Precipitator tests were then performed to study the effect of
variable influent compositions on operating results.
1. Parameters Studied in the Precipitator Tests
The parameters which were found to be important for
copper removal in the jar tests were tested in the
Precipitator. These were: (1) ferrous sulfide dosage,
(2) ferrous sulfide sludge blanket solids concentration,
(3) ferrous sulfide sludge blanket mixing time with the
liquid waste, and (4) use of a cationic polyelectrolyte
to aid sedimentation of precipitated solids. Other
parameters which were also studied included (5) a.gitator
mixing speed, (6) effluent pH, (7) sludge recycle, (8)
high iron concentrations in the complexed copper influent,
(9) different concentrations of copper in the influent,
and (10) the inclusion of five heavy metals in the influent
30
-------
a. Ferrous Sulfide Dosage
The dose of freshly precipitated ferrous sulfide
introduced into the influent has a significant effect on
the level of residual copper in the effluent. The
Precipitator was operated using 2 to 4 times the number
of equivalents of ferrous sulfide which are theoretically
needed to precipitate 20 mg/1 of copper. The Precipitator
tests show, for the simulated waste containing 20 mg/1 of
copper complexed with Rochelle salt and EDTA, that the
optimum dosage is about three times the theoretical
(stoichiometric) requirement.
Initially, the Precipitator was operated at FeS
dosages of 2 times theoretical and 4 times theoretical.
The pH range for these initial tests was maintained between
7.0 and 8.0. In this pH range, a ferrous sulfide dose of
2 times theoretical (55 mg/1 as FeS) is not sufficient to
continuously reduce the copper conconcentration to the level
which had been set as a goal (-0.1 mg/1 as Cu). In the
same pH range (7.0-8.0), but using an increased ferrous
sulfide dose of 4 times theoretical (110 mg/1 as FeS),
the effluent copper is about 0.25 mg/1 as Cu (~7.0), 0.15
mg/1 as Cu (pH -7.5), and 0.03 mg/1 as Cu (pH ~8.0). A
ferrous sulfide dose of 3 times theoretical was not tested
in the pH range 7.0 to 8.0 because, as shown above, the
results with a ferrous sulfide dose of 4 times theoretical
were only acceptable at the upper boundary of this pH
range. However, 3 times the theoretical ferrous sulfide
dosage was extensively studied in the pH range 8.5 to 9.0.
For these conditions, the average effluent copper level
is consistently less than 0.1 mg/1 as Cu.
Table 6 shows the effects of the different influent
rates, ferrous sulfide dosages, effluent pH, and solids
concentrations, on both effluent copper and effluent iron
concentration. It can be seen that at a given set of
conditions, increasing the pH not only reduces the residual
copper concentration but also the residual iron concentra-
tion. Although iron is only present in the Precipitator
influent at about 0.1 mg/1, a significant concentration of
iron (as FeS) is introduced into the treatment process.
The ability of the Sulfex process to reduce not only the
heavy metal concentration but also the iron concentration
is significant since no additional equipment is needed for
iron removal.
31
-------
b. Sludge Blanket Solids Concentration
The sludge blanket concentration is defined as
the volume of settled solids in a sample of the sludge
blanket that has been allowed to settle undisturbed for
10 minutes. This concentration is designated as volume
percent solids and represents the percent of the original
sample volume that is occupied by the settled portion
after 10 minutes of settling.
The results obtained in the jar tests show lower
effluent copper concentrations occur when the concentration
of sludge blanket solids is increased. The influence of
sludge blanket concentration on copper removal was carefully
observed in the Precipitator tests. The trends observed in
the jar tests are also seen in the Precipitator tests.
Increased copper removal coincides with an increased
concentration of settleable solids in the mixing zone.
The volume percent solids of both the settling
zone and mixing zone of the Precipitator were routinely
measured. The data shows the volume percent solids in the
mixing zone is more critical than in the settling zone.
The critical level of mixing zone solids concentration is
about 15 volume percent. If the mixing zone solids
concentration is less than 15 volume percent, the effluent
copper concentration is usually higher than 0.1 mg/1 as Cu.
Values above 15 volume percent give copper residuals which
are lower than 0.1 mg/1 as Cu. The sludge blanket solids
concentration in the mixing zone is generally higher than
15 volume percent for influent rates of 20 gpm and 30 gpm.
However, at an influent rate of M-0 gpm, the mixing zone
solids concentration drops below 15 volume percent when a
higher rotational speed of the agitator is not maintained.
Table 6 shows the increased effluent copper
concentrations that result at M-0 gpm as the mixing zone
solids concentration decreases while the FeS dose and pH
remain constant. Figure 5 also shows the relationship
between residual copper concentration and mixing zone
volume percent solids. This figure was developed by
averaging many data points collected during the
Precipitator tests at 30 and 40 gpm flow rates and
constant pH.
32
-------
Table 6. PILOT PLANT EFFLUENT RESIDUALS OF TOTAL COPPER
AND IRON AT DIFFERENT LEVELS OF OPERATING
PARAMETERS FOR COMPLEXED Cu INFLUENT
Influent Rate
(gpm)
20
20
20
20
20
20
20
20
30
30
30
30
30
30
40
40
40
40
40
40
40
FeS
Dosage*
2X
2X
2X
2X
4X
4X
4X
4X
2X
2X
2X
3X
3X
3X
3X
3X
3X
3X
3X
3X
3X
Effluent pH
Value
7.4
7.6
8.0
9.0
6.7
7.4
8.0
8.7
8.6
8.7
8.8
8.6
8.8
8.9
8.7
8.8
8.9
8.8
8.9
8.9
9.0
Avg. Mixing Zone
Solids (vol. %)
35
35
35
35
45
45
45
45
22
22
22
26
26
26
18
18
18
8
6
13
11
milligrams
Effluent
Avg . Cu
2.9
2.9
2.0
0.25
0.25
0.15
0.03
0.01
0.71
0.70
0.62
0.07
0.05
0.03
0.19
0.09
0.05
0.40
0.40
0.21
0.25
per liter
Quality
Avg. Fe
10
5.0
2.4
<0.1
25
10
1.0
0.4
<0.1
<0.1
<0.1
0.1
0.1
0.2
0.1
0.3
0.1
0.1
0.1
<0.1
<0.1
* X = Theoretical concentration required to react with 20 mg/1 Cu
33
-------
0.6
0.5
0.4
0.3
0.2
complexed Cu influent at 30 to 406PM and effluent pH at 8.8 to9
1 r 1
o.
Q_
O
O
<
13
O
LU
cr
O.I
0.05
0.02
T
FILTERED EFFLUENT COPPER
vs
MIXING ZONE % SOLIDS
T
T
12 15
VOLUME % SOLIDS
18
21
24
Figure 5. Influence of mixing zone solids concentration on
copper conentration of filtered effluent.
27
34
-------
c. Sludge Blanket Contact Time with Liquid Waste
Changing the influent flow rate to the Precipitator
changes the detention time in both the mixing zone and
settling zone and, therefore, the length of sludge blanket
contact time with the liquid waste. At 20 gpm (1.0 gpm/ft^
surface rate), the detention time in the mixing zone is
about 25 minutes followed by 75 minutes detention in the
settling zone. At 30 gpm (1.5 gpm/ft2 surface rate), the
mixing zone contact time is about 17 minutes followed by
51 minutes of detention in the settling zone. At 40 gpm
(2.0 gpm/ft2 surface rate), the mixing zone contact time
is about 12.5 minutes followed by 37.5 minutes of detention
in the settling zone. An effluent containing 0.1 mg/1 or
less copper is produced, provided the other operating
parameters are maintained at their optimum conditions, at
all three flow rates. This shows a total contact time of
approximately one hour is sufficient to achieve the
desired results.
In addition to studying the copper removal as a
function of the total sludge blanket contact time, copper
removal at different stages (contact times) in the
Precipitator was determined. The results of these tests
are shown in Table 7. This data shows that 95 to 98 percent
of the complexed copper is removed during the first 10
minutes of contact (i.e. mixing zone). During the remainder
of the contact time in the Precipitator, less than 1 mg/1
of copper is removed. However, this last period of extended
sludge blanket contact is Important because it reduces the
quantity of suspended and colloidal metal sulfides that
will be carried over in the effluent.
d. Use of a Coagulant Aid
The cationic guar gum derivative polyelectrolyte
which was found most effective in the jar tests was
successfully used in the Precipitator tests. The polymer
was fed (as a 0.05 to 0.1 weight percent solution) directly
into the lower part of the mixing zone with a metering
pump. A distributor tube running across the Precipitator
uniformly distributed the polymer solution in the mixing
zone. The supernatant was very turbid when no coagulant
aid was used during the jar tests. Therefore, no
Precipitator tests were made without the polyelectrolyte.
35
-------
Table 7. COPPER REMOVAL IN VARIOUS STAGES OF THE
PRECIPITATOR FOR COMPLEXED Cu INFLUENT
(Stage 1 to 4 Cu concentrations are for 0.45 micron filtered samples)
Run A
Stage
1
FeS Dose
Effluent
Influent
PH
Cu++
2X
7.4
18
4X
7.0
18
4X
7.3
20
4X
6.8
20
2.4X
8.9
23
(contact time = 0 min)
Upper Mixing Zone Cu++ 3.2 0.9 0.89 0..76 0.45
mg/1 (contact time =
5-10 min)
Lower Settling Zone 2.8 0.3 0.54 0.44 0.10
Cu++ mg/1 (contact
time = 30 min)
Precipitator Effluent 2.9 0.2 0.41 0.31 0.10
Cu++ mg/1 (contact
time = 100 min)
Sand Filter Effluent 2.9 0.2 0.41 0.22 0.05
(total contact time = 115 min)
36
-------
The Precipitator was started at 20 gpm using
3 mg/1 of polymer to build the sludge blanket. Once
the sludge blanket is formed, the effluent turbidity
is low (generally less than I JTU). When the polymer
dose is reduced to 1.5 mg/1, the turbidity remains less
than 1 JTU, but the settleable solids concentration in
the mixing zone decreases significantly. When the
polyelectrolyte concentration is subsequently increased
to U mg/1, at influent flow rates of both 20 gpm and 30
gpm, very clear Precipitator effluents are obtained.
Under these conditions, it is possible to operate the
sand filter for periods of 40 hours or more before
backwashing is needed. When the influent flow rate is
increased to 4-0 gpm, the polyelectrolyte dose has to be
increased to 5 or 6 mg/1 in order to maintain low
effluent turbidities (less than 1 JTU) and a sufficient
mixing zone solids concentration (greater than 15 volume
percent).
e. Agitator Mixing Speed
The speed of the agitator in the mixing zone of the
Precipitator is not critical at the 20 to 30 gpm flow rates,
but it is important at the 40 gpm flow rate. The tip
speed of the agitator used is 0.13 ft/sec, per rpm.
Initially, when the Precipitator was operating at 20 gpm,
more rapid sludge blanket formation occurred at agitator
speeds of 10 to 12 rpm than at 5 rpm. Then, increasing
the agitator speed from 12 rpm to 14 rpm increased the
average mixing zone solids concentration from about 35
volume percent to about 45 volume percent. This was
beneficial. However, further increases in agitator speed
from 14 rpm to 22 rpm resulted in only a temporary increase
in the mixing zone solids. Over an 8 hour period it
returned to 45 volume percent solids.
Agitator speeds of 11 to 14 rpm at 20 and 30 gpm
flows are sufficient to maintain a mixing zone solids
concentration above 15 volume percent solids, once the
sludge blanket is formed. Lowering the agitator speed
(i.e. to 7 rpm) at either 20 gpm or 30 gpm has no
immediate or significant influence on the mixing zone
solids concentration. Therefore, the agitator speed
does not appear to be a critical influence on effluent
quality at flow rates below 30 gpm. When testing agitator
speeds at Precipitator flow rates which are greater than
30 gpm, it is found that the speed of the agitator has more
influence on the mixing zone solids concentration and, hence,
on the effluent quality. At the 40 gpm flow rate (the
highest flow rate tested), it is found that the speed of
37
-------
the agitator had to be 'kept sufficiently high in order
to maintain the mixing zone settleable solids concen-
tration above the critical value of 15 volume percent.
For example, lowering the agitator speed from 11 rpm
to 7 rpm allows the mixing zone solids concentration to
gradually decrease from about 20 volume percent to about
6 volume percent. When the agitator speed is then
increased back to 11 rpm, the mixing zone solids concen-
tration increases to 20 volume percent where it levels out
This increase of agitator speed, therefore, accelerates
growth of metal sulfide precipitates into settleable
sized particles.
For this pilot plant Precipitator, the 11 rpm
agitator speed maintains an average solids concentration
of about 20 volume percent at 40 gpm, 25 volume percent at
30 gpm, and 40 volume percent at 20 gpm. Although the
required agitator speed at any flow rate will vary with
the design of the Precipitator, it has to be fast enough
to bring about rapid agglomeration of the precipitates
without causing shearing of settleable sized particles
into smaller particles that will be swept out of the
Precipitator by the flowing liquid.
f . Effluent pH Value
As already indicated in the discussion of FeS
dosage, the pH of the reaction has a significant effect
on the effluent quality. The pH of the Precipitator
effluent is essentially the same as the pH in the mixing
and settling zones. However, it is more stable and.,
therefore, is used in this study to represent the
equilibrium pH of the system.
When operating the Precipitator in the pH range
of 7.0 to 8.0, there is no clearcut relationship between
complexed copper removal and effluent pH. However, the
removal of iron improves with increasing pH in this range.
When the pH is increased to values of 8.5 or greater, the
importance of pH on effluent quality becomes very apparent
Figure 6 shows that operating the process in the 8.5 to
9.0 pH range gives significantly lower copper and iron
residuals in the effluent than operating in the 7.0 to
8.0 pH range, all other conditions remaining equal. Since
both complexed copper and iron are reduced to very low
residuals in the pH range of 8.5 to 9.0, this range is
established as the optimum range for the given influent.
38
-------
30
20
-------
g. Sludge_ Recycle
One way of Improving the effluent quality without
increasing the FeS dosage was expected to be by recycling
the sludge from either the settling zone or concentrator
back to the influent line. To make more FeS available
during the mixing stage of the process, it was proposed
that the partially reactive sludge from the settling zone
and/or concentrator could be re-exposed to the influent
stream. It also was thought the recycled sludge would
increase the settleable solids concentration in the mixing
zone which has already been discussed as being beneficial
to copper removal.
Sludge recycle was tested at 20, 30, and 40 gpm
influent rates. The volume of sludge recycled was
approximately 1 gpm, and it was pumped from the lower
settling zone into either the influent line directly or
into the FeS slurry tank. The lower settling zone was
selected as the source of the sludge recycle because it
contained a lower percent of precipitated copper than the
other zones of high sludge accumulation. Thus, it is
apparently the fresher source of sludge. Table 8 shows
the relative percentages of copper and iron in the sludges
found in the settling zones and concentrator.
With the sludge recirculation, the volume percent
of sulfide solids introduced into the Precipitator influent
was increased by 50 to 125 percent over the amount added by
the fresh FeS. The sludge recycled to the influent is not
the same as increasing the fresh FeS slurry by 50 to 125
percent, since the recycled sludge contains unreactive
solids such as copper sulfide and iron hydroxide.
At an influent rate of 40 gpm, there were periods
when the sludge recycle appeared to give lower effluent
copper values. However, most of the time the use o::
sludge recycle did not improve the effluent quality at
any of the flow rates tested when all other operating
conditions were maintained at optimum levels. Since
good results are obtained without sludge recycle, it is
not used in the Precipitator tests using the other
influent compositions.
40
-------
Table 8. SLUDGE COMPOSITION FROM VARIOUS
PRECIPITATOR ZONES
Upper Lower
Settling Zone Settling Zone Concentrator
Copper % (wt/wt)
6.0
3. 2
"4.5
Iron
(wt/wt)
30
Suspended Solids
(mg/1)
3000
6500
11,500
41
-------
2. Sulfex Precipitator Optimization Test
After the optimum level of each operating parameter
was determined in the Precipitator for treating a 20
mg/1 complexed copper feed, a continuous 10-day test run
was made. The optimum levels were maintained during
this test run. These levels were:
a. The Precipitator was operated at 30 gpm.
This flow rate represented a moderate surface rate
(1.5 gpm/ft2) through the Precipitator.
b. The FeS dosage (85 mg/1 as FeS) was maintained
at 3 times the stoichiometric requirement.
c. The pH value in the Precipitator was controlled
between 8.5 and 9.0.
d. The mixing zone settleable solids concentration
was maintained above 15 volume percent.
e. The cationic polyelectrolyte concentration was
maintained at 4- to 6 mg/1 in the mixing zor>? ,
Figures 7 and 8 show the conditions and results,
respectively, of the operating parameter study and
optimization test. In these figures, the levels of
copper and iron in the Precipitator influent and in
the filter effluent (Figure 8) and other pertinent
operating parameters (Figure 7) are plotted versus
the operating time. The results show that when the
operating parameters are maintained at the optimum
levels, the effluent copper concentration is maintained
below 0.1 mg/1 total copper. During the optimization
test, the effluent quality under optimum conditions never
increased above 0.1 mg/1 total copper. The two occurrences
of Cu levels greater than 0.1 mg/1 were caused by a failure
of the NaHS feed and a breakthrough of the filter.
Therefore, these points should not be included in the
evaluation. However, it should be noted that the
filtered effluent quality dropped below 0.1 mg/1 total
copper again after the filter was backwashed. In
Figures 7 and 8, the operating period between 415 hours
and U75 hours represents the optimization test.
42
-------
z 50
UJ
£ 30
& 10
400
£ 350
Q 300
** 250
CO
« 200
- 150
1* 100
50
0
uj 10
3 9
> 8
a. 7
6
CO
e 80
° 60
3« 40
^ 20
3 0
o
INFLUENT FLOW RATE vs OPERATING TIME
FERROUS SULFIDE DOSAGE
vs
OPERATING TIME
c-
NoHS
STOPPED
pH VALUE vs OPERATING TIME
INFLUENT
r-rJ-''A.
>.'
-' %-A
EFFLUENT pH
MIXING ZONE SETTLEABLE SOLIDS
vs
OPERATING TIME
OPTIMIZATION
TEST
1
0 40 80 120 160 200 240 280 320 350 400 440 480 520 560 600
HOURS OF OPERATION
Figure 7. Precipitator operating data for removal of complexed copper.
43
-------
or
LiJ
Q_
0.
o
o
100
50
10
5
I
0.5
O.I
0.05
0.01
100
50
10
5
o
a:
en
E
0.5
O.I
0.05
0.02
VV
INFLUENT COMPLEXED Cu
COPPER CONCENTRATION vs OPERATING TIME
FILTERED
EFFLUENT COPPER
i <\ '1
FILTER
BREAKTHFEU ,-x
NaHS
STOPPED
\ i
V
I i I
IRON CONCENTRATION vs OPERATING TIME
FILTERED
EFFLUENT
IRON
INFLUENT
I
I
I
_L
I
J_
I
I
I
OPTIMIZATION
TEST
10 40 80 120 160 200 240 280 320 360 400 440 480 520 560 600
HOURS OF OPERATION
Figure 8. Precipitater operating results for removal of complexed copper.
44
-------
3. Conclusions of Sulfex Tests with 20 mg/1 Copper
Complexed with Rochelle Salt and EDTA
The 5.5 ft. long x 5.7 ft. wide x 11 ft. high Permutit
Package Precipitator followed by a filter can be success-
fully used for removal of complexed copper ions from an
influent stream at flow rates up to 40 gpm. The following
optimum conditions have to be maintained for successful
operation.
a. FeS dosage at least 2.5 to 3 times stoichiometric.
b. Mixing zone settleable solids concentration equal
to or greater than 15 volume percent.
c. Mixing zone cationic polyelectrolyte concentration
equal to 4 to 6 mg/1.
The Precipitator effluent can be successfully polished
with a dual media sand and anthracite filter to contain
less than 0.1 mg/1 total copper at the optimum operating
conditions. Recycling of the sulfide sludge from the
sludge blanket to the Precipitator influent does not
significantly improve the effluent quality when the
operating conditions are maintained at optimum levels.
4. Study of Optimized Sulfex Process on Influent Containing
20 mg/1 Copper and 50_mg/1 Ferrous Iron Complexed with
Rochelle Salt and EDTA
a. Precipitator Tests
Figures 7 and 8 include the results of the test
with high dissolved iron concentrations in the influent
in addition to the complexed copper (i.e. influent shown
in Column II, Table 3 + 50 mg/1 Fe++). The operating
period between 476 hours and 496 hours represents this
test. During this period, the influent copper averaged
21 mg/1 as Cu,and the influent dissolved iron averaged
more than 50 mg/1 as Fe. The operating parameters were
maintained at optimum levels with about 3 times theoretical
FeS dosage (75 to 80 mg/1 as FeS), 20 to 22 volume percent
solids in the mixing zone, and an effluent pH in the range 8.6
to 9.0. Due to the increased amount of iron added to the
influent, the influent pH was lower (6.1 to 7.6) than for
the influent composition shown in Column II, Table 3
(pH 8.8 to 9.0). The effluent pH was raised to the optimum
level (8.5 to 9.0) by increasing the lime dosage to the FeS
slurry tank. The influent flow rate to the Precipitator
was maintained at 30 gpm during this test.
45
-------
The goal of an effluent Cu concentration of equal
to or less than 0.1 mg/1 was adhered to during this test.
However, the effluent iron level (about 1.5 mg/1) had
increased to levels higher than those resulting from tests
with much lower influent dissolved iron levels (10.2 mg/1
as Fe).
b. Conclusions
The conclusions made from the Precipitator test,
with an influent containing both complexed copper and a
high concentration of complexed iron, is that the influent
iron concentration has no deleterious influence on the
removal of the copper. However, the presence of the
complexing agents with the high influent dissolved iron
does tend to give higher than expected effluent iron
concentrations. In the cases where high dissolved iron
concentrations together with the presence of strong
complexing agents is a problem, there are two possibilities
for obtaining lower effluent iron concentration.
(1) Increase the pH in the Precipitator (i.e.
to 10-10.5) so that the hydroxide of ferrous
iron becomes less soluble in the system.
(2) Add an oxidizing agent to the filter effluent
to oxidize the ferrous iron to ferric iron since
ferric hydroxide is less soluble at pH 8.5 to 9.0
than ferrous hydroxide. Then, pass the effluent
through a second filtering step to remove the
precipitated iron.
C. Sulfex Tests with 5 mg/1 Copper in Influent Complexed
with Rochelle Salt and EDTA
For these tests, a Precipitator influent was used contain-
ing one-fourth the salt concentrations shown in Column II,
Table 3. Since this influent composition was reduced to
one-fourth of the concentration shown in this table, the influent
pH was lower than 8.8-9.0 (i.e. pH 7.6 to 7.8).
1. Tests with 3 Times the FeS Dosage
The first test on this influent stream was run with
3 times the theoretical FeS requirement (about 21 mg/1
as FeS), 30 gpm influent rate, pH 8.7 to 8.9, and the
mixing zone settleable solids concentration maintained at
18 to 22 volume percent. The results of this test, shown
in Figures 7 and 8 for the 23 hour operating period 507
46
-------
through 530 hours, indicate that these operating conditions
are not sufficient to reduce the effluent copper concen-
tration to the desired level (i.e. < 0.1 mg/1 as Cu).
For the first 8 hours of this test, the effluent copper
level was within the desired range since it averaged
Id.01 mg/1 as Cu during this period. However, after
this period, the effluent copper level rose in a 1
hour period to about 0.85 mg/1 as copper and then leveled
off.
2. Test with 7.8 Times the JFeS Dosage
Another test on this influent was run with a higher
FeS dosage (i.e. 55 mg/1 as FeS or 7.8 times the
theoretical FeS dosage). This test is also shown in
Figures 7 and 8 for the 23 hour operating period 551
through 574 hours. With the exception of the FeS dose,
the operating conditions were maintained as before with
30 gpm influent rate, pH 8.7 to 8.9, and the mixing zone
solids concentration maintained at 18 to 25 volume percent.
The results show an effluent copper level lower than
the maximum desired level of 0.1 mg/1 as copper (i.e.
about 0.04 mg/1 as Cu).
3. Conclusions
a. Influent complexed copper concentrations in the
5 mg/1 as Cu range require higher than 3 times the
theoretical FeS requirement. However, the total dosage
required is less than the total amount of FeS required
for complexed copper influents in the 20 mg/1 as Cu range.
b. With the exception of the FeS dosage, the other
optimum operating parameters are the same at the 5 mg/1
complexed copper level as they are at the 20 mg/1
complexed copper level.
c. The sludge blanket in the mixing and settling
zones of the Precipitator has a reserve capacity to
remove on the order of 0.5 mg/1 of complexed copper
for 8 hours (at the 30 gpm flow rate) before an
increase in the effluent copper occurs. This increase
in the effluent copper level is due to an insufficient
dosage of fresh FeS over that 8 hour period.
47
-------
D. Sulfex Precipitator Tests with Influent Containing About
80 mg/1 Copper Ion Complexed with Rochelle Salt and EDTA
A Precipitator influent containing about 4 times the salt
concentrations shown in Column II, Table 3 was used. The pH
of this influent was higher than the 8.8 to 9.0 pH shown in
this table. The pH of this more concentrated influent was
9.6 to 9.8, and the lime dose to the FeS slurry tank was
lowered in order to maintain the pH in the Precipitator in
the 8.5 to 9.0 pH range. Since the concentration of all the
constituents in the Column II, Table 3 influent were increased
4 times in going from 20 mg/1 Cu to 80 mg/1 Cu, the concen-
trations of the complexing agents were proportionately
increased.
1. Test with 2 Times the FeS Dosage
In this test, the average FeS dosage to the Precipitator
was maintained at about 225 to 230 mg/1 as FeS (i,e. about
2 times theoretical). The other operating conditions were
maintained at 30 gpm flow rate, pH 8.7 to 8.9, and the
mixing zone settleable solids concentration at 23 to 25
volume percent. The results of this test, shown In
Figures 7 and 8 for the 20 hour operating period between
530 and 550 hours, indicate that a higher FeS dosage is
necessary because the residual effluent copper levels are
high, 1.0 to 1.5 mg/1 as Cu.
2. Test with 3 Times the FeS Dosage
The FeS dosage was then increased to 3-3.2 times the
theoretical requirement for the influenc containing about
80 mg/1 of complexed copper. Again, the influent flow
rate was maintained at 30 gpm, the pH at 8.8 to 8.9, and
the mixing zone settleable solids concentration above 15
volume percent (i.e. 17 to 24 volume percent). Figure 7
shows these test conditions and Figure 8 shows the
results for the 21 hour operating period between 574 and
595 hours. The effluent residual copper level is
maintained below 0.1 mg/1 as Cu during this period.
In both test periods where the influent copper levels
averaged about 80 mg/1 as copper, the effluent iron
levels are high (2 to 8.5 mg/1 as Fe) even though the
influent dissolved iron levels are still relatively low
(0.5 mg/1). This condition is different than in the
tests where high effluent iron levels resulted from high
influent dissolved iron levels (50 mg/1 as Fe). In the
48
-------
previous tests with low iron influents, the iron added
to the process as FeS was removed at pH 8.5 to 9.0. In
this test (with low iron and high copper influent), the
high iron residuals in the effluent have to be due to the
addition of high concentrations of FeS to the influent.
The reason for the high iron residuals is probably due
to the fact that the very high concentrations of complexing
agents (508 mg/1 Rochelle salt and 180 mg/1 EDTA) have
more influence on complexing the soluble iron than they
do at one-fourth these concentrations.
3. Conclusions of Sulfex Tests_with 80; mg/1 of
Complexed Copper in the Influent
a. The ferrous sulfide dosage has to be maintained
at about 3 times theoretical with the higher influent
copper levels.
b. High influent complexed copper levels can be
removed as easily as low influent copper levels in the
Precipitator provided the FeS dosage and other conditions
are maintained at optimum levels.
c. Very high concentrations of complexing agents
will interfere with complete precipitation of ferrous
hydroxide more strongly than the precipitation of
cupric sulfide. The filtered effluents from Sulfex that
contain high iron residuals due to the presence of high
concentrations of strong complexing agents will have to
be treated by oxidation and a second filter or at higher
pH values during the sulfide precipitation process (i.e.
maintain Precipitator at pH 10.0 to 10.5 or higher).
This last alternative (high pH in the Precipitator)
would not be desirable when chrome is to be removed.
E. Sulfex Tests with Influent Containing Copper, Cadmium,
Chromium, Nickel, and Zinc Complexed with EDTA and
Rochelle Salt
In these tests, the five heavy metals were dissolved in
the influent at levels of about 4 mg/1 each. The composition
of this influent is listed in Column III, Table 3. The
Precipitator was initially operated with no metals added to
the influent for about 16 hours. The purpose of this was to
build up a concentration of fresh FeS solids in the mixing
zone to bring it to at least 15 volume percent solids before
adding the heavy metals.
49
-------
1. Test with 3.9 Times FeS Dosage
a. Test Conditions
In this test, the Precipitator was operated at a
30 gpm influent rate with 83 mg/1 FeS dosage. The
theoretical (stoichiometric) requirement for this solution
which contains 4 mg/1 of each of these five metals is only
21 mg/1 of FeS. This requirement is based only on the
presence of copper, cadmium, nickel, and zinc (4 mg/1 of
each). The sulfide of chromium does not occur in aqueous
solutions, but chromium does precipitate as the hydroxide
of chrome III and does not consume FeS (unless chrome VI
is present which is reduced to chrome III by FeS).
Therefore, three times the theoretical FeS for these 5
metals, at the concentrations present in the influent, is
actually 3 x 21 mg/1 or 63 mg/1 FeS. The 83 mg/1 FeS dose
is equal to 3.9 times the theoretical requirement. The pH
in the Precipitator was maintained between 8.8 and 9.1,
and the solids concentration in the mixing zone was
maintained between 17 and 23 volume percent settleable
solids.
b. Results
After about 24 hours of operation, the effluent
metal concentrations leveled out. The effluent metal
concentrations obtained at these steady state conditions
are shown in Column A, Table 9. These test conditions
and results are shown in Figures 9 and 10 for the 22
hour operating period between 617 and 639 hours. It was
apparent from these results that the complexing agents
interfered with the removal of most of these metals more
than they interfered with the removal of copper. Therefore,
it was decided to try a test with an increased sulfide
dosage.
2. Test with 5.5 Times the FeS Dosage
a. Test Conditions
The test conditions were maintained as before
except for increasing the FeS dosage. The Precipitator
was operated at 30 gpm influent rate with 5.5 times the
theoretical FeS requirement (117 mg/1 as FeS), and the
pH was maintained at about 8.8 during this period. The
mixing zone solids concentration was maintained between
20 and 30 volume percent during this period.
50
-------
Table 9. EFFLUENT METAL RESIDUALS FROM PRECIPITATOR
TESTS WITH COPPER, CADMIUM, CHROMIUM (III),
NICKEL, AND ZINC IN INFLUENT. FOUR MG/L OF
EACH METAL PRESENT
A.
Sulfex Test
With EDTA and
Sulfex Test With
C.
Hydroxide Test
With Only
Influent Rochelle Salt
Tested (Col. Ill Table 3)
Metal
Copper
Cadmium
Chromium (III)
Nickel
Zinc
Iron
0.22
1.6
3.4
3.6
0.87
0.42
Only Rochelle Salt
(Col. IV Table 3)
Effluent
0
<0
3
<0
-------
b. Results
The results obtained are shown in Figure 10 for
the 7 hour operating period between 639 and 646 hours.
After 6 to 8 hours of running, the effluent residual copper
level is significantly reduced from 0.22 mg/1 Cu (at 3.9 x
FeS dosage) to 0.14 mg/1 Cu (at 5.5 x FeS dosage). However,
the other heavy metals do not show a significant reduction
in their residual effluent concentrations during this same
period of time. This test was terminated in order to conduct
jar tests to determine the influence of the individual
complexing agents on the heavy metals. The results of
these tests have already been discussed and are shown in
Table 5.
F. Tests with Cu, Cd, Cr(III), Ni, and Zn Complexed with
Only Rochelle Salt (Tartrate)
1. Precipitator Test with Sulfex Process
a. Influent Compositions
The composition of the influent for these tests is
shown in Column IV, Table 3. This influent had the same
concentration of Rochelle salt (126.9 mg/1) as the initial
Precipitator tests, but the EDTA was left out because of
the jar tests indicated that EDTA forms stronger complexes
with most of these metals than does Rochelle salt. Before
introducing this influent into the Precipitator, the
Precipitator was run with only copper and Rochelle salt
in the influent for about 24- hours in order to build a
sludge blanket. The old sludge blanket, produced during
operation with EDTA in the influent, was disposed of in
order to insure formation of a sludge blanket representative
of the new influent conditions.
b. Test Conditions
The Precipitator was operated at 30 gpm influent
rate with about 3 times the theoretical FeS requirement
for 20 mg/1 copper (i.e. 80 to 90 mg/1 as FeS). During
the test period, the pH range was maintained at 8.5 to 9.0
while the solids concentration in the mixing zone was
maintained between 22 and 35 volume percent settleable
solids.
52
-------
LU 40
u. 30
- 20
^ I0
o
^ 350
1 300
o 250
2 200
cr
_ 150
^> 100
INFLUENT FLOW RATE vs OPERATING TIME
FERROUS SULFIDE DOSAGE
vs
OPERATING TIME
n
V)
o
58
10
8
7
6
80
60
40
20
LIME DOSAGE
vs
OPERATING TIME
pH VALUE vs OPERATING TIME
EFFLUENT pH
- --\
y v, X,
INFLUENT
MIXING ZONE SETTLEABLE SOLIDS vs OPERATING TIME
600 615 630 645 660 675 690 705 720 735 750
SULFEX
HYDROXIDE
HOURS OF OPERATION
Figure 9. Precipitator operating data for removal of complexed Cu,Cd,
Cr,Ni and Zn (from influent containing 4mg/l of each metal).
53
-------
INF I =no metals in influent to build sludge blanket
INF 31 =5metals +Rochelle Salt + EDTA in influent
INFIU =only Cu + Rochelle Salt in influent to build sludge blanket
IN F12 = 5 metals + Roche lie Salt in influent
filtered effluent metal concentration MB operating time
-------
c. Results
The conditions and results of this test are shown
in Figures 9 and 10 for the 30 hour operating period
between 669 and 799 hours. When the quality of the
filtered Precipitator effluent reaches constant levels
under the given conditions, the effluent concentrations
shown in Column B, Table 9 are obtained.
Comparing the results in Column B, Table 9 with
those in Column A, Table 9, it is seen that the Rochelle
salt is a weaker complexing agent for copper, cadmium,
nickel, zinc, and iron than is the EDTA-Rochelle salt
combination because better metal removal generally
occurred in the solution containing only Rochelle salt.
On the other hand, it is seen that chromium is complexed
at least as much if not-more strongly by Rochelle salt
(tartrate) than it is by EDTA. It is reasonable to assume
that of the metals tested, the chromium precipitation is
most influenced by the Rochelle salt because its removal
is accomplished only by hydroxide precipitation. The
other metals are more readily precipitated as the sulfides
which are much less soluble than the hydroxide species of
these metals. Since the tartrate, by forming metal
tartrate complexes, tends to be particularly strong in
inhibiting the precipitation of metal hydroxide, the metals
which can precipitate as sulfides are less influenced by
its presence.
2. Precipitator Test with Hydroxide Precipitation
The object of this test was to determine the
effectiveness of removal of the 5 metals by precipi-
tation as hydroxidet in a single stage precipitation
process.
a. Influent Composition
The influent composition in this test was the same
as that used in the previous Precipitator test with Sulfex
(i.e. Column IV, Table 3). No EDTA was contained in this
influent. Using the same influent in the hydroxide process
and the Sulfex process tests , a direct comparison was made
between the effectiveness of both processes for the removal
of Cu, Cd, Cr, Ni, and Zn when a complexing agent (tartrate)
is present.
55
-------
b. Test Conditions
Except that no ferrous sulfide was added to the
influent, the Precipitator was operated at the same
conditions for the hydroxide precipitation test as for
the Sulfex test. The Precipitator was operated at a 30
gpm influent rate and the pH was controlled at about
pH 9-9.2 by adding a slurry of hydrated lime, Ca(OH)2,
to the influent via a metering pump. The pH value of
the hydroxide process was controlled near pH 9.0 for the
following reasons:
(1) The pH value where the solubility of the
hydroxides of these five metals is at a minimum
is different for each metal. For copper, cadnium,
and nickel, the solubility decreases with increasing
pH. The higher the pH, the lower the residual metal
concentrations will be (i.e. pH should be 9.0 or
higher for best results). Chrome III and zinc
hydroxides are amphoteric. The minimum solubilities
are obtained within rather narrow pH ranges. The
pH ranges of minimum solubilities for Cr and Zn
differ somewhat in the literature because the
composition of the solution containing these metals
has an influence on the solubilities. Generally,
the most appropriate theoretical pH range to achieve
the lowest possible solubilities of these metal
hydroxides, in treatment of most squeous solutions,
would be about pH 9. The EPA seminar publication
Waste Treatment (Upgrading jjetal-Finishing Facilities
to Reduce Pollution)13 shows the solubilities "of
these metal hydroxides at different pH values and
how they can vary from one source to another.
(2) The regulations on the maximum permissible
pH values for waste disposal are generally about
9.5; therefore, higher pH values would usually
require a subsequent neutralization step after
filtration.
REFERENCES
Lancy, L. E. and R. L. Rice, Upgrading Metal-Finishing
Facilities to Reduce Pollution: Waste Treatment,
Environmental Protection Agency, Technology Transfer
Seminar Publication No. 2, Fuly 1973, pp. 28-30
56
-------
(3) Common practice in the treatment of metal
finishing wastes (containing these metals) by
hydroxide precipitation is to operate somewhere in
the pH 8.0 to 9.0 range.
In addition to the reasons listed above for
operating our test at pH 9.0, this test gave us the
opportunity to compare Sulfex with hydroxide precipitation
in the same approximate pH range. The same polyelectrolyte
used in the Sulfex tests was added to the hydroxide test at
about the same concentration (4 mg/1). The hydroxide test
Precipitator effluent was polished by the same dual media
sand anthracite filter that was used in the Sulfex tests.
Before adding the 5 metals to the influent, the Precipi-
tator was run with a copper containing influent for about
16 hours to build up an adequate sludge blanket. The
solids concentration in the mixing zone was maintained
at 18 to 22 volume percent during the actual test with
the 5 metals.
c. Results
The conditions and results of this test are
shown in Figures 9 and 10 for the 27 hour operating
period between 717 and 744 hours. When the filtered
Precipitator effluent quality reaches equilibrium, the
average metal concentrations shown in Column C, Table 9
are obtained.
C. Conclusion of Precipitator Tests with Cu , Cd, Cr, Ni,
and Zn in the Influent
1. EDTA interferes with the precipitation of the 5
metals in the Sulfex Process more severely than the Rochelle
salt.
2. EDTA hinders the precipitation of cadmium and nickel
more than it does copper, chromium, zinc, and iron with the
Sulfex Process.
3. With Rochelle salt present, but no EDTA, about 3-4
times the theoretical requirement of FeS removed the metals
to or below the maximum desired levels for all metals except
chromium which was not removed effectively.
57
-------
4. When Rochelle salt is present, the trivalent chromium
is not effectively removed because chromium III is only removed
as the hydroxide. The optimum pH range for lowest CrCOH)^
solubility is in the 8 to 9.0 range. The presence of hydroxyl
ion enhances the formation of the Cr-tartrate complexes that
hinders the formation of Cr(OH)3-
5. As expected, in the presence of a complexing agent
such as tartrate, Sulfex is significantly more effective
in removing the metals than conventional lime (hydroxide)
treatment.
6. Chromium III is not removed any more effectively in
the Sulfex process than in the hydroxide process. However,
in cases where Chrome VI is to be removed, Sulfex can reduce
the Cr+6 to Cr+3 and remove it in one step. Therefore, for
the removal of total chrome, the Sulfex process is superior to
the conventional hydroxide process.
7. Further studies are necessary to evaluate Sulfex on
actual electroplating plant effluents where various other
complexing agents (and other compounds) are present at
various levels of concentration.
58
-------
SECTION VI
DCG (SLUDGE DEWATERING) TESTS
A. Laboratory Sludge Dewatering Tests
1. Description of Laboratory Sludge Dewatering Tests
The settled sludges from the Sulfex jar tests and the
initial Sulfex Precipitator tests were saved and used as
samples for the laboratory sludge dewatering tests. These
small scale tests are routinely performed to predict the
operating conditions and operating results of Permutit's
Dual Cell Gravity Filter (DCG) and Multi Roller Press (MRP),
The tests consist of a simulated rapid mixing step in which
the best type and dose of polyelectrolyte, needed to
condition the sludge, is determined. The operation of the
DCG is then simulated using the conditioned sludge as feed
for a bench top roll test. This test provides information
on the percent solids in the dewatered cake (sludge roll)
and on the suspended solids in the subnatant (water drained
from sludge). Following simulated DCG operation, the MRP
is simulated by a simple sludge pressing test in order to
give a rough estimate of the percent solids in the further
dewatered cake that results from Multi Roller Press opera-
tion.
2. Results of the Laboratory Sludge Dewatering Tests
The laboratory dewatering tests were made on Sulfex
sludges of about 1.0 weight percent solids. These tests
show that either cationic or mild anionic polyacrylamide
polyelectrolytes condition the sludges satisfactorily.
The following results are obtainable:
a. DCG test cake = 5 weight percent solids
b. MRP test cake = 8 weight percent solids
c. Polyelectrolyte requirement = 100 to 150 mg/1
of a cationic polyacrylamide
d. Consistency of MRP cake is that of a
bladeable material
59
-------
In the actual pilot plant tests, the DCG was fed
sludge from a decanting tank in which the solids
concentration of the Precipitator blowoff sludge was
increased from 1.0 weight percent to over 3 weight
percent solids. As a result, the DCG and MRP cakes
obtained had a higher solids concentration than the
laboratory tests predicted. The results of the pilot
plant dewatering tests are discussed below (Section IV,
Part B, "Full Scale DCG-MRP Tests").
B. Full Scale DCG and MRP Tests
1 Objective
The full scale Dual Cell Gravity Filter (DCG) and
Multi Roller Press (MRP) units were run in combination.
These tests were made to determine the effectiveness and
operating requirements for a full-scale DCG-MRP sludge
dewatering plant.
2. Operating Procedure
The 1.0 to 1.5 weight percent blowdown sludge from
the Precipitator operation was collected in a decanting
tank of about 1500 gallons total capacity. The Precipi-
tator blowdown rate was approximately 0.5 to 1.0 gpm at
the 30 gpm influent rate and the sludge was removed
intermittently. During periods of no blowdown, especially
at night, a high degree of settling took place and it was
possible to decant the supernatant liquor. After about
2 weeks of operation, the settled sludge level was such
that the decanting tank could not accept more Precipitator
blowdown sludge without solids escaping in the overflow.
The volume of decanted sludge was then about 1000 gallons.
When the decanted sludge reached this level, the clear
water on top of the settled sludge was siphoned off.
When, the sludge in the tank was agitated with a mechanical
agitator to evenly distribute the solids in the holding
tank so that a homogeneous sludge was fed to the DCG via
a sludge pump. The concentration of this sludge was about
3 to 3.5 weight percent solids. The 1000 gallons of sludge
could be dewatered by the DCG-MRP combination within 4 to 6
hour periods which includes time needed for homogenizing of
the holding tank sludge and start up of the dewatering equip-
ment. A high molecular weight slightly anionic polyacrylamide
sludge conditioner was mixed with the DCG influent (in a separate
mixing tank) before it entered the DCG. See Section III b
for additional description of the assembly and operation of
the Precipitator pilot plant.
60
-------
3. Results of the Full Scale Dewatering Tests
a. The sludge from the holding tank is a very thick
but still fluidized 3 to 3.5 weight percent. This sludge
is pumped to the DCG at about 5 gpm. An average poly-
electrolyte concentration of 125 to 150 mg/1 is necessary
to condition the sludge prior to the DCG.
b. When the treated sludge passes through the DCG,
the sludge solids concentration increases to about 14- to
15 weight percent.
c. The sludge cake from the DCG is immediately
conveyed to the MRP where -more water is pressed out
leaving a sludge containing about 26 weight percent
solids. This final sludge has a bladeable dryness that
is suitable for final disposal.
d. On the basis of the final sludge from the MRP,
the polyelectrolyte requirement is calculated at 7.7
pounds of polymer per ton of dry solids.
e. The standard DCG units have speed controls for
varying the rotation speed of the DCG screen which
enables operation at higher influent rates. The DCG
unit used in this test did not have a speed control.
It is estimated that the sludge can be dewatered at a
rate of about 7 gpm in the same size unit equipped with
a speed control.
4. Conclus ions
Dewatering the sludge from the Sulfex process to a
bladeable solid can be routinely accomplished by
conventional sludge dewatering operations.
61
-------
SECTION VII
COST ESTIMATES
A. Chemical Operating Costs
A comparison of the estimated chemical operating costs for
the hydroxide and Sulfex processes is shown in Tables 10 and 11.
Table 10 shows the difference between the chemical costs of
hydroxide and Sulfex for wastes containing various concentrations
of copper. In each case, the copper is complexed with tartrate
and EDTA. This table shows that, although conventional hydroxide
treatment (pH 9) chemical costs are lower than Sulfex costs, the
effluent quality obtained does not meet the EPA standards.
Improved copper removal is obtained by hydroxide precipitation
at very high pH (12), which raises the chemical costs to
considerably more than Sulfex, but the effluent quality is
still not good enough to meet the EPA standards. Table 10
also shows that although the EPA standards are met with Sulfex
at about neutral pH (7.5), it is less costly to obtain even
better effluent quality by operating at pH values of about 8.5.
The reason for this is that the required dosage of ferrous
sulfide reactant is less at higher pH value.
Table 11 shows the difference between chemical costs of
hydroxide and Sulfex for wastes containing mixtures of
common heavy metals in both complexed and uncomplexed solutions.
Waste composition a. in Table 11 contains a mixture of five
common heavy metals complexed with tartrate (Rochelle salt).
At pH 9 hydroxide precipitation is less costly, but the
residual metals obtained in the effluent at this pH are all
higher than the EPA standards. Even operation at pH 12, which
adds high chemical costs due to the additional lime required
plus the acid for neutralization, does not meet the EPA standard
for copper. Sulfex, however, removed all the metals, except
chromium, to levels below the EPA standards at a pH value within
the EPA standards for waste disposal.
A comparison between hydroxide precipitation and Sulfex is
shown for jar tests made on waste composition b_ in Table 11.
These jar tests show that the uncomplexed combination of metals
in solution are reduced to meet the EPA standards by Sulfex or,
at pH 10, by lime. However, the cost of hydroxide at pH 10 is
about the same as Sulfex which gives equal or better effluent
quality. Also, the hydroxide process at pH 10 has the disadvan-
tage of requiring a final acidification step after filtration
in order to meet the EPA standard of pH 6.0 to 9.5 for discharge.
62
-------
Table 10. CHEMICAL COSTS OF HYDROXIDE VS. SULFEX
FOR VARIOUS COPPER CONCENTRATIONS
Waste
Composition
a. 5 mg/1 Cu
complexed with
Tartrate & EDTA
Effluent
Quality (mg/1 Cu) p_H_
<0.10 8.5
<0.25 7.5
*EPA 0.5 6.0-9.5
1.5 12
^4.3 9
Process
Sulfex
Sulfex
Hydroxide
Hydroxide
Cost C/IOOO Gal,
8.8
9.5
31.8
2.4
b. 20 mg/1 Cu
complexed with
Tartrate & EDTA
<0.25
*EPA 0.5
4.3
12
8.5
7.5
6.0-9.5
12
9
Sulfex
Sulfex
-
Hydroxide
Hydroxide
13.1
15.8
32.1
2.7
c. 80 mg/1 Cu
complexed with
Tartrate & EDTA
<0.10
<0.25
*EPA 0.5
>4.3
8.5
7.5
6.0-9.5
12
9
Sulfex
Sulfex
Hydroxide
Hydroxide
45.6
57.6
33.9
4.8
* Maximum daily average allowed by EPA. Based on rinse water usage of
160 1/sq. meter for existing plants electroplating common metals
according to 4-24-75 Federal Register.
63
-------
co
w
EH
co
is
o
X
w
CO
co
w
Q
H
X
§
co
fa
O
co
EH
CO
O
CJ
Z:
H
w
K
CJ
r-l
0)
rH
X!
rd
EH
u
rd
O
O
o
o
i-H
-O-
-P
in
cS
33
ftl
o<
u
G
O
rH
0) -P
4J -H
in in
rd O
00
X
o in r-i o
G . . . .
N O O O rH
-v
31 in H O
o o o n
CN) in H o
ro o o H C
rd
EH ft
g -
O T3
rd U U
U tsl
U
(0
pq
m
0)
TJ
H
X
o
0)
H
X X
0) O
4-1 ^H
r-H i}
CO K
in
in CTI in
01
co H o r-
o rH in n
o o o H
v
m
o rH m ro
o o o
v
m
o rH m m
o p o o
V v v
oo
o o o
v
I-H m
O rH rH
. . C
o o o
v <
TJ
0) ^ -P
X ro rd
tt) + ,C
rH (-1 C O
ft O N rd
g W
O - Ti
U T! G rH
C U rd \
D tn
3 -H
A O 2 «*
CM
X
0)
4-1
CO
H
X
O
ffi
in
in cr> m
cxi o co
rH in co
C
isi o o o
vo m m m
+ ooo
to '
u p o o
v
CM
W
*
U
G
N
in tji
CM g
*
* &
0)
in
G
H
to
G
0
0)
in
rd
PQ
.
in
rH
rd
-P
d)
g
G
O
G
o
u
tn SH
G 0)
-H -P
-P ttl
rd -H
rH t7>
ft fl)
O Pi
!H
-P rH
U rd
fl) to
H Q)
0) TJ
0)
in fa
-P
C m
rd r-
rH 1
ft m
CM
tn i
G ^3*
H
-P 0
tn -p
H
X en
0) C
r-j
H T3
0 to
4-1 O
0
in o
C rd
o
H SH
-p rH
rd
4-1
>i O
rH
rH 0)
rd D"1
T3 rd
in
§ ^
g in
3 0)
X -P
rd rd
S S
in
.p
rd
ft
0)
.^
-P
O
0
4->
rd
4H
0
CM
SH
4J
U
rrj
4-1
p
G
rd
g
O
4H
^1
0)
1 i
rd
s
0)
tn
rd
J5
rH
3
-p
u
G
rd
in
H
«
D
G
0
rH
-P
H
in
1
0
0
fl)
-p
in
rd
64
-------
The results shown for hydroxide effluent quality at pH 10 on
waste composition 1) is actually lower than the theoretical
solubilities for the hydroxides of cadmium, chromium, nickel,
and zinc. It is assumed that these better than expected
results are due to co-precipitation and/or other interactions
of the metal hydroxides. Therefore, it should not be concluded
that hydroxide precipitation can always reduce these metals to
the values shown in Table 11 at pH 10, since the combinations
of metals present in the influent waste can change frequently.
The results obtained for Sulfex, shown in Table 11 at pH 8.5
for waste composition b_, are well above the theoretical
solubilities and do not depend on the presence of several
different metals to meet the EPA standards. Therefore, for a
combined metal influent without complexing agents, the Sulfex
process is more predictable than hydroxide precipitation with
respect to effluent quality.
Waste composition _c shown in Table 11 is an actual plant
waste that was studied in jar tests. In order to reduce the
hexavalent chromium to permissible levels, by EPA standards,
the chrome +6 was reduced by sulfite reduction at low pH before
hydroxide precipitation at pH 9.5 was used. The zinc is not
removed to the required level with hydroxide precipitation in
this case. Sulfex reduces the Cr+6 and precipitates the Cr+3
in the same step while simultaneously removing the zinc to below
the maximum permissible level. The chemical cost is less for
Sulfex than hydroxide precipitation in this case. A detailed
breakdown of the chemical costs shown in Tables 10 and 11 is
presented next. These cost estimates are based on the chemical
requirements for the optimum conditions found during the
Sulfex studies.
Breakdown of Chemical Costs
The chemical prices used in the following cost estimates
are averages for the high and low values listed in the Oil,
Paint and Drug Reporter' for July 28, 1975. The chemical
cost for NaHS should be increased by 56.6 percent if 45
percent liquid NaHS is used instead of the 71 percent flake
NaHS shown in these estimates. The cost for sulfuric acid
at 5 £ per pound is based on purchasing 55 gallon drum lots
instead of bulk basis or carboys which varies the price from
2C to St per pound.
REFERENCES
Chemical Marketing Reporter. Oil, Paint and Drug Reporter.
Schuell Publishing Company, Inc., July 28, 1975.
65
-------
1. Chemical Operating Cost for Influent Containing Copper
Complexed with Tartrate and EDTA
a. Influent containing 5 mg/1 complexed copper
(1) Hydroxide Process
Eff. Qual.
(mg/1 Cu)
pH 9_ pH 12
54.3
1.5
Chemical
Ca (OH) 2
Polymer
H2S04
Dosage
Cost
lb/1000 gal.
pH 9 pH 12
0.5
0.025
C/lb
1.8
60
5
Total
C/IOOO gal.
pH 9 pH 12
0.9
1.5
2.4
31.8
(2) Sulfex Process
Eff. Qual.
(mg/1 Cu)
pH 7.5 pH 8.5
<0.25 <0.1
Chemical
71% Na^S
FeSO4-7H20
Polymer
Ca(OH)2
Dosage
Cost
lb/1000 gal.
pH 7.5 pH 8.5
0.19
1.58
0.03
1.13
0.17
1.38
0.03
1.13
C/lb
11.25
2.25
60.00
1.80
Total
C/1000 gal.
?H 7.5 pH 8.5
2.1
3.6
1.8
2.0
9.5
1.9
3.1
1.8
2.0
8.8
66
-------
b. Influent containing 20 mg/1 complexed copper
(1) Hydroxide Process
Eff. Qual.
(mg/1 Cu)
pH 9 pH 12
12
4.3
Chemical
Ca(OH)2
Polymer
H2SO4
Dosage
lb/1000 gal.
pH 9 pH 12
0.5
0.03
4.6
0.03
4.4
Cost
C/lb
1.8
60
5
Total
-------
c. Influent containing 80 mg/1 complexed copper
(1) Hydroxide Process
Eff. Qual.
(mg/1 Cu)
pH 9 pH 12
Chemical
Ca(OH)2
Polymer
H2S04
Dosage
lb/1000 gal.
3H 9 pH 12
0.67
0.06
4.6
0.06
4.4
Cost
C/lb
1.8
60
5
c/iooo gal.
pH 9 pH 12
1.2 8.3
3.6 3.6
22.0
Total
4.8
33.9
(2) Sulfex Process
Eff. Qual.
(mg/1 Cu)
pH 7.5 pH 8.5
<0.25 <0.1
Chemical
71% NaHS
FeS04-7H20
Polymer
Ca(OH)2
Dosage
Cost
lb/1000 gal.
pH 7.5 pH 8.5
1.4
12
0.06
6.2
1.1
9.2
0.05
5.3
C/lb
11.25
2.25
60
1.8
Total
-------
2. Chemical Operating Cost for Influents Containing Mixtures of
Common Heavy Metals
a. Influent containing 4 mg/1 each of Cu, Cd, Cr , Ni, and
Zn complexed with Rochelle salt
The required PeS dosage was the same for this influent as for
the one containing 20 mg/1 of copper complexed with EDTA and
Rochelle salt. However, due to the higher pH of the 20 mg/1
complexed copper influent, the lime requirement was higher for
this influent which had a lower initial pH (i.e. pH 6.8 instead
of pH 8 to 9).
(1) Hydroxide Process
Eff. Qual.
(mg/1)
pH 9 pH 12
Cu 1.2
Cd 1.7
Cr 2.0
Ni 3.0
Zn 1.0
0.17
Chemical
Dosage
lb/1000 gal.
pH 9 pH 12
0.65
0.05
<0.1
Ca (OH) 2
Polymer
H9SO4
1.0
0.03
-
5.0
0.03
4.4
1.8
60
5
Cost
C/lb
1.8
60
5
C/1000
pH 9
1.8
1.8
-
gal.
pH 12
9.0
1.8
22.0
Total
3.6
32.8
(2) Sulfex Process
Cu
Cd
Cr
Ni
Zn
Eff. Qual,
(mg/1)
pH 8.5
<0.1
0.01
3.2
0.4
-------
b. Influent containing 4 mg/1 each of Cu, Cd, Cr+3, Ni, and 2n
without complexing agents
The following chemical costs are based on studies using jar
tests with synthetic wastes to compare hydroxide precipitation
with Sulfex. The pH of the influent used in these tests wass
about pH 6.0.
(1) Hydroxide Process
Eff. Qual.
(mg/1)
pH 7.5 pH 10
Chemical
Dosage
lb/1000 gal.
pH 7.5 pH 10
Cost
Cu
Cd
Cr
Ni
Zn
0.1
3.8
<0.5
2.3
1.3
<0.1
<0.1
<0.1
<0.1
<0.1
Ca (OH) 2
Polymer
H2S04
0.33
0.03
-
0.92
0.03
0.61
1.8
60
5
Total
0.6
1.8
-
2.4
7.2
(2) Sulfex Process
Eff. Qual.
Cu
Cd
Cr
Ni
Zn
(mg/1)
pH 8.5
0.01
0.1
<0.05
0.05
0.01
Chemical
71% NaHS
FeS04-7H20
Polymer
Ca(OH)2
Dosage
Cost
lb/1000 gal.
pH 8.5
0.09
0.77
0.03
1.13
-------
c. Influent containing 25 mg/1 hexavalent chromium and 6 mg/1
of zinc
The following chemical costs are based on jar tests with an
actual plant waste sample that was assumed to have little or no
strong complexing agents present. Sodium bisulfite (NaHSC^) was
selected as the reducing agent in the hydroxide process for
reasons of safety, availability, and handling.
Process
Hydroxide*
Step 1
Hydroxide*
Step 2
Eff. Qual.
(mg/1)
T-Cr <0.05
Zn 0.8
Chemical
H2S04 (pH 2)
+ NaHSO3
Ca(OH)2
+ Polymer
Dosage
lb/1000 gal,
0.75
0.69
8.5) 1.35
0.04
C/lb
-------
3. Chemical Operating Costs for Sludge Dewatering
The following costs are based on the results of
running sludge dewatering tests on the Sulfex sludges
with the DCG/MRP dewatering system. The amount of sludge
resulting from the hydroxide Precipitator tests was not
enough to make a full scale evaluation of the chemical
requirements.
Chemical
Polymer
138
200
Ibs/ ton
dry solids
7.7
$/ton
dry solids
15.40
One thousand gallons of Precipitator influent produced
about 5 gallons of sludge requiring 138 mg/1 of polymer
to dewater.
The sludge dewatering chemical operating cost was:
B.
!8 g/£ x
1
Ib x
3.785
a
x 5
gal x
454g gal
Therefore, sludge dewatering w
gallons to the above Precipitator
Capital Equipment Costs for Sulfex
200
-------
Constituent mg/1 total mg/1 dissolved
Chromium (as Cr) 50 ?5 (as Cr+6)
Zinc (as Zn) 20 6
Copper (as Cu) 5 1
Nickel (as Ni) 1 0.5
Iron (as Fe) 20 1.0
Calcium (as Ca) 20 20
Suspended solids 100
pH 6 -
In addition to the heavy metals listed, other heavy
metals such as cadmium, lead, mercury, etc. may also be
present up to several mg/1 of each without changing the
equipment costs. The presence of strong complexing
agents is assumed to be low or absent.
2 . Equiprcien_t___Cos t
a. Installed Cost
The installed cost determined for an actual
proposal of the plant shown in Figure 11 is estimated
at $130,290. The installation cost includes neutraliza-
tion tanks constructed below ground level to save space
It is estimated that installation costs for tanks of
above ground construction would be substantially lower.
For spare chemical feed pumps, sump pump, and transfer
pump, $5,000 should be added to the selling price.
b. Selling Price
Equipment P r i ce
1. Precipitator with clear well, centrifuge
for dewatering, chemical feeds and
agitators, and engineering drawings $ 39,126
2. Filter with transfer pump and
engineering drawings 4 ,337
3. Neutralization system including agitators,
chemical feeds, pH controls, sump pump, and
engineering drawings 49 ,002
Total selling price $ 92,465
c. Installation Cost (estimated by
outside contractor) $ 37,825
73
-------
Figure II. 40GPM Sulfex plant for combined removal of Cr+6,Zn,Cu,Cd,Ni and Fe.
74
-------
SECTION VIII
REFERENCES
Fales, H. A. and F. Kenny. Inorganic Quantitative
Analyses. New ed. Appleton Century Co., New York-
London, p. 233, 1939.
Handbook of Chemistry and Physics, 50th ed., Robert
C. Weast, ed. The Chemical Rubber Company, p. B252,
1969.
Handbook of Analytical Chemistry, Louis Meites, ed.
McGraw-Hill, Inc., pp. 1-15, 1-19, 1963.
Metal Finishing Guidebook and Directory for 1970.
Nathaniel Hall, ed. Palmer H. Langdon, pub., p. 473,
1970.
Ward, A. E., Chapter 11, Wastewater Control and Treatment
In: Electroplating Engineering Handbook, 3rd Ed.,
Graham, A. K. (ed.), New York, New York, Van Nostrand
Reinhold Company, pp. 359-360, 1971.
Lancy, L. E. and R. L. Rice. Upgrading Metal Finishing
Facilities to Reduce Pollution: Waste Treatment.
Environmental Protection Agency Technology Transfer
Seminar Publication No. 2, July 1973. U. S. Government
Printing Office, pp. 28-30, 1974
Chemical Marketing Reporter. Oil, Paint and Drug
Reporter. Schnell Publishing Company, Inc., July 28,
1975
75
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1 REPORT NO.
EPA-600/2-77-049
2.
4. TITLE AND SUBTITLE
TREATMENT OF METAL FINISHING WASTES BY SULFIDE
PRECIPITATION
3. RECIPIENT'S ACCESSI ON-NO.
5 REPORT DATE
February 1977
issuing date
6. PERFORMING ORGANIZATION CODE
7 AUTHOR(S)
Richard M. Schlauch and Arthur C.
of The Permutit Company
8. PERFORMING ORGANIZATION REPORT NO.
Epstein
9. PERFORMING ORGANIZATION NAME AND ADDRESS
The Permutit Company, Permutit Research and
Development Center, Princeton, New Jersey 08540
for
The Metal Finishers' Foundation
10. PROGRAM ELEMENT NO.
IBB610
11. CONTRACT/GRANT NO.
Grant No. R802924
12. SPONSORING AGENCY NAME AND ADDRESS
Industrial Environmental Research Laboratory - Cin., OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/600/12
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This project involved precipitating heavy metals normally present in metal
finishing wastewaters by a novel process which employs ferrous sulfide addition
(Sulfex), as well as by conventional treatment using calcium hydroxide for
comparison purposes. These studies consisted of laboratory jar tests and bench
scale tests to determine the chemical and physical requirements for the precipitation
of the heavy metals and the subsequent dewatering of the resulting sludges.
Following the laboratory tests, pilot plant tests were made to confirm the validity
of the laboratory test results and provide realistic operating data. As a result,
it was demonstrated that Sulfex is a technically viable process that is superior to
conventional hydroxide precipitation for removal of copper, cadmium, nickel, and
zinc from a given influent. And, when operated in the pH 8-9.0 range, the Sulfex
process will remove total chromium to a concentration which is less bhan or equal
to that from a conventional hydroxide precipitation process. Hexavalent chromium
can be removed by Sulfex in a one-step operation. The effluent quality from either
process is dependent on the type and concentration of complexing agents present in
the influent.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Industrial wastes, wastewater, waste
treatment, metal finishing*, operating
costs, precipitation*, ferrous sulfide*
sludge disposal
Sulfide precipitation*,
hydroxide precipitation*
complexing agents,
sludge dewatering
13B
1S. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
86
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
76
US GOVERNMENT PRINTING OFFICE 1977757-056/5588
------- |