Removal of silica from Raft River geothermal

Removal of silica from Raft River geothermal

DISCLAIMER
This report was prepared as an account of work sponsored by an
agency of the United States Government. Neither the United States
Government nor any agency Thereof, nor any of their employees,
makes any warranty, express or implied, or assumes any legal
liability or responsibility for the accuracy, completeness, or
usefulness of any information, apparatus, product, or process
disclosed, or represents that its use would not infringe privately
owned rights. Reference herein to any specific commercial product,
process, or service by trade name, trademark, manufacturer, or
otherwise does not necessarily constitute or imply its endorsement,
recommendation, or favoring by the United States Government or any
agency thereof. The views and opinions of authors expressed herein
do not necessarily state or reflect those of the United States
Government or any agency thereof.
DISCLAIMER
Portions of this document may be illegible in
electronic image products. Images are produced
from the best available original document.
EGG-FM-5170
REMOVAL OF SILICA FROM
RAFT RIVER GEOTHERMAL WATER
D. F. Suciu
R. L. M i l l e r
Published June 1980
EG&G Idaho, Inc.
Idaho Falls, Idaho 83415
Under
I
Prepared f o r the
U, S. Department o f Energy
Idaho Operations O f f i c e
DOE Contract No. DE-AC07-76ID01570
,
.
ACKWWLEDGEMENTS
P
Special thanks. are extended t o Richard N. Wallace District
Representative for Betz Laboratories, Inc.; A. 6. Mindler; Research and
Development Division, P e m t i t Corporation; D. A. Conley , Materials
Science Branch, for h i s contributions t o the Raft River Geothermal Test
Project ; and INEL Library personnel.
6
.
ii
ABSTRACT
k c k of sufficient quantities of clean surface or near-surface water
?
i
a t Raft River for cooling purposes dictates that cooled geothermal f l u i d ,
effluent f r o m the' Raft River 5 MW(e) P i l o t Power Plant, must also be used
as condenser coolant. Prior testing revealed that a water-treatment system would be required t o reduce silica and calcium concentrations of the
cooling f l u i d . The water-treatment system specified by the Department of
:'G~ergy*SArchitect-Engineer was t o use dol t i c lime for both pH adjustment and source of magnesium. The dolomitic lime treatment was investigated and found t o be inadequate. Subsequent testing was done t o f i n d
chemical systems that. would adequately reduce silica concentrations:
Three magnesium and two iron compounds were found which reduced silica t o
acceptable concentration levels. They are magnesium bicarbonate, magneium chloride, magnesium sulfate, Iron sulfate, and iron chl
Magnesium oxide, using a two-sta
rent process, w i l l also reduce silica t o adequate levels.
-. ,
.
P
r
iii
,
a
CONTENTS
......................................................
ABSTRACT ..............................................................
UCTION .....................................................
1. I
ACXNOWLEffiEMfiNTS
ii
\
2.
LITERATURE
S i l i c a Reduction Mechanisms.
pH E f f e c t
E f f e c t o f Magnesium Type
2.3.1 Dolomitic Lime
2.3.2 Mndissolved Magnesium Compounds
2.3.3 Dissolved Magnesium Compounds (Salts)
.I..
3.
4.
5.
6.
7.
8.
9.
10.
11.
1
4
~C\RCH.L......................r.....oooo.o..oo.o.o.
. ..............................
...................................................
....................................
.......................................
.....................
................
DESIGN SeECIFICATIONS ...........................................
EXPERIMENTAL PROCEDURE ...........................................
s ...................................................
Analysis ...........................
..............
. .
EVALUATION OF SYSTEM SPECIFIED BY THE ARCHITECT-ENGINEER.
5.1 Dolomitic Lime ..............................................
5.2 E f f e c t o f pH and Magnesium ..................................
5.3 Temperature E f f e c t ..........................................
EVALUATION OF WGNESIUM OXIDE SYSTEMS ............................
6.1 E f f e c t o f Grade o f Magnesium Oxide ..........................
6.2 Temperature E f f e c t Evaluation Using Remosil..................
6.3 Dolomitic Lime P l u s Remosil .................................
6.4 High Concentrations o f Magnesium Oxide ......................
6.5 Preliminary Conclusions o f Magnesium Oxide Systems ..........
TWO-STAGE COUNTERCURRENT PROCESS UNIT ............................
SOLUBLE MAGNESIUM COM'Ou\ID SYSTEMS ...............................
8.1 Magnesium Bicarbonate .......................................
8.2 Magnesium Sulfate ...........................................
8.3 Magnesium Chloride ..........................................
PREACIDIFICATION .................................................
IRON SYSTEMS .....................................................
DISCUSSION OF RESULTS ............................................
2.1
2.2
2.3
iii
iv
CI
4
6
6
6
7
7
9
10
10
11
13
13
13
14
16
16
16
17
18
18
19
20
20
21
24
26
27
28
it
12.
CONCLUSIONS AND RECOWNDATIONS
13.
REFERENCES
APPENDIX A
...................................
30
.........................................e..............
- ANALYSIS OF WATER TREATED WITH VARIOUS SYSTEMS ............
32
37
FIGURES
1.
2.
3.
4.
5
.
6.
.
Log percent transmittance versus concentration of silica .........
Water treated with lime a t
.&... ... ........ .... ........ .
Volume of make-up required a t various cycles of
ncentration
60°C
Concentration of residual silica versus concentration of
magnesium added as (a) Epsom s a l t , (b) Epsom s a l t p l u s
soda ash, and (c) Epsom s a l t plus sodium bicarbonate
Concentration,of residual silica versus concentration of
(a) magnesium oxide plus sulfuric acid and (b) magnesium
oxide plus soda ash and sulfuric acid..............................
3
12
14
\
22
,
23
...... .........
Concentration of residual silica versus concentration of
magqesium oxide p l u s hydrochloric acid a t fixed pH
25
f
TABLES
P
Make-up and Blowdown Rates for Different Cycles
of Concentration
A-1.
................................................
Analysis of Filtered Samples o f Magnesium Hydroxide
Slurries After Bubbling Carbon Dioxide to Constant pH ...........
Water Treated w i t h Dolomitic Lime a t 6OoC .......................
A-2.
Water Treated with Dolomitic Lime a t Four Temperatures
A-3.
Water Treated with Different Grades o f Magnesium Oxide
a t 60°C .............................o.......b.......~...o.......
41
1.
2.
A-4.
A-5.
................
Water Treated with Dolomitic Lime and Remosil a t Constant
Magnesium Dosage a t
.......................................
Water Treated with Remosil a t
....... .....................
Water Treated with Remosil t o Simulate a Two-Stage’
Countercurrent Process .........................................
Water Treated with Remosil a t Four Temperatures
60OC
A-6.
A-7.
..........
60°C
2
20
39
40
42
43
43
44
A43
.
.
Water Treated with Carbonated Solutions o f Magnesium Oxide
a t 6OoC
...............................................
45
Water Treated with Mac$esium Sulfate (Epsom S a l t ) a t 6OoC
46
.......
A1
.0 . Water Treated with Magnesium Sulfate and Soda Ash a t 60% .......
A.ll . Water Treated
Equimolar Quantities o f Magnesium Sulfate
and Sodium Bicarbonate a t 6OoC ..................................
A-12 . Water Treated w i t h Magnesium Sulfate a t 6OoC ..... .............
9
A.
46
\
with
A-13
.
A-14
.
A-15
.
A1
.8
.
.
.
A1.9
.
A-16
A1
.7
........
Water Treated with Magnesium Sulfate-and Soda Ash a t 6OoC
.......
Water Treated with Magnesium Chloride a t 60OC and pH 11.2 ......
Water Treated w i t h Magnesium Oxide Following pH Ad&
a t 6OoC .........................................................
Water Treated w i t h ferrous Sulfate a t 60°C and Varying pH .......
Water Treated with F e r r i c Chloride a t 6OoC ......................
Tabulation o f Magnesium Systems a t 60°C .........................
Water Treated with Magnesium Chloride a t ‘60°C and
vi
0.2.
47
47
1
4
48
48
49
50
51
51
c
52
REMOVAL OF SILICA FROM RAFT RIVER GEOTHERMAL WATER
'
1. INlRODUCTION
.
Ih
rc
I
?
The 5 MW(e) P i l o t Power P l a n t a t Raft River is O f conventional design
w i t h cooling provided by evaporation i n a cross-flow cooling tower. This
plant differs from other conventional power plants b y ' i t s use of geothermal water as cooling fluid because of restrictions -on the consunptive use of cold aquifer or clean surfaGe waters i n the Raft River Valley.
Raft River geothermal water has a high concentration of silica, a
major scale former, and its use requires means t o l i m i t silica scaling.
Standard procedures nominally employ a process called blowdown, by which
build-up of s i l i c
other detrimental substances can be regulated by
partial removal and dilution. Also, because cooling is by evaporation,
water loss must be supplemented by additional water which, i n the present
system, would be geothermal water with its attendant high silica
ns. Additional losses occur from wind blowing a portion of
the water mist out o f the cooling tower. In steady-state operation, the
amount of geothermal ter added, termed make-up, is the sum of blowdown,
evaporation, and win
Thus, t o use geothermal water as coolant for the 5 MW(e) P i l o t Power
Plant w i l l require means t o effectively control silica levels within both
coolant and make-up water.
1.
T
The ratio of silica concentration i n the recirculating coolant t o its
concentration i n the make-up is termed 'Icycles of .concentration." This
ratio w i l l always be greatdr than unity'owing t o the build-up of silica i n
the recirculating f l u i d fr6m evaporation and s o l u b i l i t y ch
cooling.
A communication (April 20, 1977) from E3etz Laboratories, Inc., 1
stated that it 'was not possible to control s i l i c a scaling by use of dispersants alone and that the best way to handle t h e problem is t o keep-
i
silica a t low concentrations (<lo0 ppm). Because s i l i c a concentrations i n
Raft River geothermal water range from 160 t o 200 ppm, use of geothermal
water requires that s i l i c a concentration be reduced. If the silica
content after i n i t i a l reduction is sufficiently low, some of the costs of
appcoach can be tempered by use of high cycles of concentration which
t i v e l y reduce quantities of make-up required. The volume o f make-up
required as a function of cycles of concentration has been calculated by
Nguyen;2 h i s data are shown i n Table 1 and Figure 1
.
I
.
*
TABLE 1. MAKE-UP AND BLOWDOWN RATES FOR DIFFERENT CYCLES OF CXINCENTRATION
Number
of c yclesa
1
2
3
4
5
6
7
8
9
10
a.
L/s
38.82
29.11
25.87
24.26
23.28
22.64
22.83
21.83
21.56
Blowdown rate
Make-up
E
L/s
E
616.00
462.00
410.67
385.00
369.60
359.33
352.00
346.50
342.22
18.43
8.73
5.50
3.88
2.91
2.26
1.80
1.46
1.19
292.60
138-60
187.27
61.60
46 20
35.93
28.60
23.10
18 82
.
.
Temperature change is 36 K (20OF).
Conventional water-treatment systems use magnesium oxide t o remove
silica, and lime t o control pH. Dolomitic lime ik frequently used because
it provides both MgO and Ca(OH)2. This method, using a warm-lime softeni n g u n i t , was proposed by t h e Architect-Engineer for t h e Raft River plant.
This report addresses t h e problem of removing silica from-geothermal
water by using t h e system described above. The Architect-Engineer’s recommendation was tested and results conclusively demonstrated it t o be totally
2
.
8
a
Cycles of concentration
INEL-A-15 049
&
Figure 1. Volume o f make-up required a t various cycles o f concentration.
*
i
inadequate. lhereafter, other sorption-based schemes were investigated t o
determine effective silica removal techniques, especially those which could
reduce silica levels t o 10-15 ppm and thereby permit five t o ten cycles o f
concentration t o be used.
15
r
_I
2.
LITERATURE SEARCH
Both Chemical Abstracts and the Engineering Index were researched t o
gather information on the removal of silica from water. Owing t o high
silica levels and the chemical composition of geothermal water, only a few
of the articles had direct application t o the Raft River chemical pretreatment program. Thus, the Literature provided general guidelines only
for interpreting data and establishing test parameters.
1
* . -
2.1
1
*
Silica Reduction bkchanisms
Silica is removed by s o r p t i o n onto magnesium hydroxide particles as
they precipitate from solution. The silica sorption reaction is complex;
the exact mechanism is not known. The reaction varies w i t h pH, temperature, and form or composition o f magnesium compound; however, no s i l i c a
reduction occurs below the pH a t which magnesium hydroxide normally pre-'
cipitates. 3
According t o Krauskopf ,4 the s i l i c i c acid monomer (H4Si04) is
the dominant species i n solution over a pH range o f 0-9 and has a solub i l i t y of 100-140 ppm a t 298 K. Above pH 9, the solubility-increases as
the s i l i c i c acid dissociates according t o the following equilibrium react i o n s a t 298 K.
= %SiOi* + H+, pK2 = 11.7 (Ref. 7)
H2SiOi2 = HSiOi3
H S ~ O - ~= Si04
-4
4
By
.
+ H+, pK1 = 9.7 (Ref. 5)
H4Si04 =
H3SiOi1
1
+
+ H+,
pK3
= 12.0 (Ref. 7)
H+, pK4 = 12.0 (Ref. 7)
convention the equilibrium constant is
4
c
I
=
concentration of products
concentration of reactants
and the pK term is defined as
For weak acids, at the point where the molar concentrations of the
ionic species become equal, the pK and pH values coincide, For
example, using the equation above, w e have
. .
which reduces to
at pH = 9.7, since IHISiOi2] and [H,SiO;;l]
become equal.
Thus,
1 = IO
pK = loglo K
The temperature
determined by Rythenko8 to be
~
I
PK1 = 3405*9
T
Y"2
=
*'
T
#
- 6.368 + 0,0163436T
- 33.000
+ 0.049580T
5
for T varying from
o
to 2 5 0 ~ ~ .
The first ionized form of s i l i c i c acid (%SiO;.')
silica reduction.
2.2
is the! active species i n
pH Effect
Any increase i n hydroxide alkalinity above pH 9 w i l l increase the
concentration of ionized forms o f s i l i c i c acid. Therefore, increasing pH
should increase s i l i c a reduction up t o some optimum pH value. Betz e t
al.9 treated water samples having silica contents varying from 7.1 t o
26.2 ppm with magnesium oxide. Their tests covered a pH range o f 9.5 t o
10.6. They concluded that the optimum pH for silica reduction is iO.1.
They also stated that as residual silica i n the treated water decreases,
control of pH becomes less important.
Wey and Siffert'' treated solutions containing 140 ppm silica w i t h
equivalent amounts of magnesium chloride and determined that the maximum
precipitation of s i l i c a occurred i n the pH range 11.0-11.5.
Wjeriego6 treated water containing 100 and 140 ppm silica with
several metals, including magnesium, and concluded that reaction pH i s the
single most important parameter governing silica removal by sorption on a l l
of the hydrous metal oxides studied. He states that i n the temperature
range from 303 t o 313 K (86-104*F), the optimum silica removal occurs a t a
pH given by the expression
2.3
2.3.1
Effect o f Magnesium Type
Dolomitic Lime
The use of dolomitic lime for removal of s i l i c a has several advantages. It is inexpensive and provides both magnesium (37 weight percent
6
MgO) for s i l i c a removal and lime (58 weight percent @(OH),)
for pH
i
f.
Behrman and Gustafson'l treated identical water samples containing
17 p p m silica as follows: one sample was treated with a high calcium
lime, the second with dolomitic lime i n the amount required t o produce the
same quantity of calcium hydroxide as i n the first sample. They reported
that t h e concentration of residual silica i n both samples was 10 ppm and
concluded that (a) dolomitic lime is ,not effective i n reducing silica and
(b) magnesium hydroxide t o be really effective must be formed i n s i t u .
--
Ektz e t a l O 9indicated that dolomitic lime has merit for removal of
silica under controlled conditions, although the process alters the water
chemistry i n such a way as t o make the process impractical.
'
f
Crossan'' treated Mississippi River water with dolomitic lime and
found that dolomitic lime (on the basis of equivalent magnesium) was a t
least one-half a s effective as ionic magnesium. For water samples with
silica content over 4.5 ppm, even better results were obtained.
2.3.2
Nondissolved Magnesium Compounds
I
There are a number of commercially available sources of magnesium.
They include soluble s a l t s of magnesium, such as magnesium chloride, and
insoluble magnesium compounds, such a s dolomitic lime, magnesium carbonate, and various grades of magnesium oxide. The difference between t h e
two forms is that when a soluble magnesium s a l t is used, the magnesium
hydroxide must be generated by a pH adjustment.
r,
2.3.3
Dissolved Magnesium Compounds (Salts)
The concept of generating the magnesium hydroxide
i n s i t u for in--
creased effectiveness i n silica reduction has sound basis i n the literature.
There is, however, some disagreement as t o the actual percentage increase i n silica removal attributable t o soluble magnesium salts
over t h a t o f insoluble magnesium compounds. The previously noted literature generally concurs with Wohlberg and Buchholz, 13 It... silica i n water is a metastable system whose behavior is d i f f i c u l t t o predict."
8
8
3.
c
*
I
DESIGN SPECIFICATIONS
Facility Systems Engineering Corp
s Angeles was selected
a s the Architect-Engineer for the 5 MW(e) Pilot Po
recommendation^ f their consultant, Garrett Energy
Development, Inc )' specified that a conventional warm-lime softener u n i t
would be adequate for water-treatment requirements.. A Cochrane Warm
Process
. Water Softener Design WW311or equivalent, o f 21.8 L/s (346 gpm)
capacity w i l l be the system used for -the water pretreatment operation.
.
Chemical reagents specified include dolomitic lime for both pH adjustment and silica-%removal
and soda ash for caicium reduction. Design
specifications for inlet and outlet water quality are:
Inlet Water
Total silica, as S i
Outlet Water
10-15
P
Total calcium, as CaCO3 (ppm)
93.9.
342.8
Total bicarbonate, as CaCO3 (ppm)
49.8
0
0
156.9
Total (OH) alkalinit
4.
.
EXPERIMENTAL PROCEDURE
e principal method used to study silica removal was to mix cooled
geothermal f l u i d with chemicals used i n the treatment test i n amounts.and
sequence required t o form the magnesium silicate-magnesium oxide-calcium
rry. Flocculation of precipitated materials was used t o
.
simulate typical industrial practice. The apparatus and technique used
are described below.
I
+
The treated and clarified water fraction of the s l u r r y was analyzed
t o determine changes i n chemical composition. The chemical components
) determined were silica, pH, calcium concentrations, and alkaMethods used are described below.
4.1
Jar Tests
Tests were conducted i n accordance w i t h ASTM 0-2035-74
(Coagulation-Flocculation Jar Test of Water) ,14 using a Phipps and Bird
Model 7790-300 six-paddle stirrer. Test apparatus consisted of a constant
temperature bath, s i x 2-litre beakers, and the Phipps and Bird test
stirrer. The beakers, each containing 1.8 litres of geothermal water,
were placed i n the constant temperature bath and were allowed t o come to
thermal equilibrium. Chemical reagents, as solutions or slurries, were
added t o each beaker i n the quantities and sequence required for each
test. Deionized water was added t o b r i n g the total volume i n each beaker
t o two litres. Each system was flash-mixed for one minute by using a
paddle velocity of 120 rpm, with velocities then reduced t o 20 rpm or just
rapid enough t o keep the resulting floccules i n suspension and stirred for
fifteen minutes. The beakers were removed from the constant temperature
bath and the water samples filtered. The water samples were chemically
analyzed for silica, calcium, pH, and alkalinity. Analyses were performed, according t o procedures recommended by the Had7 Company15 for
water analysis, using reagents that are marketed commercially.
10
6
*
4.2
Chemical Analvsis
Silica concentrations were determined as silicomolybdates, using a
Beckman Model B spectrophotometer. A calibration curve (Figure 2) was
generated by using standard solutions of 1 and 10 ppm silica. A sample of
untreated water diluted with distilled water, i n the ratio o f 1 t o 25,
gives a percent transmittance between the two standards. The i n i t i a l
silica concentration was then determined and used i n subsequent dilutions
(2:25, 5:25, etc.). A plot of the l a g s f percent transmittance versus
concentration was then used t o determine the silica concentration.
Calcium was determined by titrating ~O-CII? samples of treated water
w i t h a standard solution of 0.02 N EDTAOa A l k a l i n i t y was determined by
titrating SO-cn? samples of treated water with 0.02 N sulfuric acid. :The
latter titration was performed by using an automatic t i t r a t e r t o a pH of
4.8. Calcium concentrations are reported as ppm Cam3; alkalinities are
reported as ppm CaC03.
a. EDTA = (Ethylenedinitri1o)tetraacetic acid or
ethylenediaminetetraacetic acid.
\
I
11
..
.~
.
.
.. ..
._.
..
-
.. .
.
-
.
. . ..
.
.-.
..
.
. .
.
.
._
a
1 ,
I
1
I
I
0)
0
J
2
t
I
I
10
20
I
30
Concentration of silica (ppm)
I
40
I
INEL-A-15 05C
Figure 2. Log percent transmittance versus concentration o f s i l i c a .
12
5.
EVALUATION OF SYSTEM SPECIFIED BY- THE &CHITECT-ENGINEER
As noted i n Section 3, FSEC specified a dolomitic lime treatment
l
c
system using a conventional warm-lime softener u n i t for the silica removal
operation. This section addresses test results o f the above-mentioned
system.
5.1 Dolomitic Lime
I
i
0
h
4
-
The use o f dolomitic lime for silica reduction has the d i s t i n c t advantage o f being an inexpensive source o f magnesium. It also contains
calcium hydroxide which is used t o adjust pH. The disadvantage is that
the magnesium oxide concentration cannot be varied without changing the
pH. To evaluate dolomitic lime, the concentration i n t h e geothermal f l u i d
was varied from 0 t o 1200 ppm. Table A-1 summarizes the data w i t h the
respective concentrations of magnesium oxide and calcium hydroxide listed
as ppm M g O and CaO, respectively.
5.2
Effect of pH and Magnesium
Results (Figure 3) show that silica reduction t o desired levels was
not achieved, a t least i n the concentration and pH range covered. The
increase i n residual s i l i c a a t pH 11.0 was unexpected; it had been anticipated that any increase i n magnesium would result i n some additional
s i l i c a removed. This increase i n residual s i l i c a a t pH 11.0 was observed
i n other tests sun with dolomitic lime and other forms o f magnesium oxide,
and it was noted i n those tests that a pH of 10.9 or 11.1 was never
This would indicate that a t pH 11.0 an equilibrium is being
established with the s i l i c a i n solution a
partially soluble complex, or
that the magnesium is complexed i n some
that makes it unavailable t
react with the silica. The decrease i n s i l i c a above pH 11.0 is
t o silica precipitation as a calcium compound.
13
.
150
r
\
50
I
0
I
I
100
200
I
300
I
400
I
500
I
600
I
700
Dolomitic lime (ppm)
800
900
lo00 1100 1200
,
INEL-A-15 425
Figure 3. Water treated with lime a t 60OC.
5.3
Temperature Effect
Increases i n temperature are known t o increase the effectiveness o f
magnesium i n reducing silica. If increased temperature resulted i n a
significant increase i n silica reduction, the hot (403 K, 266OF) geothermal water could be mixed w i t h the power plant effluent t o raise the
temperature prior t o its entering the lime softener. Test results show
that a temperature increase from 323 t o 353 K (50 t o 8OoC) resulted i n an
increase i n silica removed, from 42 t o 55 percent o f the i n i t i a l silica
concentration. The lowest residual silica, 65 ppm, occurred a t a pH o f
10.5 and a temperature o f 353 K (80OC); the highest, 143 ppm, a t a pH of
10.8 and a temperature o f 323 K (5OOC). Data for the four temperatures
(50, 60, 70, and 8OoC) evaluated are listed i n Table A-2. The magnesium
oxide concentrations were held constant a t 200 ppm MgO for plant operating
economy as well as t o reduce the time required for
14
z
a
"
I
testing and chemical analysis. Also, i n this test, and many subsequent
tests, the pH did not exceed 10.8. It is not economical, on a large
scale, t o operate i n the higher pH range as this would require additional
soda ash for hardness reduction and sulfuric acid for pH reduction t o 6.5
'
as specified for the recirculating coolant.
rn
0
15
6.
EVALUATION OF MAGNESIUM OXIDE SYSTEMS
Results of the design basis water-treatment system tests (Tables A-1
t o A - 6 ) suggest that additional testing be done t o determine what modifications of the chemical pretreatment system would be required t o reduce
silica concentration t o acceptable levels. In this regard, additional
systems were investigated t o f i n d a pretreatment program capable of producing water of required quality.
-"
6.1
Effect of Grade of Magnesium Oxide
Three commercially marketed grades of magnesium oxide (Remosil,a
heavy, and l i g h t ) were obtained t o determine what effect the grades of
magnesium have on silica reduction. The magnesium oxide concentration was
held constant a t 200 ppm and the pH was varied, by lime additions, from
10.2 t o 10.8 i n each test set.
The Remosil is s l i g h t l y more efficient i n silica removal than either
the l i g h t or heavy grades of magnesium oxide (Table A-3).
The difference
between Remosil and heavy grade is only five percent, with a residual
s i l i c a content of 65 and 74 ppm, respectively, a t pH 10.5. Remosil i s
approximately 20 percent more effective i n silica reduction than dolomitic
lime a t equal magnesium oxide concentration and pH. The heavy and l i g h t
grades are comparable t o each other.
6.2
Temperature Effect Evaluation Using Remosil
Remsil was selected for evaluation of the temperature effect, since
it was s l i g h t l y more effective i n silica reduction than the l i g h t and
heavy grades o f magnesium oxide (Table A-3).
The same test procedure of
holding magnesium oxide addition constant and varying pH by lime addit i o n was followed. The temperature range covered was 323 t o 353 K
(50-8OOC).
a. Remosil is a l i g h t grade of magnesium oxide marketed by Betz
Laboratories
.
16
c
8
Maximum silica reduction occurred a t 353 K (8OOC) and a pH of 10.2;
t h e residual s i l i c a was 68
and corresponded t o a silica reduction of
56 percent from the untreated water. The minimum silica reduction
occurred a t 323
5OOC) and a pH o f either 10.2 or-10.5, the residual
m, representing a reduction of only 11 percent. The
ed water was 155 ppm. Data are listed i n
silica content o
Table A-4. The most significant feature of the data is the large amount
o f lime required for pH adjustment a t 353 K, compared t o that required a t
323 K; the
ately 3 ta.3. A similar obs ation was seen
for dolomit
2) and indicates that any i
ase i n silica
reduction with increasing temperature would be offset by the cost of
additional lime d soda ash required for pH adjustment
removal.
I
5
i
,
Due t o the large amount of Ca(OH)2 i n the dolomitic lime, the poss i b i l i t y exists
djusting pH w i dolomitic lime, and using a second
source of magnes
oxide t o achie the desired reduction i n silica.
Table A-5 shows results o ests for t h i s hypothesis. In these tests, the
magnesium oxide concentra ns were held constant a t 250 ppm and the pH
varied from 10.2 t o 10.
Test samples 1 and 4, and 5, 3 and 6 have
ons. Duplicate tests w
run t o determine the
I
I
system w i l l not adequately remove silica. The silica reductions i n t h i s
test series are w i t h i n 3 percent of each other. The chemical analysis
data indicate that possibly a slight excess of dolomitic lime may have
been added t o Sample 2. Acceptable levels of silica could not be achieved
when using geothermal f l u i d , no matter what level o f magnesium oxide was
i
1
I
used.
~
17
I
1
6.4
I
H i m Concentrations of Magnesium Oxide
A final magnesium oxide test was run a t higher magnesium’conc
trations, t o 750 ppm M g O , t o determine what level is required t o reduce
silica t o acceptable levels. Data (Table A-6) show that silica reduction
t o 73.5 ppm (64 percent removal) a t a pH 10.2 and magnesium o
625 p p m is neither efficient nor economical.
Y
,
Reduction o f silica t o required levels cannot be achieved w i t h either
.
dolomitic lime, t h e various grades of magnesium oxide, or any
combinations. This suggests that either further testing be done w i t h
ener be increased i n size t o
other chemical reagents or that the l i m
fewer cycles of concenhandle the larger volumes o f water requ
t r a t i o n than t h e design basic o f ten.
Data indicate that the primary factor affecting silica reduction is
the pH reached with additions o f magnesium oxide. The large quantities of
magnesium oxide added for water having high silica contents automatically
adjusts the pH t o approximately 10.2. A t t h i s pH, too l i t t l e magnesium
oxide dissolves for efficient silica sorption. Because the extent of the
sorption reaction is dependent on the available surface area of the magnesium hydroxide floccules and because the surface area is pH-dependent,
the sorption is pHdependent. The result of these interactions is reduced
silica removal a t high pH values.
18
7.
TWO-STAGE COUNTERCURRENT PROCESS UNIT
!
-
This test was s e t up t o simulate a two-stage countercurrent pro-
*
8
cessing u n i t t o evaluate a change i n the mechanical design system.
Remosil was selected as the.source of magnesium nd hydrated lime was
used f o r pn adjustment. Equal quantities of a l l reagents were simultaneously added t o each beaker a t the s t a r t of the test. A t the end of the
test, water samples were filtered i n t o two-litre beakers and ,returned t o
the constant temperature bath. RemosiLwas then added t o each beaker i n
concentrations varing from 25 t o 200 ppm. The systems were stirred slowly
for 15 minutes, then allowed t o settle.
s were qiltered and+analyzed for silica.
Data ,(Table A-7) show that a two-stage countercurrent process
i
unit
,
ely reduce silica concentration. The addition of 400 ppm MgO
reduces s i l i c a t o 22 ppm and thereby allows 4.6 cycles o f
concentration o the recirculating coolant, Thus, i n a continuous flow
processing system silica reduction is enhanced.
8.
SOLUBLE MAGNESIUM COIUPOUND SYSTEMS
8.1
,
l
Magnesium Bicarbonate
The decision t o evaluate magnesium bicarbonate system was the res u l t of a conversation w i t h Mr. A. 8. Mindler of Permutit who indicated
that prior to’1940 it was standard practice o bubble carbon dioxide
through a magnesium hydroxide slurry prior t o addition t o the softener
u n i t , t o enhance silica removal. This agrees with the preliminary indications that the magnesium should be added i n ionic form. To evaluate
t h i s hypothesis, four samples of ten grams per l i t r e of magnesium oxide
slurry were prepared. These systems used the l i g h t , heavy, Remosil, and
dolomitic lime forms of magresium oxide. Carbon dioxide was bubbled
through each of these systems u n t i l no further pH change was observed over
a fifteen-minute period. The l i g h t , heavy, and Remosil magnesium oxide
systems stabilized a t pH 6.9, and a l l were clear solutions. The dolomitic
lime system reached a constant pH value of 7.1. This system resulted i n a
clear solution above a deep bed of precipitate.
Samples from each system were filtered and analyzed for magnesium t o
determine t h e amount of magnesium dissolved. Data are shown i n Table 2.
TABLE 2. ANALYSIS OF FILTERED SAMPLES OF MAGNESIUM HWROXIDE
SLURRIES AFTER BUBBLING CARBON DIOXIDE TO CONSTANT pH
pH After
UI2 Bubbling
Magnesium Source
Dissolved
Magnesium i n
Samples
%
1. L i @ t magnesium oxide system
6.9
100
2. Heavy magnesium oxide system
6.9
09
3. Remosil magnesium oxide system
6.9
81
4. Dolomitic lime magnesium oxide system
7.1
100
20
?
4
.
9
,.
The dolomitic lime system was analyzed for calcium and was found t o
contain less than 0.06 percent i n solution, indicating that the precipi t a t e was calcium carbonate,
Two additional jar tests were run w i t h each of the four bicarbonate
systems i n which the magnesium oxide dosage was held constant and the pH
varied from 10.2 t o 10.8. The 200 ppm magnesium oxide added is based on
magnesium i n solution, as determined from data o f Table 2.
..
.I
Results indicate that silica reduction using magnesium bicarbonate is
a significant improvement over the corresponding magnesium oxide systems
Residual s i l i c a concentrations indicate that a l l four systems are equivalent i n s i l i c a removal efficiency (variations only t o s i x percent) as
shown i n Table A-8. The best silica reductions occurred~atpH 19.2. The
small differences i n siliua reduction a t different pH values indicate
l i t t l e , i f any dependence on pH when large amounts of the i n i t i a l silica
are removed.
.
.
8.2
Magnesium Sulfate
Based on improved s i l i c a reduction with magnesium bicarbonate, five
additional tests were run, t o evaluate magnesium sulfate for silica
removal. Five magnesium sulfate systems were evaluated. The first
three tests used Epsom s a l t ( M g S 0 4 ~ 7 H ~as) the source of magnesium.
The variable i n these three tests was the amount o f carbonate added as
sodium bicarbonate, NaH033. Sodium bicarbonate, rather than sodium
carbonate, was added, prior t o magnesium oxide addition, to maintain a low
of 8, Data and systems used are listed i n Tables A-9, A-10, and A-11.
c
1
Results are shown i n Figure 4. Maximum silica reduction is achieved
when no soda ash is added. The least silica is removed w i t h the magnesium
sulfate-sodium b
Mujeriego6 who
fou
fect
c
Figure 4. Concentration o f residual silica versus concentration o f magnesium added as ( a ) Epsom salt, ( b ) Epsom salt plus soda ash,
and ( c ) Epsom salt plus sodium bicarbonate.
22
I
.
19
I
:-
17r
*
161
I
I
I
I
I
I
---
Magnesium added as:
Magnesium oxide plus sulfuric acid
Magnesium oxide plus soda ash and
sulfuric acid
I
11
144
-
I
12€
i
Q
3
0
v)
4
2
80
64
48
32
16
4
Figure 5 . Concentration o f residual s i 1 ica versus concentration o f ( a )
magnesium oxide plus sulfuric a c i d and Jb) magnesium oxide plus
soda ash and s u l f u r i c acid.
23
The next two tests were run as magnesium sulfate systems using
Remosil as the source o f magnesium. The f i r s t system was a 10 g/L slurry
of Remosil which was acidified w i t h concentrated sulfuric acid u n t i l the
solution cleared. The second system used equimolar quantities o f Remosil
and soda ash, and the slurry was acidified with concentrated sulfuric acid
u n t i l a clear solution resulted. The results o f these two tests are given
i n Tables A-12 and A-U and shown i n Figure 5. Both systems were superior
i n silica removal t o any previously tested system. The reason f o r the
enhancement over Epsom s a l t systems is not-clear; however, it appears that
acidification frees a l l the magnesium from any chemical complex and, thus,
increases silica-reduction activity.
8.3
Maqnesium Chloride
Two additional tests were conducted, using magnesium chloride
systems. A s l u r r y o f Remosil i n water was acidified w i t h concentrated
hydrochloric acid u n t i l clear. Tests were run a t pH 10.2 (Table A-14) and
11.2 (Table A-15).
Results are shown i n Figure 6 . Silica reduction a t pH
10.2 i s equivalent t o that achieved with magnesium bicarbonate systems.
Silica removal a t pH 11.2 is most sicpificant and concurs with data of Wey
and Siffert” that the optimum pH for silica reduction with magnesium
chloride is 11.0-11.5.
Since chloride concentrations i n the geothermal water are
1000-1400 ppm Cl’, a l l treatment systems should behave as magnesium
chloride systems, suggesting that additional jar tests a t pH values near
11.2 be conducted t o evaluate t h i s hypothesis.
24
9.
F'REACIDIFICATION
Jar tests were conducted using magnesium oxide after first a d j u s t i n g
pH w i t h concentrated s u l f u r i c acid. Magnesium was added i n the form of a
10 s/L s l u r r y o f magnesium oxide. Results (Table A-16) show that pH adjustment t o between 6.2 and 8.3 p r i o r t o treatment had no s i g n i f i c a n t
effect when using 200 ppm MgO. Even w i t h p r e a c i d i f i c a t i o n , pH reached 10
following s l u r r y addition. For i n i t i a l pH values as low as 4.0, no sign i f i c a n t difference i n silica reduction wasachieved. However, for prea c i d i f i c a t i o n t o a pH of 3.0, magnesium oxide a d d i t i o n s increased the pH
t o 3.8 and with pH adjustment t o 10.2. I n t h i s case, r e s i d u a l silica was
reduced t o 17.6 ppm as compared t o a r e s i d u a l silica value of 82 ppm i n
t h e other samples. This f i n d i n g s u b s t a n t i a t e d previous data which demonstrated t h a t magnesium hydroxide must be generated i n s i t u and must ' n o t be
the agent for pH adjustment.
*
P
--
c
26
10.
\
*
IRON SYSTEMS
e use of iron as a silica removal agent was previously demonstrated
.I6 TWO tests were run using iron s a l t s t o reduce silica. These
tests used Fe(I1) i n the form of ferrous sulfate (FeS04) and Fe(II1) i n
the form o f ferric chloride (FeCl,.).
Table A-17 shows results using
e silica was reduced from 185 t o 27 ppm
ource of iron.
pm Fe(I1) a t pH 8.6 w i t h mo significant increase a t
higher pH values. Table A 4 8 shows silica reduction using ferric chloride
(Feel3) as the source of Fe(II1). These data show that iron w i l l adequately reduce s i l i c a levels t o the 15-20 ppm Si02 range when using
approximately 180 ppm Fe(II1). The primary difference between mapesium
and Fe(II1) is that iron reduces the pH t o about 2.4. This eliminates the
use of soda ash for calcium removal and dictates the use o f sodium hydroxide rather than the less expensive lime for p~ adjustment.
3
4
I
27
11. DISCUSSION OF RESULTS
Various grades of magnesium oxide and dolomitic lime w i l l adeq
reduce silica if carbon dioxide is bubbled through the magnesium slurries
t o a pH of 7. Dolomitic lime, i n t h i s type of system, poses a huge sludge
problem. Each 100 l b of dolomitic lime used w i l l produce 78 l b o f calcium
carbonate i n the chemical feed tank. Also, there is a large capital investment for carbon dioxide production. The magnesium oxide systems w i l l
not produce sludge, but they require carbon-dioxide. The i n i t i a l capital
investment to.purchase a carbon dioxide generation u n i t which w i l l produce
91 kg (2000 l b ) o f carbon dioxide per day (the requirement for the 5
Pilot Power Plant) is estimated a t $150,000. The cost of using l i q u
carbon dioxide stored i n an on-site tank holding 2730 kg (30 tons) is
$2,200 per month ( a t early 1979 prices).
magnesium sulfate and magnesium chloride w i l l reduce silica
concentration t o acceptable levels. The use o f magnesium s a l t s w i l l increase total dissolved s o l i d s and may reduce the nurrber of cycles of concentration achievable i n t h e recirculating coolant. The increased sulfate
from the magnesium sulfate presents a potential gypsum scaling problem on
the tubes i n the condenser. The excess chloride, from magnesium chloride
addition, may not o n l y reduce cycles of concentration i n the system, b u t
may significantly increase the corrosion rate, particularly p i t t i n g of the
condenser tubes, thereby effectively reducing the service lifetime of the
condenser.
Both
Other magnesium oxide systems tested were preacidification o f make-up
water w i t h sulfuric acid prior t o addition of a magnesium oxide s l u r r y .
Two iron systems were also tested: Fe(I1) from ferrous sulfate and
Fe(II1) from ferric chloride. These iron systems pose the same scaling
and corrosion problems as their corresponding magnesium systems. They
may, however, be more economical. Ferric chloride differs s l i g h t l y from
magnesium chloride i n that sodium hydroxide rather than lime is used to
adjust pH. The problem w i t h t h i s system is its inability t o reduce the
calcium concentrations w i t h soda ash.
28
*
4
4
The use of any chemical systems to reduce silica t o acceptable levels
i n the make-up water w i l l naturally create significant changes i n the
o r i g i n a l l y anticipated water chemistry of the cooling tower-condenser
system. The effect o f these8chemical changes must be evaluated and i n tegrated with demands o f recirculating and blowdown treatment systems
before sound recommendations for pretreatment can be made.
12.
CONCLUSIONS AND RECOMMENDATIONS
Pnalyses of water treatment systems tested lead t o the following
conclusions.
The design specification of 10-15 ppm silica i n the
effluent water from the pretreatment system cannot be
achieved by the use of dolomitic lime, nor with any of the
other conventional water-treatment systems simulated i n the
tests reported here.
3
4
The 10-15 ppm residual silica level can be achieved with
either a soluble magnesium or iron s a l t system, where the
metal is added as a solution.
The zero-level bicarbonate i s automatically met
adjustment.
w i t h pH
Y
The maximum calcium concentration of 340 ppm as Cam3 is
readily achieved through the addition of soda ash.
(5) The hydroxide concentration, reported as Cam3, is i n the
150-250 ppm range, s l i g h t l y higher than the 157 ppm
specified.
Recommendations are:
\
(1) That dolomitic lime should not be used as the source of
magnesium for s i l i c a reduction, and
S
(2)
That no specific chemical system can be recommended based
on the data reported here.
Several systems can adequately reduce silica, yet each of these
systems poses potential problems which have yet t o be evaluated.
It is recommended that the chemicals which reduce silica t o req u i r e d levels be tested and evaluated for scaling effects and corrosion rates on carbon steel.
/
9
9
Chemicals recommended for testing for scaling and corrosion on
carbon steel after s i l i c a removal are Mg(HC03)2, MgS04, and
MgC12. Data i n Table A-19 show that these systems are effective i n
silica removal and also have a low ratio of MgO t o Si02 removed.
The most effective treatment is the +MgO/Na2C03/H$04 system;
however, it is not considered economical and would substantial$y i n ’
crease the s o l i d s content. &cause of the common anions (Cl-,
SOi2) of the iron systems, evaluation of the magnesium systems
should be indicative of how the iron s a l t systems would behave with
respect t o scaling. However, for data confirmation, tests t o determine chemical loss and corrosion rates on low-carbon steel using water
treated with ferric chloride also should be conducted.
31
l3. REFEFtECES
,
1.
R. N. Wallace, private communication, Betz Laboratories, Inc.
April 1977.
2.
V. Thanh Nguyen, Raft River Thermal Loop Design o f t h e Cooling
Water Treatment Plant Using Geothermal Water Supply, Garrett
Energy Research and Engineering, Inc., August 15, 1977. Reviewed
and updated February 17, 1978, by L.--E. Hiebert.
3.
L. 6. Owen, Precipitation o f Amorphous Silica from
HigbTemperature Hypersaline Geothermal Brines, Lawrence
Livermore Laboratory, Vtiversity o f California, M.S. dated 1975.
4.
B. K. Krauskopf, "Dissolution and Precipitation of Silica a t LOW
Temperatures, Geochimica e t Cosmochimica Acta, 10, 1956,
pp. 1-26.
5.
S. A. Greenberg and E. W. Price, "The Solubility of Silica i n
Solutions of Electrolytes,t1 Journal of Physical Chemistry, 61,
11, 1957, pp. 1539-1540.
6.
Rafael Wjeriego, Silica Removal from Industrial Water,
Lhiversity blicmfilms International, Ann Arbor, Michigan, Ph. D.
thesis, University of California, Berkeley, 1976.
7.
Handbook of Chemistry and Physics, 53rd edition, Cleveland:
Chemical Rubber Publishing Conpany, 1971.
0.
B. N. Ryzhenko, "Determination of Hydrolysis of Sodium Silicate
and Calculation of Dissociation Constants of Orthosilicic Acid a t
Elevated Teweratures," Geochemistry International, 4, 1967,
pp. 99-107.
32
m
i
9.
L. D. Betz, C. A. Noll, and 3. J. Maguire, "Adsorption o f Soluble
S i l i c a from Water," I n d u s t r i a l and Engineering Chemistry, 33, 6,
1941, pp. 814-821.
11.
R. Wey and B. S i f f e r t , Col. o f Intern, Centre N a t ' l Recn. Sci.
(Paris) NO. 115, 1962, pp. 11-23.
11.
A. S. Behrman and H. Gustafson, "Removal o f S i l i c a from Water,"
I n d u s t r i a l and Engineering Chemistry, A p r i l 1941, pp. 468-472.
,
12.
H. L. Tiger, 5 i l i c a Removal by an Improved Magnesia Process,n8
Trans. ASME, (Discussion Section), January 1942, pp. 49-63.
I
13.
Cornel Wohlberg and Jerry R. Buchholz, 5 i l i c a i n Water i n
Relation t o Cooling Tower Operation,## Paper No. 143,
Corrosion 75, Toronto, Apri 1 14-18,
14.
1975.
Annual Book o f ASTM Standards, Part 31, Philadelphia: Anerican
Society f o r Testing and Materials, 1978, pp. 1172-1176.
15.
Hach Manual
Ames, Iowa:
Water and Waste Water Analysis, 3rd edition,
Hach Chemical Company, 1975, pp. 2-5,
2-48,
and
2-1 19.
ings,"
Journal of
September 1948, pp. 981-988.
o f S i l i c a from
Water by Hot P r
ss," I n d u s t r i a l and Engineering Chemistry, 32, 11,
1941, pp. 1323-1329.
33
i
Betz Handbook o f I n d u s t r i a l Water Conditioning, 7 t h e d i t i o n , Trevose,
Pennsylvania: Betz Laboratories, Inc. , 1976.
R. N. Wallace, p r i v a t e communication, Betz I n d u s t r i e s , Inc., J u l y 1978.
Nordell Eskel, Water Treatment for I n d u s t r i a l and Other Uses,
New York: Reinhold Publishing Corporation, 1961.
t
Ghanshyam 0. Sharma, Ynfluence of C02 on Silica i n Solution,t1
Geochemical Journal, 3, 1970, pp. 213-227.
G. 8 . Alexander, W. M. Heston, and R. K. Iler, *#TheS o l u b i l i t y
Amorphous Silica i n Water," Journal of Physical Chemistry, 58, 6,
1954, pp. 453-455.
K. 8. Krauskopf, The Geochemistry o f Silica i n Sedimentary
Environments, Special Publication of Society o f Economic
P a l e o n t o l o g i s t s and Mineralogists Symposium, 7, pp. 4-19.
h e r , The Colloid Chemistry of Silica and Silicates, New York:
Cornell University Press, 1955 ,
R. K.
A. V. Karyakin, Yu 6. Kholina, and N. V. Soboleva, "The I n t e r a c t i o n of
Water w i t h Silica,Il Geochemistry I n t e r n a t i o n a l , 1975, pp. 176-178.
G. 6. Alexander, Vhe Reaction o f Low Molecular Weight Silicic.Acids
with Molybdic Acid," Journal of the American Chemical Society, 75,
1953, pp. 5655-5657.
R, M. Garrels, '%ilica: Role i n t h e Buffering of Natural Waters,Il
Science, 148, April 1965, p. 69.
*I
I
G. 0. atamto, a u r a Takeshi, and Gota KatSumi, IIProperties of Silica
i n Water," Qeochimica e t Cosmochimica Acta, 12, 1957, pp. 123-132.
34
G. R. &l
J.l,
P. Leineweber, and
J. C. Yang, Characterization and
Removal of S i l i c a f r o m Webster, South Dakota and Roswell, New Mexico
Well Waters, Lhited States Department of I n t e r i o r , Contract No.
14-01-0001-854,
Research and Development Progress Report No. 286,
January 1968.
John E. Schenk and Walter 3. Weber, Jr., IIChemical Interactions
E
Dissolved S i l i c a with Iron(I1) and (111),It Journal o f the American
Water Works Association, February 19@, pp. 199-212.
i
c
of
.
f
Y
-1
4
4
4
APPENDIX A
ANALYSIS OF WATER TREATED WITH VARIOUS SYSTEMS
37
.
t
J
TABLE A-1.
WATER TREATED WITH DOLOMITIC LIMEa AT 60°C
Sanple Nurrber
Chemica1
Ad d i t ions (ppm)
- - - - - 1
2
3
4
5
6
CaO
88
176
264
352
440
528
MgO
74
140
222
296
370
440
4-I
Adjusted pH
9.2
10 .O
-
10.6
11.o
11.3
11.6
Untreated
Water
8.1
Analysis (ppm)
. Cab
152
160
18.5
138
5.7
110
4.1
140
12.3
114
6.9
164
108
7.9
,
lime is 58 w t
% Ca((w)2 and 37 w t % MgO.
a.
Dolomitic
b.
Calcium is reported
c.
Silica
d.
M@/Si@ i s the r a t i o o f MgO added t o Si@ removed.
Note:
is reported as Si@.
Data representation i n succeeding tables identical.
TABLE A-2.
WATER TREATED WITH DOLOMITIC LIME AT FOUR TWERATURES
n
Sanple Nurrber
Chemical
Additions (ppm)
-
- - 1
2
3
6
Tenperature 8OoC
Temperature 7OoC
.
CaO
216
216
682
216
250
603
MgO
200
200
200
200
200
200
Na2c03
346
346
792
- 346
384
1020
Adjusted pH
10.5
10.5
10.8
10.5
Untreated
Water
t
'
10.5
10.8
8.2
Analysis (ppm)
ca
26
8
6
8
36
12
113
Si02
79
75
113
85
81
133
165
M@/Si02
Chemical
Additions (ppm)
2.3
2.2
3.9
2.5
2.4
3
Temperature 6 O O C
Temperature 5o°C
CaO
216
216
412
216
216
450
MgO
200
200
200
200
200
200
Na2m3
346
346
660
346
346
560
Adjusted pH
9
6.3
10.5
10.5
10.8
56
44
10
103
96
113
10.6
10.8
8
12
10
113
130
123
143
165
10.6
8.2
Analysis (pprn)
3.2
2.9
3.9
40
5 07
4.8
9.1
TARE A-3.
.
WATER TREATED WITH DIFFERENT GRADES
OF MWlESIUM OXIDE AT 6OoC
Sample Number
Chemical
Additions (ppm)
;6
Y
la
2a
3a
7
qb
7
Sb-
CaO
200
370
400
284
321
435
MgO
200
m
200
200
200
200
180
540
510
235
265
355
Na2m3
Adjusted pH
1,
lo02
10.5
10.8
4
6
6
68
65
66
--
Untreated
- Sb
10.2
10.5
10.8
4
14
78
74
12
60
Water
8.1
Analysis (ppm)
ca
Si02
1.9
MgO/SiO,
Chemical
Additions (ppm)
CaO
MgO
7c
2.0
8'
->
2.0
1.9
lld
1Zd
387
537
637
200
200
200
200
125
170
310
200
145
285
375
ca
SiO2
Untreated
. Water
lo02
10.5
10.8
10.2
10.5
6
6
4
6
4
16
113
76
81
73
92
88
90
176
2.0
M@/Si02
176
7
380
Adjusted pH
i
1.8
208
200
152
*2a3
1.8
1l3
2.1
1.9
\
I
4
a. As Remosil.
b.
As heavy grade.
c.
As l i g h t grade.
d.
As dolomitic lime.
41
2.7
2.9
10.8
2.9
8.1
-
TABLE A-4.
e.
WATER TREATED WITH REMOSIL AT FOUR TWERATURES
Sample Number
Chemical
Ad ditions ( ppm)
- - 4
Temperature 8OoC
CaO
483
200
600
Adjusted pH
920
200
1070
5
6
Untreated
Water
Temperature 70OC
l320
200
1600
.
189
200
208
I
265
200
288
-t
605
200
774
10.2
10.5
10.8
10.2
10.5
10.8
10
68
2.3
16
73
2.4
6
4
6
6
80
80
2.7
91
100
a
8.0
Analysis (ppm)
ca
Si02
MgO/Si02
Chemical
Additions (ppm)
Adjusted pH
2.7
3.1
120
155
9
3.6
Temperature 500C
Temperature 60°C
189
200
265
605
121
200
200
300
397
900
200
181
200
363
200
180
271
545
10.2
10.5
10.8
16
123
6.3
12
103
3.9
4
12
113
4.8
I38
11.8
10.2
10.5
10.8
20
138
11.8
8
l30
8.0
8.0
Analysis (ppm)
ca
Si02
M@/Si02
42
120
155
-?
1
TABLE A-5.
WATER TREATED WITH DOLOMITIC LIME AND REMOSIL AT COEGTANT
MAGNESIUM DOSAGE AT 6OoC
Sample Number
Chemical
Additions (ppm)
1
-
2
3
-
4
5
-
296
129
208
175
296
250
0
140
75
0
CaO
129
MgO
110
208
175
140
75
MgO
(Remsil)
Adjusted ptl
-.
10.2
*
10.5
6
250
10.8
Untreated
Water
10.2
10.6
10.8
186
98
1.6
308
80
2.6
332
75
2.4
117
-
-
untreated
625
750
8.5
Analysis (ppm)
ca
182
103
1.7
so2
MgO/Si02
290
84
2.7
330
77
2.5
178
b
,
TABLE A d .
WATER TREATED WITH REMOSIL AT 60%
(pH ADJUSTED WITH NaOH)
Sample Number
MgO
NaOH as
=%
125
0.014
2
3
4
250
375
500
3
0.010
0.002
5
6
Water
0
b
.2
t
10.2-
10.2
10.2
76
74
8.4
Analysis
Si@ (ppm)
M@/Si@ ( P V )
pH
142
2.0
165
6.2
76
2.9
3.9
after
MgO
addition
80
6.0
4.7
(
9.4
9.8
10
10.1
10.2
10.2
206
.
TABLE A-7.
WATER TREATED WITH REMOSIL TO SIMULATE
A TWO-STAGE COUNTERCISRENT PROCESS
Chemical
Additions (ppm)
Adjusted pH
1
2
-3
4
5
6
302
302.
302
302
302
302
200
200
2al
200
200
200
428
428
428
420
428
420
lo .4
10.4
lo .4
10.4
10.4
10.4
92
92
86
97
84
94
Untreated
Water
8.4
Analysis (ppm)
Si02
MgO/Si02
2.7
2.7
2.5
2.9
2.5
165
2.8
P
(Samples were f i l t e r e d and returned t o constant temperature bath. )
3
Chem i ca1
Additions (ppm)
MyO
25
50
75
100
150
200
50
40
29
34
25
22
Analysis (pprn)
Si@
M@/sis
0.6
1.0
1.3
1.6
2.5
2.9
2.0
2.0
2.0
2.3
2.5
2.8
(Step 2
M@/SiQ
(Process)
f
44
-I
TABLE A-8.
WATM TREATED WITH CARBONATED SOL JTIONS
OF MAGNESIUM OXIDE AT 60%
Sample Nurber
Chemical
Additions (ppm)
i
Ca0-
306
4 03
569
MgO
200
200
200
334
427
540
200
200
200
_
.
I
Adjusted pH
10.2
10.5
10.8
10.2
10.5
10.8
32
33
38
32
34
36
0.3
Analysis (pprn)
SiO2
1.4
M@/Si02
*
1.4
1.4
1.4
1.4
180
1.4
Untreated
Water
Chemical
Additions (pprn)
CaO
357
439
MgO
200
2m
Adjusted pH
556
340
436
550
200
200
200
10.2
10.5
10.8
10.2
10.5
10.8
32
39
41
35
36
30
8.3
Analysis (pprn)
SiO,
Mg0/SiO2
.
1.4
1.4
1.4
1.4
1.4
4
\
a.
As l i g h t grade.
b.
As heavy grade.
c.
As dolomitic Lime.
d.
As Remosil.
t
45
1.3
180
,
TABLE A-9.
I
WATER TREATED WITH MAGNESIUM SULFATE
(E?SCN SALT) AT 60oC
Sanple NJrrber
Chemical
Additions (ppm)
1
2
3
4
5
6
CaO
740
740
740
MgO
50
100
150
Adjusted pH
10.2
740
200
815
890
225
250
Ultreated
Water
.t
10.2
10.2
10.2
10.2
10.2
0.3
Analysis (ppm)
ca
so2
M@/Si02
P
a06
300
85
0.6
50
0.9
480
25
1.1
520
12
1.3
560
5
1.4
600
3.6
1.6
140
3.63
,
TABLE A-10.
WATER TREATED WITH MAGNESIUM SULFATE
AN) SOOA ASH AT 60%
T
Sanple Wnber
Chemical
Additions (pprn)
1
2
3
4
5
6
350
350
500
620
50
100
375
150
200
225
250
600
750
1
m
1200
375
Adjusted pH
3
10.2
10.2
10.2
780
Ultreated
Water
900
10.2
10.2
10.2
16
24
20
8.3
Analysis (ppm)
ca
ma
UO,
M@/Si02
95
0.7
32
74
1.1
2
43
1.3
23
1.4
9
1.5
7 ’
1.8
140
163
t
a.
NO denotes ncne detected.
3
i
WATER TREATED W I T H EQUIMCLM WANTITIES OF
TABLE A-11.
MAGNESIUM SULFATE AN) SODIUM BICARBONATE AT 60%
Chemical
Additions (ppm)
B
Sanple Mrber
1
2
3
.4
-5
6
CaO
io37
1133
urn
1333
1451
1667
MgO
50
im
l50
200
225
250
Adjusted pH
10.2
10.2
10.2
10.2
10.2
10.2
Ultreated
water
8.3
,
Chalysis (ppm)
40
40
10
40
20
16
140
124
102
1.6
71
1.6
50
1.8
35
1.6
31
1.7
161
1.3
T K E A-12.
WATER TREATED WITH MAGNESIUM SULFATEa AT 6OOC
3
Chemical
Additions (ppm)
CaO
M@
b
c
4
‘
Adjusted pH
1
2
378
503
so
im
658
874
3
4
5
6
631
150
11M
923
1134
200
m
BO4
400
1710
1975
70
10.4
10.4
10.4
10.4
63
42
22
u
10.4
m.4
Ultreated
Water
8.4
Analysis (PPI)
Si%
M@/Si02
a.
0.5
0.8
1.0
1.3
6
1.8
System made by a c i d i f i c a t i o n o f MgO suspension w i t h H$i04.
47
5
’ 3.0
173
TABLE A-13.
7
WATER TREATED UITH MAGNESIUM SULFATE AND SODA ASHa AT 60OC
Sample Number
Chemical
Additions (ppm)
1
2
3
4
5
6
CaO
121
50
Adjusted pH
10.4
129
100
10.4
129
150
57
300
118
200
10.4
10.4
10.4
Untreated
Water
400
10:4
8.6
Analysis (ppm)
Si02
MgOlS iO2
a.
t
P
29
0.3
14
0.6
5
*.4
0.9
.
1.2
3
1.7
3
2.3
176
System produced by addition of equimolar quantities o f MgO and Na2CO3. then
H20 and H2S04 u n t i l solution clears.
E
TABLE A-14.
WATER TREATED WITH MAGNESIUM CHLORIDEa
AT 6OOC AND pH 10.2
Sample Number
Chemical
Additions (ppml
1
2
3
CaO
MgO
Na2C03
2 60
25
332
Adjusted pH
10.2
296
50
3 72
10.2
335
75
426
4,-5
6
407
100
518
10.2
10.2
453
125
577
10.2
3
Untreated
Water
475
150
605
10.2
8.3
Analysis (ppml
Ca
Si02
M@/S 102
16
14
12
10
12
8
140
106
0.4
94
0.7
62
0.7
55
0.9
39
1.0
30
163
1.1
7
‘J
a.
System produced by addition o f HC1 t o MgD suspension u n t i l s o l u t i o n clears.
48
i
TABLE A-15.
Chemical
Additions (ppm)
1
,
WATER TREATED WITH MAGNESIUM CHU3RIDEa
AT 60% AND pH 11.2
2
3
4
loo0
1000
100
loo0
125
5
I
I
6
-
Untreated
Water
c
CaO
4
MgO
loo0
40
80
loo0
140
1000
155
*.-
Adjusted pH
11.2
11.2
11.2
11.2
11.2
11.2
Analysis (pprnj
8.3
\
I
SO2
Mg0/SiO2
23
0.3
15
0.5
14
0.7
12
0.8
11
0.9
9
1.0
0
a. System produced by addition o f concentrated HC1 t o MgO suspension u n t i l
solution clears.
163
v
TABLE A-16.
WATDi TREATED WITH MAGNESIUM OXIDE
FOLLRWING pH ADJUSTMENT AT 6OOC
!
Chemical
Additions
Sanple Mrber
1
2
3
8.2
8.1
7.6
4
5
6
Untreated
Water
7.1
6.6
6.2
8.3
---
@-i
following
H2SQ
I
1
addition
I
MgO ( P P d
I
200
200
-2 m
200
200
pH following
MgO addition
10.1
10.0
10.0
10.0
10.0
10.0
Adjusted pH
10.2
10.2
10.2
10.2
10.2
10.2
94
93
92
90
Analysis (ppm)
Si@
M@/SiO2
Chemical
Ad d i t ions
!
200
112
2.7
7
-
95
2.2
'
2.2
2.2
2.1
185
2.1
F
8
11
- - 12
9
10
3
pH after
"2SQ
addition
5.6
5.0
4.0
3.0
2.4
2.2
8.3
:
MgO ( P P N
I
200
200
200
XI0
200
200
pH after
MgO addition
10.0
10.0
10.0
10.0
10.0
10.0
Adjusted pH
10.2
10.4
10.2
10.2
10.2
10.2
9o
82
84
18
18
18
Analysis (ppm)
Si02
2.1
1.9
2.0
1.2
1.2
1.2
185
v
3
50
TABLE A&
WATER TREATEDWITH FERROUSSULFATE
AT 60% AND VARYING pH (pH ADJUSTED WITH NaOH)
Sanple Number
t
Chemical
Addit ions (ppm)
1
250
Fe
- .2
3
4
5
250
250
250
i50
6
250
i
NaOH as
CaCO3
0.066
0.071
0.078
0.083
Adjusted pH
8 .O
8.6
9 .o
9.6
Untreated
Water
.
0.102
10
b
o
0.138
10.4
8.3
Analysis (ppm)
Si@
65
Fe/SiO2
23
27
1.6
2.1
20
1.5
1.5
21
1.5
22
185
1.5
I T H FERRIC CHLORIDE AT 60oC
TABLE A-18.
Sanple Nurber
Chemical
Additions (ppm)
1
2
3
4
5
6
Fe
72
,108
144
180
226
CaO
333
394
707
1157
1483
Adjusted pH
8.2
7
7.1
6.5
8.8
0 .0
llncreated
Water
8.3
> .
4 -
4
Si02
Fe/Si02
i
125
0
1l3
1.0
93
1.2
47
1.0
17.0
1.1
12
1.3
185
Original
si02 (ppm)
Addit ions
MgO
Final
CaO
MgO
(PPm)
-
PH
Source (ppm)
MgO/Si02
Removed (ppm)
T
164
74
88
160
9.2
18.5
164
148
176
138
10.0
5.7
164
222
264
110
10.6 -
4.1
164
296
352
140
11.0
12.3
164
370
440
114
11.2
6.9
164
444
528
108
11.3
7.9
176
200
200
60
lo 02
1.9
176
200
370
65
10.5
1.8
176
200
400
66
10.8
1.a
176
200
284
78
10.2
2.0
176
200
321
74
lo .5
2.0
176
200
435
68
10.8
1.9
176
200
152
76
10.2
2.0
176
200
208
81
10.5
2.1
176
200
380
73
10.8
1.9
176
200
387
92
10.2
2.7
176
200
537
88
10.5
2.9
176
200
637
90
10.8
2. i
178
250
129
103
lo .2
1.7
7
178
250
208
84
10.5
2.7
a
I
,
0
F
3
52
TABLE A-19,
(Continued)
Addit ions
Original
si02 (ppm)
MgO
(ppm)
Final
CaO
c
178
250
296
77
c
?
178
250
129
98
178
250
208
178
250
206
i
ppm)
DL
10.8
MgO/SiO2
Removed (ppm)
2.5
10.2
DL+M(R)
1.6
.80
u3.6
DL+M(R)
2.6
296
75
10.8
DL
2.4
125
0
142
lo02
M(R)
2.0
206
250
0.
165
10.2
M(R)
6.2
206
375
0
lo 02
M(R)
206
500
0
10.2
M(R)
3.9
625
0
74
1002
M(R)
4.7
M(R)
6.0
3
c
MgO
PH
(ppm)
.
--
.
2.9
206
750
80
10.2
180
200
32
10.2
M(L)+C02
1.4
180
200
33
10.5
M ( L) +co2
1.4
180
200
38
10.8
M(L)+C02
1.4
180
200
32
10.2
M(H)+%
1.4
180
200
34
10.5
M(N+CO2
1.4
180
200
36
10.8
1.4
180
200
32
10.2
1.4
180
200
39
10.5
DL+Q
180
200
41
10.8
DL+C02
1.4
1
200
35
10.2
M(R)+cO2
1.4
36
10.5
M(R )+CO2
1.4
180
334
.
.
1.4
\
.
(Continued)
TABLE A-19.
180
200
550
30
10.8
M(R)+CO2
1.3
163
50
740
85
10.2
EPS
0.6
163
100
740
50
10.2
EPS
0.9
163
150
740
25
10.2
EPS
1.1
163 . 2 0 0
740
12
10.2
EPS
1.3
163
225
015
5
10.2
EPS
163
250
890
4
10.2
EPS
1.6
163
50
350
95
lo .2
EPS+SA
0.7
163
100
350
74
10.2
EPS+SA
1.1
163
150
500
43
10.2
EPS+SA
1.3
163
200
620
23
10.2
EPS+SA
1.4
163
225
780
9
10.2
EPS+SA
1.5
163
250
900
7
10.2
WS+SA
1.8
163
50
lo37
124
10.2
VSNAB
1.3
163
100
ll33
102
10.2
WS+W
1.6
163
150
1200
71
10.2
EPS+NAB
1.6
163
200
l333
50
10.2
EpS+NAi3
1.8
163
225
1451
35
10.2
EPS+NAB
163
250
1667
31
10.2
EpS+W
173
50
378
63
10.4
173
100
503
42
10.4
54
,
1.6
1.7
0.5
MgS04( R)
0.8
cr
F
9
TABLE A-19
.
(Continued)
~
Additions
MgO/SiOZ
Removed (ppm)
PH
li
10.4
1.0
i
10.4
1.3
1.8
-6
173
400
l304
5
10.4
MgS04( R)
3.0
176
50
121
29
lo 04
M( A)
0 03
176
100
129
14
10.4
M(A)
0.6
176
150
129
5
lo 04
M(A)
0.9
176
200
110
4
10.4
M(A)
1.2
176
300
57
3
lo 04
M(A)
1.7
176
400
57
3
10.4
M(A)
2.3
163
25
260
106
10.2
MSC12(R)
0.4
163
50
296
94
10.2
MSC12(R)
0.7
335
62
10.2
MSC12(R)
0.7
407
55
10.2
MSC12( R)
0.9
4
163
'
I
75
!
163
100
163
125
453
39 .
10.2
MSC12(R)
1.0
163
150
475
30
10.2
MSC12( R)
1.1
163
40
lo00
23
11.2
MSC12(R)
0.3
163
80
loo0
15
11.2
M$12(R)
0.5
4
163
100
loo0
14
1102
M&(R)
0.7
4
163
125
la30
12
11.2
0.8
3
140
loo0
11
11.2
0.9
\
t
I
TABLE A-19.
(Continued)
PH
'
163
155
loo0
9
185
200
0
112
185
200
0
95
MgO
Source (ppm)
MgO/SiO2
Removed (ppm)
11.2
1.0
8.2,a
10.2b
2.7
8.1,
2.2
10.2
185
200
0
94
7.6,
10.2
2.2
185
200
0
93
7.1,
10.2
2.2
185
200
0
92
6.6,
10.2
2.1
\
185
200
0
90
6.2,
10.2
2.1
185
200
0
90
5.6,
10.2
2.1
185
200
0
82
5.0,
10.2
1.9
185
200
0
84
4.0,
2.0
10.2
185
200
0
18
3.0,
10.2
1.2
185
200
0
18
2.4,
10.2
1.2
185
200
0
18
2.2,
10.2
1.2
a.
pH following preacidification.
b.
F i n a l pH.
/
56
.
TABLE A-19
.
(Continued)
OL = Dolomitic Lime.
-G
M(R) = Remosil.
M(H) = t"eavy*t MgO.
$
M(L) = "Light" M g O .
EPS
= Epsom S a l t (MgS04).
SA = Soda Ash (M2C03).
NPB = Sodium Bicarbonate (NaHCDj).
MgS04( R) = Acidified Remosil using Sulfuric Acid (H2S04).
M(A) = Acidified Mixture of Remosil and Soda Ash.