WATER POLLUTION CONTROL RESEARCH SERIES • 17010DHK 08/69
CHEMICAL EXFOLIATED
VERMICULITE FOR
REMOVAL OF PHOSPHATE
FROM WASTE WATERS
U.S. DEPARTMENT OP TOT INTERIOR • FEDERAL WATER QUALITY ADMINISTRATION
-------
WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Reports describe the results
and progress in the control and abatement of pollution of our
Nation's waters. They provide a central source of information on
the research, development, and demonstration activities of the
Federal Water Quality Administration, Department of the Interior,
through in-house research and grants and contracts with Federal,
State, and local agencies, research institutions, and industrial
organizations.
Water Pollution Control Research Reports will be distributed to
requesters as supplies permit. Requests should be sent to the
Planning and Resources Office, Office of Research and Development,
Federal Water Quality Administration, Department of the Interior,
Washington, D. C. 20242.
-------
CHEMICALLY EXFOLIATED VERMICULITE FOR
REMOVAL OF PHOSPHATE FROM WASTEWATERS
by
Jacob Block
W. R. Grace & Company
Clarksville, Maryland 21029
for the
FEDERAL WATER QUALITY ADMINISTRATION
DEPARTMENT OF THE INTERIOR
Contract #14-12-485
August, 1969
For sale by the Superintendent of Documents, U.S. Government Printing Office
Washington, D.C. 20402 - Price 60 cents
-------
FWQA Review Notice
This report has been reviewed by the Federal Water
Quality Administration and approved for publication.
Approval does not signify that the contents neces-
sarily reflect the views and policies of the Federal
Water Quality Administration, nor does mention of
trade names or commercial products constitute
endorsement or recommendation for use.
-------
ABSTRACT
Many aluminum-vermiculites were prepared and tested for phosphate
removal. The maximum capacity obtained was 10 mg P04/gram of treated
vermiculite. The highest phosphate capacities were obtained with mate-
rials prepared at relatively low pH (3.0) and dilute aluminum solutions
(0.1-0,05M), materials prepared at higher pH values (3.8) and more con-
centrated aluminum solutions (0.33M) had lower phosphate capacities. Ex-
perimental results have shown hydroxylated aluminum vermiculite is in-
active towards phosphate ion, and that our original concept of hydroxyl
ion replacement was not valid. The mole ratio of phosphate adsorbed to
exchanged aluminum seemed to approach 0.33> or an exchange capacity of
13.7 mg P04/gm. The highest experimental capacity obtained was 12.2 mg/gm
for a Li-expanded Poole vermiculite.
Various regeneration schemes were attempted, the most successful
being one in which a dilute sulfuric acid solution containing a small
amount of A12(S04)3 was used as the regenerant. The addition of the
A12(S04)3 prevented the loss of aluminum from the vermiculite.
No significant differences in capacity were found between thermally
and chemically exfoliated vermiculite. In addition, an aluminum vermicu-
lite prepared directly from a vermiculite ore without prior exfoliation
had a capacity of 7.3 mg P04/gm (compared to about 9 mg/gm for exfoliated
material).
Vermiculites containing cations other than Al were also prepared.
These included Fe(lll), Fe(ll), La(lll), and Cu(ll). None, however,
appeared more promising than Al-vermiculite.
The adsorption isotherm of aluminum-vermiculite was obtained, and
the data were found to fit both the Langmuir and Freundlich plots. How-
ever, adsorption at phosphate concentrations of around 25 ppm was extremely
low ( ~ 0.k mg/gm), and for this reason work on this project was terminated.
lit
-------
TABLE OF CONTENTS
;e No.
FOREWORD ii
ABSTRACT v
STATEMENT OF THE PROBLEM vi
RECOMMENDATIONS vii
IOTRODUCTION 1
EXPERIMENTAL WORK 3
Chemicals 3
Equipment 3
Analytical 3
Exchange Capacity Determination 3
Exfoliation 3
Chemical 3
Thermal if
Preparation of Al-Vermiculite k
Determination of Phosphate Capacities k
Batch Method k
Column Method 5
Regeneration 5
Batch Method 5
Column Method 5
Adsorption Isotherm 5
RESULTS AND DISCUSSION 6
Vermiculite Ore Characterization 6
Exfoliation 7
Preparation and Phosphate Capacity of
Aluminum Vermiculite 7
Use of Cations other than Al 11
iv
-------
TABLE OF COMEEHTS (cont'd)
Page No.
Regeneration 12
NH^OH 12
HC1 13
Various Salts 1^
A12(S04)3 + HgS04 15
Working Capacity of Al-Vermiculite 15
Economic Evaluation 17
REFERENCES 21
APPENDIX A - Economic Evaluation A-l
APPENDIX B - Assumption and Data for Calculations B-l
APPENDIX C - System Design C-l
APPENDIX D - Capital Cost Estimate D-l
APPENDIX E - Cost Estimate of Phosphate Removal from
Wastewater E-l
APPENDIX F - Breakdown of Total Capital Requirements F-l
-------
STATEMENT OF THE IROBLEM
The primary objective of the work described herein was to prepare
a chemically treated vermiculite with a high phosphate capacity that
could be economically regenerated. Our original concept was to replace
the exchangeable cations in an expanded vermiculite with hydroxylated
aluminum ions. Phosphate removal could then be achieved by exchange
with the hydroxyl groups, resulting irj the aluminum being bonded to the
vermiculite and phosphate. Regeneration was to be attempted with an
alkali such as ammonia, which would replace the phosphate groups with
hydroxyls.
-vi-
-------
RECOMMEKDA.TIOHS
In studies on the use of activated alumina to remove phosphate from
wastewater, Yeei used NaOH as the regenerant and observed an 8$ alumina
loss per cycle. Neufeld and Thodos2 found that an activation step with
nitric acid is needed "before recycle. In the current work on aluminum-
vermiculite we found that a dilute sulfuric acid solution containing a
small quantity of AlaCsO^a served simultaneously as a good regenerant
and "activator." We recommend that this regenerant he applied to acti-
vated alumina to achieve a lower raw material cost, decreased alumina
losses and a highly active alumina.
We feel that the idea of producing a selective inorganic anion ex-
changer has some merit. The replacement of the exchangeable cations in
vermiculite with aluminum does not seem to be the answer, however. A
better approach might be the direct synthesis of an inorganic anion ex-
changer, rather than trying to convert a cation exchanger into jan anion
exchanger.
-vii-
-------
INTRODUCTION
Phosphates occur in municipal effluents mainly from human wastes
and synthetic detergents. The phosphates used in detergents are sodium
tripolyphosphate and tetrasodium pyrophosphate. However, these con-
densed phosphates break down to ortho-phosphates in sewage, and it can
he assumed that after 2k hours at 23°C most of the condensed phosphates
are in the ortho form.3 Phosphates are a problem, because they create a
nutrient imbalance in natural waters which leads to algae blooms (growths)
and eutrophication.
Phosphate presently is removed with lime, iron salts and alum. These
treatments, however, are not selective for phosphate and result in large
quantities of sludge. In addition, highly alkaline effluents are produced
in the lime treatment, and high sulfate or chloride effluents in the other
treatments.
Adsorption methods for phosphate removal include the use of .ion ex-
change resins and activated alumina 1'2. The resins, however, suffer
from the disadvantage of being non-selective for phosphate, whereas acti-
vated alumina tends to show losses of 8-10$ per cycle.
Our objective was to develop a selective anion exchanger for phosphate
by chemically treating vermiculite to convert it from a cation exchanger
to an anion exchanger.
The mineral is a layer silicate, consisting of thin aluminosilicate
platelets held together by hydrated, exchangeable cations (Ca and Mg).
When vermiculite ore is passed through a hot furnace, the water of hydration
associated with the exchangeable cations is converted to steam. As a re-
sult, the ore exfoliates, i.e., the platelets are separated by expansion
of the steam. Thermally exfoliated vermiculite is sold commercially as
an insulation material.
Thermal exfoliation ruptures the chemical bonds which hold the platelets
together and also partially decreases the available surface area. In addition,
some of the mobile cations become "fixed" to the surface.
On the other hand, vermiculite can be exfoliated chemically without the
use of excessive temperature. Simple cation exchange of the exchangeable
divalent cations with monovalent cations followed by hydration with HgO,
will readily separate the platelets. On chemical exfoliation, platelet
size and integrity is retained, maximum surface area is developed and,
most importantly for our current purposes, the calcium and magnesium ions
are still exchangeable.
-1-
-------
At present, no chemically exfoliated vermiculite is marketed com-
mercially; however, tlie process of chemical exfoliation is inherently
simple and lends itself to being performed on tonnage scale.
Cation exchange capacity of thermally exfoliated vermiculites ranges
from 30 to 70 meq/100 gm. Upon chemical exfoliation, this capacity can
be raised as high as 130 meq/100 gm.
Anion exchange capacities for vermiculites are very low, never being
more than 30-^0 meq/100 gm. It is apparent that the anion exchange capacity
of vermiculite is much too low to be considered for removal of phosphate
from wastewaters. On the other hand, we developed a concept by which the
higher cation exchange capacity of chemically exfoliated vermiculite may
be converted to phosphate anion exchange capacity, and this phosphate
anion exchange capacity can, at the same time, be greater than the origi-
nal cation exchange capacity.
Our concept was based on the replacement of the exchangeable cations
with hydroxylated aluminum ions such as Al(OH)+2 and Al(OH)2+. We reasoned
that the aluminum ions would be firmly held by the vermiculite while the
hydroxyl groups would be exchangeable with phosphate, thus converting the
cation exchange capacity to an even greater anion exchange capacity.
-2-
-------
EXPERIMENTAL WORK
CHEMICALS
All chemicals were reagent grade. Vermiculite was obtained from
the Zonolite Division of W. R. Grace & Co. and consisted of three types,
a South Carolina blend, a Libby-Montana ore, and a Boole vermiculite
from South Carolina. The first two vermiculites are normally used to
form packaging materials, while the latter exfoliates poorly but has a
high exchange capacity.
EQUIPMENT
All colorimetric measurements were made with a Bausch and Lomb
Spectronic 20 spectrophotometer. A Beckman zeromatic pH meter was used
for pH measurements, and for some sodium ion determinations.
ANALYTICAL METHODS
Phosphate was determined colorimetrically with ammonium molybdate
and l-amino-2-naphthol-4-sulfonic acid as the colorimetric reagents.
Aluminum and iron were determined colorimetrically with aluminon
and o-phenanthroline,respectively.
Calcium and magnesium determinations were made by EDTA titration,
using either Murexide or Eriochrome Black T as indicators.
Potassium was determined by flame photometry, and sodium either by
flame or with a sodium ion electrode.
EXCHANGE CAPACITY DETERMINATION
Vermiculite exchange capacities were determined by reacting 25 grams
of previously washed and air dried vermiculite with one liter of 3.5M
NaCl at 80°C for at least 18 hours. The amount of Ca, Mg and K released
was then determined.
EXFOLIATIONS
Chemical
Pour different reagents for chemical exfoliation were tried: NaCl,
LiCl, Ha02 and HaSO^ (Exfoliation is needed to facilitate the ion exchange
-3-
-------
reactions of vermicullte). The exfoliation is caused by the hydration
of the sodium, lithium or sulfuric acid. The peroxide-induced expansion
is a result of a catalytic decomposition of the peroxide which causes
the vermiculite platelets to separate.
NaCl - 1.0 kg of ore was placed in a k-liter beaker. Then 2.0
liters of 20$ NaCl solution were added. The beaker was covered and heated
overnight at 80°C. The solution was decanted and the vermiculite was
water-washed until the filtrate was negative to an AgNOs test. The vermicu-
lite was then left in 2.0 liters of distilled water overnight, filtered
and air-dried.
This procedure was later modified by substitution of a 6M NaCl solu-
tion followed by a boiling water treatment.
Lid - Same as the original NaCl procedure except that 15$ LiCl was
used. The material was difficult to filter and was left as a slurry.
HgS04 - 1.0 kg of ore was treated with one liter of concentrated
H2S04 for l6 hours. The acid was decanted and the vermiculite was water-
washed until negative to BaCl2. The material was then allowed to stand
overnight in distilled water, then filtered and air-dried.
Thermal
Portions of vermiculite ore (75 grams each) were put in a porcelain
dish and placed in a furnace, at a preset temperature, for 4 5 minutes.
PREPARATION OF AL-VERMICULITES
Fifty grams of exfoliated vermiculite were treated with one liter of
Al(N03)s solution at the desired concentration and pH. In order to avoid
the addition of foreign ions, pH adjustments were made either by the addition
of HN03 or by the addition of Amberlite IR ^5 ion exchange resin (a weak
base in the hydroxyl form). After the desired pH was reached, the resin
was filtered off. The vermiculite and the aluminum solution was stirred
for a minimum of 18 hours. The vermiculite was then filtered, washed
and. air-dried. The filtrates were gravimetrically analyzed for Al.
The vermiculites containing other cations were prepared in a similar
manner.
DETERMINATION OF PHOSPHATE CAPACITIES
Batch Method
A stock pH 7 sodium phosphate solution was prepared by dissolving 280
gm of sodium hydrogen phosphate (Na2HP04«7 HgO) and 50 gm of sodium dihyflrogeTa
-------
phosphate (HaH2P04'H20) In water and diluting to ^-liters. (Total phos-
phate concentration is 0.37 moles/liter). To determine PC>4 capacity,
100 ml of this solution vas added to 5 grains of Al-venniculite and the
mixture vas stirred for a minimum of l8 hours. The vermiculite was then
filtered, washed and air-dried. A one-gram sample was then treated with
100 ml of a 2$ NaOH solution for 18 hours, and the resultant liquid was
analyzed for phosphate. (The 2% NaOH treatment was checked for complete
phosphate removal by analyzing the filtrate for phosphate and good agree-
ment was obtained).
Column Method
Four grams of vermiculite were placed in a 25 ml burette over a glass-
wool plug. A phosphate solution, previously boiled to remove C02 and
fitted with an Ascarite plug, was then passed through the column at the
desired flow rate. The effluent was analyzed for phosphate.
REGENERATION
Batch Method
A quantity of vermiculite containing a known amount of phosphate was
weighed into a flask. A measured amount of regenerant was then added, and
the mixture was placed in a shaker for 18 hours. The vermiculite was then
filtered and washed. The filtrate was analyzed for phosphate and aluminum
after first being brought to volume.
Column Method
The vermiculite containing a known amount of phosphate was placed in
a 25 ml burette over a glass-wool plug. The regenerant was passed through
the column at a controlled rate, and the effluent was analyzed for phosphate
and aluminum as in the batch method.
ADSORPTION ISOTHERM
Samples of sodium phosphate solution (20 ml), each of a different phos-
phate concentration, were added to one-gram samples of Al-vermiculite. The
mixtures were placed in a shaker for 18 hours. The vermiculite was then
filtered, washed and air-dried. Both solids and liquids were then analyzed
for phosphate.
-5-
-------
RESUUTS AND DISCUSSION
VERMICULITE ORE CHARACTERIZATION
The following properties of the vermiculite ores were determined:
ion exchange capacity, exchangeable cations, elemental composition (by
emission spectrescopy), interlayer spacing (by x-ray diffraction), sur-
face area, and particle size distribution. The results are summarized
in Table I.
X-ray diffraction patterns of the ores revealed thatothe South Carolina
ore was a typical vermiculite having a d spacing of 1.4.3 A. The Libby-
Montana ore also showed the same spacing, but an impurity of a three-layer
structure mica was also present. The Poole ore diffraction pattern was
one of a different type of vermiculite having a d_. spacing of 1J.4. i.
The surface area (by nitrogen adsorption) of all three ores was less
than 1 m2/gm.
The mesh size of the South Carolina and the Poole ores was -8 -f 30
(No. 3 size) and the Libby ore was -k + l6 (No. 2 size). Originally we
planned to work with all No. 2 ores, but these were not available from the
South Carolina mill.
TABLE I
Vermiculite Ore Characterization
A. Exchangeable Cations and Exchange Capacity (meq/lOO gm)
Ore Ca Mg_ K Na Total
Libby 26 20 10 O.U 56
South Carolina 22 27 10 O.U 59
Poole 3 110 6 0.7 120
B. Differences in Ores Found by Emission Spectroscopy
Element Libby South Carolina Poole
Ba 0.1-1 0.1-1 0.01-0.1
B 0.01-0.1 0.01-0.1 n.d.*
Ca 0.1-1 1-5 0.01-0.1
Na 0.1-1 0.1-1 n.d*
Ti 1-5 1-5 0.1-1
* None detected
-6-
-------
The following elements were detected, but had the same value for
each ore: Al(l-5W, Cr(0.1-l#), Fe(l-5$), Mg (> 5$), Mn (0.1-1$), Ni
(0.01-0.1#), Si (> 5$) and V(0.01-0.1$.
EXFOLIATION
The sulfuric acid exfoliation resulted in partial decomposition of
the Libby and the Poole vermiculites as evidenced by 30-40$ weight losses
and large amounts of iron in the acid phase. The volume change after ex-
pansion, however, was between ^0-I.OO^> (measured by comparing the volumes
of 10-gram samples of the ore and the expanded material). Because of the
obvious attack of the vermiculite in two cases and the difficulties in-
volved in handling concentrated H2S04, work with this material was discon-
tinued .
The LiCl expansion resulted in a product that was so highly hydrated
it could not be filtered. It was extremely slimey and remained suspended
in the aqueous phase. Because of these unfavorable properties, plus the
relatively high cost of LiCl, work with Li-expanded material was also halted.
The H202 reaction resulted in excellent expansion for the Libby ore
(> 200$) but only fair expansion for the other two ores (U0$ for the South
Carolina and 25$ for the Poole).
The expansion with 3.^M NaCl was less than 20$ in all cases. However,
further tests with 6M NaCl (nearly saturated solution) followed by a boiling
water treatment, resulted in expansions of 25, 60 and 65$ for the Poole,
Libby and South Carolina ores, respectively. This material was used in
subsequent experiments with Al solutions.
PREPARATION AMD PHOSPHATE CAPACITY OF ALUMINUM VERMICULITE
The results of Al-vermiculite preparation and phosphate capacity are
summarized in Table II.
From the sodium values in Table II, it can be concluded that complete
exchange occurred in the Poole and South Carolina vermiculites, but not in
the Libby. Also, it should be noted that the relationship of the average
charge on the exchanging Al ion (obtained by dividing mmoles of Na released
by the mmoles of Al adsorbed) with pH, i.e., at pH 3.0, 3.5 and 3.8 the
charge is approximately 3.0, 2.6 and 2.1, respectively. This means that
at the lower pH values, we are exchanging mostly with hydrated Al+3, whereas
at the higher pH values, we have a hydroxylated species such as Al(OH)+2
(or a mixture of Al+3 and Al(OH)2+).
This is confirmed by the pH decrease after the high pH exchange re-
actions which can be expected if the hydrolysis product is removed from
solution, i.e.,
-7-
-------
TABLE II
Phosphate Removal vith Al-Vermiculite
Aluminum Exchange
Vermicu- ml Al(N03)3
lite Solution
A. 0.05M Al(N03)3
25 gm Tbole 1000
25 gm Pbole 1000
B. 0.1M Al(KO3)3
50 gm Poole 1000
50 gm Tbole 1000
50 gm Tbole 1000
50 gm Ibole 1000
50 gm Tbole 1000
50 gm Tbole 50
50 gm Tbole 100
50 gm Tbole 150
50 gm S.C. 1000
50 gm S.C. 1000
50 gm S.C. 1000
50 gm Libby 1000
50 gm Libby 1000
50 gm Libby 1000
C. 0.333M Al(N03)3
50 gm Poole 500
50 gm Ibole 500
50 gm Poole 500
PH
Initial
3.29
3.79
2.82
3.05
3.25
3.55
3.81
3.22
3.22
3.22
3.22
3.48
3.75
3.19
3.50
3.80
3.19
3.49
3.79
PH
Final
3.20
3.41
3.05
3.18
3.10
3.12
3.22
3.71
3.70
3.52
3.14
3.38
3.69
3.21
3.51
3.76
—
mm Al Adsorbed/
100 gm
48*. 9
43.3
42.0
47.5
49.7
62.8
9.6
18.4
25.2
21.2
21.6
25.1
14.5
13.8
18.0
57^3
57.0
mmoles Na
Released/100 go
119
114
132
127
130
130
126
26.2
55.8
81.4
59.2
62.0
59-2
36.0
36.0
38.4
127
117
103
Effective*
a Charge on Al
2.5
2.3
3.0
3.0
2.7
2.6
2.1
2.7
3.0
3.2
2.8
2.8
2.4
2.6
2.7
2.1
2.6
2.0
1.8
Phosphate Adsorption
mg P04/gm (mmoles P04/
mmoles Al)
7.8
4.9
8.2
7.2
7.4
6.0
4.7
4.3
4.8
0.18
O.ll
0.20
0.18
0.16
0.10
2.6
3.2
4.7
1.8
2.0
1.7
2.2
0.28
0.18
0.19
0.09
0.08
0.12
0.13
0.10
0.08
0.09
* mnoles Na released/mmoles Al adsorbed
-------
Al+3 + H20 ** A1(OH)+2 + H*,
removal of Al(OH)+2 would lower the pH. Thus, the Al-vermiculite pre-
pared at the higher pH values should have a greater number of hydroxyl
groups.
At low pH values, Al+3 is removed, and the pH should increase due
to decreased hydrolysis. This is confirmed by the results at initial pH
of 2.8 and 3.0.
The amount of phosphate adsorbed is surprisingly low, and is incon-
sistent with our idea of hydroxyl group replacement. In fact, better
phosphate removal is obtained with Al-vermiculites prepared at low pH
and dilute aluminum solutions, conditions which result in a high concen-
tration of unhydrolyzed Al+3 ions. This leads to the conclusion that
hydroxyl groups on the aluminum are "inactive" towards phosphate.
Further evidence of this can be seen from the following: we have
observed that some of the aluminum on vermiculite is readily exchangeable
with Na, while the remainder is not. According to the literature4, unhy-
drolyzed Al is readily exchangeable, whereas the hydrolyzed Al is firmly
held. We therefore reacted some pH 3.8 Al-Ibole vermiculite with 1M NaCl
solution in order to leave only hydroxyl aluminum on the vermiculite.
The resultant vermiculite was then reacted with phosphate as in previous
experiments. The results, summarized in Table III show a decrease in
P04/A1 ratio, indicating a lower phosphate adsorption with hydroxy-alumi-
num.
TABLE III
Preparation of, and Reaction with, Hydroxy-Al Vermiculite
(mg Al/gm) (mg Al/gm) (mg Al/gm) mg P04/gm (Moles P04/
(Initial) (Removed) (Remaining) Moles Al)
7-0 3-30 17.0 4.5 12.5 2.2 0.05*
* Initial ratio was 0.10
The original phosphate capacity of the above-mentioned vermiculite was
6.0 mg P04/gm. We can then assume that 6.0-2.2 or 3*8 mg of P04 were origi-
nally adsorbed by 4.5 mg of Al (the unhydrolzyed aluminum), for a mole ratio
of 0.24. In the experiment with partially exchanged vermiculite we obtained
a mole ratio of 0.28 (Table II) and the mole ratio obtained with the Li-ex-
panded vermiculite described in the following paragraphs was ~ 0.29. These
mole ratios seem to be approaching a maximum of 0.33, i.e., one mole of ad-
sorbed phosphate for every three moles of adsorbed aluminum. If we then
assume a maximum ratio of 0.33 and an exchange capacity of 130 meq/100 gm,
the maximum phosphate capacity attainable is 13-7 mg/gm.
-9-
-------
Other experiments were carried out with highly expanded vermiculites
to determine if greater interplatelet distances, would improve phosphate
adsorption. In one experiment, a H202 expanded Llbby ore (vol. 21056 greater
than ore) was treated with 0.1M Al(N03)3 at pH 3.10. The phosphate capacity
of this material was 2.3 mg P04/gm, compared to 1.7 mg/gm for an equivalent
Ha-expanded vermiculite (volume 60$ greater than the ore), a 35$ increase
in capacity.
In another experiment, a Li-expanded Poole vermiculite (as a slurry)
was treated with 0.1M Al(H03)3 solution at pH 3.15- (The Li-expanded vermic-
ulites are highly expanded and cannot be filtered). The resultant Al-vermic-
ulite was then reacted with P04 according to our standard procedure, and
the capacity was 12.2 mg/gm (compared to f.k mg/gm for an equivalent Ka-
expanded material), a 65$ increase in capacity. Thus both the peroxide and
Li-expanded vermiculite showed significant increases in phosphate adsorption,
the latter value approaching the predicted value of 13.7 mg/gm mentioned
above.
Still other experiments were carried out with thermally expanded vermic-
ulites. It is known that heating vermiculite ores to about 9^0 C will re-
sult in expansion of the ore into the well-known, highly exfoliated commercial
packing material. However, the high temperature also causes a loss in ion
exchange capacity of the vermiculite. In our experiments, we hoped to expand
the vermiculite at lower temperatures thereby obtaining significant expansion
but at a minimum loss in exchange capacity. The results are summarized in
Table IV.
TABLE IV
Preparation and Adsorption - Thermally Exfoliated Vermiculite
Exfoliation
Temperature, °C
25
200
250
300
350
kOO
1*50
500
600
700
Volume
of 10 gm (cc)
10.0
10.0
10.0
10.0
10.5
11.0
13.5
15.5
2k
26
Ca and Mg Released
(meg/LOO gm)
85
95
109
113
109
105
101
96
79
30
Phosphate
Adsorbed (mg/gm)
7.3
7.1*
6.7
7.8
8.9
8.9
8.3
8.6
4.8
-10-
-------
As expected, vermiculite expansion increases with temperature. How-
ever, the exchange capacity first increases with temperature as the plate-
lets expand, and then decreases at high temperatures as the mobile cations
become fixed to the surface. Phosphate adsorption on the other hand, nearly
follows the exchange capacity, and is at about the same level as the sodium
expanded vermiculites.
In efforts to improve phosphate adsorption the following experiments
were carried out:
(l) Reaction with 0.1M Al(N03)3 at 6o°C
Na-expanded Pbole vermiculite was reacted with 0.1M Al(N03)s solution
at 60°C instead of room temperature. All other conditions were identical
to previous experiments. The results show no significant change over pre-
vious experiments, i.e., capacity = 7.^ mg/gm.
(2) Reaction with P04 for 70 hours
In order to be certain that P04 removal was complete, a sample of
pH 3.2 Al-Poole vermiculite was reacted with phosphate solution for 70
hours instead of the usual 18 hours. The phosphate adsorption was 8.3
mg/gm, compared to J.h- found previously, only slightly higher than the l6
hour test results we are now using.
(3) Use of Powdered Vermiculite
A sample of pH 3.2 treated Al-vermiculite was ground to a fine powder
and then reacted with the standard phosphate solution. The phosphate capacity
was 6.8 mg/gm (compared to 7.^), showing that an increase in external sur-
face area is not significant, and that internal surface area may be more
important.
Use of a Mixed Al-Fe Vermiculite
A mixed Al-Fe(lll) vermiculite was prepared by reacting Na-expanded
Poole vermiculite with a 0.1M mixed nitrate solution containing 80$ Al and
20$ Fe. The mixed vermiculite was then reacted with phosphate solution and
its capacity was 6.7 rag/gnij about the same as Al and Fe vermiculites (see
below).
USE OF CATIONS OTHER THAN Al
The following cations were exchanged with Na-expanded Poole vermicu-
lite in a manner similar to Al: Fe(lll), La(lll), Ca(ll), Cu(ll), Fe(ll).
These vermiculites were then tested for P04 adsorption. The results, sum-
marized in Table V, show no significant improvement over the Al-vermiculite
results.
-11-
-------
TABLE V
The Use of Cations Other than Al
Cation pH.
Pe(IIl) 1.75
Fe(lll) 2.10
La(lll) 2.98
Ca(ll) 5.31
Cu(ll) U.32
Fe(ll) 3.^5
pHf mmoles M/
100 gm Vm
1.79
2.02
3.69
5.21
3.7^
3.12
U7.6
50.6
Vf.9
mmoles Wa
Released/
100 gm
127
127
127
117
127
Avg. Chg. mg P04/
on M gm Vm
2.7 5.9
2.5 6.U
3^0
-"- 5!6
mmoles P04/
mmoles M
0.13
0.13
0.07
—"f
* P04 removal may be due to the formation of
M represents metal
Vm represents vermiculite
In the copper experiment, a blue precipitate was found in the 'flask
after reaction of the copper vermiculite and phosphate, indicating removal of
copper from the vermiculite followed by precipitation. A precipitate was
also observed with Ca -vermiculite. However, no precipitates were observed
with the trivalent cations, confirming that these ions are held much more
strongly than the divalent cations.
REGENERATION
Several regeneration schemes were attempted by both batch and column
techniques.
Batch experiments with various concentrations of ammonia solutions
and phosphate-containing vermiculite were set up. The results are summarized
in Table VI.
-12-
-------
TABLE VI
Batch Regeneration with
Vm* Taken NH4OH Initial
(gm) Cone.(M) ml P0^(mg/gm)
5.00
5.00
5.00
1.00
1.00
1.00
1.00
1.00
1.5
0.75
0.50
OA5
0.375
0.30
0.15
0.075
100
100
100
20
20
20
20
20
U.9
P04 Removed
(mg/gm)
U.7
U.8
5.1
U.7
3.2
0.7
Initial Al**
(mg/gm)
11.7
11.7
11.7
11.7
11.7
11.7
11.7
11.7
Al Removed
(mg/gm)
2.8
2.3
2.2
1.7
1.7
1.3
0.6
* Vermiculite
** Exchangeable Aluminum
It is apparent that phosphate can be removed from vermiculite vith ammonia;
however, an appreciable quantity of aluminum is also lost.
Some of the vermiculite that had been ammonia treated vas recycled with
our standard phosphate solution. One sample had been treated with 0.5M and
the other with 0.3M NIUOH. The recycle phosphate capacity was 1.2 and l.lj-
mg/gm respectively, indicating that the ammonia treatment had left the vermicu-
lite fairly inert to phosphate. This can be explained by the lost aluminum,
and the probable conversion of the aluminum groups to hydroxy-aluminums.
Column regeneration with dilute ammonia solution was also tried, but
these results appeared to be no more promising than the batch results, so
all work with NHs regeneration was terminated.
HC1
In order to avoid formation of hydroxy-aluminum groups, regeneration
was tried with HC1. The results of preliminary batch experiments are sum-
marized in Table VII.
TABLE VTI
Batch Regeneration with HC1
Vm* Taken
HC1 Initial
Cone. (M) ml P04(mg/gm)
1.00
1.00
1.00
1.00
0.032
0.020
0.010
0.002
20
20
20
20
>4 Removed
(mg/gm)
3.2
1.9
0.9
0.5
Initial** Al Removed
Al(mg/gm) (mg/gm)
11.7
11.7
11.7
11.7
1.1
* Vermiculite
** Exchangeable Aluminum
-13-
-------
It is apparent from these results that either larger quantities or
more concentrated HC1 will be needed to obtain sufficient phosphate re-
moval. Also, the loss of aluminum will be a problem.
Column regeneration was tried with 0.032, 0.05 and 0.08M HC1, and
the results showed again that more concentrated HC1 would be required and
that appreciable Al was being lost.
Various Salts
To avoid extreme acidity (or alkalinity), and consequently the loss
of aluminum, batch experiments were set up with a variety of salts. Our
aim was to remove phosphate by replacement with the salt anion. The re-
sults are summarized in Table VIII.
TABLE VIII
Regeneration with Various Salts
Salt Added Cone.* ml pH gm Vm Initial P04 P04 Removed
(M) (mg/gm) (mg/gm)
NaCl 1.0 100 5.7 1.00 5.8 0.3
NaHC03 1.0 100 8.0 1.00 5.8 0.9
A1C13 0.33 100 3.0 1.00 5.8 1.5
Na2S04 1.0 100 8.k 1.00 5.8 0.6
A12(S04)3 0.33 100 2.9 1.00 5.8 3.2
* All 1M in anion concentration.
The most promising of the salts was A12(S04)3, and this may be due
in part to the low pH. However, the sulfate salts were all more promising
than the corresponding chlorides, probably because of complexing between
sulfate and aluminum. A possible advantage of using an aluminum salt for
regeneration would be the prevention of aluminum losses by the vermiculite.
The experiment with A12(S04)3 regeneration was repeated at lower pH
values. (H2S04 was used for pH adjustment). These results, summarized
in Table IX, show improved phosphate removal at lower pH.
TABLE IX
Batch Regeneration with Alg(S04)3
Vm Taken pH Initial P04 P04 Removed
(gup (mg/gm) (mg/gm)
1.00 2.5 5.8 U.2
1.00 2.0 5.8 5.2
1.00 1.5 5.8 5.^
-------
The vermiculite sample that had been treated at pH 1.5 was water-
washed and recycled with our pH 7 phosphate solution. Its capacity was
8.1 mg/gm, which agrees with previous values for an Al-vermiculite. This
test showed that A12(S04)3 treated vermiculite can be recycled for further
phosphate removal.
Alg(S04)3
Experiments were first tried with H2S04 alone. As with HC1, these
tests showed significant loss of aluminum from the vermiculite. Column
experiments were then set up with solutions of A12(S04)3 plus HfeS04 as
the regenerant. The phosphate -containing vermiculite was placed in a 25-ml
burette. Flow rates were controlled with fine control metering valves.
The effluent was collected in 5 -ml portions and each was analyzed for P04.
After passage of the final portion of regenerant, the vermiculite was water-
washed, air-dried and tested for phosphate removal by our standard test
using pH 7 phosphate solution.
The results, summarized in Table X, show that a pH of about 0.9 is
needed for complete regeneration. In addition, the material can be recycled
for further phosphate removal. In fact, the recycle capacity is greater
than the initial capacity. This result is not yet fully understood, but
our experiments show that it is not due to insufficient removal of excess
aluminum. The_increase seems to be a result of either the low pH or the
presence of S04 rather than N03~.
WORKING CAPACITY OF Al-VERMICULCTE
Up to this point, all P04 capacities were determined by a batch pro-
cedure using relatively concentrated P04 solutions. It was therefore
necessary to determine the actual working capacity of Al-vermiculite with
dilute P04 solutions in a column operation,
A column containing ^ grams of Al-vermiculite was prepared, and a
25.6 ppm P04 solution at pH 7 was passed through at a rate of 0.76 cc/min.
The effluent was collected and analyzed for phosphate. The results showed
that only 26.7$ of the phosphate was removed from the first 100 ml of
effluent, and 16.9$ from the second 100 ml. The run was stopped, and the
column washed with H2S04-A12(S04)3 regenerant. The dilute phosphate so-
lution was again passed through the column, but only lU.4$ of the phos-
phate was removed from the first 100 ml of effluent.
To be certain that the problem was not one of bicarbonate competition,
the solution was boiled to remove C02. The cooled solution was passed
through a fresh column, and P04 removal from the first 100 ml of effluent
was 33-2$, just slightly better than the previous run.
-15-
-------
TABLE X
Column Regeneration with HgS04 -
Cone. H2S04
(ml/100)
0.96
0.86
0.80
0.66
0.58
0.57
0.53
A12(S04)3'18 HgC
(gm/100)
0.
0.
0.
0.
0.
0.
0.
833
833
833
833
4l6
208
416
) pH
0.7
0.75
0.8
0.9
0.9
0.9
0.95
Vm*
(gm)
4.0
4.0
4.0
4.0
4.0
4.0
4.0
Bed Ht.
(cm)
7.7
7.8
7.8
7.8
7.7
7.7
7.8
Bed Vol.
(cc)
4.7
4.7
4.7
4.7
4.7
4.7
4.7
Flow Rate
(cc/min)
0.3
0.3
0.3
0.15
0.15
0.15
0.15
Initial
P04(mg)
21.8
21.8
21.8
21.8
21.8
21.8
21.8
Total P04 Total Vol. P04 Ad-
Removed Regenerant sorbed
(mg) (cc) on Re-
cycle
(mg/gm)
21.9
22.3
22.0
21.4
20.8**
20.9
21.1
60
60
60
50
50
50
50
9-7
9.6
9.5
10.3
8.5
8.4
9.0
* Vermlculite
** Initial flow rate for first 15 cc was about 0.23 ml/min, and is probably the cause of this low value
-------
Batch experiments were set up to determine the phosphate capacity
of Al-vermiculite in dilute solution. A one-gram sample of Al-vermiculite
vas reacted with 500 ml of the 25 ppm P04 solution for 16 hours. The phos-
phate capacity was Q.k mg/gm, indicating that the problem in the column
experiments was one of low capacity at low phosphate concentrations. The
experiment was repeated at 100°C, and no improvement in capacity was ob-
tained, showing that the problem was not one of slow kinetics.
Finally, an adsorption isotherm was obtained by reacting one-gram
samples of Al-vermiculite with a series of phosphate solutions. The re-
sults, plotted in Figure i, show a typical adsorption isotherm, which
results in a straight line when plotted as 1/c vs 1/x (Langmuir5 - Figure 2)
or log c vs log x (Freundlich6 - Figure 3)> where c = solution concentration
and x = solid concentration. Unfortunately, the adsorption at low concen*-
trations drops off very sharply, i.e., the equilibrium constant for the
reaction:
Al-Vm + P04~3 2 Al-Vm-P04
is relatively low, and the material is therefore impractical as a P04 ad-
sorbent.
ECONOMIC EVALUATION
During the course of the investigation, a cost estimate was made of
a phosphate removal process using Al-vermiculite. The evaluation was
made on the following initial parameters which were thought to be applicable
at that time:
• A P04 capacity of 10 mg/gm
• An attrition rate of 3$ per cycle
• Ammonia regeneration requiring k-J moles of NHa per one mole
of P04
It later became apparent that the material could not be regenerated
with ammonia. However, the use of 1^304, plus Al2(S04)s, might be more
economical.
The cost estimate, plus an earlier one made prior to the study are
included in Appendices A and E.
-IT-
-------
500 1POO
Phosphate liq. (ppm)
100
Figure 1
Adsorption Isotherm, Al-Vermiculite + P04
-------
1 i I I I
L—J 1 1
150
VcXIO4
Figure 2
Langmuir Plot
-------
10.0
S i.o
0.1
10
I I
100 1,000
Phosphate liq. (ppm)
Figure 3
Freundlich Plot
-20-
-------
REFERENCES
1. Yee, W. C., J. Am. Water Works Assoc., 58, 239
2. Neufeld, R. D., and Thodos, G., Environ. Scl. Technol. 3., 667 (1969).
3. Misslngham, G. A., J. Am. Water Works Assoc., 59, 183 (1967).
k. Rich, C. I., Soil Sci. Soc. Proceed. 2^ 2k (1960).
5. Langmuir, I., J. Am. Chem. Socf, 40, 1361 (1918).
6. Freundlich, H., Colloid and Capillary Chemistry, Methuen, London, 1926.
-21-
-------
APPENDIX A - ECONOMIC EVALUATIONS
COST ESTIMATE - PQ4 REMOVAL WITH VERMICULTTE 6/11/69
by R. J. Bettacchi
SUMMARY
You have requested a revision in the cost estimate for phosphate
removal from waste water using an aluminum treated vermiculite.
A base case was established based on the information in the origi-
nal estimate (DeCicco 2/28/68). The data and assumptions for the base
case are appended (Appendix B). Using these data and assumptions, the
cost of treating 1000 gallons of water (25 ppm K>4 influent - 2.5 ppm
effluent) was determined and the results were analyzed to determine the
important variables.
The results of the analysis (Basis: 1,000 gallons of water treated)
are listed below:
• For the base case the cost to remove the phosphate is l4^/thousand
gallon.
• Nearly 75$ of the cost is contributed by the raw materials (vermicu-
lite make up 30^ ammonia regenerant ^5$)»
• The cost of the vermiculite make up is determined by -
1. The assumed percentage loss per cycle,
2. the capacity of the vermiculite,
5. the price of the vermiculite.
• The regenerant cost is the most critical variable and this cost is
dependent upon -
1. The lbs/ft3 of HH3 required to regenerate,
2. the capacity of the adsorber.
• The other operating costs and the fixed capital investment are de-
pendent on -
1. The vermiculite bed depth,
2. the volumetric flow rate through the bed (GPM/ft3).
• All raw material costs are directly proportional to the amount of
P04 removed.
A-l
-------
We can conclude from these results that if further work is to "be
done on this project, it must be directed towards determining -
1. The operational capacity of the vermiculite,
2. the operational regeneration level,
3. an accurate estimate of the vermiculite loss per cycle.
DISCUSSION
Capital Cost
In order to determine the cost of removing phosphate from wastewater,
with an aluminum treated vermiculite, a treatment system was designed
based on the data and assumptions in Appendix B. The design calculations
are shown in Appendix C and indicate that the system (under base case condi-
tions) will require nine columns, each 11 ft diameter with a bed depth of
5 ft of vermiculite. The columns will operate in a 15-hour cycle (ll to
exhaust and ^ to regenerate) with an average flow of 2.kk GHVI/ft3.
The capital was estimated for this system and is summarized in Appen-
dix D. The capital estimate is dependent on the assumptions of volumetric
flow rate and bed depth as shown in the following Table A-l.
TABLE A-l
Capital Cost ($M)
Flow Rate
1.73 2.44 4^1
908 685 ^92
688 534 361
% 6 604 470 340
&
•8
A-2
-------
Although the capital estimate will vary between 3^0 and 900 $M, these
variables are not critical as the operating cost is relatively unchanged
with these fluctuations.
Operating Cost
Based on the information known about vermiculite and on the assumptions
in Appendix B, an operating cost summary was calculated for the base case
and is shown in Table A-2. As the results in this table show, the greatest
cost is contributed by the raw material components with the ammonia regen-
erant contributing about 45$ of the cost and the vermiculite make up con-
tributing 3056. Examination will show that a 100$ increase in the fixed
capital would result in only a 7»5$ increase in the operating cost.
TABLE A-2
Operating Cost Summary Base Case
Direct Operating Cost //Thousand Gal.
Raw Materials 10.3
Vermiculite make up 4.1
Ammonia @ 4{#/lb. delivered 6.2
Direct Labor 1 man-shift/12 units @ if shifts/day
$30/day 0.9
Indirect Labor 40$ D.L. 0.4
Maintenance 1% of Capital 0.2
Utilities 1^/kwh, 7430 kwh 0.7
Supplies 0.1
TOTAL DIRECT OPERATING COST 12.6
Overhead Operating Cost
G & A 30$ of D.L. 0.3
Debt Service (Amortize @ 5$ for 25 years) 1.1
Taxes and Insurance 2$ of Capital 0.4
TOTAL OVERHEAD COST TTo7
TOTAL OPERATING COST l4.4
A-3
-------
SENSITIVITY ANALYSES
Sensitivity analyses were performed to determine what assumptions and
variables affected the raw material cost. The cost of the vermiculite
make up is summarized in Figure A-l. Examination of this figure shows that
the cost of the vermiculite make up is highly sensitive to the capacity of
the vermiculite and to the assumption of what the loss/cycle will be. The
cost is less sensitive to the price of vermiculite (over the range of prices
the operating cost varies 2f#/thousand gal.).
A second sensitivity analysis was performed to show the effect of the
regeneration level (which is the largest cost component) on the cost. This
is shown in Figure A-2. As this figure shows, the regeneration cost is de-
pendent on the NH3 usage per cycle and on the capacity of the vermiculite.
GENERAL
The most critical variable in the system is the capacity of the vermicu-
lite. Both raw material cost components are inversely proportional to the
capacity. Since the regeneration costs and the vermiculite make up costs
are calculated per cycle, and since the volume of water treated per cycle
is dependent on the capacity, the cost/thousand gallons is dependent on the
capacity. This is important since the Base Case capacity is the total
capacity of the vermiculite which may be greater than the column or opera-
tional capacity.
-------
2.0 3.0 5.0
VERMICULITE LOSS
(%/CYCLE)
2.0 3.0 5.0
COST OF LOSS
(*/M GAL)
Figure A-l
Cost of Vermicullte Make-up
A-5
-------
14
12
~ 8
ro
0.25
Ib P04/ft;
0.5
Ib P04/ft
Ib P04/ft
3
;S
12345
NH* USAGE (Ib/ft3-cycle)
Pigijre A-2
Cost of Regeneration
A-6
-------
A.
APPENDIX B
Assumptions and Data for Calculations
Plant Information (Original Estimate)
1. Capacity
2. P04 Influent
3. P04 Effluent
B. Vermiculite Information ( Your Memo of 5/27/69)
1. Density
2. Capacity
3. Regeneration Level
C. Design Assumptions
1. Regeneration Time
2, Tower Diameter
D. Assumed Variables'
Variable
1. Attrition Rate
2. Vermiculite Price
3. Bed Depth
0-5$
per cycle
10MM GPD
25 ppm
2.5 ppm
0.83 gm/cc
0.5 #P04/ft3
.15 #NH3/ft3
k hrs
11 ft
Base Case
per cycle
2-6 ft
k. Volumetric Flow Rate 1.5 -5.0 GPM/ft3
E. Rationale or Source for Assumed Variables
U ft
2.kk GPM/ft 3
1. Attrition Rate - our conversation 6/6/69.
2. Price - based on DeCicco's memo of 2/28/68 giving price
@ 8.7^/lb and assumptions that a different method
of manufacture could lower the price.
3. Bed Depth - based on knowledge that almost all ion exchange beds
are in range of 2-6 ft with most at k ft. Depth is
determined by pressure drop and kinetic considerations.
k. Flow Rate - our conversation of 6/6/69.
B-l
-------
APPENDIX C
System Design
Plow Bate = 10Y GPD = 6950 GPM
Volume of Verraiculite
= 6950 GPM 7 2 GPM/ft3 = 3^75 fta
Volume per Tower
= (11 ft diameter)2 x TT x k ft deep = 390 ft3/tower
j.
No. of Towers
= 3^75 ft3 1 390 ft3/tower = 9
Flow Rate per Tower
= 6950 GPM '- 9 towers = 773 GPM/tower
P04 Removal per Tower - hr
• 773 GPM/tower x 60 min/hr x 8.3^ Ib/gal x 22.5 lbs/106 Ib
8.7 lb P04
tower - hr
P04 Removal per Tower-Cycle
= 0.5 lb/ft3 x 390 fts = 195 its/tower cycle
Cycle Time
= 195 Ibs/cycle 7 13 Ibs/hr = 15 hrs/cycle
Flow Rate
= 2 GPM/ft3 x 15 hrs = 2.U4 GPM ft3
C-l
-------
APPENDIX D
Capital Cost Estimate
Vermiculite Towers, 9 @ $27,000 (Chem. Engr. 12/6/65)
390 ft3 tower, rubber lined carbon steel
Slowdown Tank for Regenerant Effluent
95M gal, 5 days storage
Water Peed Pumps, 4 @ $3500
2,000 GPM, iron and bronze
Motors for Water Pumps, 4 @ $2800
100 HP enclosed
Transfer Pump 400 GPM
Iron and bronze
Motor for Pump, 15 HP
Total Installed Equipment Cost (lEC)
Instrumentation - (ChemL Engr. 12/6/65)
Piping - 5% of IEC
E-lectrical - 5$ of IEC
Service Facilities and Buildings, 10$ of IEC
Total Physical Cost
Engineering and Construction, 1056 TPC
Direct Plant Cost
Contingency, 10$ of DPC
Fixed Capital
Vermiculite Charge 3475 ft3 @ $3.64/ft3
Total Fixed Capital
23,200
14,000
11,200
900
900
$293, 000
77,000
1^,700
lit-, 700
29,300
Working Capital,
Total Capital
Fixed
42,900
$471,800
47,200
$519,000
12,600
$531,600
20,000
$551,600
D-l
-------
APPENDIX E
Cost Estimate of Phosphate Removal from Wastewater
with an Aluminum Treated Vermiculite 2/28/68
by Robert DeCicco
I. Summary and Conclusions
The cost of a phosphate removal process utilizing aluminum treated
vermiculite as the exchange resin has "been estimated. The purpose of the
estimate was to provide research guidance by establishing the economics
of a "speculative" water treating system and to identify the factors con-
trolling the economics of the system. The results of the cost estimate
may be summarized as follows:
Step 1 - Conversion of vermiculite ore to aluminum vermiculite
Production - 4,000 Ib/day
Total Operating Cost - 8.7{£/lb
Capital Investment - $99,600
Fixed - $93,600
Working - 6,000
Step 2 - Treatment of water with aluminum vermiculite to remove
phosphate and regeneration of vermiculite with ammonia.
Plant Capacity - 10MM gal/day
Total Operating Cost - 10.7j#/M gallons
Capital Investment - $910,000
Fixed - $880,000
Working - 30,000
The following major conclusions were drawn from the analysis:
A. Ore Processing
1. With the exception of direct labor, the raw material cost of the
vermiculite ore is the major cost component of the ore processing step.
This is primarily the result of the assumed 25$ ore loss. The high direct
labor cost (j^/lb basis) is due to the relatively low production volume
and long retention times which require two-shifts-per-day operation.
E-l
-------
2. The large volume of water necessary for washout of the A12(S04)3
and NaCl (2.k gal/lb vermiculite for each step) requires the reaction
vessel be of sizeable proportions. Reduction of the water requirement
would reduce the volume of the vessel and thereby lower the fixed capital.
Also, the wash water effluent (~ 20,000 gal/day) may be a disposal problem
in itself if it cannot be discharged into the river. Should be be the case,
an evaporator system would be needed to purify the water such that it can
be either recycled or dumped.
B. Water Treatment and Vermiculite Regeneration
1. The cost of make-up vermiculite accounts for 30$ of the total
operating cost. The vermiculite loss is a function of the following three
interdependent variables:
a. Flow rate of water per unit volume of vermiculite.
This variable determines the volume of vermiculite charged to
all towers - V^,.
b. Exchange capacity of vermiculite. This variable determines the
number of cycles per day - Cd.
c. Vermiculite loss per cycle - L-.
The loss per day is then: Ld = V^ x Cd x LQ (ft3 vermiculite). Cost of
make-up vermiculite: C
-------
necessary to know either the time for regeneration or the time for adsorption.
This analysis assumed the cycle time was divided equally between the two
steps. Consequently, for every adsorption tower there must be a tower for
regeneration.* As the cost of the towers represent the most significant
part of the fixed capital, the cycle time split must be determined if a
more accurate cost estimate is desired.
U. The by-products of ammonia formed during regeneration of the ver-
miculite may have some resale value. No value has been considered in this
analysis.
The process specifications upon which the co.st estimate was based are
provided in Section II of this report. The detailed economics are shown
in Section III. A breakdown of fixed capital investment for both ore pro-
cessing and water treatment-regeneration is provided in Appendix F.
II. Process Specifications - 10MM GPD Wastewater Treatment for Phosphate
Removal •
A. Ore Processing
Ore Processing Loss - 25$
NaCl Treatment (20$ solution) - 9.2 Ib NaCl/100 Ib ore
A12(S04)3 Treatment (16.7$ solution) - 20 Ib A12(S04)3/100 Ib
vermiculite
Washout of Wad - 2^0 gal/100 Ib vermiculite
Washout of A12(S04)3 - 2^0 gal/100 Ib vermiculite
Retention Time - 3 hours mixing for each step
Temperature - ambient for each step
B. Phosphate Adsorption and Regeneration of Vermiculite
[P04~3] in influent water - 25 ppm
[P04" ] in effluent water - 2.5 ppm
Exchange capacity of vermiculite - U.15 Ib P04/ft3- cycle
Bulk density of vermiculite - 1.33 g/cc
Water Flow Rate/volume of vermiculite - 1.17 gpm/ft3
Water Flow Rate/x sectional area of towers - 11.7 gpm/ft3
* The second tower would be used for absorption of phosphate while the
charge in the first tower is regenerated.
E-3
-------
Vermiculite Bed Height/Tower 10 ft
Tower Diameter 10 ft
Total Volume of Vermiculite Charged 11,870 ft3
Volume of Vermiculite/Tower 7^2 ft3
Number of Towers 16
Cycles/Day/Tower 13.2 days
Cycle Time/Tower 13.2 days
Adsorption and Draining -6.6 days
Regeneration and Wash - 6.6 days
Vermiculite Loss/Cycle 5%
Vermiculite Loss/Tower/Day 0.3856
Total Vermiculite Loss/Day 1*-5.1 ft3
Vermiculite Turnover 263 days
Water/Vermiculite Contact Time 6.4 minutes
Volume of Vermiculite Regenerated/Day 899 ft3
Ammonia Regenerate Required (3$ solution) - k.15 lb NHa/ft3 Vermiculite
Wash Water Required - 100 gal/ft3 vermiculite
III. Detailed Economics
A. Ore Processing - Operating Cost for ^M Ib/day Production of
Vermiculite
Direct Operating Cost ^/Ib Vermiculite
Raw Materials 2.1
Ore @ 0.8^/lb f.o.b. - 1.06^/lb verm.
NaCl @ 1.25^/lb delivered - 0.15^/lb verm.
A12(S04)3 @ 3.8l^/lb delivered - 0.76<#/lb verm.
Water at 25^/M gallons - O.ll^/lb verm.
-------
Direct Labor - h men/day & $30/man-day 3- 0
Payroll Extras - 15$ of Direct Labor 0.5
Maintenance - 5$ of Direct Capital/Year 0.3
Utilities - 1790 kwh @ 1
-------
Payroll Extras - 15$ of Direct Labor 0.3
Maintenance - 156 of Direct Capital/Year 0.2
Utilities - 7^30 kwh @ 1^/kwh 0.7
Supplies - 50$ of Maintenance 0.1
Total Direct Operating Cost 8.0
Overhead Operating Cost
General and Administrative Expense - 30$
of Direct Labor 0.5
Debt Service (Amortization of Direct
Capital @ k-l/2% over 25 years) 1.6
Taxes and Insurance - 2$ of Direct Capital/
yr 0.5
Interest on Working Capital @ 4-l/2$/yr 0.1
Total Overhead Operating
Cost 2.7
Total Operating Cost 10.7
Direct Capital - $880,000
Working Capital - 30,000
Total Capital - $910,000
E-6
-------
APPENDIX F
Breakdown of Total Capital Requirements
A. Ore Processing - Design Capacity = 5M lb/day of Vermiculite
Installed Equipment
Item Cost
Conveyor for Vermiculite Ore - 2V belt x 50' L $8,800
Water Pump - 400 GPM, Iron and Bronze 1,120
Motor for Water Pump - 15 HP Enclosed 950
Vermiculite Slurry Pump (Transfer of Vermiculite
to Storage) 30 GPM, Iron and Bronze 600
Motor for Slurry Pump - 5 HP Enclosed 360
Vermiculite Slurry Pump (Transfer of Vermiculite to
Towers) 30 GPM, Iron and Bronze 600
Motor for Slurry Pump - 5 HP Enclosed 360
Reaction and Wash Vessel - 12M gal, agitated,
carbon steel 16,000
Storage Vessel for Vermiculite - 2M gal for two
days storage, agitated, carbon steel 7,000
Total IEC $35,790
Delivery of Equipment - 2% of Purchased Equipment Cost 570
Total IEC Delivered $36,360
Instrumentation - 18$ of IEC 6,550
Piping - 35$ of IEC 12,730
Electrical - 20$ of IEC 7,270
Service Facilities and Buildings - kO% of IEC 1*1,
Total Physical Cost $77,350
Engineering and Construction - 10$ TPC 7,7^0
Direct Plant Cost $65,090
F-l
-------
Installed Equipment
Item (cont'd) Cost
Contingency - 10% of DPC 8,510
Total Direct Fixed Capital $93,600
Working Capital
Raw Material Inventory - 30 days @ EMC + Freight $3,360
Work-in-Process Inventory - 1 day® Direct
Operating Cost 290
Finished Vermiculite - 2 days @ Total
Operating Cost 690
Cash - Payroll for one week 1;220
Total $5,560
Say 6,000
Total Capital Investment $99,600*
Contains no provision for land site.
F-2
-------
B. Water Treatment and Vermiculite Regeneration - 10MM GPP
Installed Equipment
Item Cost
Vermiculite Towers, 16 required @ $29,500
7^2 ft3 bed volume, rubber lined carbon steel $Jf72,000
Slowdown Tank for Regenerate Effluent
95M gal for 5 days storage, carbon steel 23,200
Water Feed Pumps, k pumps required @ $3^500
2,000 GEM, Iron and Bronze 14,000
Motors for Water Pumps - 4 required @ $2,800
100 HP Enclosed 11,200
Transfer Pump for Regenerate Effluent
UOO GPM Iron and Bronze 900
Motor for Transfer Pump, 15 HP Enclosed 950
Total IEC $522,250
Delivery of Equipment - 2% of Purchased Equipment
Cost 8,350
Total IEC Delivered $530,600
Instrumentation - 5$ of IEC 26,500
Piping - 5$ of IEC 26,500
Electrical - 5$ of IEC 26,500
Service Facilities and Buildings - 10$ IEC j?3, OOP
Total Physical Cost $663,100
Engineering and Construction - 10$ of TPC 66,300
Direct Plant Cost $729,^00
F-3
-------
Installed Equipment
Item (cont'd) Cost
Contingency - 10$ of DPC $72,900
Total Fixed Capital $802,000
Vermiculite Charge - 11,870 ft3 @ $7.18 ft3 85,250
Total Direct Fixed Capital $887,530
Say $880,000
Working Capital 30,OOP
Total Capital Investment $910,000*
* Contains no provision for land site.
F-U
* U.S. GOVERNMENT PRDrTDKI OFFICE ; 1970 O - 40J-1JO
------- |