EPA-660/2-74-086
DECEMBER 1974
                       Environmental Protection Technology Series
Mercury Recovery From Contaminated
Waste Water  and  Sludges
                                    National Environmental Research Center
                                     Office of Research and Development
                                     U.S. Environmental Protection Agency
                                            Corvallis, Oregon 97330

-------
                      RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development,
U.S. Environmental Protection Agency, have been grouped into
five series.  These five broad categories were established to
facilitate further development and application of environmental
technology.  Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in
related fields.  The five series are:

          1.   Environmental Health Effects Research
          2.   Environmental Protection Technology
          3.   Ecological Research
          4.   Environmental Monitoring
          5.   Socioeconomic Environmental Studies

This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY STUDIES series.  This series describes research
performed to develop and demonstrate instrumentation, equipment
and methodology to repair or prevent environmental degradation from
point and non-point sources of pollution.  This work provides the
new or improved technology required for the control  and treatment
of pollution sources to meet environmental quality standards.

This report has been reviewed by the Office of Research and
Development, EPA, and approved for publication.  Approval  does
not signify that the contents necessarily reflect the views and
policies of the Environmental Protection Agency, nor does  mention
of trade names or commercial products constitute endorsement or
recommendation for use.

-------
                                 EPA-660/2-74-086
                                 December 1974
     MERCURY  RECOVERY FROM CONTAMINATED WASTE

                  WATER AND SLUDGES
                         By

                    Richard Perry
                  Project 12040 HDU

               Program Element 1BB037

              ROAP/TASK No.  21 AZX/022
                   Project Officer

                   Ralph H. Scott
Pacific Northwest Environmental Research Laboratory
      National Environmental Research Center
              Corvallis, Oregon  97330
      NATIONAL ENVIRONMENTAL RESEARCH CENTER
         OFFICE OF RESEARCH & DEVELOPMENT
       U.S.  ENVIRONMENTAL PROTECTION AGENCY
              CORVALLIS, OREGON   97330
     For sale by the Superintendent of Documents. U.S Government Printing Office
             Washington, D.C. 20402 - Stock No. 5501-00972

-------
                       ABSTRACT
         A system was designed, installed and operated to
recover mercury (Hg) from waste water and sludge produced by
a mercury cell chlor-alkali plant.  Hg content of the waste
water ranged from 300 - 18,000 ppb while Hg content of the
brine sludge ranged from 150 - 1500 ppm Hg.  Deposits from
the waterway near the plant outfall were also processed.

         From a variety of removal techniques evaluated,
sulfide precipitation was selected for process water treat-
ment and high temperature roasting for sludge treatment.

         The sulfide precipitation system steps include
collecting the process water streams, adjusting the pH to 5-
8 with spent sulfuric acid, settling the large particles in
a surge tank, adding sodium sulfide to a 1 - 3 ppm excess,
adding diatomaceous earth at the rate of 0.7 gpl (0.62 lb/
1000 gal), and filtering through an R. P. Adams pressure
filter.  The effluent Hg levels range from 10 - 125 ppb and
average 50 ppb Hg, an 87 - 99% removal, averaging 97%•  The
44.8 m2 (169 ft2) filter processes up to 380 1/min (100 gpm)
with an approximate 48-hour cycle time between backwashings.
Capital costs totaled $143,900 and operating costs average
50 /3785 1 (1000 gal).

         The sludge treatment system includes a collection
system, a 3.7 m (12 ft) diameter thickener, a 1.8 m (6 ft)
diameter rotary vacuum filter, a 1.37 m (4.5 ft) i.d. multi-
ple hearth furnace, and 3 stainless steel condensers 21 m2
(224 ft2 ) each.  Processing rate for the sludge is 140 -
320 kg/hr (300 - 700 Ib/hr), dry basis.  At present, ap-
proximately 18 m tons (20 s tons) of sludge per month are
processed.  Operating temperatures range from 540°C - 760°C
(1000°F - 1400°F), feed Hg content ranges from 290 - 440
ppm Hg (dry basis), and clinker Hg content after treatment
varies from 0.5 - 7.2 ppm Hg, for a removal rate of 98.3 -
99-8$  Waterway sediments containing 12.8 ppm were roasted
at 750°C (1350°F) and the clinker contained 0.95 -1.7 ppm
Hg, for an 87 - 92$ removal.  Capital costs totaled $364,500
and operating costs are $32/m ton ($35/s ton) of dry sludge
treated.

         This report was submitted in fulfillment of
Project Number 12040 HDU by the Georgia-Pacific Corporation,
under the partial sponsorship of the Environmental Protec-
tion Agency.  Work was completed in April, 1974.
                             111

-------
        UNITED STATES ENVIRONMENTAL PROTECTION AGENCY

                   NATIONAL ENVIRONMENTAL RESEARCH CENTER
                              300 S.W. 35TH ST.
                           CORVALLIS, OREGON 97330


                             February 7,  1975
Dr. Ho L. Young, Chemist
U.S. Environmental Protection Agency
6ZO Central Avenue
Alameda, CA   94501

Dear Dr. Young:

     In response to your recent request, I am enclosing a copy of
our publication entitled "Mercury Recovery From Contaminated Waste
Water and Sludges," hPA-660/2-74-086.

     Thank you for your interest and please call us if we can be of
further assistance.
                                 Sincerely,
                                 Chris L. West,  Director
                                  Public Affairs Office
Enclosure

-------
                             CONTENTS






Sections                                             Page




I     Conclusions                                      1




II    Recommendations                                  2




III   Introduction                                     3




IV    Process Design                                   7




V     Construction                                    44




VI    Operation and Evaluation                        48




VII   Discussion                                      55




VIII  References                                      76




IX    Patents and Publications                        81




X     Glossary                                        82




XI    Appendices                                      84

-------
                        FIGURES
No.                                                    Page

1    The Bellingham Chlor-Alkali Plant and Brine          4
     Sludge Pond

2    The Bellingham Chlor-Alkali Plant with Sludge        5
     Pond and Hg Recovery Structure

3    Lab Kiln Test Assembly for Roasting Sludge          10

4    Hg Contents of Untreated Brine Sludge After         12
     Roasting in Lab and Pilot Kilns
                                                         1 R
5    Cut Away Drawing of the Internals of a Multiple     ±0
     Hearth Furnace Showing the Rake Arms and Hearth
     Assembly

6    Acid Treatment of Brine Sludge Before Roasting      20
     in Lab and Pilot Kiln

7    Eimco Rotary Vacuum Pilot Filtration of Brine       2^
     Sludge

8    Proposed Brine Sludge Handling System                '

9    Solubility of HgS in Excess S=                      33

10   Schematic of the Reduction Method of Hg Re-         ^5
     moval from Water
                                                         •?£
11   Schematic of the Ion exchange Process for Hg        J
     Removal from Water

12   Schematic of Activated Carbon Hg Removal from       3'
     Waste Water

13   Schematic of the Sulfide Precipitation Process      3
     for Waste Water

14   Proposed Sulfide Precipitation System               39
                                                         IIP
15   Data from Pilot Tests with 11 Sq. Ft. Niagara
     Filter

16   Pilot Filter Tests to Determine Cycle Length         ^
     with and without Precoat and Body Feed

-------
                       FIGURES  (cont.)

No.                                                  Page

17   The Model of the Hg Recovery System               ^6

18   Installed Brine Sludge Handling System            51

19   Lab Tests on Mercury Removal from Water Using     62
     Sulfide ppt., Activated Carbon, Ion Exchange
     Resins and Reduction Chemicals

20   Installed Sulfide Precipitation System for Water  64
     Treatment

21   Installed 6' x 6' Eimco Rotary Vacuum Filter      65
     for Brine Sludge Dewatering ,

22   Installed 54" i.d. BSP Multiple Hearth Furnace    66

23   Installed R. P. Adams Filter for HgS Removal      6?

24   The 12' x 6' Sludge Thickener Prior to the        68
     Rotary Vacuum Filter in the Sludge Treatment
     System

25   The Full Scale Hg Recovery System as Installed    70
     at the Bellingham Chlor-Alkali Plant

26   Mix Tank in the Sulfide Precipitation System      71
     where the D.E. and Sulfide are Added

27   Particle Distribution in  Brine Sludge which was   73
     Washed and Screened to Remove Particles 0.007"
     diameter  (Experiment 34)

28   Bench Test Set-ups for Chemical Oxidation  of      87
     Sludge

29   Effect of Staging on Mercury Recovery by          89
     Tokawa
                             VI

-------
                         TABLES


No.                                                    Page

1    Kiln Treatment of Brine Sludge  & Graphite  .          8

2    Kiln Roasting of Sludge, No Chemical Treatment      H

3    Rotary Calciner Roasting of Brine Sludge            13

4    Multiple Hearth Furnace Roasting of Brine Sludge    1^
     BSP 76 cm (30 in) Pilot Furnace

5    Multiple Hearth Furnace Roasting of Brine Sludge    15
     BSP 76 cm (30 in) Pilot Furnace

6    Multiple Hearth Furnace Roasting of Brine Sludge    17
     BSP 76 cm (30 in) Pilot Furnace, ICI Data

7    Kiln Roasting of Chemically Treated Brine Sludge    19

8    BSP Test on Batch Kiln                              21

9    Test Results from Barrett Centrifuge on De-         23
     watering of Brine Sludge by Centrifuging

10   Data from the Eimco Pilot Rotary Vacuum Filter      24

11   Lab Test Data on Sulfide Precipitation for Hg       30
     Removal from Water

12   Hg Levels in Water After Exposure to Sulfide Ion    31
     for 30 Seconds to 10 Minutes

13   Partial List of Data from Start-up of the Full      50
     Scale Sulfide Precipitation System

14   Data from Start-up of the Full  Scale Brine Sludge   53
     Treatment System

15   Names and Addresses of the Companies Contacted      56
     for Information by Direct Communication During
     the Project

16   Data from Literature on Ion Exchange Resins         ^

17   Lab Test Data on Ion Exchange Resin for Hg          59
     Removal from Water
                              vn

-------
                        TABLE'S, (cont. )
No.                                                  Page
18   Lab Test Data on Activated. Carbon for Hg Removal
     from Water

19   Lab Test Data on Hg Removal from Water by Reduc-
     tion

20   Cost Estimate - Water Treatment System            ^

21   Cost Estimate - Sludge System                     ^5

22   Oxidation of Brine Sludge Using Sodium Hypo-      ^°
     chlorite

23   Oxidation of Brine Sludge Using Sodium Hypo-      88
     chlorite "Work Performed at University of
     British Columbia

24   Oxidation of Brine Sludge Using Chlorine Gas      91

25   Oxidation of Brine Sludge Using Combinations      92
     of Hypo, Chlorine, Electrolytic Acid Treatment
     and Roasting

26   Hg Analysis of Brine Sludge Size Fractions        94

27   Hg Removal Rates Necessary for Various Size
     Chlorine Plants to Achieve 45 gm (0.1 Ib)
     Per Day Mercury in the Effluent

28   Comparison of Substances Used or Considered
     for Reducing Mercury Ion in Solution
                               Vlll

-------
                   ACKNOWLEDGEMENTS
         The research, process selection, pilot tests,
analytical work and report preparation were performed by
a team of chemists, engineers and technicians at Georgia-
Pacific Corporation, Bellingham Division, consisting of
Dr. Scott Briggs, Ed Dahlgren, Luther Dunn, Karen Hulford,
Dick McLeod, Dick Perry and Don Rachor.  Research assist-
ance was also provided by Dr. Bill Groves of Vancouver,
B.  C.,  Canada.

         Engineering design and construction supervision
of the full-scale plant were provided by Lynn Baker, Ivan
Campbell and Hal Henkel of G-P.

         The start-up team for the full-scale system con-
sisted of Steve Baklund, Steve Earp and Bruce Swanson of
G-P.  Don Elliot and Don Wines coordinated operational
aspects.

         The support of the Project Officer, Ralph H.
Scott, and Director, N. A. Jaworski, EPA Pacific Northwest
Environmental Research Laboratory in Corvallis, Oregon is
gratefully acknowledged.
                             IX

-------
                      SECTION I

                     CONCLUSIONS
1.    Sulfide precipitation offers several advantages over
     other methods of Hg removal from water:  (a) fewer
     process steps, (b) pH range compatible with total
     plant effluent, (c) concentrated Hg products, (d) in-
     expensive chemicals used, and (e) minimal environ-
     mental stress.

2.    In the laboratory, sulfide treatment achieved   99.9$
     removal of Hg from solutions containing 10 - 100 ppm Hg.

3.    In the plant, the sulfide process achieved 87 - 99-2%
     removal from solutions containing 0.3-6 ppm Hg.  The
     average effluent Hg content was 50 ppb.

4.    The major problem experienced with the sulfide process
     was pH control.  With concentrated sulfuric acid, a
     two-stage addition system was needed.

5.    Sulfide system capacity is 380 1/min (100 gpm).  Capital
     costs totaled $1*13,900, and operating costs of the
     sulfide system are 13
-------
                      SECTION II

                    RECOMMENDATIONS
1.    Sludge from other chlor-alkali plants should be
     roasted to determine the efficiency of this process
     on various wastes.   Also, sludges from other indus-
     tries and municipal sewage plants which contain Hg
     should be tested.

2.    The Hg recovery from the air leaving the furnace re-
     quires further work to solve the dust removal problem.

3.    To offset the operating cost of the sludge process,
     potential uses for the calcium and magnesium oxide in
     the clinker should be investigated.

*J.    A further step in the water treatment process would
     be the design of a polishing filter to remove most of
     the remaining Hg in the filter effluent.

-------
                    SECTION III

                    INTRODUCTION
         The industrial hygiene problems associated with
Hg vapor, and both inorganic and organic Hg compounds have
long been recognized and safeguards have been developed to
avoid harmful exposures.  The situation changed dramatic-
ally with the publication in 1968 of the biological con-
version of inorganic Hg to methyl Hg and similar compounds.
The toxicity, persistence, and concentration of methyl Hg
in food chains caused concern for any discharge of Hg into
the environment.

         Extensive analyses in North America indicated
high Hg levels in fish and sediments associated with cer-
tain Hg cell chlor-alkali facilities.  All facilities in
North America rapidly took steps to reduce total Hg dis-
charges to less than 0.23 kg/day (6.5 Ib/day) to the re-
ceiving waters at each installation.  In most instances,
this involved stockpiling of Hg-containing materials such
as process sludges (Figure 1).

         An objective of these studies was to develop a
system to reduce the Hg content of brine process sludge and
other Hg-containing solids and liquids to a level suffic-
iently low that they may be disposed of without significant
hazard to the environment.  A further objective was the re-
covery of Hg without significant loss into the atmosphere.

         The Bellingham Chlor-Alkali plant, Figure 2,
went into production in 1965 with a capacity of 122 m tons
(135 s tons) per day of chlorine.  Brine sludge averaging
1.4 m tons (1.5 s tons) per day resulted from the precipi-
tation of calcium and magnesium compounds from the incom-
ing solar salt, and erosion of graphite from cell anodes.
The sludge had been stockpiled in an impoundment basin
pending the development of a Hg recovery system.  Hg con-
taminated water is generated at the rate of 110 - 190 1/min
(30 - 50 gpm).

         The major sources of the brine sludge are:  (1)
the brine clarifier, (2) the brine filters, and (3) the
salt saturator residue.  Other Hg-containing solids in-
clude:  caustic filter backwash, cell residue, and caustic
storage tank residues.

-------
Figure 1.  The Bellingham Chlor-Alkali Plant
               and brine sludge pond

-------
Figure 2.   The Bellingham Chlor-Alkali Plant with sludge
                pond and Hg recovery structure

-------
         The major sources of Hg-bearing waste water are:
(1) floor washings from the cell room, (2) purge streams
from the  cell end-box wash water recycle systems,
(3) purge streams from the brine system, (4) drainage from
the caustic filtration area, and (5) water from tank
cleaning.

         An extensive literature search was conducted
before and during the project; the results are included
in the reference section.

         This is the final report on the project and the
work performed from June 1, 1971 to April 30, 197^.

-------
                      SECTION IV

                    PROCESS DESIGN


SLUDGE TREATMENT FOR Hg REMOVAL

Lab or at or y Me thod s

         The chemical oxidation tests performed in the
lab were on the scale of 250 - 1000 ml of brine sludge
treated in beakers (Appendix A).  The sludge roasting
tests were carried out in lab furnaces with volumes of 3
liters and 100 liters.  The temperature, heating time, and
air purge were controlled as described below.  The labor-
atory phase of the project lasted 9-10 months before the
single process to be used was selected.

Chemical and Electrolytic Oxidation

         An extensive investigation was devoted to de-
veloping a chemical means of removing Hg from the brine
sludge (Appendix A).  The alternatives tried involved ad-
dition of sodium hypochlorite, chlorine, or electrolysis
of brine to generate small bubbles of chlorine gas.  These
treatments are reportedly used to treat Hg ores as well as
chlor-alkali cell wastes.  Removal rates of over 99% are
claimed for concentrated ores and residual Hg levels of less
than 0.1 ppm for chlor-alkali sludge (5, 6, 9).

         In this study, the treatments not only dissolved
Hg, but significant quantities of other components of the
sludge as well, so that Hg separation was not effective.
The maximum Hg removal was less than 88$ with a minimum
Hg in sludge after treatment of 47 ppm  (Appendix A).

         Due to problems of (1) dissolving components
other than Hg, (2) multi-staging to achieve desired percent
recovery, and  (3) difficulty of the filtration and wash
steps between the stages, the chemical alternatives  to
sludge treatment were abandoned.

Roasting at High Temperatures

         The roasting of Hg-bearing solids has been used
since ancient times to separate Hg from other material
(1).  In preliminary tests in a small lab muffle furnace,
crucibles of brine sludge were heated to several tempera-
ture levels for various lengths of time to determine the
approximate temperature and time parameters  (Table 1).  The

-------
Table 1.  KILN TREATMENT OF BRINE SLUDGE & GRAPHITE
                         (ppm Hg)


Brine
sludge



Cell
graphite
Time.
hr.
Start
1
8
16
24
Start
5
16
Temperature
°C 121 427 538
°F 250 800 1000
140 140
4.6
4.5
48
4.2
4200
20
140
3.7
3.1
	
2.5
	
:::
649
1200
140
0. 19
0.06
	
0.04
4200
6.2
                             8

-------
tests yielded residues ranging from 0.3 - 1.7 ppm Hg.  These
initial results were 20 - 100 times lower than the lowest
residuals achieved by chemical treatment.

         Following the preliminary tests, a series of trials
were conducted in a large kiln on samples ranging, in size
from 100 g to over 30 kg (Figure 3).  The air rate through
the kiln was carefully controlled to remove the vaporized
Hg to keep from saturating the vapor phase with Hg.  Re-
siduals as low as 0.02 ppm Hg were achieved (Table 2 and
Figure 4).  Temperatures, in the range of 800°C - 900°C
(1450°F - 1750°F) were required to achieve Hg residuals
below 0.2 ppm.

Furnace Selection

         Following these successful lab runs, kiln manu-
facturers were contacted to verify the data on a pilot
scale.  Tests were conducted at Bartlett-Snow, Cleveland,
and BSP Division of Envirotech, Brisbane, California.

         At Bartlett-Snow, a 15 cm (6 in) diameter rotary
calciner was operated at 800°C (1475°F) with a residence
time of 30 minutes.  The minimum Hg level achieved in the
tests was 25 ppm Hg, which was significantly higher than
the batch kiln test at the same temperature (Table 3).
The tests were shifted to a multiple hearth furnace to
gain better control over residence time and eliminate
short-circuiting.

         Two multiple hearth furnace manufacturers were
contacted; the BSP Division of Envirotech was selected to
test the dewatered brine sludge.  Tests run in April and
June 1972 in a 33 cm diameter (13 In) batch kiln yielded
clinker Hg contents of 0.32 ppm.

         From these data, a pilot run was scheduled in
July to test the procedure on a 76 cm (30 in) furnace at a
higher solids feed rate.  At temperatures of 730°C - 760°C
(1350°F - 1400°F), the Hg level in the clinker was 3.2 ppm
(Table 4).  This was not as low as desired; however, in a
second test at 870°C - 955°C (1600°F - 1750°F), residuals
of .12 - .14 ppm Hg were obtained (Table 5).

         From these data, the furnace hearth loading was
found to be a maximum of 39 kg/m2 (8 lb/ft2) per hour so
that a wet solids feed rate of 224 kg/hr (600 Ib/hr)
would require a 7 m2 (75 ft2) furnace.  This corresponds
to a standard 1.37 m (4.5 ft) i.d. 6-hearth unit with
7-9 m2 (84 ft2) hearth area.

-------
Figure 3.  Lab kiln test assembly for roasting sludge
                w///////////////,
    MANUAL AIR
    CONTROL VALVE
                                                   TEMPERATURE
                                                   CONTROLS FOR
                                                   ELECTRIC
                                                   HEATING
                                                   ELEMENTS IN
                                                 KILN TEMPERA-
                                                 TURE INDICATOR
    COMPRESSED AIR
                         10

-------
                             Table 2. KILN ROASTING OF SLUDGE, NO CHEMICAL TREATMENT
Exp.
no.
31
32
M
33
38
42
51
tl
1 1
It
57
61
61
63
65
65
70
72
72
73
76
70
80
87
89
90
81
91
91
91
Sample, treatment
Brine sludge, rotated
it
tf
Brine sludge
tr
ir
ir
it
it
M
it
Metal anode sludge
Our brine sludge
Pond sludge
Brine sludge
it
ii
ti
ii
it
it
it
it
M
It
tt
II
II
Brine sludge from
filter tests
Residence Temperature Agitation, Hg content, ppm
time, hr. 'C T min. Start End
8
8
8
8
8
8
5
6
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
7
7
7
7
649
677
677
649
824-38
774-793
149-260
232-343
649-760
649-760
649-760
649-760
649-760
649-760
649-760
649-760
649-760
649-774
649-774
649-760
593-649
593-649
677-718
649-774
649-663
538-571
802-830
941-9G9
802-830
941-969
1200
1250
1250
1200
1525-100
1425-1460
300-500
450-650
1200-1400
1200-1400
1200-1400
1200-1400
1200-1400
1200-1400
1200-1400
1200-1400
1200-1400
1200-1425
1200-1425
1200-1400
1100-1200
1100-1200
1250-1325
1200-1425
1200-1225
1000-1060
1475-1525
1725-1775
1475-1525
1725-1775
0
0
0
3
4
4
9
9
9
9
2
2
2
2
i
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
...
	
	
...
500
	
1340
880
1340
1340
...
158
250
...
	
	
246
822
1100
822
822
822
822
1735
1735
1735
1735
1735
2250
2250
4.4
0.4
1.3
1.7
0.18
0.12
880
50
0.5
4.8
0.53
0.02
0.95
0.69
0.97
1.7
0.37
0.47
0.75
0.07
5.21
5.6
2.0
1.7 )
5.6 )
14 )
0.08 }
0.03 )
0.07
0.02
Comments

6 ceramic balls (fine powder)
n ii

NaCl appeared to fuse

Still slightly wet
Still slightly wet
Top 1" of 9" depth *
Bottom of 9" depth *






pH 10


Excess of air




Sludge shipped to Envirotech
7-10-72, 5 barrels for 30"
kiln test.



* Volume decreased to 2/3 upon roasting.

-------
to
a


E
P.
a.
          figure 4.  Hg contents of untreated brine sludge after roasting

                             in lab and  pilot kilns
      10
° r
01 L
[TI • ~
:
H
J
O
CK
0 1
h
0
O
§


—
-
KEY
• Lab
1 1 ° ub
g 1 (5 Lab
0.1 1— 3 Lab
E • Lab
r A Lab
\y

%


•
kiln, agitated
kiln 1/2" layer
kiln 2" layer
kiln 4" layer
kiln 9" Layer
kiln, metal anode sludge
L Q Bartlett-Snow pilot rotary


.01
V BSP
A BSP
A BSP
r id
i •*• i
batch kiln, 13"
30" kiln, 200 Ibs/hr.
30" kiln, approx. 300 Ibs/hr.
Australia Ltd. 30" kiln tests
i i 	 i 	 1 	 1 	 L_
                                                     9
                                                          V   1
                                                           A




                                                           A
                                                                O
AGO       600       800       1000       1200      1400



                      KILN TEMPERATURE, 'F
                                                                    1600
                                                                     1800
                                           12

-------
                       Table  3.  ROTARY CALCINER ROASTING OF BRINE SLUDGE
GO

Hg content of starting material, ppm dry basis
Maximum temperature at steady state, °C.
°F.
Retention time, min.
Water content of feed, %
Solids feed rate, gm/min.
Ib. /min.
3
Purge air rate, standard m /hr.
scfh
Screen analysis of solids collected
in first 2 hours, % +8 mesh
-8 +60 mesh
-60 mesh
Hg content of screened fractions, ppm
+8 mesh
-8 +60 mesh
-60 mesh
Hg content of steady state
5 minute sample, ppm
Hg removed, %
Run no. 1
1540
768
1415
30
30
57
0.125
7.4
80

40
31
29

14
92
284

50
96. 8
Run no. 2
(drying only)
1540
357
675
30
30
57
0.125
7.4
80

	
	
----

	
	
	

89
94.2
Run no. 3
89
802
1475
30
0
57
0.125
7.4
80

59
33
8

6.6
19
93

25
98. 4 Over;
                                                                                                      72 This  pass

-------
           Table 4. MULTIPLE HEARTH FURNACE ROASTING OF BRINE SLUDGE
                              BSP 76 cm (30 in^ PILOT FURNACE
Test
no.
1
2



3



Sample
number
1
2
3



4



Hearth
sampled
3
4
5
6
3
4
5
6
3
4
5
6
Hg, Wet feed rate, Retention time,
ppm kg/hr Ib/hr min.
5.7, 5.4 91 200 30
11.4, 5.8
4. 8, 4. 7
3.2, 3.4
5.6 91 200 45
4.7
4.1
3.2
4.1 136 300 30
3.7
4.3
4.4
Test date:  7-24-72
Feed  moisture content: 37.6%
Feed  Hg content: 1735 ppm,  dry basis
Furnace temperature:  760° C.  (1400° F.)
Average gas  consumption per pound of feed: 0.068 m  (2.64 ft. )

-------
           Table 5.  MULTIPLE HEARTH FURNACE ROASTING OF BRINE SLUDGE
                                 BSP 76 cm (30 in) PILOT FURNACE

Test
no.
1



2



3



4




Hearth
sampled
3
4
5
6
3
4
5
6
3
4
5
6
3
4
5
6
Retention
Hg, Wet feed rate, Temperature, time,
ppm kg/hr Ib/hr °C. ° F. min.
1.8 91 200 870 1600 30
1.2
0.31
0.14
1.5 91 200 955 1750 20
0.69
0,74
0,12
5.6 136 300 870 1600 20
5.1
2.5
1.3
182 400 870 1600 20
10.7
4.7
2.3
Test date:  8-10-72
Feed moisture content:  about 47%
Feed Hg  content:  1, 735 ppm,  dry basis

-------
          Since the tests at BSP were run, contacts were
 made with investigators from an Australian chlor-alkali
 plant who were also searching for a brine sludge treating
 method for removing Hg.  They also tested the multiple
 hearth furnace at our suggestion after experiencing un-
 favorable results from chemical treatment methods.  The
 data from their runs in the 76 cm (30 in) pilot kiln fur-
 nace show results similar to ours (Table 6 and Figure 4).
 Minimum values of 0.1 ppm Hg in clinker were achieved at
 gas temperatures  >800°C
          The multiple hearth furnace is  shown in Figure
 5.   The feed material is conveyed into the top and is
 carried across the top hearth slowly by the rabble arm
 plows,  then falls to the next hearth.   This continues from
 hearth  to hearth until the clinker falls out the bottom
 of the  furnace to be cooled and/or discarded.  The heat
 for the furnace is supplied by gas jets  on 2 - 4 hearths
 and the temperature is controlled by thermocouples and gas
 flow control valves.  Smooth furnace operation with minimum
 attention is dependent on a constant feed of uniform mois-
 ture sludge from the filtering step.

          The roasting method should involve the least op-
 erator  attention of any of the methods considered.   The
 chemical methods studied required many more processing
 steps with more equipment and more .critical control points.

 Roasting with Acid Treatment

          During the roasting tests a number of variations
 were tried, including reducing the volume of the sludge so
 that a  smaller kiln could be used to treat the sludge.  To
 reduce  the basic sludge,  acids were tried successfully.
 Surprisingly,  when acid-treated sludges  were roasted, even
 lower final Hg levels were achieved than for untreated
 sludge  at the  same temperature:   0.02  ppm Hg was achieved
-in  the  clinker below 760° C (1^00°F)  (Table 7 and Figure
 6).   The mechanism is not  known although the phenomenon
 was  observed in 31 separate tests.

          A patent application has been submitted on this
 process to the EPA Office  of the General Council.

          In pilot tests  at BSP Division  of Envirotech on
 June 21,  1972,  Hg residuals as low as  0.10 ppm were found
 after 30 minutes at  730°C  (1350°F)  in  their 33 cm (13 in)
 batch kiln (Table 8).
                              16

-------
Table 6.  MULTIPLE HEARTH FURNACE ROASTING OF BRINE
         SLUDGE BSP 76 cm (30 in.) PILOT FURNACE, Id DATA
Test
no.
1
2a
2b
2c
3
4
5
Furnace temperature
Solids Gas
°C. °F. °C. °F.
699
610
538
610
599
599
599
1290
1130
1000
1130
1110
1110
1110
___
840
849
840
921
799
799
— _-
1544
1560
1544
1690
1470
1470
Hg,
Feed
640
640-1200
it
it
640
667
450
ppm
Clinker
3.8 to 0.5
1.0 to 0.3
8.6 to 0.3
2.2 to 0.6
4.0 to 0.1
5.7 to 0.3
0.1
Test date:   8-29-72
Conditions:
   Feed moisture content:               23-38%
   Feed rate, dry material discharged:  100-300 Ib/hr.
   Estimated retention time:            10-30 minutes
   Data from ICI Australia Limited Plant Pilot test in Australia.
                               17

-------
                  Figure 5.  Cut away drawing of the internals of a multiple
                            hearth furnace showing the rake arms and hearth assembly
                                                                     FEED
FURNACE SHELL OF
SHEET METAL LINED
WITH 6" - 9" OF FIRE
BRICK
                                                                        DISCHARGE
                                    STEP OL'
                                              DXVC-. NB
                                             18

-------
                                         Table 7.  KILN ROASTING OF CHEMICALLY TREATED BRINE SLUDGE
CO
Exp. Residence Temperature
no. Treatment time. hr. • C • F
35
53
61
63
63
65
65
70
70
70
70
72
72
72
72
73
73
73
73
73
73
74
74
74
74
74
74
74
76
76
79
7!)
79
73
79
80
*
**
Chlorine pretreated *
HCl treated
HCl treated in crucible
C12 treated
HCl treated
HCl treated
HCl treated
Acetic acid to pH 2
H.SO to pH 2
HCl fo pH 2
HCl to pH 2; NaOH to pH 10
H,SO.
Acetic acid
HCl
HCl then NaOH to pH 10
Ar:ntic acid pll 2-8
H2S04pH 2
HCl pH 2
HCl pH 0
HCl pH 3
HCl pH 0 (80- C.)
H2SO4
M
M
11
M
II
t|
HCl treated
H2S04 treated +*
HjSO4 treated
tl (I
It II
M It
II M
«2S04 **
Drum rotated 1st hour
Sent to Envirotech 5/24/72
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8


677
649-746
649-760
649-760
649-760
649-760
649-760
649-774
649-774
649-774
649-774
649-774
649-774
649-774
649-774
640-760
649-760
649-760
649-760
649-760
649-760
663-732
663-732
603-732
663-732
663-732
663-732
663-732
593-649
593-649
593-649
593-649
593-649
593-649
593-649
677-719


12SO
1200-1375
1200-1400
1200-1400
1200-1400
1200-1400
1200-1400
1200-1425
1200-1425
1200-1425
1200-1425
1200-1425
1200-1425
1200-1425
1200-1425
1200-1-1UO
1200-1400
1200-1400
1200-1400
1200-1400
1200-1400
1225-1350
1225-1350
1225-1350
1225-1350
1225-1350
1225-1350
1225-1350
1100-1200
1100-1200
1100-1200
1100-1200
1100-1200
1100-1200
1100-1200
1250-1325


Agitation,
min.
0
2
2
2
2
i
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2


Hg contentt ppm
Start End
113
230
3660
495
3660
2280
2280
246
246
246
246
58
855
563
764
862
206
634
284
669
120
586
379
418
503
364
364
312
585
585
120
503
364
379
418
...


244.
0.065
0.21
0.48
0.25
0.26
0.36
0.26
0.05
0.14
0.37
0.02
0.17
0.58
0.10
0. 08
0.02
0.03
0.16
0.15
0.06
0.06
0.02
0. 01
0.05
0.10
0. 002
0.22
0.64
4.84
0.46
0.10
-.
0.13
0.04
0.09



-------

U
        Figure 6.  Acid treatment of brine sludge before roasting in

                         'lab and pilot kiln
     10
60



t    1.0
P*




8
H
CO
w

I
     0.1
     .01
1 1
-


™
-


—
.
-
-

-
—
"

1 1 1 1 1 1 1 -
KEY
Lab, HC1 O
Lab, H2S04 Q
Lab, Acetic A
Lab, Cl ^ "
2
BSP 13", •
o 2 k —_
0 o. :
D »J
8 ;
A ^h
W •
D ^
B ° ~
D
O
i
i i i t i i i
        800
1000        1200        1400


          KILN TEMPERATURE,  °F
1600
                                                                     1800
                                    20

-------
Table 8.  BSP TEST ON BATCH KILN
Test
no.
1



2





Sample
Date treatment
4-16-72 Untreated
t>
it
it
6-21-72 Acid
treated
M
ti
ii
it
ii
Temperature Retention
0 C. ° F. time, min.
760 1400 30
60
120
180
732 1350 15
30
45
60
75
90
Hg content,
Start
874
874
874
874
620
620
620
620
620
620
ppm
End
1.40
.41
.42
.32
.55
.10
.17
.45
.15
.11
                  21

-------
         Acid pretreatment was not included in the final
process design because heating the untreated sludge an
additional 1650° (300F°) achieves the same residual Hg
levels at less capital and operating expense.  However,
the acid pretreatment may be incorporated into the process
at a later date if it appears to be necessary- due to in-
creased sludge volume or higher than expected residual Hg
levels.

Solids Dewatering

         The starting material for the sludge processing
system is sludge as it comes from the sludge pit.
Typically, this sludge is only 5 - 10$ total suspended
solids as it is pumped from the sludge pit; a 45 - 60/E
solids feed to the furnace is desirable for economic op-
eration.  The dewatering methods tried were:  (1) gravity
settling, (2) centrifuging, and (3) filtration.

         Gravity settling was not satisfactory for this
sludge since the maximum concentration achieved was ap-
proximately 30% solids.

         In lab tests at Barrett Centrifuge, the sludge
was dewatered to 12% solids in the first stage and the
liquid from the second stage contained 0.2 - 1% insoluble
solids (Table 9).  However, there may be disadvantages to
centrifuging over filtration for this application.

         Possible disadvantages are:  (1) higher capital
cost for equipment for a given capacity, (2) difficulty of
obtaining corrosion-resistant material for wetted parts in
other than stainless steel, (3) two- to four-stage centri-
fuging is needed to achieve solids-liquid separation, (4)
centrifuges are generally made in only 1 or 2 sizes in each
style so that multiple units must be used to achieve re-
quired throughputs, and (5) the high speeds of centrifuges
with such abrasive and corrosive material as brine sludge
could lead to high maintenance.  For these reasons, rotary
vacuum filtration was selected instead.

         The high solids loading indicated that pressure
filtration was not feasible.  Vacuum filtration was ef-
fective on a bench scale.  Two standard rotary vacuum fil-
ters were tried:  a Komline-Sanderson 0.9 m (3 ft) diameter
x 1.5 m (5 ft) unit, and an Eimco 0.9 m (3 ft) x 0.3 m
(1 ft) unit.  As shown in Table 10 and Figure 7, drum rate
and solids feed concentration are critical for filter ca-
pacity and cake dischargeability.  Since filters are avail-
able in standard sizes, a filter with an area greater than
                             22

-------
       Table 9.  TEST RESULTS FROM BARRETT CENTRIFUGE ON
               DEWATERING OF BRINE SLUDGE BY CENTRIFUGING
Sample
No.
   Test           Solids        Insoluble   Flow rate thru  unit,
description    by volume, %    solids,  %	
              As received
                   40
25.6
                         MODEL 912 CENTRIFUGE
   2
   3
1 pass
   sludge
   supernatant
                                 21
71.8
 6.0
                        MODEL 125 CENTRIFUGE (912 Supernatant)
   4

   5

   6
   7
1 pass              0. 7
   supernatant
3 pass              0. 24
   supernatant
5 pass              0. 11
   supernatant
   sludge
 1.06

 0.4

 0.2
21.4
1

1

1
                              TEST METHOD
                              sludge
feed fc
'
'
Model
912
1

supernatant
t
V
Model
125


1 sludge
                                                              supernatant
                                   23

-------
                        Table 10.  DATA FROM THE EIMCO PILOT ROTARY VACUUM FILTER
to
Test
no.
1
2
3
4
5
6
7
8
9
10
11
12
13
Drum rate,
rpm
0.22
0.32
0.41
0.41
0.22
0.47
0.30
0,22
0.41
0.30
0.22
0.22
0.22
Feed Cake dry
suspended solids, solids rate
weight % kg/hr Ibs/hr
18.7
17.6
18.3
17.0
15.5
15.2
15.7
6.3
5.7
4.9
21.0
15.5
18.4
34.2
41.1
44. 1
32.3
16.2
21.5
19.1
8.3
24.3
16.6
56.2
38.6
48.5
75.4
90.5
97.3
71.3
35.8
47.3
42.0
18.4
53.6
32.1
124.0
85.0
107.0
Filter size
for 3.5 ton/day
m sq. ft.
3.4
2. 9
2.6
3.6
7.2
5.4
6.1
13.9
4.7
8.0
2.0
3.0
2.4
37
31
28
39
77
58
66
150
51
86
22
32
26
Cake thickness,
mm in.
5.
3.
3.
3.
2.4
1.6
2.4
1.6
1.6
1.6
8.
6.
4.
0.19
0.13
0.13
0.13
0. 09
0.06
0.09
0.06
0.06
0.06
0.06
0.25
0.16

-------
                     Figure  7.   Eimco rotary vacuum pilot filtration of brine sludge
     150

-------
7 m2  (75 ft2 ) was selected.  The filters 0.9 m (3 ft) x
2.4 m (8 ft) and 1.8 m (6 ft) diameter x 1.2 m (4 ft) are
approximately 7 m2  (75 ft2 ) in area; the larger diameter
is preferred since there is greater control of drying time
and cake formation time.  A 1.8 m (6 ft) x 1.8 m (6 ft)
filter would be a good investment since the small addi-
tional capital cost would provide 50$ more filter capacity
in nearly the same space.  The filter area would be approx-
imately 10.4 m2  (112 ft2 ).  At a filtration rate of 10 -
18 dry kg/m2/hr (2.3 - 4 dry lb/ft2 /hr) with a 33% sub-
mergence, the 1.8m (6 ft) x 1.8m (6 ft) rotary vacuum
filter will discharge 2.8 - 4.9 m tons (6,200 - 10,800 Ib)
of solids per 24-hour day.  The extra capacity permits the
filter rpm to be slowed to build a thicker, drier cake if
cake discharge becomes a problem.

Equipment Sizing

         The system was sized on the basis of 3-2 m tons
(3.5 s tons) per day to handle the expected solids from
the chlor-alkali plant at a chlorine production of l8l m
tons/day (200 s tons/day).  This corresponds to a
sludge production of 1.37 m tons (1.5 s tons) per day from
the plant plus 1.83 m tons (2.0 s tons) per day from stock-
piled sludge and other Hg-containing solids.

         The major pieces of equipment include a 3-7 m
(12 ft) diameter x 1.8 m (6 ft) high thickener, a rotary
vacuum filter and a 1.37 m (4.5 ft) l.d. 6-hearth multiple
hearth furnace (Figure 8).  All decanted and filtrate brine
is recycled to the settling pond so that the small amounts
of solids remaining in these streams will not load up the
water handling system.  In addition, if shower water is
needed to clean the filter cloth or sluice out sludge build-
ups around the filter, brine will be used and returned to
the sludge pit.  No fresh water will be used for wash-down
to maintain the water balance.

WATER TREATMENT FOR Hg REMOVAL

Laboratory Methods

         The test methods used to find the optimum water
treatment process involved many standard laboratory pro-
cedures.  Solid particles such as the ion exchange resins,
activated carbons, and metal particles were packed in a
glass column of 1.9 cm (3/4 in) i.d. with a packing depth of
approximately 30 em (12 in).  A constant liquid flow was
maintained by a head of liquid 7.6 cm - 15 cm (3 - 6 in)
above the top of the packing.  Chemical tests to reduce Hg


                             26

-------
                                        Figure 8.   Proposed  brine  sludge  handling system
NJ
                                                                                                                            FLUE CAS TO
                                                                                                                            REFRIGERATION
                                                                                                                            SYSTEM
                                                                                                                   HRINE FILTRATE TO
                                                                                                                   CURiriEX
I ^ sonns
 X"-7*"-^v
<&&£s
                                                                                                                    UKINEMTED SOLIDS
                                                                                                                    TO IANDMLL
                                                                                           taZMRAlED
                                                                                           SOLIDS BIN

-------
ions with sodium borohydride, to precipitate Hg sulfide with
sodium sulfide, or to reduce Hg ions with powdered metals
such as zinc or aluminum were performed in 500 ml beakers
with magnetic mixer agitation.

         The untreated Hg-contaminated water was first
added to the beaker.  The pH was adjusted with hydrochloric
acid or sodium hydroxide to the desired point.  The ap-
propriate chemicals were added and mixed for varying periods.
The liquid was then filtered through a 10 cm (4 in) or 15
cm (6 in) Buchner funnel precoated with about 6 mm (0.24
in) of diatomaceous earth.  The untreated and treated so-
lutions were then analyzed for Hg by flameless AA.

         To determine the solubility of mercuric sulfide
in solutions of varying pH and excess sulfide, solutions of
mercuric chloride and sodium sulfide were combined.  The
resulting precipitate was collected, washed and weighed
into equal amounts.  These samples of HgS were placed in
sealed containers of water at various pH and excess sulfide,
agitated and allowed to come to equilibrium.  The super-
natant was then analyzed for Hg.

Alternative Methods Investigated

         During the laboratory phase of the project, sev-
eral of the methods proposed for Hg removal from water were
tried including:

         1.   Ajinomoto ion exchange resin

         2.   Billingsfors - Langed ion exchange resin

         3.   Nuchar 722 activated carbon

         4.   Pittsburgh HGR activated carbon

         5.   Calgon Filtersorb 400 activated carbon

         6.   Zinc particles

         7.   Sodium borohydride

         8.   Stannous chloride

         In general, these methods were not able to achieve
effluent Hg levels below 0.10 ppm for starting solutions of
2-20 ppm, or their capacity was limited so that their
effective life was greatly shortened by concentrated Hg
feeds.  The ion exchange resins and activated carbons ap-
                              28

-------
pear to be most effective as polishing steps after the first
stage of treatment has removed the bulk of the Hg.  They
are able to treat solutions in the range of 40 - 100 ppb
down to 1 - 5 ppb consistently.  Appendix C details the
results of these tests .

         Using another metal to reduce and adsorb Hg ions
while dissolving the second metal tends to trade one ef-
fluent problem for another, e.g. the zinc reduction method.
         Of the methods tested, sodium borohydride,
appears to be the best alternative to the sulfide precipi-
tation for a primary Hg removal process.  The equipment
necessary is very similar to the sulfide addition process
and with careful engineering a system could be built to
use NaBHn - sulfide interchangeably with only minor modifi-
cations.  The only drawbacks found for the NaBHjj method for
this application appeared to be the slightly lower effic-
iencies found in the lab tests and the higher cost of NaBHjj.

         Ventron Corporation holds a patent on the use of
NaBHjj for heavy metal removal  (2).

         In the last decade, the literature includes sev-
eral less common ways to. remove Hg from water or brine.
Among the methods proposed are:  (1) solvent extraction with
high molecular-weight amines  (3, 4); (2) electrolytic means
by passing the solution  through a type  of diaphragm  cell
 (5, 6);  (3) hydroxide flocculation and  filtration  (7); and
 (4) adsorption of Hg compounds by CaCp  slag and flocculation
with PeSOh   (8).  Of these methods, the latter two seem to
have some promise although we  did not investigate these
techniques.

 Sulfide  Precipitation

         A number of publications have  described the
 use of  sulfide ions for  precipitation of Hg from water so-
 lutions  (9,  10,  11,  12,  13,  14,  15,  16,  17, 18,  19,  &  20).
 It is generally  agreed that  a removal rat.e  of  99-9$  can be
 achieved with  sulfides and this  has  been  confirmed in  our
 test work on laboratory  and  pilot  scales  (Table  11).

         The drawbacks to this method include:  (1) the
 formation of soluble  sulfide complexes  at high levels  of
 excess  sulfide,  (2)  the  difficulty  of monitoring  excess sul-
 fide  levels, and (3)  the problem of sulfide residue  in the
 waste water  discharge.
                               29

-------
                               Table 11.   LAB TEST DATA ON SULFIDE PRECIPITATION FOR Hg REMOVAL FROM WATER
CO
o
Sample number
pH Range 7-14
1
2
pH Range 5-7
4
5
6
7
pH Range 1-5
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
PH
12.0
10.0

5.5
5.5
5.5

4.5
4.5
4.5
3.5
3.5
1.5
4.0
4.0
4.0
4.0
1.0
3.0
4.0
4.0
4.5
4.5
Reaction Filtration
time, Filter
min pad Precoat
60
60

1/2
10
60

60
60
60
60
60
120
30
60
60
50
60
60
60
20
30
30


Paper
Paper
Paper

Glass-Fiber
Glass-Fiber
Glass-Fiber
Glass-Fiber
Glass-Fiber
Glass-Fiber
Glass-Fiber
Glass-Fiber
Glass-Fiber
Glass-Fiber
Glass-Fiber
Glass-Fiber
Glass-Fiber
Paper
Paper
Paper


D.E.
D.E.
D.E.

No
No
No
No
No
No
No
No
No
No
No
No
No
D.E.
D.E.
D.E.
Body
feed


No
No
No

No
No
No
No
No
No
No
No
No
No
No
No
No
D.E.
D.E.
D.E.
Acid
for pH
adjust.
None
None
HC1
HC1
HC1
HC1

HC1
HC1
HC1
HC1
HC1
HC1
HC1
HC1
HC1
HC1
HC1
HC1
HC1
HC1
H2S04
HC1
Initial
Hg content,
ppm
4.2
10.0
7.0
16.0
16.0
16.0

52.0
52.0
60.0
60.0
60.0
14.0
36.0
36.0
36.0
36.0
36.0
10.0
10.7
27.0
17.5
17.5
Final Hg ,
ppm
1.8
.36
0.11
0.63
.04
.03

5,6
.44
.03
.02
.01
.18
.002
.015
.006
.002
.012
.008
.12
.31
.15
.1
Hg removal
57.0
96.4
98.4
99.6
99.9
99.8

89.2
99.2
99.95
99.97
99.98
98,7
99.994
99.96
99.99
99.994
99.96
99.92
98.9
98.8
99.1
99.4

-------
         From our work we have found that sulfide excess
is less critical so long as the HgS precipitate is filtered
out of the solution as soon as it is formed, as shown in
Table 12.
   Table 12.
HG LEVELS IN WATER AFTER EXPOSURE TO
SULFIDE ION FOR 30 SECONDS TO 10 MINUTES
           Mercury content   Exposure time to  Hg removal,
           of solution, ppfa  sulfide ion, min	%	
Starting
  solution
 51,500
After precipi-
  tation &      9.7
  filtration   15.0
               15.0
               14.0
                   0
                     0.5
                     2.0
                     5.0
                    10.0
                                99-98
                                99.97
                                99.97
                                9-9.97
Probably the sulfide comlex is formed more slowly than
the mercuric sulfide particle.
Hg
Hg
      +2
          2S
                          ~2
                                  > 2HgS + 2Hg
           TT 0   _-2 ,      .  slower  TT  -2
           HgS + S   (excess)        > HgS0
                            •^          ti
         In laboratory sulfide precipitation tests, 99-98$
of the Hg was reacted after only 30 seconds of contact.
For all practicle purposes, the reaction can be said to
go to completion in 30 - 60 seconds.  As Figure 9 indi-
cates, even large excesses of sulfides did not reduce the
recovery of Hg significantly.  Therefore, the accurate
monitoring of the exces's sulfide level is not necessary.

         The third problem stated above, that of sulfides
discharged with the treated water, is easily solved in
a chlorine plant.  There is normally a small amount of
residual chlorine in the main cooling water stream which
effectively oxidizes the sulfide ion.  If residual
chlorine is not available in the treated discharge, then
                              31

-------
a small amount of sodium or calcium hypochlorite can.
be added to the filtered waste water to eliminate the re-
maining sulfide ion.

         The most critical parameter in controlling the sul-
fide precipitation of Hg was found to be pH.  Lab tests were
performed using mercuric sulfide which was precipitated from
mercuric chloride and sodium sulfide solutions (Figure 9).
A standard solution of sodium sulfide was prepared and
placed in full bottles of water containing various concen-
trations of sodium sulfide.  The mixtures were agitated sev-
eral times and then allowed to settle 24 - 48 hours, enough
time for the Hg in the solid phase to approach equilib-
rium with the Hg in solution.  The results appear in Figure
9 as initial excess sulfide level versus final dissolved Hg
level as a function of starting pH.

         Since this experiment allowed a lengthy contact
time, it is likely, that the soluble mercuric sulfide ions,
HgS| , are formed at the higher initial sulfide levels so
that high dissolved Hg levels are due to both pH and excess
sulfide.

         From the data presented, it appears that within
the pH range of 3 - 8 the level of excess sulfides does not
affect the amount of Hg sulfide which is redissolved.  Less
than 0.1$ of the mercuric sulfide added to the sulfide con-
taining water redissolved at these pH conditions in 48 hours.
However, as high as 75% of the mercuric sulfide redissolved
above a pH of 10.

         Possible methods of adding a controlled amount of
sulfide to the waste water stream are: (1) meter the desired
concentration from a concentrated NaHS solution, (2) add
the solid NaHS or Na2$ particles from a dry feeder, or (3)
pass the waste water through a bed of less toxic metal sul-
fide which has the proper solubility to release sulfides at
the 1-5 ppm level.

         The first method has been used exclusively in our
test work because it is the easiest in batch tests in the
lab or pilot plant.  However, on a continuous plant.scale
the second method would eliminate the need for an operator
to mix an exact solution concentration.  Instead, he
could simply add a bag or two of the solid sulfide to the
dry feeder hopper when necessary.  A possible problem with
this method is the rate of dissolution of the sulfide par-
ticles in the cold waste water.  If sufficient agitation
and residence time were not provided, the sulfide would not
dissolve completely.

                             32

-------
                                             Figure  9. Solubility of HgS in excess S
           1000
CO
           en

           B.
           (X
3
            100
                                                 <
                                              pH 6
                                             pH 6 pH  7 pF8
                                   pH  ft  pH  7
                                    pH 8  pH 4,5 pH  5  pH pH
                                                        3   6
                                                       A
                                                       pH 6
                     Q-—
                                         pK  13
            0.1
                                                                                          12
                                                                                                   KEY

                                                                                            0  G.P.  data -  R.P.

                                                                                            £D  Chlorine Institute
                                                                                                     data

                                                                                            A  G.P.  data -  D.R.
                                                                                      At 0 ppm S
                .001
                      .01
.1               1.0              10

"  IN SOLUTION AT  EQUILIBRIUM, ppm Hg
                                                                                                     100
                                                                                                           1000

-------
         The third method proposed has been discussed but
not tested.  In theory, since a certain excess sulfide must
be maintained, it should be possible to control this excess
chemically.  By finding a sulfide compound of the correct
solubility and forming a packed bed of this material,, the
waste water passing through the bed would pick up the metal
and sulfide ions to saturation.  As the sulfide ions are
consumed in the HgS precipitate, more sulfide would dis-
solve according to the solubility product of the specific
metal sulfide used.  Possible problems with this method are:
(1) toxicity of the metal ion selected, (2) coating or de-
activation of the ted by contaminants in the waste water,
or (3) suppression of the solubility of the metal sulfide
if the metal ion concentration is high in the waste water.

The Selected Water Treatment Process

         Of the alternatives considered, it appeared that the
sulfide precipitation method was the best choice for sev-
eral reasons.  First, only five processing steps are needed
to achieve 99% Hg removal.  These steps are: pH adjustment
to 5 - 7, addition of the sulfide ion, addition of the fil-
ter aid, filtration, and solids feed to the furnace.  Most
other methods require 1-3 more steps (figures 10 - 13).

         Secondly, the pH adjustment which is required of
the waste water changes the pH slightly to the acidic side
which will help to neutralize the basic cooling water ef-
fluent.  Thirdly, the precipitated Hg is in a very concen-
trated form in the filter cake, 15 - 30$ Hg, and can be fed
into the furnace without adding significantly to the solids
load.  Finally, the sodium sulfide is inexpensive and no
additional contaminating ions will be in the effluent after
the excess sulfides are eliminated by available chlorine in
the remaining effluent.

         The process proposed is shown in figures 13 & 14.
The process begins with the collection of all the Hg con-
taminated waste water in a 170,000 1 (45,000 gal) agitated
tank.  The pH is adjusted in this tank automatically and
continuously by the addition of spent sulfuric acid from
pH 11 down to pH 5 - 6.  The waste water flows from this
vessel into a container in which the concentrated sodium
sulfide solution (or sodium hydrosulfide solution) is
added by a metering pump.   The treated waste water is then
pumped through a system which adds a measured amount of
filter aid, diatomaceous earth, to the stream.  It then
flows through a pressure filter which separates the Hg-
bearing solids from the water.  The solids are discharged
in a slurry and pumped into the sludge dewatering filter.
                             34

-------
                         Figure JO.  Schematic of the reduction method of Hg removal from water
                                  REDUCING AGENT
                                  LIQUID OR POWDER
w
en
WATER
                                                                           'Hg VAPOR TO COND.
                                              r*i
BED OF
I1EDUC-  	
ING PAR-j    ^FILTER
TICLES
                                                              •»< NACE
                                         ION EX.  I
                                           OR
                                         ACT. C.
                                                                                         EFFLUENT
                                                                             SOLIDS TO LANDFILL

-------
                                  Figure 11,   Schematic of the ion exchange process  for Hg  removal  from water
CO
                                       RECYCLE TO BRINE
                                           PRECIPITATION CHEM.
                                                                                                        SOLIDS TO
                                                                                                        LANDFILL

-------
                                   Figure  12.   Schematic of activated carbon Hg removal from waste water
                                                                      REPLACEMENT CARBON
GO
                                                                                      SECONDARY
                                                                                      CARBON BED
                                                                                                             SOLIDS
                                                                                                             TO
                                                                                                             LANDFILL

-------
                              Figure 13.  Schematic of the sulfide precipitation process  for waste water
co
oo
                                                               FILTER
ION EXCHANGE
OR ACTIVATED
CARBON
EFFLUENT
                                                                                                       Hg  VAPOR TO
                                                                                                       CONDENSER
                                                                                                      SOLIDS TO
                                                                                                      LANDFILL

-------
                  Figure 14.  Proposed  sulfide precipitation  system
lit COKTAM-  ^^mm^
INATED  -X"
WATER   1
    DM
      I	
                                    K.jl
       MISTING 40,000 |>1
       Ftl TANK
                                                                                                   Hx
                                                                                              TILTI*
                                                                                              1ACXUASR
                                                                                              RECEIVER
                                                                                                                  KF.S1S'
                                                                                                                  .\.~T1VA'
                                                                                                                  TED
                                                                                                                   IMC
                                                                                                                   BED

-------
The filtered water might then either go to the outfall or
pass through an optional activated carbon or resin bed
for a final polishing step before discharge.

Liquid/Solid Separation Techniques

         This liquid-solids separation consists of removing
22 kg (60 Ib) per day of fine particles from a 150,000 1
(40,000 gal) water stream.  Thus, the solids make up
0.02J6 by weight of the liquid stream or<0.01jS by volume.
A pressure filter is normally used, since centrifuges handle
high solids loading, rotary vacuum filters handle medium
solids loading, and pressure filters normally handle the
solids loading below 1035 solids by volume.

         A number of alternatives were evaluated:

         1.   Dry or slurry solids discharge?

         2.   If dry discharge, vibrating or centrifugal
              action cake removal?

         3.   Should the filtering elements be vertical
              cylinders, vertical plates or horizontal
              plates?

         4.   Should a precoat be added and what type?

         5.   Is continuous filter aid addition necessary?

         6.   Should the entire chamber open or one large
              bottom outlet be provided?

         In selecting the proper filter, the method of
cake removal is the most important feature to be examined.
Almost any filter will build up a suitable cake; the prob-
lem is frequently to remove the cake completely and restore
the original clear filtering area (21).

         It is desirable for further handling of the Hg-
containing cake that the driest cake possible be discharged
from the filter.  However, because of the small solids load-
ing, 22 kg (60 Ib) per day, the difficulty of cleaning the
filter properly without washing, and the extra cost in-
volved for the more complex equipment, $5,000 - $10,000 ad-
ditional for a 9 m2  (100 ft2  ) filter, we believe the
sluicing and settling technique is best for this applica-
tion.
                             40

-------
         It is more difficult to sluice clean a horizontal
filter.  Also, filtration occurs only on one side of the
horizontal leaf whereas both sides of a vertical are avail-
able for filtration, producing a smaller filter for the
same total area.  For these reasons a vertical cylinder
filter was selected.

         For the efficient filtration of fine Hg sulfide
particles, a precoat is advisable. Our  pilot runs on a
batch pressure filter with horizontal leaves showed that
diatomaceous earth, cellulose fibers and activated carbon
are all effective at removing the Hg sulfide, although the
diatomaceous earth was slightly more effective (Figure 15).
Since the Hg sulfide particles are so fine, they tend to
form an impervious layer on the filtration media rather
rapidly.  However, the addition of small amounts of dia-
tomaceous earth as a filter aid continuously to filter feed
water extended the cycle time by a factor of 10 - 20.  The
diatomaceous earth was added at the rate of 0.7 gpl
(Figure 16).

         The diatomaceous earth filter aid and Hg sulfide
will continue to build up on the filter elements until the
cake space between elements is full.  It is preferable to
backwash the filter before the cake volume is completely
full for a more complete cleaning action.

-------
              Figure )5.  Data from pilot tests with Jl sq. ft.  Niagara filter
CUJ
ffi
X)
a.
a.
H
W

O
u
ffi
H
D
J
EK
fc
W
     10,000
      J.OOO
        JOO
         10
                                                           KEY
                                                Diatomite pll 5-7    0
                                                Diatomite pH 1-3    D
                                                Diatomite pH n-13  X
                                                Pre-co-floc         O
                                                Nuchar 722          A
                      8
                      X
               »  0
                                                                99%
                  O
                  O
                  9
                      a
                      a
                                                                          t
                                                                99.9%	
                                        ]0
                                                 12
14
J6
18
                 Hg LEVEL IN UNTREATED WASTE WATER, ppm Hg
                                     42

-------
                     Figure 16,  Pilot filter testa to determine cycle length with and without precoast and body feed
GO
                                                                          No body feed or flow control
                                                                               - coarse paper, no precoat
                                                                               - fine paper, precoat of D.E«
                                                                               - fine paper', precoat of
                                                                                    Nuchar
                                                                          Body feed and flow control
                                                                               - fine paper, precoat of D.E.
                                            2000
              4000

TOTAL FLOW THROUGH FILTER, gal
                                                                                            6000

-------
                        SECTION V

                       CONSTRUCTION

PROCEDURE

         Following the selection of sulfide precipitation
for water treatment and sludge roasting for solids
treatment, the construction phase of the project began in
October, 1972.  Construction consisted of several
phases:

         1.   Selection and ordering of long delivery
              time items.

         2.   Design and construction of the supporting
              structure.

         3.   Installation of major pieces of equipment.

         4.   Process piping.

         5.   Supplying of necessary utilities.

         6.   System start-up and modifications.

         The equipment ordering began in October, 1972;
construction began in March, 1973.  Start-up of  the water
treatment system began November, 1973> and start-up of the
sludge system began in January, 197^.

Selection of Major Equipment

         The longest delivery time item appeared to be the
multiple hearth furnace with a 32-week delivery.  Three
suppliers were investigated: MSI Industries, Envirotech,
and Nichols Engineering.  Envirotech was selected
on the basis of equipment quality and personnel experience
in Hg ore roasting.

         The rotary vacuum filter was the other major piece
of equipment in the sludge system and the alternative manu-
facturers of plastic or rubber-lined rotary vacuum filters
for corrosive liquids were Ametek and Eimco.  The Eimco unit
was selected because a one-month old unit of all plastic
construction was available at reduced cost with immediate
delivery.
                              44

-------
         The only long delivery time item for the water
treatment system was the pressure filter.  There are a
large number of domestic and foreign manufacturers of pres-
sure filters in various designs.  The manufacturers investi-
gated were:

         1.   Buffalo Filters

         2.   De Laval

         3.   Durco - Enzinger

         4.   Niagara Filters

         5.   R. P. Adams

         6.   U. S. Filters

         7.   Votator - Schenck

         The R. P. Adams filter was selected based on:
(1) operator and maintenance experience with this model,
(2) standardization of parts with existing filters, (3) ease
of cake removal with few moving parts, and (U) price per
square meter of filter area.

Const'ru'c'tion Mode 1

         In order to assist in arranging the equipment for
this project, a student engineer was assigned the task of
constructing a model of the entire system (Figure 17).  Due
to the solids handling problems, the elevations and equip-
ment and piping layouts were critical.

         The model was also useful during the operator
training phase of the project.  Before the construction was
complete the operators could see the location of the pipes
and valves during the training classes.

Modifications During Construction and Start-up

         Originally, the design called for an anode crusher
and conveyor to process the spent cell anodes.  However,
during the project, the decision was made to convert over
to metal anodes so the anode crusher was eliminated.
                             45

-------
Figure 17.  The model of the Hg Recovery System
                        46

-------
         Between the plant sludge collection system and the
rotary vacuum filter it was planned to have, hatch settlers.
However., since the other equipment was continuous, the
batch settlers were replaced with a continuous thickener.
In practice the thickener is not needed since the sludge is
fed in batches from it anyway.
                             47

-------
                      SECTXON VI

               OPERATION AND EVALUATION

START-UP OP WATER SYSTEM

Objective

         Following the construction stage, the start-up
crew began checking out the individual parts of the water
treatment process and training the operators.  This phase
began in mid December, 1973 for the water treatment part of
the project.  The goal was to reduce the chlorine plant
effluent to less than 45 g (0.1 Ib) Hg per day by January
1, 1974* in compliance with discharge permit T-3456, and
have the bugs worked out of the system.

Start-up

         The start-up phase of the water treatment system
lasted approximately one month, from mid December, 1973 to
mid January, 1974 before control was turned over to the op-
erators.  The major changes to the original design were:
(1) rerouting of the Hg-containing water through the pond,
(2) changes in the acid addition system, and (3) a change
in the sulfide storage and addition mechanisms.

         Untreated water flow from the cell room was in-
termittent, from-a collection sump.  This made pH adjustment
difficult.  Therefore, the existing pond was used as a surge
tank to give a constant flow to the pH adjustment system.
The acid addition system was also modified to provide two-
stage dilution for pH adjustment of the waste water with
concentrated spent sulfuric acid.  At present, the system
controls in the range of 6 - 8 pH.

         Originally, the sodium sulfide was to be mixed
in a storage tank and then transferred to a small metering
tank to be metered into the process through a pump.  This
was altered so that the sulfide flows directly from the
large storage tank to the mix tank through a rotameter.

Evaluation

         At present, the water treatment system is oper-
ating as expected and requires approximately 30 - 45 min-
utes of operator time each shift.  Following the initial
shakedown period, the operators seem pleased with the
                             48

-------
system.  The only operator attention needed normally is to
add diatomaceous earth once per shift and backwash once
every 1-2 days.

         Prom a design viewpoint, the system is oper-
ating as desired.  The 50 ppb average effluent from the
system at an average flow; of 150 1/min (.40 .gpm) accounts
for only 20% of the maximum allowable 45 g (0.1 Ib) per day
Hg discharged.  The Hg in the filter cake is fed back to
the sludge system as a slurry after each backwash for dis-
posal by incineration.  As shown in Table 13, the average
Hg removal has been 97% the first 3 months of operation
and the effluent Hg content has averaged 49 ppb.

START-UP OP SLUDGE SYSTEM

Objective

         As the construction neared completion, the start-
up crew began checking out each piece of equipment by moving
the sludge through each stage of the process.  As expected,
the main problems encountered in this start-up were solids
handling.  The start-up began in early March, 1974 and
the operators took over two months later.

Problems Encountered In Start-up

         The majority of changes in the sludge system
during start-up were in the conveying system between the
rotary vacuum filter and the furnace.  Originally a Moyno
pump was installed, but when the cake was dry enough to
discharge well from the filter, 65$ total solids, the cake
was too dry for the Moyno pump to handle.  Next, a screw
conveyor was tried but it was still necessary to add a
little water to the filter cake to keep it from sticking
to the screw.  Finally, a small belt conveyor was installed
which appears to be working well.  At present, a 5-vane star
valve is being tested as an air seal where the sludge enters
the top of the furnace.  The sludge did not plug this star
valve in the one-hour tests run to date.

         Other changes have been made during the operation
to improve the system (Figure 18).  It was found that a
large number of sticks were being pumped out of the clari-
fier and plugging the line to the thickener.  Therefore,
the strainer was moved from just before the thickener to
just after the clarifier.  A flow indicator was also in-
stalled in the line to the thickener so that the operators
could readily tell when this flow stops for any reason.
                              49

-------
      Table 13. PARTIAL LIST OF DATA FROM START-UP OF FULL
                   SCALE SULFIDE PRECIPITATION SYSTEM
Date
12-9-73
12-9-73
1-2-74
1-8-74
1-9-74
1-15-74
1-23-74
3-12-74
3-15-74
3-20-74
3-26-74
Average
Hg content, ppb
Feed Filtrate (
820
740
2000
1400
1400
140
800
5000
1300
6000
5800
*
Minimum: 300
Max.
6000
16
40
48
125
50
18
42
68
96
50
51
49 ppb
10
125
Excess Flow rate
% removal pH Na2S, ppm 1/min. gpm
98.
94.
97.
91.
96.
87.
94.
98.
92.
99.
99.
96.
87
99.
0
5
6
0
4
1
7
6
6
2
1
8%

2
6.5
6.2
5.8
8.0
7.6
5.9
6.0
6.9
6.8
7.2
6.0
5.1
8.2
2
2
2
0
1
-
2
2
-
-

0
3
330
310
330
310
310
388
290
310
310
310
330
290
388
85
80
85
80
80
100
75
80
80
80
85
75
100
* The average values for 30 sets of data
                                  50

-------
                        Figure l€.   Installed brine sludge handling  system
BRINE
FILTER
BACKWASH
MISC.
SLUDGE
       PUMP
                                 SLUDGE THICKENER
                                               BELT  CONVEYOR
DECANTED
BRINE
FILTER
u
1/1 ;
, -j

                                       110"
                                       RECVLE
                                       WATER
                                        RINE  SHOWER
                                    PUMP
                                               MULTIPLE
                                               HEARTH
                                               FURNACE
                      1NE CIARIFIER
                                                                                            Hg
                                                                                         CONDENSERS
                                                            FLUE
                                                        Hg   CAS  TO
                                                            REFRIG-
                                                            ERATION
                                                            SYSTEM'
                                                                         BRINE FILTRATE TO
                                                                         CLARIFIER
                                                                       DUST
                                          INCINERATED SOLIDS
                                          TO LANDFILL
                                INCINERATED
                                SOLIDS BIN

-------
         Due to the small size and plugging potential of
the control valve between the thickener and the rotary
vacuum filter, the small continuous valve was replaced with
a larger intermittent valve.  Rather than holding the filter
at a constant level, it is allowed to cycle over a 7.6 -
10 cm (3 - 4 in.) range.

         The rotary vacuum filter was able to pick up a
thicker cake 6 - 10 mm (1/4 - 3/8 in.) than was found in our
pilot tests.  As a result, the drive sprocket on the rotary
vacuum filter was reduced in size so as not to overload the
furnace.  The Eimco rotary vacuum filter minimum rotation
speed was reduced from one revolution in 14 minutes to one
revolution in 26 minutes.  The filter apparently has ample
capacitv for future needs.

Evaluation

         At present, from a design standpoint, the system
is working better than expected.  It runs consistently at
a feed rate of 6.4 m tons (7 s tons) of sludge-per 24 hours
and will remove 99. 8£ of the Hg from the sludge at 730°C -
760°C (1350°F - 1400°P).  The system operates with a feed
content of 345 ppm Hg and a discharge (clinker) Hg content
of 0.5 - 0.8 ppm (Table 14).  This feed rate is twice the
design rate.  As a result of the high throughput, lower
sludge output from plant than expected, and no return of
sludge from the pond to date, it is only necessary to
operate the furnace 1 or 2 shifts every 3 days.  At other
times, the temperature is lowered to 370°C - 480°C (700°P -
900°F) to reduce refractory stress.  As lower quality salt
is processed, producing more sludge, and as sludge inventory
is reprocessed, the filter and furnace will operate for
longer periods.

         Prom furnace tests, it has been found that
dredged material from bark sludge beds can only be fed at
about one half the rate of brine sludge through the furnace
because of its different handling characteristics.  There-
fore, the furnace capacity drops to 3.5 m tons (3.8 s tons)
per day for dredged cellulosic material.

         Prom an operator's standpoint, the process is
working well at this stage.  The biggest problems seem to
be plugging of the strainer between the clarifier and
thickener, and conveying problems between filter and fur-
nace.  The wood problem in the strainer may be eased by
                              52

-------
Table 14.  DATA FROM START-UP OF THE FULL SCALE
           BRINE SLUDGE TREATMENT SYSTEM
Sludge
Source
Brine Sludge
it it
it it
it ii
ii it
Bay dredging
Addition
rate
kg/hr Ib/hr
226
255
264
205
309
137
540
560
580
450
680
300
Temperature
•F °C
1400
1250
1350
1350
1386
1350
760
677
732
732
752
732
Feed
345
255
290
438
370
128
Hg
ppm.
Clinker
0.5-0.8
1.6-3.1
1.7-2.6
2-7.2
1.6
0.95-1.7
% removal
Range Avg.
99.8
98.7-99.2
99.1-99.4
98.3-99.5
99.6
86.7-92.1
99.8


98.9


                          53

-------
more careful wood removal by the screens in the brine flow
ahead of the clarifier.  The sludge conveying problem into
the furnace is being solved by the start-up and construc-
tion crews as problems appear.

         The Hg recovery from the air leaving the furnace
still requires modification to solve the dust plugging
problem in the condensers.  At present the condensers must
be cleaned every 2-7 days.

         The air leaving the condensers has been analyzed
to contain approximately 0.5 lb Hg per day.  The stream is
routed to an existing chilled water heat exchanger and a
Brink demister for Hg recovery.  This recovery system re-
covers over 90/5 of the mercury.
                             54

-------
                       SECTION VII


                        DISCUSSION

DESCRIPTION AND ANALYSIS OF WORK PERFORMED

Literature Search and Company Contacts

         At the beginning of the project, an extensive
literature survey was conducted utilizing the Chemical Ab-
stracts, the Dow Chemical Company Keyword Index onEnvi-
ronmental Aspects of Mercury Usage, and others (22, 23, 24,
25).Pertinent articles from various periodicals and pa-
tents were collected for study.  Also, 22 organizations in
or associated with the chlor-alkali industry, were contacted
by phone, mail, or in person to gather information on the
methods used or contemplated for Hg removal from solids and
liquids.  These contacts are listed in Table 15.

Laboratory Tests, Process Design & Equipment Selection

         As discussed in Section IV, a series of labora-
tory and pilot tests were conducted on brine sludge to
learn which method was. the most effective and practicle to
scale up to a full size plant.  Of the methods tried, sludge
roasting resulted in the lowest Hg in the clinker by 2 - 3
orders of magnitude.

         Concurrently with the sludge trials, tests were
performed in the laboratory on Hg contaminated waste water
to select a process capable of removing Hg to meet the Jan-
uary 1, 1971* limit of 45 g (0.1 lb) per day maximum Hg
discharged in the water effluent.  The literature survey
revealed a large number of alternate water treatment methods
tried, proposed, or potentially effective.  Several samples
of ion exchange resins and activated carbons claimed to re-
move Hg from water were purchased.

         As the data show (tables 11, 16, 17, 18, 19 and
Figure 19) a large number of tests were performed varying
the parameters of concentration, reaction or residence time,
?H and filtration, methods.  In addition to resins and car-
sons, reducing agents (both metallic and chemical), sulfide
Ion precipitation, and flocculating agents were tried.  The
nost consistently effective and practical method from these
sxperiments was a combination pH adjustment and sulfide
precipitation followed immediately by filtration on a
Jrecoated filter.
                             55

-------
       Table 15.  NAMES AND ADDRESSES OF THE COMPANIES
                 CONTACTED FOR INFORMATION BY DIRECT
                  COMMUNICATION DURING THE PROJECT
Company name

a) Aktiebolaget Billingsfors-Langed


b) FMC Corporation


c) Weyerhaeuser Company



d) Stauffer Chemical

el Chemapec, Inc.


f)  Crawford & Russell, Inc.

g) Rohm and Haas Company



h) Sobin Chlor-Alkali,  Inc.


i)  Ventron Corporation


j)  Hoechst-Uhde  Corporation



k) Diamond Shamrock
      Address

S-660 11 Billingsfors
Sweden

Squamish, B.C.
Canada

Chlorine Plant
P. O. Box 188
Longview, Wash.   98632

Axis. Alabama 36505

1 Newark Street
Hoboken, N. J.  07030

Stamford. Conn. 06904

Ion Exchange Dept.
Independence Mall West
Philadelphia, Penn.  19105

P. O. Box 149
Orrington, Maine   04474

Congress Street
Beverly.  Mass.  01915

550 Sylvan Avenue
Englewood Cliffs
New Jersey  07632

Deer Partk,  Texas  77536
                                  56

-------
                          Table 15.  CONT.
Company name

1)  Wyandotte Chemical Corp.


m) B. F. Goodrich Chemical Co.


n)  Monsanto Company

o)  British Petroleum (BP) Chemicals


p)  Mo Och Domsjo


q)  Finnish Chemicals Oy
     Address
P. O. Box 161
Port Edwards, Wise.  54469

P. O. Box 527
Calvart City. Kentucky  42029

Sanget.  Illinois  62201

Murgatroyd's Works
Sandbach, Cheshire

Husum, Sweden and
Ornskoldsvik,  Sweden

Aetsa, Finland
                                  57

-------
  Table 16.  DATA FROM LITERATURE ON ION EXCHANGE RESINS
Literature source
             Hg level after this series of tests.
Initial Hg                  ppb
  level.        Pre-                 Polishing
   ppb       filtration    I. E, resin     resin
 (5) Osaka Soda process     «w 20, 000      /u 5. 000
(18) Dow Chemical patent
       #3.083,079

(28) A.B. BiUingsfors -
    Langed

(41) Terraneers process
    (ion exchange or ad-
    sorbant material not
    specified)
(62) Ajinomoto  Co.  of N. Y.    1.000-       Yes
                             15, 000   Not measured
15. 000
2. 000-
5,000
29. 000-
70, 000
None
None
None
                              150
2-5
                             300

                            100-200    10-20


                            110-1500
                            1-10
                                   58

-------
        Table 17.  LAB TEST DATA ON ION EXCHANGE RESIN
                      FOR Hg REMOVAL FROM WATER
Resin Company
Ajinomoto
Ajinomoto
Ajinomoto
Ajinomoto
Ajinomoto
Ajinomoto
Billingsfors-Langed
Billingsfors-Langed
Billing sf or s - Lang ed
Billingsfors-Langed
Billingsfors-Langed
PH
11
11
1.5
6.0
1.5
6.0
11
11
1.5
6.0
6.5
Initial Hg.
ppm
13.5
1.8
.06
.087
189.00
205.00
13.5
1.8
189.00
205. 00
0.035
Final Hg,
ppm
0.38
.99
.005
.003
1.9
0.4
1.8
2.0
51.5
15
.001
Hg removal,
%
97.2
45.0
92.0
96.5
99.0
99.8
86.7
0
63
92.7
97.0
NOTE:  All tests were performed in a  glass  column 3/4" i. d. with a
        packing depth of 12 inches.   Flow was controlled by  main-
        taining a head of liquid  3-6"  above top of packing.
                                    59

-------
Table 18..  LAB TEST DATA ON ACTIVATED CARBON FOR Hg REMOVAL FROM WATER
Activated Carbon
Nuchar 722
Nuchar 722
Nuchar 722
Huchar 722
Nuchar 722
Nuchar 722
Pittsburgh HGR
Pittsburgh HGR
Pittsburgh HGR
Pittsburgh HGR
Pittsburgh HGR
Calgon Filtersorb 400
Calgon Filtersorb 400
Initial
pH
11.5
11.5
6.0
1.5
6.0
1.5
11.5
11.5
6.0
6.0
1.5
11.5
11.5
Initial
Hg,
ppm
13.5
1.8
.087
.060
205
189
13.5
1.8
.087
205
189
13.5
1.8
Final
Hg,
ppm
0.23
.02
.006
.0045
.37
.1
0.43
.47
.020
3.1
16
0.73
.03
Hg
removal ,
%
98.3
98.9
93.0
92.5
99.8
99-. 95
96.8
73.9
77.0
98.5
92.0
94.6
98.3
  NOTE:  All tests were performed in a glass column 3/4" l.D. with a packing
         depth of 12 inches.  Flow was controlled by maintaining a head of
         liquid 3-6" above top of packing.
                                    60

-------
Table 19. LAB TEST DATA ON Hg REMOVAL FROM
               WATER BY REDUCTION
Reduction agent
Zinc particles




Sodium borohydride


SnClo
Material
form
10 mesh




Liquid


Solution
PH
11.5
10.0
6.0
2.5
6.2
12.2
	
	
____
Initial Hg,
ppm
1.8
12.5
12.5
12.5
52.0
10.7
4.0
26.0
2.8
Final Hg,
ppm
0. 14
.83
.75
.47
0.09
0.22
.42
.82
.5
Hg removed,
%
92.2
93.4
94.0
96.2
99.83
98.0
89.5
96.85
82.0
                       61

-------
Figure 19.   Lab tests on mercury removal from water using sulfide ppt.,
            activated carbon, ion exchange resins and reduction chemicals










•a
a.
•
2
^H
w
H
**
05 g
K3 ^
10 a
F-
2
H
D
J
fa
fa
W








X X
-, '' y
• NallS pH 7-14 T X y
0 NaflS pll 5-7 	 Sulfide X .' A
- O NallS pll 0-5 Jj ^X / /
• Nuchar 722 ~~1 V X X
B Pittsburgh HGR — Carbon ^X QX ^X
D Filtersorb — J X X X
A Ajinomoto ~1_ Resin / / */
A Billingsfors-Langed_J x yr >
+ Zinc ~"| /f • AX .*
X SnCl2 1— Reduction xx xX /
- - NaBH4J / A x' / /'
X/ r* <* / '' ''
' / .1 x/° ^ /
v X / _— X X
x^ / x °x x
^ ./* °/°° +/'
&* / s S 9 / x
x x / x x
/ „ / °/ ";/ • /
/ B rA<>X • X 0 X
XX ^X /X X ° X
XX« Af/ / ° o XX

xxxj V x /
X A XX QAV Q^°X
X / CX' X CX'X
^X x ^ X ^>^ 00
/ A l/X IX Ix'' 1
         STARTING Hg CONTENT OF THE SOLUTION, ppm

-------
         Following the selection of the water and sludge
treatment processes, the preliminary process design was
drawn up and cost estimates made (figures 8 and 14) .  During
the interval between the preliminary process design and full
scale plant start-up, there were several equipment and pro-
cess changes.  The system in operation is essentially that
shown in figures 18 and 20.

         The major pieces of equipment consisted of a
filter and furnace in the sludge system and a filter in the
water system.  Due to the need for nearly complete solids
removal from the water phase and the high solids, content,
vacuum filtration was selected (Figure 21).
         A number of different furnace designs were      ^
 sidered but this was narrowed to two basic designs for pilot
 tests due to temperature limits and solids handling problems.
 Tests were conducted on a rotary calciner, indirect fired
 and on a multiple hearth furnace (MHF), direct fired.  The
 multiple hearth furnace was selected because it produced
 lower Hg levels in the clinker (Figure 22).

         The filter in the sulfide precipitation  system
 called for a type that would remove a small amount of fine
 solids from, a water stream with minimum losses.   This
 narrowed the filter selection to a pressure filter and prob-
 ably a precoated pressure filter due to the fine  particle
 size (Figure 23).

 Const ruct ion

         As equipment selection was made  the first  stages
 of construction took place.  Engineering  drawings and a
 plant model  (Figure 17) were made.  The support  structure
 and foundation were designed and construction  was started.
 As the major pieces of equipment arrived  they  were  installed.
 The multiple hearth furnace is. shown in Figure 22,  the rotary
 vacuum filter in Figure 21, 'the R.  P.  Adams  filter in Figure
 23, and  the  sludge thickener in Figure  24.

          Once the major equipment,  pumps  and various  tanks
 were in  place, the piping was  laid  by  the contractor, and the .
 electrical contractor was  called in to wire  the  process. A
 separate contractor was hired  to design and  install the  in-
 strumentation for the multiple hearth furnace.  One major
 piece, of construction  involved piping  the natural gas  from
 the nearest  location  to  the  furnace,  a distance  of 350  m
 (1150 ft).
                              63

-------
                                  Figure 20.  Installed sulfide precipitation system for water  treatment
OJ
                         PRESSURE
                         FILTER
                                TO SLUDGE
                                TREATMENT
                                FILTER
                                                       TO SEWER
                                                           DIATOMACEOUS
                                                              EARTH
                                                       DRY
                                                       FEEDER
    Na2S
    STOW
    TANK
                                                FILTER FEED PUMP
FILTER
AID AND
PRECOAT
MIX TANK
GE
9

MERCURY
WASTE
WATER
STORAGE
                     pH ADJUST
                     MIX TANK
                                                                                                              CELL
                                                                                                              ROOM
                                                                                                              SUMP

-------
en
               Figure 21.  Installed 6' x 6' Eimco rotary vacuum filter for brine sludge dewatering

-------
CJ>
05
                         Figure 22.  Installed 54" i. d.  BSP multiple hearth furnace

-------
Figure 23.  Installed R. P. Adams filter for HgS removal
                             67

-------
Figure 24.  The 12' x 6' sludge thickener prior to the rotary vacuum
                     filter in the sludge treatment system
                                68

-------
         Our plant instrument department designed  and  in-
stalled the remainder of the instrumentation for the sludge
and sulfide system.  The nearly completed plant is  shown in
Figure 25.

'Start-up-

         The start-up team was selected during the  con-
struction phase and the start-up leader began, operating and
checking out the equipment as soon as each piece was com-
pleted.  An operating manual was written for use in operator
training (Appendix D).

         As the start-up date approached, the start-up
leader held several oner-hour training classes with  each op-
erating shift, going over the process, the model, the de-
sired operating procedures and the installation.  Feedback
from the operators was valuable in correcting minor problems
apparent before start-up began.

         Numerous problems were encountered during  the
start-up and were corrected, as discussed earlier.  The
start-up was divided, into two parts since the water and
sludge systems are nearly independent.  The water treatment
system was  started up about two months prior to the sludge
system.

INNOVATIONS AND NEW TECHNIQUES

Waste Water Treatment Innovations

         In our laboratory work, and confirmed on a plant
scale, the  critical operations in the sulfide removal
system are  the control of pH and rapid filtration following
the sulfide addition.  The sulfide and filter aid addition
equipment is shown in Figure 26.

Sludge Treatment Innovations

         For our sludge, the furnace roasting process
achieved Hg levels, in the clinker 2 to 3 orders of magnitude
lower than  the lowest Hg levels possible after chemical
treatment (Table 2 and Appendix A).  Still further reduc-
tions could be achieved by acid treating the sludge before
roasting, as shown in Table 7,  although, this  method had the
disadvantage of generating foam and lowered the fusion
temperature of the clinker.
                             69

-------
Figure 25.  The full scale Hg Recovery System as installed at
                the Bellingham Chlor-Alkali Plant
                              70

-------
Figure 26.  The mix tank in the sulfide precipitation system
               where the D. E.  and sulfide are added "
                             71

-------
          Prior to this work, it was reported that the
brine sludge was too sticky to dewater with a rotary vacuum
filter alone; a precoat filter would have been required,
causing greater operating expense and requiring more opera-
tor attention.  The pilot and full scale plants have
demonstrated that our sludge dewaters easily on a rotary
vacuum belt filter.

          One further discovery was that the Hg present in
the sludge was concentrated in the graphite particles
present; the smaller particles had a much higher Hg content,
990 ppm, than the larger particles, 100 ppm,(Figure 27).

ECONOMIC ANALYSIS

Water Treatment System

          The economics of the water treatment system are
shown in Figure 20 and Table 20.  The system as installed
cost $1^3,900 to construct and will handle up to 570,000 1
(150,000 gal) per day.  The operating costs include
chemicals, electricity, operator time and maintenance cost.
These costs total $510 per week or 13
-------
  Figure 27.. Particle distribution in brine sludge
  which was washed and screened to remove particles
  0,007" diameter (Experiment 34)







80
§
H
•U
S5
W
1 40
Q
g
H
w 20
o
3
0
CO
0


\
\
\
\
\
\
\
\
\
\
\
i N
^ ^
X.
\ ^-v^
\ *"*•.
\ -—^
\. ^--_, -

"
•
-



- ^—

-

^



•
,— r~
''
































I 	 .
i\t\j\/
60
se
o.
ta
N
M
en
a
o
h-l
H
500 <
Ou
W
H
H
g
U
bO
0















i   r
            .05
i    i   r   i   i   i
  .10               .15
             PARTICLE DIAMETER, inches
                          73

-------
Table 20.  COST ESTIMATE- WATER TREATMENT SYSTEM
Item Description
Labor
Materials
Total
Filter and installation
Pumps
$ 3,000
4.700
Instrumentation and controls 9, 000
Tanks and vessels
Piping and valves
Electrical
Painting
Structure, ladders and


3, 100
41,000
7,000
1, 100
platforms 8, 700

Engineering at MH @
$16,000
4.000
6,700
5.600
14, 600
4.000
460
4,400
Subtotal
$10.00 MH
Total investment required
$19,000
8,200
15, 700
8,700
55,600
11, 000
1,560
13, 100
132,860
11,000
$143,900
                            74

-------
Table 21.  COST ESTIMATE - SLUDGE SYSTEM
Item Description
Multiple Hearth Furnace
Rotary vacuum filter
Incinerated solids screw feeder
Sludge feed
Instrumentation and controls (System)
Furnace instrumentation and controls
(Union Heating)
Sludge piping and thickener
Heat exchangers and associated
off gas piping
Structure, ladders and platforms
Natural gas and water service
Site preparation and foundation
Pumps
Painting
Electrical
Labor
$ 9.000
2,700
1,200
4,000
6,200

28.000
6,300
31,000
14, 000
8.000
3.000
8,000
12,000
Engineering at MH
Materials
$58, 000
26, 000
4.800
7.500
2. 600

19.000
22,000
14. 000
11, 000
4,000
8,000
2.200
6,000
Subtotal
@ $10.00 MH
Total investment required
Total
$ 67,000
28,700
6, 000
11,500
8,800
18, 000
53, 000
28, 300
45, 000
25, 000
12, 000
11,000
10, 200
18,000
342,500
22,000
$364, 500
                      75

-------
                       SECTION VIII

                        REFERENCES

1.   Botwick, E. J. and D. B. Smith.  Mercury Recovery.
     U.  S.  Patent 3,600,285.  1971.

2.   Ventron Mercury Removal Process.  Koertrol, subdi-
     vision of Ameteck, cata. 7R. 1974. p. 4lB.

3.   Caban, R. and T. W. Chapman.  The Extraction of HgCl
     From Acid Chloride Solutions With Trioctylamine.  A.
     I.  Ch. E. Journal.  18(5):904, 1972.

4.   Moore, F. L. Solvent Extraction of Hg from Brine So-
     lutions with High Molecular-Weight Amines.  Environ-
     mental Science and Technology.  6:525-9, June, 1972.

5.   Edwards, G. E. and N. J. LePage.  Treatment of Brine
     Solutions Containing Both Free and Available Chlorine
     Plus Mercuric Chloride.  U. S. Patent 3,102,035-  1963.

6.   Edwards. G. E. and N. J. LePage.  Treatment of Brine
     Solutions.  Great Britain Patent 885,818.  1961.

7.   Hg Removal by Hydroxide Floe and Filtration.  Hayward
     Filter Co.  Santa Ana, California.  Lit. No. M-10,
     No. 127, 44-A, 119. 1971.

8.   Suhara, I., et al.  Removal of Mercury from Waste
     Liquid Containing Mercury.  Japan Patent 3405.  1964.

9.   Bergeron, G. L. and C. K. Bon.  Mercury Recovery from
     Electrolytic-Cell Brine Effluents.  U. S. Patent
     2,860,952.  1958.

10.  Bouveng, H. 0. and P. Ullman.  Sweden Tightens Up On
     Mercury Wastes.  Industrial Water Engr.  pp. 24-6,
     June,  1969.

11.  Chlorine Institute, May publication on mercury state
     of the art.  Published by the Chlorine Institute.  1971.

12.  Chlorine Institute.  Minutes of the ad hoc mercury
     committee.  Published by the Chlorine Institute.  1970,
     1971.
                             76

-------
13.  Deriaz, M. G.  Recovery  of  Mercury from Waste Brine
    by Sulfide Precipitation.   U.  S.  Patent 3,213,00.6.
    1965.

14.  Fulcher, R. A.  Sulfide  Ion Electrode.   Personal Corn-
    muni cat ion.  1971.

15.  Gardiner, W. C.  Mercury Problems in the Chlor-Alkali
    Industry.  Presented  to  Niagara Palls section of
    Electro-Chemical Society.   Published by the Electro-
    Chemical Society.  November 9, 1970.

16.  Hirs, G.   Private  Communication  to W.  H.  Hunt.  July
    28, 1970.

17.  Hazards of Hydrogen Sulfide.   Safety News  Letter Pulp
    and Paper Section.  National Safety Council.  November,
    1971.

18.  Wilkes, A.  Private Communication.  1971.

19.  Knepper, W. and S. Austin.   Process for Recovering
    Mercury from Waste Waters of Industrial Process.  U.
    S. Patent 3,695,838.   1972.

20.  Dean, W. E. and C. M.  Dorsett. Mercury Removal.
    U. S. Patent 3,67^,428.  1972.

21.  Maloney, G. F.  Selecting and Using Pressure Leaf
    Filters.  Chemical Engineering.  79(11):88-94, 1972.

22.  Bouveng, H. 0.  Control  of  Mercury in Effluents from
    Chlorine Plants.  Presented at International Congress
    on Industrial Waste Water  (Stockholm),  November 2-6,
    1970.  Butterworths Pub.  (London).  1972.

23.  Bouveng, H. 0. and P.  Ullman.   Reduction of Mercury in
    Waste Waters from Chlorine  Plants.  Swedish Air and
    Water Pollution Research Laboratory (Stockholm).
    April, 1969.

24.  Smith, S. B., et.al.   Mercury Pollution Control by
    Activated Carbon:  A  Review of Field Experience.  West-
    vaco Corp.  No. M1002.01.   1971.

25.  Yokota, Y.  Recovery  of  Mercury.   Soda and Chlorine.
    19(3):87-95, 1968.

26.  Tokawa, D. T.  Treatment of Mercury Cell Waste.  B. S.
    Thesis.  University of British Columbia.  1971.

                             77

-------
27-  Glaeser, W.  Method of Producing Mercury.  U. S.
     Patent 1,637,481.  1924.

28.  Parks, G. A. and R. E. Baker.  Mercury Process.  U. S.
     Patent 3,476,552.  1969.

29.  Parks, G. A. and N. A. Fittinghoff.  Mercury Extrac-
     tion Now Possible Via Hypochlorite Leaching.  Engin-
     eering and Mining Journal,  pp. 107-109, June, 1970.

30.  Town, J. W. and ¥. A. Sttckney.  Cost Estimates and
     Optimum Conditions For Continuous-Circuit Leaching.
     U. S. Dept. of the Interior Bureau of Mines report of
     investigations.  6459.  1964.

31.  Osaka Soda Mercury Recovery Process.  Crawford &
     Russell, Inc.  Private Publication.  November 5, 1970.

32.  Gardiner, W. C. and P. Munoz.  Mercury Removed from
     Waste Effluent via Ion Exchange.  Chemical Engineering.
     79(19):57-59, August 23, 1971.

33.  Scheiner, B. J., R. E. Lindstrom, D. E. Shanks, and
     T. A. Henrie.  Electrolytic Oxidation of Cinnabar
     Ores for Mercury Recovery.  U. S. Dept. of the Interior,
     Bureau of Mines Metallurgy Research program.  Technical
     Progress Report - 26.  June 1970. 11 pp.

34.  Anon.   Process Removes Hg in Plant Wastes.  Chemical
     & Engineering News.  48:48, December 14, 1970.

35-  Allenbach, C. R.  Reduction by Gallium, Aluminum,
     and Mercury in Aqueous Solution.  University Microfilms.
     Ann Arbor, Michigan.  No. 3927.  1952.

36.  Gilbert, J. F. and C. N. Rallis.  Recovery of Mercury
     from Brines.  U. S. Patent 3,039,865-  1959.

37-  Karpink. R. S. and J. J. Hoekstra.  Recovery of Hg
     from Brine.  U. S. Patent 3,029,143.  1962.

38.  Karpluk, R. L. and J. J. Hoekstra.  Mercury Recovery
     Using Liquid Alakll Metal.  U. S. Patent 3,029,144.
     1962.

39.  Neipert, M. P. and C. K. Bon.  Reduction of Mercury
     Ion to Metallic with Aldehyde.  U. S. Patent 2,885,282.
     1959.
                              78

-------
40.  Rickard, M.  D.  and G.  Brookman.  Metal Reduction Aid
    for Mercury.  Water and Wastes Engineering, p.  D-2,
    July 1971.

41.  Rhodes, D. W. and M. W. Wilding.  Reduction of  Hg in
    Solution.  TJ. S.  Patent 3,^63,635.  1969-
42.   Smith. W.  W.   Reducing Hg"*"1" and Hg+ to Hg with
     and NaP.   Chemical Engineering (London).  Vol.  216, 1968.

43.   Anon.  Chelating Resin Separates Ions.   Chemical and
     Engineering News.  37:^9, January 26, 1959.

44.   Anon.  Paring Mercury Pollution.  Chemical Engineering.
     78(5):70-71,  1971.

45.   Anon.  Application of Resinous Mercury Adsorbent (Che-
     lating Resin  for Heavy Metal) to Sewage  Disposal.
     Ajinomoto  Co., Inc.  Tokyo, Japan.  1971.

46.   Anon.  Resinous Mercury Adsorbent.  Ajinomoto  Co., Inc.
     Tokyo, Japan.  Technical Data Bulletin No. 71927.
     1970.

47.   Calkins, R. C., et al.  Removal of Mercuric  Ions Prom
     Electrolytic  Solutions.  U. S. Patent 3,083,079.  1963.

48.   Grain, G.  E.  and R. H. Judice.  Electrolytic Process
     for  the  Recovery of Mercury.  U. S. Patent 3,213,006.
     1965.

49.   Walton,  H. P. and J. M. Martinez.  Reactions of
     Mercury  (II)  with a Cation-Exchange Resin.   J. Phys.
     Chem.   63:1318-19-  1959-

50.   Puxelius,  L.   Ion-Exchange Resin for Removal of Heavy
     Metal Ions in Waste Water.  Presented at International
     Congress  on Industrial Waste Water (Stockholm), November 2-6,
     1970.  Butterworths Publ. (London).  1972.

51.   Hokuetsu Tanso Kogyo, K. K.  An Introduction to Mer-
     cury Adsorbing Resins.  Hikari Kogyo K.  K.,  11 Takara-
     cho  2-chome chuo-ku, Tokyo, Japan.

52.   Kraiker, H.  Micro-Ionic Systems for Mercury Pollution.
     L. A.  Water Conditioning Bulletin No. 1100.  1970.

53.   MacMillian. A. L.  Private Communication.  1971.


                             79

-------
54.  Morissette, B. G.  Recovery of Hg from Brine.  Cana-
     dian Patent 595,813.  I960.

55.  Percival, R. W.  Private Communication.  1970.

56.  Rohm and Haas Co.  Recovery of Mercury by Ion Exchange.
     Private Communication from C. T. Dickert.  1970.

57.  Rosenzweig, -M. D.  Paring Mercury Pollution.  Chemical
     Engineering.  78(5):70-71, 1971.

58.  Scholten, H. G. and G. E. Prielipp.  Hg Removal by Ion
     Exchange Resins.  U. S. Patent 3,085,859.  I960.

59.  Selezrieva, N. A., et al.  Separation of Selenium and
     Mercury on Anion Exchangers.  Inst. Yad. Fiz., Alma-
     Ata (USSR).  19(!*):76-77S 1969.

60.  Tsujiya, T.  Private Communication.  1971.

61.  Law, S. L.  Methyl Mercury and Inorganic Mercury Col-
     lection by a Selective Chelating Resin.  Science.
     17^:285-286, 1971.
                              80

-------
                       SECTION IX

                PATENTS AND PURIFICATIONS

1.    Patent Application - "Removal of Mercury from Mercury
               Cathode Sludge", Donald A, Kaciior
               and Richard A.  Perry, Patent Ap-
               plication Serial No.  354,983',
               filed April 27, 1973.

2.    Publication - Perry, R. A^,  Mercury Recovery from
          Process Sludges,   Chemical Engineering
          Progress, 70'(3') :73-86, 1974.
                             81

-------
                       SECTION X

                       GLOSSARY
1.    Body feed - A filter aid added continuously to the sus-
          pension to he filtered to keep the filter from
          plugging.

2.    Brine sludge - Sludge resulting from chemical addition
          to sodium chloride brine to precipitate calcium
          and magnesium compounds and other impurities.

3.    Cell anode - One of the electrodes in a Hg cell, made
          of graphite or metal.

4.    Chlor-alkali plant - A plant producing chlorine and a
          metal hydroxide.

5.    Diatomaceous earth - A meterial used to precoat filters
          and as a filter aid.

6.    Effluent - The waste liquid discharged from a process.

7.    Electrolytic oxidation - The generation of chlorine in
          a brine with electricity to cause oxidation of a
          desired material.

8.    Hg - The chemical symbol for the element mercury.

9.    Hg contaminated waste water - The waste water which
          comes in contact with Hg or Hg-containing material
          in a chlor-alkali process.

10.'  Mercury cell - The unit producing chlorine and a metal
          hydroxide from electricity and brine cathode.

11.  Multiple hearth furnace - A direct fired furnace with
          trays on a vertical shaft.

12.  Precoat - A filter aid added to coat the filter ele-
          ment before filtration begins.

13.  Pressure filter - A filter which uses pressures
          greater than atmospheric pressure on the unfiltered
          side.

14.  Polishing filter - A final filter which removes the last
          traces, following a preliminary filter.
                             82

-------
15.  Rotary calciner - An Inclined  cylinder  heated  and  ro-
         tated; material is passed through  the  cylinder.

16.  Rotary vacuum  filter - A  filter utilizing vacuum inside
         a eye Under to pick  up  and de water a cake on  the
         outside of a drum.

17.  Star valve - A rotating paddle wheel  that allows solids
         to pass through but  seals the  opening  against air
         leakage.

18.  Thickener - A  large continuously  fed  tank which concen-
         trates or thickens a sludge  to a higher total
         solids.

19.  Untreated sludge - Chlor-alkali plant sludge as it comes
         from the  process.
                             83

-------
                       SECTION XI


                       APPENDICES

                                                 Page
A.   Hypochlorite, Chlorine and Electrolytic      35
          Oxidation

B.   Determination of Hg by Flameless AA          95

C.   Experimental Data for Alternate Hg Re-      102
          covery Methods from Water

D.   Operating Manual                            108
                           84

-------
                       APPENDIX A
   HYPOCHLORITE,  CHLORINE,  AN0 ELECTROLYTIC OXIDATION

         The  object  of  the chemical oxidation trials was
to convert Hg  to  the  soluble mercuric ion in the presence
of chloride  ions  to form the soluble mercuric tetrachloro
complex.  The  overall reactions involved are as follows:
      Hg + CIO" + 3Cl" + H2O        ^  HgCll + 2OH~

    2Hg+ + CIO" + 7Cl" + H20         v  2HgCl* + 2OH~

              Hg++ + 4C1"          v  HgCl^


Procedure

          The trials were  conducted  using 250 ml to one 1
of brine  sludge  for each test.   Trial conditions and results
are shown  in Table  22.

          The hypochlorite oxidation trials  involved mix-
ing liquid sodium hypochlorite  with  a brine  sludge in an
agitated beaker  for the time  period  specified.  The treated
sludge  was filtered in a Buchner funnel and  washed with
200 - 400  ml of  distilled  water before analyzing the washed
sludge  for residual Hg (Figure  28).

          The chlorine oxidation trials were similar ex-
cept that  gaseous chlorine was  sparged into  the sludge from
a cylinder of liquid chlorine.

Results

          From the  results of bench  scale trials, the hypo-
chlorite  appeared to dissolve nearly as much sludge solids
(60 - 80$) as Hg (60 - Q6%).  Thus,  up to 90% of the Hg
was removed, but the solid residue contains  100 - 300 ppm
Hg.  In the work of Tokawa (26), a one-stage hypochlorite
extraction removed  40  - 1Q% Hg  and 4-8 stages were needed
to removed over  95% of the Hg (Table 23 and  Figure 29).

          Other  investigators have suggested that the use
of chlorine gas  as  the oxidizing agent would reduce re-
sidual  Hg  in the treated sludge to 1 ppm.  In the 9 experi-
                            85

-------
                                      Table 22.  OXIDATION OF BRINE SLUDGE USING SODIUM HYPOCHLORITE
CO
Exp.
No.
1A
IB
1C
ID
2A
213
3
4A
4B
5
6A
6B
6C
7
8

12
15
17
29
Hypo
concentration,
gpl C12
30
30
30
3D
30
30
30
90
90
90
90
90
90
87
89

190
150
150
100
PH
6.2
6. 2
8.5
8.5
5.7-10
5.2-10
7.3
11.2
11.2
6.8-10. 8
10.0
9.0
8.0
8.0
8.0

8.0
9.0
8.0
8.0
Reaction
time.
hr.
4
4
4
4
3
3
4
3
3
4
1
1
1
23
1

1
1
1
3
Temp..
•c
60
22
60
22
60
60
60
60
25
60
60
60
60
60
60

60
60
60
8-11
Initial
Hg content,
ppm
270
270
270
270
270
270
270
270
270
270
270
270
270
270
270

150
150
4200
260
Solids Remaining
dissolved. Hg removal. Hg in solids,
% % ppm, dry
304
	
	
197
169
175
110
A; 100
A^lOO
93
62%
64%
73.5%
168
«vl258

76% 77% 140
78% 86% 282
47
70% 76% 130
Comments






2 stage


4 stage



2 stage
Pressure
reaction


cell graphite


-------
Figure 28.  Bench  test set-ups for chemical oxidation of sludge
        HYPO  HCl
            6
       MAG MIXER
        HYPO METHOD
                                      Cl,
     6
MAG MIXER
                                                    CHLORINE GAS METHOD
 GRAPHITE PIATES
                       6
                    MAG MIXER
                ELECTROLYTIC OXIDATION

                        METHOD
                                              POWER SUPPLY
                                 87

-------
                        Table 23.  OXIDATION OF BRINE SLUDGE USING SODIUM HYPOCHLORITE
                                   WORK PERFORMED AT UNIVERSITY OF BRITISH COLUMBIA
oo
00
Run No.
Stage No.
START
1
2
3
4
5
6
7
8
F
Hg „
recovered,
%
--
40%
67
81
87
92
96
98
99
Hg
remaining in
sludges, ppm
3280
1970
1080
620
430
260
130
.66
33
G
Hg
recovered,
%
-_
70
86
91
94
95



Hg
remaining in
sludges, ppm
931
280
130
84
56
47



I
Hg
recovered,
%
_ _
70
88
93
96
97.5



Hg
remaining in
sludges, ppm
146
44
18
10
6
4




-------
           Figure 29.  Effect of staging on mercury recovery by Tokawa
100
 50
                                         89

-------
ments performed in these trials, half of the residues
contained more Hg than the starting material (Table 2*1.)...
The chlorination of the sludge dissolved 70 - 9Q% of the
solids and approximately the same quantity of Hg so there
was no net reduction of Hg concentration in the residual
solids.

          A. sample of brine sludge from a chlor-alkali plant
utilizing metal anodes was oxidized with hypochlorite and
chlorine gas.  Over $8% of the Hg was removed, leaving a
solids residue containing 9.2 ppm Hg (Table 25).  This
suggests that metal anode sludges are susceptible to
chemical conversion of Hg to the tetrachloro complex.

ALTERNATIVE Hg RECOVERY METHODS

A.   Hypochlorite and Chlorine Oxidation

          The initial work on extracting Hg from brine
sludge was performed using sodium hypochlorite.  This
method was known to remove Hg as early as 1924 from
Glaeser's work (27).  Extensive work has been done on
sodium hyopchlorite leaching of Hg from low grade ores
by Parks (28, 29), Town (30) and others. Although Parks
achieved a 96-4$ recovery, there still as 10 ppm Hg left
in the residue.  Town was able to achieve a 99.8$ removal
by leaching  of very concentrated ores (79#) but
there remained 24,000 ppm Hg in the residues.

          In Japan, this process has been used by 4 chlor-
alkali plants for up to 5 years to remove Hg from brine
sludge.  This process has been marketed in the United
States since 1970 by Crawford & Russell, who claim the
process will remove 95% or more of the Hg from brine sludge
(31).  In a more recent publication, Crawford & Russell
claimed a reduction of Hg in the dry sludge from 50 - 4000
ppm to 0.1. ppm via the hypochlorite leaching process with
pH adjustment (32).  Prom our work, the 95$ Hg removal
stated in their sales literature is more realistic than the
0.1 ppm Hg residual claimed.

          The basis for this process is the conversion of
elemental Hg and insoluble Hg compounds to water soluble
mercuric Ions with the OC1.  The soluble stable complex
HgCl]j  is formed.

          Tokawa found that multistaging the extraction
process could increase the Hg recovery to 99% with 8
stages (26).  However, the maximum achieved in our labora-
tory in one stage was 86% recovery with a minimum final
                            90

-------
Table 24. OXIDATION OF BRINE SLUDGE USING CHLORINE GAS
Exp.
No.
14A
14B
23
24
25
26
27
28
30
Chemical
Chlorine/ Flotation
Chlorine / Flotation
Chlorine
Chlorine
Chlorine
Chlorine
Chlorine
Chlorine
Chlorine
Time, hr
	
...
2.5
5.0
5.0
1.75
7.0
7.0
7.0
Start
10.0
10.0
10.0
5.0
10.4
10.3
10.3
1.7
10.0
End
5.0
5.0
4.7
4. 1
4.5
4.0
3.6
1.5
4.5
Initial Hg
content, % solids
ppm reduction
150
150
150
150
150
150
150
150
150
-._-
	
70.0
86.5
86. 0
--..
	
	
88.0
% removal
...
...
74
'83
78
45


...
Remaining
Hg in solids.
ppm, dry Comments
114 Foam Hg content
102 Solids
200
163 Reducted pH to 5
,,o before Clo
.1 jO
314
A, 400
A/400 Acid addition to
117 PH 7

-------
CO
to
                          Table 25.  OXIDATION OF BRINE SLUDGE USING COMBINATIONS OF HYPO,

                                    CHLORINE, ELECTROLYTIC ACID TREATMENT AND ROASTING
Exp.
No.
9
10
11
18
29
20
21
22
36
54
Conditions of test
H2SO4 + Hypo
Electrolytic oxidation
Hypo + electioxidation
Hypo + chlorine gas
HC1 only
Hypo + chlorine gas
Hypo after acid treatment
Hypo of roasted
Leaching of roasted
Hypo + chlorine
Initial
Hg content,
ppm
137
150
150
150
150
4200
81
0.26
1.7
158
% solids
reduction
...
	
	
31
53
53
59
50
	
72.4
% recovery
88
—
—
56
72
85
54
17
--
98.4
Final
Hg in solids,
ppm. dry
65
111
120
43
81
150
100
0.2
3.0
9.2
Comments





Cell graphite
Acid treated
Roasted
Roasted
Metallic anode sludge
(Wyandotte)

-------
Hg content of 47 ppm.  Due  to the  great difficulty and
expense required to  separate the  liquid from the fine solids
at each stage In a multistage extraction, a practical
system would be restricted  to a one- or two-stage operation.
It became evident that the  chemical equilibrium was not
favorable to remove  the  residual  5' - 1Q£ of the Hg in the
sludge in one stage.  The use of  chlorine injections did not
improve Hg conversion.

         In the laboratory, graphite in the sludge was
concentrated as the  sludge  was  digested.  In addition,
the Hg concentration in  the graphite particles was
found to be 6-8 times  higher  than the Hg level in
the remainder of the sludge (Table 26).  Prom these
and other data, we hypothesize  that the graphite from
the anodes and decomposer packing is a major contributor
to the residual. Hg after sludge digestion.  This is further
supported by the results from  a sample of metal anode
sludge (little graphite  present)  treated with hypo
and chlorine  (experiment 54 and Table 25).  The final
Hg level-was-' 9.2 ppm compared  to  a sample of our sludge
treated  exactly the  same (experiment 18, Table 25)
which contained  43 ppm  Hg  after treatment.

         The distribution  of  Hg content in our sludge
of various graphite  particle size groups is shown in
Figure 27.  Over  80  weight  percent of particles are 0.76 mm  -
1.8 mm (0.03 -  0.07  in.) in diameter.  Moreover, the Hg
content  increases  as the particle size decreases.  This is
consistent with  an adsorption mechanism  of Hg on the
graphite,  since  the  surface area increases with decreasing
particle size.

          Attempts were made to remove graphite from sludge,
but abandoned  for  3 reasons:   (1) the  sludge would  still
have to be treated for Hg removal;  (2) removal  of the
very fine graphite particles with the  highest Hg content
was difficult  and costly; and  (3) to remove the graphite
quantitatively from the sludge was not practical.
                             93

-------
CO
                      Table 26. Hg ANALYSIS OF BRINE SLUDGE SIZE FACTIONS
Sample Description
Brine sludge before separation
Large graphite particles in sludge
Remaining fines in sludge
Hg content
wet basis,
ppm
45
646
26
Total
solids,
%
35
40
10-15
Hg content
dry basis,
ppm
270
1600
175-250

-------
B.   Electrolytic Oxidation

         A modification of the  hypochlorite and chlorine
oxidation methods for treating brine  sludge is generation
of  hypochlorite ions during sludge treatment. This
method has been used successfully  by  Scheiner (33) of the
U. S.  Bureau of Mines to extract 90 - 95$ of the Hg from
ores containing 300 - 10,000 ppm Hg.   The electrolytic
method simply uses dissolved sodium chloride in the sludge
mixture to liberate chlorine gas when a DC voltage is
applied across 2 graphite electrodes.   Scheiner believes
the tiny chloride gas bubbles formed  and the reaction
between Hg and chlorine at the surface of the electrodes
provides a more effective oxidation than simple injecting
chlorine gas or hypo into the slurry.

         Our tests were performed in an apparatus consist-
ing of 2 graphite plates with an area of 13 cm2 (2 in2) in
a 1000 ml beaker spaced 2.5 cm (1  in)  apart (Figure 20).  A
DC power source supplied current to the brine-sludge solution
to generate chlorine gas from the  sodium chloride present".
The tests were not significantly more  successful than the
chemical oxidation methods; residual  Hg levels in the re-
maining solids averaged 100 ppm.
                             95

-------
                       APPENDIX B


                   DETERMINATION OP Hg

                     BY FLAMELESS AA

          The procedure used to analyze  samples for Hg
during the course of this  project was  modified from an EPA
method published in 1970.   The  complete  procedure is de-
scribed below.

I.   Sampling

     1.   Rinse all glassware and polyethylene containers
          with dilute nitric acid and  then  with distilled
          water prior to use.

     2.   Acidify samples  of water and effluent if they
          will stand more  than  one day prior  to determin-
          ation of Hg.   Add 10  ml concentrated HNOo per
          1000 ml sample.

II.  Equipment

     1.   Rinse all glassware with dilute HNOo  and then
          with distilled water  prior to  use.

     2.   Store glass beads  in  a small amount of concen-
          trated HNO^  and rinse with  distilled water prior
          to use.

     3.   Modified A.O.A.C.  digestion  apparatus.   Substi-
          tute single-neck 250  ml or 300 ml boiling flasks
          for the 3-neck digestion flask.

     4.   Perkin-Elmer Atomic Absorption Spectrophotometer,
          Model 303.   AA settings:   wave length,  25^.5;
          range,  UV;  slit  width,  3;  source  current, 10 ma;
          meter response,  1;  scale,  1.   Perkin-Elmer Re-
          corder Readout:   noise suppression, 2;  scale ex-
          pansion,  x3.   Airflow meter, set  at *JO.   Align
          the gas  absorption cell to allow  maximum light
          to pass  through.   Allow equipment to warm up at
          least 20  minutes before using.

     5.    Aeration  apparatus:

          a.    Use  Anhydrone (magnesium perchlorate)  as
               the  drying  agent;  change weekly or  more
               often  as  needed.

                            96

-------
              CAUTION:  If Anhydrone  comes in  contact with
              skin  or clothing, wash  area immediately with
              water.  Magnesium perchlorate may  cause
              severe burns to  skin  or may cause  fire when
              in  contact with  clothing  or combustible
              material.

         b.   Clean the gas washing bottle biweekly with
              a small amount of HF  acid, rinse with water,
              and clean again  with  dichromate  cleaning
              solution.  Clean the  sparger biweekly with
              boiling dilute HC1.   Rinse apparatus thor-
              oughly with distilled water prior  to use.

              CAUTION:  Hydrofluoric  acid liquid and vapor
              may cause severe burns  which may not be im-
              mediately painful or  visible.  Do  not leave
              glassware in contact  with HF longer than  is
              absolutely necessary.

         c.   Do  not allow moisture to  collect in the 17
              cm  gas absorption cell.   If moisture does
              collect, dry cell thoroughly in  a  105°C
              oven  and change  the drying agent in the dry-
              ing tube.

III.  Procedure

 A.   Preliminary treatment of sample:  Use modified A.O.A.C.
     method for  organic and solid samples and effluent:  use
     modified F.W.Q.A. method for inorganic aqueous samples,
     caustic, and  sulfuric acid.

     1.   Modified A.O.A.C. method:

         Take suitable amounts of  sample  (not  more than
         100 ml or  5 g dry) to provide  0.1 - 1.5 ug Hg,
         place  in a single-neck flask and treat  each ac-
         cording  to type  of sample.

         a.   Samples

              i.    Mud,  sludge, etc.
                    Add  10 ml distilled  water to  sample  and
                    then  add  10 ml  concentrated HNOo  per  g
                    dry  sample. Proceed with digestion  as
                    below (IIIAlb).

              ii.  Effluent
                    Proceed with digestion  as below  (IIIAlb)
                             97

-------
     b.    Digestion procedure

          To the single-neck flask containing the
          sample, add 20 - 25 ml 1:1 HNCU  -H2SOjj and
          3-4 glass beads.   Attach flask to modi-
          fied A.O.A.C.  digestion apparatus.   Care-
          fully heat sample until it refluxes steadily;
          avoid losing gaseous NO  too  rapidly.   Col-
          lect condensate in extraction unit  until
          digest reaches incipient boiling or goes to
          acid fumes.

               If sample darkens or turns  black, cool,
               and add more concentrated HNOo.

          Allow digest to cool;  drain collected
          liquids back into flask, and  reflux for 10
          - 15 minutes to rid apparatus and sample of
          gaseous N02-  (Add 25  ml distilled  water to
          sample through condenser if N02   is difficult
          to remove.  Reflux again for  10  - 15 minutes.)
          Cool sample and rinse  condenser  with two 10
          ml portions of water.

               It may be necessary here to dilute the
               sample to volume  and take an aliquot of
               sample before  proceeding.

               Proceed with F.W.Q.A.  sample treatment
2.   Modified P.W.Q.A. method:  Take  suitable  amounts
     of sample (not more  than  100 ml)  to provide  0.1
     - 1.5 yg Hg,  place in  a 150 ml beaker  containing
     7 ml 1:2 HNOg -HpSOjj plus  distilled water to make
     a final  volume of 100  ml.  Treat  each  according
     to type  of sample.

     a.    H20, C12 plant  effluent, NaOCl, samples
          Proceed  with modified P.W.Q.A. treatment as
          below (IIIA2b).

     b.    Modified F.W.Q.A. sample treatment

          Dilute sample aliquot to 100 ml with dis-
          tilled water.   Add 1  ml 5%  KMnO^  and let
          sample. stand for  at  least 15 minutes.   Add
          "2 ml 5%  K2S2C>8, allow sample to stand at
                           98

-------
               least  30 minutes  and proceed with aeration
               step as  below.

 B.  Aeration Procedure

     Connect  aeration apparatus  to spectrophotometers ;
     adjust spectrophotometer, flow meter,  etc., as  in
     After allowing apparatus to warm up,  adjust baseline
     and 100% absorption line with stopcock in bypass po-
     sition.   Proceed with aeration of sample, treating each
     sample individually as below.   Carry  out  each step with
     as little delay  as possible between steps:
     1.    Destroy excess  permanganate with 2 ml 10%
          HC1,  and immediately wash the clear sample  into
          gas washing bottle.

     2.    Add 5 ml 10% SnCl2  to gas washing bottle.  Im-
          mediately replace gas washing bottle in the aer-
          ation apparatus and turn stopcock to aeration
          position.

     3.    After pen has returned to within 2% absorption,
          turn  stopcock to bypass; rinse gas washing  bottle
          and proceed with next sample.

 C.  Calculations

     A series of 6 standards  ranging from 0.10 - 1.5  ug Hg
     is  treated as for H20 and C12 plant sewer samples and
     is  run each time the spectrophotometer is operated.
     Plot  a calibration graph on semi-log paper with  yg Hg
     on  the linear scale  and  percent absorption on the log
     scale. Convert percent  absorption of the sample to
     pg  Hg and  determine  Hg content as follows :

                    (yg Hg from graph) (dilution in ml)
          Hg, ppm = _

                    (g sample, note 1)

IV.  Notes

     1.    Assume specific gravity for volumetric samples
          to be 1.0 for dilute liquids, 1.5 for 50% caustic
          and 1.84 for sulfuric acid.

V.    Reagents

     1.    Nitric acid-sulfuric acid, 1:1 mixture.  Slowly
                             99

-------
          add 250 ml concentrated H2SOij to 250 ml HN03
          with constant stirring.  Allow to cool before
          using;  store in glass  container.   Caution:   Wear
          safety  glasses and gloves at  all times during
          preparation of acid solution.

     2.    Nitric  aeid-sulfuric acid, 1:2 mixture.  Follow
          procedure above using  150 ml  concentrated HNO^
          and 300 ml concentrated H2SC>4.

     3.    Potassium permanganate, 50 gpl.   Weigh 50 g re-
          agent grade KMnOjj  into a 150  ml tall-form beaker.
          Add approximately  70 ml distilled water and stir
          for about 20 seconds.   Allow  the KMnOjj crystals
          to settle, and decant  the supernatant liquid into
          a one liter volumetric flask.  Repeat the oper-
          ations  of dissolving and decanting until all the
          KMnOjj has dissolved.   Dilute  to volume, mix, and
          store in a brown bottle in a  dark place.

     4.    Potassium persulfate,  50 gpl.  Dissolve 20 g
                 in 400 ml distilled water.
     5.   Hydroxylamine hydrochloride ,  100  gpl.   Dissolve
          40 g NH2OH-HC1 in 400 ml distilled water.

     6.   Stannous chloride,  100 gpl.   Dissolve  20 g SnCl2-
          2H20 in 20 ml concentrated HC1 on the  hot  plate.
          Cool and add 180 ml distilled water.   Prepare
          weekly or more often as needed.   If solution
          becomes discolored, cloudy,  or turns  the sample
          solution cloudy upon addition (prior  to aeration),
          discard and prepare a fresh  solution.

     7.   Stock Hg solution,  1000 ppm.   Dissolve 0.6768 g
          mercuric chloride (HgCl2) in a 500 ml  volumetric
          flask.  Add 5 ml concentrated HNOo and dilute to
          mark with distilled water.

     8.   Working Hg standard, 10 ppm.   Dilute  5.0 ml 1000
          ppm Hg to 500 ml with distilled water  plus 5 ml
          HNOo.  Prepare bimonthly.

     9.   Working Hg standard, 0.5 ppm.  Dilute  25.0 ml
          10 ppm Hg to 500 ml with distilled water plus
          10 ml HNOo.  Prepare monthly.

VI.  References

     1.   William Horwitz, Ed. "Official Methods of Anal-


                              100

-------
     ysis  of  the  Association of Official Agriculture
     Chemists,"  9th edition, Association of Official
     Agriculture  Chemists,  Washington, D.C., I960,
     pp.  327-330.

2.    Federal  Water Quality  Administration, Provisional
     P.W.Q.A. Method for Hg Determination by Flameless
     AA,  1970.

3.    Dow  Chemical Company,  Determination of Mercury
     by Atomic Absorption Spectrophotometric Method,
     1970.
                          101

-------
                       APPENDIX C


              ALTERNATE Hg RECOVERY METHODS

Reduction Methods

          A method much discussed in the "literature, and in
commercial operation, is the reduction of the mercuric ion
to the metallic state followed by physical removal of the
Hg particle by filtration (2, 25, 3^, 42).  Diverse
materials may be used to perform this reduction but all
rely using a suitable reducing agent.  Some of the chemicals
proposed or used are:

          1.   Hydrazine hydrate

          2.   Aldehydes

          3.   Sodium borohydride

          4.   Sodium amalgam

          5.   Metals:  zinc, iron, bismuth, tin, nickel,
                    magnesium, manganese, copper,
                    aluminum, tin chloride.

          The Ventron process utilizes sodium borohydride
as the reducing agent.  This process was installed at the
Sobin Chlor-Alkali Plant in Orrington., Maine, and at the
Ventron Plant in Wood^-Ridge, New Jersey. A. 99-5/6 Hg removal
efficienty was reported.   In lab tests, we were not able to
achieve Hg removals as great (Table 19).  The differences
may be explained by varying conditions between our tests
and Ventronfs or by the difference between our waste water
and the Sobin waste water.

          In any case, the equipment required is similar
to that needed for sulfide precipitation: a pH adjustment
system, a reducing chemical addition, and a filtration
step with or without filter aid (Figure 10).

          The main advantage of the Ventron process is
that Hg can be recovered in the metallic state and reused
without further processing.  However, to achieve the
99>5% recovery as claimed, the reduction step must be
followed by a carbon bed and a resin bed for polishing.
In our laboratory tests, the reduction step using sodium
borohydride alone produced recoveries in the 95 - 98?
range (Figure 19).  With these efficiencies, the reduction


                             102

-------
and filtration process could be used alone in plants
producing 100 - 200 tons per day of chlorine but large
plants would have to add the polishing step (Table 27) .

          The cost of the sodium borohydride is about
$l6.50/kg ($7. 50/lb).  Excess addition of chemicals or con-
centrations of other ions which consume NaBH^ could create
high operating costs.  Theoretically, one kg of NaBHjj could
reduce up to 21 kg of Hg if no intefering substances are
present.   However, any oxidizing chemicals such as avail-
able chlorine or metal ions capable of being reduced would
consume
          Other reduction methods tried successfully have
involved a number of chemicals.  One of the most common is
zinc.  In work performed at Merck, Sharp and Dohme by
Rickard and Brookman (^0), a 99% Hg removal was reported
using a dosage level of 3.8 kg zinc per kg of Hg.  In
our laboratory work, we have achieved recoveries of 95 -
99.8/E using zinc particles in a column, followed by filtra^
tion.  To separate the Hg from the zinc, a distillation step
is required, in common with most other methods of Hg pre-
cipitation or adsorption.  An additional problem associated
with this method is residual dissolved zinc in the effluent,
ranging from a few to a few hundred ppm zinc depending on
the pH of the effluent.  The background level of zinc in
seawater is 0.01 ppm, and as with other heavy metals, bio-
logical concentration has been reported up to 1500 ppm.
Therefore, if the zinc process is to be used, some method
of zinc ion removal would be required.  Such a process
would add to the cost and complexity of the system;
therefore, no further studies are contemplated on zinc
treatment systems .

          Many other metals have been tried with results
similar to those reported for zinc, but the toxicity prob-
lem of dissolved metal ions is present to varying extents
for each alternative.  Of the least toxic metals tried,
such as magnesium and iron, the cost of the metal is high
or its effectiveness low  (Table 28).

          Laboratory studies are reported with other
reducing chemicals such as hydrazine hydrate, aldehydes
and others.  Although we have not studed these, we believe
the same problems and advantages hold as for sodium
borohydride.
                              103

-------
      Table 27.  Hg REMOVAL RATES NECESSARY FOR VARIOUS SIZE
                CHLORINE PLANTS TO ACHIEVE 45 gm (0. I Ib) PER
                        DAY MERCURY IN THE EFFLUENT
Plant size
Cl2/day
m ton
90
180
360
740
1450
s ton
100
200
400
800
1600
Estimated1 Hg
contaminated
water volume.
I/day gpd
75,000 20.000
150. 000 40. 000
300. 000 80. 000
600.000 160.000
1.200.000 320.000
Calculated2
final Hg
level.
ppb
600
300
150
75
38
Reduction^
through
treatment.
%
94.0
97.0
98.5
99.3
99.6
 Volume estimated on the basis of 75. 000 I/day (20, 000 gpd^ of
 contaminated waste water per 90 m ton (100 s ton) chlorine production
 per day.
2
 Maximum effluent concentration to achieve level of 45 gm/day (0. 1 Ib/day)
 Hg in effluent.
 3
 Assuming the starting Hg level in waste water was 10 ppm Hg.
                                  104

-------
   Table 28.  COMPARISON OF SUBSTANCES USED OR CONSIDERED
                 FOR REDUCING MERCURY ION IN SOLUTION
Compound
or metal
Sodium borohydride
Bismuth
Tin
Nickel
Hydrazine hydrate, 85%
Magnesium
Copper
Maganese
Aluminum
Zinc
Iron
4
Sodium sulfide
Effectiveness
High
?
Low
Low
High
High
Low
High
High
High
Low
High
Cost2
$/kg
16.50
19.80
9.35
3.56
1.50
.84
1.50
.84
.68
.77
.22
.15
$/lb.
7.50
9.00
4.25
1.62
.68
.38
.68
.38
.31
.35
.10
.07
3
Toxicity
potential
--
--
Medium
High
	
Low
High
Low
Medium
High
Low
High
  Based on Standard Oxidation Reduction  Potentials.

2 From"Chemical Marketing Reporter",  April 8,  1974;
  "Metals Week", April 29, 1974.

3 Subjective information from  Water Quality Criteria.  F. W. P. C. A.,
  U. S. Department  of the Interior, April 1968.

  Not a reducing agent.
                                  105

-------
          The use of sodium amalgam to reduce Hg in brine
or waste water has been tried by Karpink with limited
success (35, 36).  The 78% reduction reported is too low.
for this system.

Ion Exchange & Che'lating "Resins

          Another method for removing Hg from waste water
that appears frequently in the literature and has been
used in several plants in Japan is the use of ion exchange
or chelating resins (43 - 6l).

          The literature states that starting with Hg levels
in the 2 ^- 30 ppm range, after, one stage of resin treatment,
the effluent contains 0.1 - CL.5 ppm Hg.  With the addition
of a polishing resin step, the effluent can reach 0.001 to
0.020 ppm (Table 16.).  Similar results have been achieved
in our laboratory tests.  With a starting solution of
10 ppm, the effluents range in concentration from 0.3 -
1.8 ppm.  But when the starting solution is low in Hg,
less than 0.1 ppm, which simulates a polishing step, the
final Hg levels are 1 - U ppb (Figure 19 and Table 17).
Resins tested in our laboratory work were from the
Billingsfors-Langed and AJinomoto companies and were
specifically designed for Hg removal.  Of the two resins
tested, the Ajinomoto resin gave more consistent results.

Activated Carbon

          Another means of removing Hg from waste water
streams is to pass the water through a bed of activated
carbon to adsorb the Hg onto the carbon particles.  This
principle has been used extensively for. the removal of Hg
from caustic soda using a  finely divided carbon,
such as Nuchar KD Special, as a precoat on a pressure fil-
ter.  In this application, the Hg concentration is lowered
from 2000 ppb down to 100 ppb.

          Although the literature contains fewer references
to work with activated carbon than resin, the experience in
our laboratory indicates that activated carbon achieves
nearly the same Hg removal rates as ion exchange resins
(Figure 19).  Effluent levels of 100 - 300 ppb Hg were
achieved with starting Hg levels  >10 ppm.  However, with
starting Hg levels below 0.1 ppm, the effluent contained
5-7 ppb.  Of the 3 carbons tried, the Westvaco Nuchar 722
gave the lowest Hg levels in the effluent (Table 18).  The
bed capacities of the carbons were not determined.
                             106

-------
          As with, the ion exchange resin, there are several
problems which must be considered with such a system.  They
include: CD periodic regeneration or replacement of carbon,
(2) Hg recovery process from regenerant or spent carbon,
(3) prefiltration of the treated stream to minimize bed
plugging, and (4) determination of bed capacity and Hg
leakage point.

          Thus, the use of activated carbon for Hg re-
moval seems more appropriate as a secondary polishing
step rather than a primary process.
                             107

-------
                       APPENDIX D

                     START-UP MANUAL
SLUDGE TREATMENT SYSTEM
Start-up
          Note:  Start-up requires a controlled sequence
                 to have each piece of equipment ready
                 when needed.  It :takes 8 hours after the
                 starting the sludge pump to the thickener
                 before the rotary vacuum filter and.furnace
                 will have to handle product.   The furnace
                 takes 48 hours to preheat to operating
                 temperatures, so plan your time accord-
                 ingly.
I.   Sludge Dewatering (Assume brine clarlfier is in opera-
               tion. )
     A.   Brine Sludge Thickener
          1.   Close drain valve on thickener.
          2.   After determining there are no potential ob-
               structions, start the rake on the thickener.
          3.   Open manual valves before and after sludge
               pumps.
          4.   Start pump by adjusting air valve on pump.
               Adjust valve on oil reservoir.
     B.   Gas Cooler Condensers
          1.   Open all manual valves so the gas can
               pass through bodies 1, 2, and 3 from top to
               bottom.
          2.   Start cooling waiter to each body.
          3.   Start induced draft fan.
     C.   BSP Envirotech Furnace
          Upstairs Control Room
                            108

-------
1.   Turn on master control switch, at the remote
     station.

2.   Turn all burner control swi'tchs to "on"
     positions at the remote station.

3.   Manually adjust controller valves to "0"
     supply.

Downstairs at Furnace

4.   Start shaft rotation.

5.   Start shaft cooling fan.

6.   Start combustion air fan.

7.   Push reset switch, (indicating shaft rotation
     has been reset).

8.   Turn master gas control switch to "automatic"
     Purge timer light should come on; timer is
     set for 5 minutes.

9.   Reset low and high pressure gas meters.
     When purge complete light comes on, proceed
     to next step.

10.  Open manual gas valve.

11.  Start the burner on low fire on No. 6
     hearth.

12.  Adjust temperature controller to 400 - 500°F.
     When stabilized, adjust controller upwards
     slowly (about 100F° per hour) until No. 6
     hearth has a temperature of 10QO°F.

13.  Start burners on No. 5 hearth.

14.  Adjust temperature controller to 400 - 50Q°F.
     When stabilized, adjust controller upwards
     slowly (about .10OF0 per hour) until No. 5
     hearth has a temperature of 1QOO°F.

15.  Start main burners on No. 4 hearth.
                   109

-------
     16.   Adjust .temperature controller to 400 - 500°F.
          When stabilized., adjust controller upwards
          slowly (about 100F° per hour) until No. 4
          hearth has a temperature of 1000°P.

     17.   Start burners on No... 3 hearth.

     18.   Adjust temperature controller to 400 - 5000F,
          When stabilized, adjust icontroller upward
          slowly (about lOOF0 per hour) until No. 3
          hearth has a temperature of 1000°F.

     19.   Start increasing temperature on all four
          hearths at the rate of 5QF° per hour.
          Operating temperature is between 1400 -
          1500°F.

     Upon reaching operating temperature of 1400 -
     1500°P, prepare rotary vacuum filter for
     operation.

D.   Eimco Rotary Vacuum Filter

     1.   Close the filter vat drain valve.

     2.   Open the wash water line to the cloth and
          rolls.

     3.   Begin taking up the slack in the filter
          belt, being sure to adjust the ends of the
          takeup rolls equally.

     4.   Start the filter drum drive and completely
          soak the cloth, while retensioning the
          belt.  When the takeup roll is at normal
          operating position, stop the filter drive.

     5.   Start the vat agitator.  This should be op-
          erated at all times when the sludge is in
          the vat.

     6.   Turn on the seal water to. the vacuum pump
          and filtrate pump.

     7.   Start discharge conveyor from furnace.

     8.   Start furnace feed conveyor.

     9.   Turn on feed to filter.
                          110

-------
          10.   When sludge level in the vat reaches 30 #
               full, start the vacuum pump, filtrate pump,
               and filter drive.

          11.   Adjust the filter drum speed and rate of
               feed as may be necessary for cake thick-
               ness, cake dryness, and removal from cloth.

          It may take *IO - 60 minutes before sludge begins
          coming out the discharge conveyor.  Check clinker
          for dryne&s and plugging of furnace.
SHUTDOWN

I.   Sludge Dewatering

     A.   Brine Sludge Thickener

          1.   Shut manual valve from bottom of clarifier
               before pump.  Flush fresh water through
               pump inlet and outlet to thickener.

          2.   After clear water appears at thickener.
               shut off sludge pump and drain water from
               line.

          3.   Continue to dewater sludge until consistency
               is too low for good filter performance.   Di-
               vert the rest of the sludge to pond by shut-
               ting off filter feed and opening line to pond,

     B.   Eimco Rotary Vacuum Filter

          1.   Fully open all wash water to the filter belt.

          2.   Open vat drain valve.  Flow will divert to
               pond.

          3.   Stop filtrate pump and vacuum pump, and shut
               off seal water.

          4.   Rotate drum drive at least 5 revolutions
               until filter cloth is clean.  Then stop
               filter drive and shut, off wash water.
               Note:  Never leave a cloth to dry unless it
               is washed thoroughly.

          5.   Release the tension on the filter cloth by
                             111

-------
     turning the takeup roll cranks equally.
     Note the numfrer of turns of the adjusting'
     cranks so that the takeup roll can be re-
     turned to its. original position at start-
     up.

6.   Plush the wash trough with a small amount of
     fresh water.  If shutdown will be longer
     than 8 hours, flush the filtrate tank and
     wash down the filter.

     Note:  It will take 2 hours after the belt
     conveyor has delivered the last little
     bit of solids into the furnace, before
     the last clinker is discharged by the
     clinker conveyor.  When the clinker
     conveyor is empty, begin shutting down
     the furnace.

BSP Envirotech Furnace

1.   Start reducing the temperatures on all four
     hearths at the rate of 50°P per hour.  When
     temperatures have stabilized at 900 - 1000°F
     continue to next step.

2.   Adjust the temperature controller on No. 3
     hearth so the temperature drops at the
     rate of 100°P per hour.  When temperature
     stabilizes at 400 - 500°F turn off burner
     on No.  3 hearth.

3.   Adjust the temperature controller on No. 4
     hearth so the temperature drops at
     the rate of 100°P per hour.   When tempera-
     ture stabilizes at ^00 - 500°F turn off
     burner on No. 4 hearth.

4.   Adjust the temperature controller on No. 5
     hearth so the temperature drops at
     the rate of 100°F per hour.   When tempera-
     ture stabilizes at 400 - 50Q°F turn off
     burner on No.5 hearth.

5.   Adjust the temperature controller on No. 6
     hearth so the temperature drops at
     the rate of 1QO°F per hour.   When tempera-
     ture stabilizes at ^00. - 500°F turn off
     burner on No. 6 hearth.
                   112

-------
               Furnace temperature should be between 300 -
               liOO0?.  A decision at that time will be made
               whether to shut off the pilot or not.  It
               is advantageous to keep the furnace at this
               temperature if .possible.

     D.   Gas Cooler Condenser

          As long as furnace is running on pilots or burn-
          ers, the cooler condensers will remain in oper-
          ation.
WATER TREATMENT

Start-up

I.   Water Treatment

     A.   Acid Mix Tank

          1.   Open valve on the inlet to Hg waste water
               storage tank.

          2.   Open recycle valves to pH mix tank.

          3.   Open valve from spent acid stream.

          4.   Adjust level controller.

          5.   Open valves on inlet and outlet of the pond
               pump to mix tank.

          6.   Start flow from pond.

          7.   Adjust pH controller and start acid flow.
               Liquid in tank will continue to recycle
               until appropriate level is reached.

          8.   Start the agitator in the Hg waste water
               storage tank when the level is above the
               agitator.

     B.   Na2S Storage Tank

          1.   Open valves on the outlet of the
               storage tank.

          2.   Adjust rotometer to required gph.
                             113

-------
     C.   R. P.  Adams Pressure Filter

          1.   Open accept valve to sewer.

          2.   Open valve for filter feed.

          3.   Make sure bottom valve to backwash, tank is
               closed.

     D.   Filter Aid and Precoat Mix Tank

          1.   Open valves on outlet and inlet of feed
               pump to filter.

          2.   Adjust level controller in mix tank.

          3.   Adjust BIF feeder to the rate of 1 oz/100 gal.

          4.   Adjust flow indicator from Hg waste water
               storage to mix tank.  Remember, the flow
               should agree with the setting on the Na2S
               addition system.

          5.   Start flow from Hg waste water storage tank.

          6.   Adjust Na2$ flow on rotometer.

          7.   Start BIF feeder.  initial start-up requires
               50 Ib. of diatomaceous earth.

          8.   Start agitator.

          Note:   When level has reached the controller set
          point, it will begin feeding filter at the rate
          you set at the flow indicator from the Hg waste
          water storage tank.


WATER TREATMENT

Shut down

I.   Water Treatment

     A.   Acid Mix Tank

          1.   Shut off pond pump to mix tank.

          2.   Shut off other flows to mix tank.
                             114

-------
     3.   Shut off acid flow.

     Note:  If shutdown is only temporary, you can
     pump for a short time to: the waste water storage
     tank. Be sure to treat with acid first.

B.   Waste ¥ater Storage Tank

     1.   When level is- bellow agitator, shut off
          agitator.

     2.   Close valve out. of tank.

     Note.:  If waste water is to be stored in tank,
     leave on agitator.

C.   Filter Aid and Freeoat Mix Tank

     1.   Turn off BIP feeder.

     2.   Turn off rotometer from Na2S storage tank.

     3.   Adjust level controller such that you can
          pump the remainder of tank through the fil-
          ter.  Then turn off pump.

     4.   Add fresh water and flush lines and pumps.

D.   R. P. Adams Pressure Filter

     1.   Main objective now is to backwash filter:

          a.   Close the feed valve from precoat mix
               tank.

          b.   Open dump valve to backwash tank. Fil-
               ter will drain to decant tank and ma-
               jority of precoat should fall off by
               reverse flow.

          c.   Refill tank 3/4 full with fresh water.
               Turn off water flow.

     2.   Pressurize filter with about 50 psi, then
          turn off air.

     3.   Open backwash valve quickly.  The compressed
          air head should push the liquid in the
          reverse direction, thoroughly purging the
          filter tubes of any remaining cake.
                        115

-------
Note:  The receiver dump will be full of water
and filter aid.  Decant off water and open dump
receiver to filter.  Time required will be de-
termined after initial sfrart^up.
                       116

-------
OPERATING NOTES

I.   Sludge Dewatering

     A.   Thickener

          1.   A low level alarm on the thickener may in-
               dicate an Insufficient pumping rate.  If
               problem cannot be quickly- corrected, notify
               the Tour Foreman.  Note time period of
               trouble in log book.

          2.   There is a screen on top of the thickener
               to prevent larger size particles from enter-
               ing the thickener.  This should be cleared
               once a shift.

     B.   Rotary Vacuum Filter

          1.   In general, vacuum will be kept at a maximum
               and not varied in order to achieve maximum
               dryness.

          2.   Vat level and drum speed determine cake thick-
               ness and production rate.  Level will gener-
               ally- be kept cons-tant and speed varied.
               High speeds will tend to decrease dryness.

          3.   If filter will not pick up cake:

               a.   vat consistency may be too high

               b.   vat level may- be too low

               c.   vacuum may be too low

               d.   the cloth may not be getting cleaned
                    properly

          4.   In order to obtain dryer cake:

               a.   slow down the filter

               b.   decrease cake thickness  (lower at  vat
                    level)

               c.   increase vacuum
                             117

-------
     5.   The vacuum pump can be severely damaged if
          it is operated" without .seal water.  Thus,
          the seal water should be adjusted or checked
          every 4-6 hours.

     6.   If the filtrate is not removed from the re-
          ceiver, it will carry over to the vacuum
          pump.  Check to. see that the valve on the
          filtrate pump is wide open at all times.

     7-   Check cloth appearance frequently.  Improper
          slack in cloth or misalignment can cause lack
          of vacuum or tearing of cloth.  It is very
          important to correct these problems immedi-
          ately.

C.   BSP Envirotech Furnace

     1.   Monitor temperatures on all four hearths.
          It is important that we maintain an oper-
          ating range of 1*100 - 1500°F.  Problems of
          insufficient temperature could be:

          a.   combustion air fan

          b.   improper gas to air ratio

     2.   Check shaft cooling fan regularly.  This
          is vital for good operation of furnace.

     3.   It is important that we maintain an even
          flow to and from the furnace.  If feed rate
          is too fast, plugging of upper hearth can
          and will be a problem.  If this does hap-
          pen, discontinue feed to furnace until in-
          cinerated solids conveyor is empty - pos-
          sibly 4 hours - then continue operation.

D.   Cooler Condensers

     1.   It is important that we receive the maximum
          amount of cooling from each condenser.  If
          scale begins building up within the condens-
          ers, poor heat transfer will result in higher
          air temperature at the induced draft fan.
          These  temperatures will be monitored each
          shift until a temperature range for opera-
          tion is established.
                          118

-------
          2.   Check Induced draft fan frequently.  Note
               excessive vibrations, or other problems which.
               might develop.

II.  Liquid Treatment

     A.   pH Adjustment

          1.   If pH becomes a problem, check the following:

               a.   acid feed pump

               b.   acid fitters

               c.   automatic control valve

               d.   make sure agitator is operating

          Note:  pH should be maintained between 5-8.
          As you approach the higher pH's, Hg becomes more
          soluble and tends to pass through the filter
          media more easily.  Also, lack of pH control tends
          to disrupt sewer pH.

     B.   Sulfide Precipitation

          1.   Excess sulfide is needed to precipitate Hg
               in the mix tank.  Therefore, It is important
               that we maintain a proper flow and the right
               concentration of Na?S to the mix tank.  Check
               these often.

     C.   R. P. Adams Filter

          1.   It is important that the tubes within the
               filter are properly precoated.  Improper
               precoating can cause tubes to plug and
               eventually break when backwashing.
                             119

-------
SELECTED WATER
RESOURCES ABSTRACTS

INPUT TRANSACTION FORM
                                              1. Report No.
                                               2.
                                                        w
 4, Title
         MERCURY RECOVERY FROM CONTAMINATED WASTE
         WATER AND SLUDGES
 7. Author(s)
         Richard Perry
         Georgia-Pacific Corporation, Bellingham Division
                                                         5. Report Bat*,

                                                         6.
                                                         8 Performing Organization
                                                           Ri-"cr! Ma.
                                                                      12040 HDU
12,
                                                                       v p? of Report and
                                                                      ;',:dod Covered
organization    EPA, Water Quality Office
        _.___._______:--^—. was Designed, built and operated to  remove Hg from
waste water and sludge produced by a mercury cell  chlor-alkali plant.  Mercury content
of  the waste water ranged from 300 - 18,000 ppb mercury while mercury content of the brii
               fr-nm 150 to 15QQ ppm Hg.  Other sludges  processed include sludges from
              near our plant outfall with a Hg content  of 10 - 25 ppm Hg.
      From a variety of removal techniques tried in the lab, the methods selected were
sulfide  precipitation for the water treatment and  high temperature roasting for the sludj
treatment.  The sulfide precipitation consists of  collecting the various water streams,
adjusting the pH from 5 - 8 with spent sulfuric acid,  settling the large solid particles
in  a surge tank, adding sodium sulfide to a 1-3 ppm excess, adding diatomaceous earth at
the rate of 0.07 gpl in an R. P. Adams pressure filter.  The effluent Hg levels range
from 10-125 ppb with an avesage of 50 ppb Hg for an 87-99% removal, averaging 96.8%.
The 4.8m3 filter handles 280-280 liters/min adequately with an approximate 48 hour cycle
time between backwashings.  Capital costs were $143,900 and operating costs were 50C/378.
      The sludge system contains a collection system, 3.7 m diameter thickener, 1.8 m
diameter rotary vacuum filter, 1.37 m i.d. multiple hearth furnace, and 3 stainless stee
condensers 21 m2 each.  Processing rate for the sludge is 140-320 kg/hr, dry basis.
At  present we are processing approximately 18 m tons of sludge per month for our Chlor-
Alkali Plant.  Operating temperatures ranged from  540 C - 760 C, feed Hg content ranged
from 290 to 440 ppm Hg (dry basis), and clinker Hg content after treatment contained
0.5 - 7.2 ppm Hg, for a removal rate of 98.3 to 99.8%.  Waterway sediments containing
12.8 ppm were roasted at 730 C and clinker contained 0.95 - 1.7 ppm Hg for an 87-92%
removal.  Capital costs were $364,500 and operating costs were $32 per m ton of dry
 ol 11/I c
 Ja. Descriptors  Wflter Pollutionj Metals,  Sludge Disposal, Waste Water Treatment,
 Sulfides, Chemical Precipitation, Hydrogen Sulfide, Filtration, Electrolysis,
 Activated Carbon, Ion Exchange, Reduction (chemical) Particle Size, Oxidation,
 Ghlorination
 i7b. Identilfer*
   Mercury, Chlor-alkali Cells,  Separation,  Recovering, Thickening, Clarificati
 Roasting, Sodium Hypochlorite, Methyl Mercury, Sodium Borohydride, Vacuum Filtration
'$,  S?*UTJVy Class,
                           . Security Class
                            (Page)
                                           21. No', of
                                             Pages

                                           22. Price
                                            Send To:

                                           WATER RESOURCES SCIENTIFIC INFORMATION CCNTKR
                                           U.S. DEPARTMENT Of THE INTERIOR
                                           WASHINGTON. OjC. 20*40
                         U. S. GOVERNMENT PRINTING OFFICE: 1974-697-649 /6I  REGION 10

-------