EPA-R2-73-150
JANUARY 1973 Environmental Protection Technology Series
Treatment of
Ferrous Acid Mine Drainage
with Activated Carbon
Office of Research and Monitoring
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and
Monitoring, Environmental Protection Agency, have
been grouped into five series. These five broad
categories were established to facilitate further
development and application of environmental
technology. Elimination of traditional grouping
was consciously planned to foster technology
transfer and a maximum interface in related
fields. The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL
PROTECTION TECHNOLOGY series. This series
describes research performed to develop and
demonstrate instrumentation, equipment and
methodology to repair or prevent environmental
degradation from point and non-point sources of
pollution. This work provides the new or improved
technology required for the control and treatment
of pollution sources to meet environmental quality
standards.
-------
EPA-R2-73-150
January 1973
TREATMENT OF FERROUS ACID MINE DRAINAGE
WITH ACTIVATED CARBON
By
Charles T. Ford
James F. Boyer, Jr.
Grant No. 1^010 GYH
Project Officer
Eugene F. Harris
Mine Drainage Pollution Control Activities
National Environmental Research Center
Environmental Protection Agency
Cincinnati, Ohio ^5268
Prepared for
OFFICE OF RESEARCH AND MONITORING
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D. C. 20^0
For sale by the Superintendent of Documelitsx U.S. Government Printing Office, Washington, D.C, 20402
Price $2.10 domestic-postpaid or $1.75 QPO Bookstore
-------
EPA Review Notice
This Deport has been reviewed by the Environmental Protection Agency and
approved for publication. Approval does not signify that the contents
necessarily reflect the views and policies of the Environmental Protec-
tion Agency, nor does mention of trace names or commercial products
constitute endorsement or recommendation for use.
ii
-------
ABSTRACT
Laboratory studies were conducted with activated carbon as a catalyst
for oxidation of ferrous iron in coal mine water. Batch tests and
continuous flow tests were conducted to delineate the process variables
influencing the catalytic oxidation and to determine the number and
types of coal mine water to which this process may be successfully
applied.
The following variables influence the removal of iron with activated
carbon: (a) amount and particle size of the carbon; (b) pH, flow rate,
concentration of iron, temperature, and total ionic strength of the
water; and (c) aeration rate. Adsorption as well as oxidation are the
mechanisms involved in iron removal by this process.
An evaluation of this process indicated technical feasibility which would
permit acid mine drainage neutralization using an inexpensive reagent,
such as Limestone. The major disadvantage is the cost of the activated
carbons since they are rendered inactive after relatively short use by ,
apparently irreversible adsorption of iron. This cost appears to be
sufficiently high to prohibit the use of this process for treating coal
mine drainage.
This report was submitted by Bituminous Coal Research, Incorporated in
fulfillment of Project Number I^OIOGYH under the joint sponsorship of
the Environmental Protection Agency and Bituminous Coal Research, In-
corporated.
iii
-------
CONTENTS
Section Page
I Conclusions •••• * 1
II Recommendations * 3
III Introduction. • • ••• 5
Objectives I 5
Nature and Scope of the Problem 5
Approach to the Problem and Research Procedure 6
IV Experimental 9
Analytical Procedures 9
General Procedures *. 9
Materials 11
Effect of Variables 13
Effect of Bacteria * 19
V Results and Discussion 23
Effect of Variables 23
Effect of Bacteria 88
VI Summary Evaluation of Iron Removal with Activated
Carbon *.... 113
VII Acknowledgments «»..* 119
VIII References 121
-------
FIGURES
Page
1 Apparatus for Batch Experiments on Catalytic Oxidation of
Ferrous Iron 3X)
2 Apparatus for Continuous Flow Experiments on Catalytic
Oxidation of Ferrous Iron
3 Comparison of Rates of Adsorption of Fe2* and Fe3* on
Activated Carton ........................................... 30
h Effect of pH on Rate of Adsorption of Fe2* on Activated
Carbon [[[ 33
5 Effect of pH on Fe2* Adsorbed per Hour on Activated Carbon ... 3^
6 Effect of pH on Rate of Adsorption of Fe3* on Activated
Carbon [[[ 37
7 Concentration of Fe2* from Reduction of Fe3* during Adsorp-
tion Tests ................................................ • 38
8 Effect of pH on Concentration of Fe2* from Reduction of Fe3*
during Adsorption Tests .................................... Uo
9 Effect of pH on Removal of Fe3* .............................. 6k
10 Effect of pH on Removal of Fej ............................... 65
11 Comparison of Effects of Various Process Variables on Removal
of Iron [[[ 67
12 Effect of Aeration, No Aeration, Nitrogen, Sparging and Oxy-
gen Sparging on Removal of Iron ............................ 71
13 Effect of Bed Depth on Removal of Fe2* ....................... 75
lU Effect of Bed Depth on Removal of Fe^ ........................ 76
15 Effect of Temperature on Removal of Fe2* ..................... 8l
-------
TABLES
1 Preliminary Tests for Effect of Aeration on Ferrous
Iron Removal ............................................... 26
2 Comparison of BCR and Bureau of Mines Results of
Continuous Oxidation of Ferrous Iron ....................... 27
oj»
3 Adsorption of Fe on Activated Carbon ....................... 32
k Adsorption of Fe3+ on Activated Carbon . 1 ................... 35
5 Adsorption of Fe3+ on Activated Carbon. II ................... 3&
6 Reduction of Iron. Batch Tests ............................... 1*1
7 Reduction of Iron. Continuous Flow Tests .................... k2
8 Comparison of Types of Carbon . Batch Tests .................. kk
9 Comparison of Types of Carbon. Continuous Flow Tests ....... . k5
10 Results of Statistical Design I Experiment on Variables
Affecting Iron Removal . Batch Tests ....................... ^8
11 Codified Design and Calculation Matrix ........ . .............. 1*9
12 Main Effects and Interactions Based on Data from Batch Tests. 50
13 Results of Statistical Design I Experiment on Variables
Affecting Iron Removal. Continuous Flow Tests ............. 52
Ik Main Effects Based on Data from Continuous Flow Tests ........ 5!*
15 Results of Statistical Design II Experiment on Variables
Affecting Iron Removal . Batch Tests ....................... $6
l6 Main Effects Based on Data from Batch Tests .................. 57
17 Results of Statistical Design II Experiment on Variables
Affecting Iron Removal. Continuous Flow Tests ............. 58
18 Main Effects Based on Data from Continuous Flow Tests ........ 6b
19 Effect of pH on Removal of Iron .............................. 63
20 Effect of No Manganese on Removal of Iron .................... 66
vii
-------
TABLES
21 Effect of No Aluminum and No Aeration on Removal
of Iron
22 Effect of Aeration, No Aeration, Nitrogen Sparging, and
Oxygen Sparging on Eemoval of Iron ......................... 70
23 Effect of No Sulfate and Various Concentrations of
Nad on Removal of Iron .................................... 72
2k Effect of Bed Depth on Removal of Iron
25 Effect of Diameter of Column on Removal of Iron from
Synthetic Coal Mine Water .......................... ....*... ?8
26 Effect of Column Diameter on Removal of Iron from
Actual Coal Mine Water ........... .......................... 79
2? Effect of Temperature on Removal of Iron ..................... 80
28 Effect of Alumina on Removal of Iron ......................... 83
29 Effect of Charcoal Briquettes on Removal of Iron. . . * ......... 8U
30 Preliminary Continuous Flow Tests With Tarrs Coal Mine
Water ............................................ ;.......*. 86
31 Analyses at Various Points Along Tributary of Little Plum
Creek ............................................ ..... ..... 8?
32 Results of Preliminary Continuous Flow Tests with Tributary
of Little Plum Creek. Sampling Point No. 1 .......... . ..... 89
33 Results of Preliminary Continuous Flow Tests with Tributary
of Little Plum Creek. Sampling Point No. 2 ................ 90
3k Results of Preliminary Continuous Flow Tests with Tributary
of Little Plum Creek. Sampling Point No. 3. ... .......... 4 . 91
35 Batch Tests with Bacteria-inoculated Water ............. . ..... 93
36 Continuous Flow Tests with Bacteria-inoculated Water. ......*. 95
37 Batch Tests with Bacteria-inoculated Carbon .................. 96
38 Continuous Flow Tests with Bacteria-inoculated Carbon. .^.*... 98
viii
-------
TABLES
Page
39 Long Term Effects of Thiobacillus ferrooxidans 99
ho Comparison of the Effect of Long Aeration Periods and Bacteria
Versus No Bacteria on Removal of Iron from Synthetic Coal
Mine Water 101
4l Comparison of the Effect of Long Aeration Periods and Bacteria
Versus No Bacteria on Removal of Iron from Actual Mine
Water. .. 102
42 Results of Statistical Design III Experiment on Variables
; - Affecting Iron Removal 104
43 Main Effects Based on Data from Continuous Plow Tests 106
44 Effect of Flow Rate on Removal of Iron from Synthetic Coal
Mine Water , , 108
45 Effect of Water Flow Rate on Removal of Iron from Tributary
of Little Plum Creek. Sampling Point No. 2 109
46 Effect of Flow Rate on Removal of Iron from Tributary of
Little Plum Creek. Sampling Point No. 1 Ill
4? Effect of Aeration on Removal of Iron. No Carbon 114
ix
-------
SECTION I
CONCLUSIONS
Studies on the treatment of ferrous acid mine drainage vith activated
carbon have led to the following conclusions:
1. A process has been developed whereby ferrous iron can be removed
from acid mine drainage at low pH with activated carbon. The use of
this process can result in almost complete removal of the ferrous iron.
2. The process essentially consists of streaming the acid mine drain-
age through an aerated column containing a bituminous coal-based acti-
vated carbon. The effectiveness of removal of ferrous iron is dependent
on the amount of carbon used, the water flow rate, and the concentration
of iron in the water.
3. The advantage of this process is that once the ferrous iron has
been removed, the acid mine drainage can be more readily neutralized
using an inexpensive reagent such as limestone.
U. The major disadvantage of the process is that the activated carbon
soon loses its activity by adsorption of the iron on the carbon. This
adsorbed iron cannot readily be removed, for example by an acid wash,
and the relatively expensive activated carbon must be discarded. Costs
for activated carbon alone might range from $62 to $165 for every 1,000
gallons of water treated. The cost would prohibit this process for use
in treating acid mine drainage.
5. The process is limited to waters having a pH below ^.0 but higher
than 2.5- Above pH ^.0, the precipitated iron compounds would physical-
ly clog the column; below pH 2.5, the use of carbon results in a signif-
icant reduction of iron back from the ferric to the ferrous state, which
is the opposite reaction desirable to this process.
6. Of relatively minor importance are the effects on ferrous iron re-
moval of particle size of the carbon, temperature and total ionic
strength of the water, and aeration rate. The surface area of the carbon
and the concentrations of aluminum, manganese, and sulfate in the water
had no effect on iron removal.
1.
-------
SECTION II
RECOMMENDATIONS
The following additional studies are recommended:
1. Larger scale continuous flow tests with activated carbon, include
ing neutralization, should be conducted with acid mine drainage to
determine if a lesser degree of iron removal than that used in the
process evaluation would be sufficient to initiate iron removal.
Adequate treatment might still be obtained since treatment with lime-
stone involves additional iron removal as well as neutralization. For
this purpose, treatment at a water flow rate of at least one gallon per
minute is recommended.
2. Examination should be made of an inexpensive method for removing
the adsorbed iron from the activated carbon.
3. Additional tests should be conducted with other materials, such as
coal, instead of activated carbon, and with the use of oxygen rather
than air.
-------
SECTION III
INTRODUCTION
This is the final and summary report on the Environmental Protection
Agency (EPA) Project No. ll*010 GYH activated June 1, 1971 > with
financial support from Bituminous Coal Research, Inc., and the
Environmental Protection Agency. The project is based in part on
exploratory work conducted on Pennsylvania Coal Research Board Project
CR-75, activated July 1, 1967, with financial support from Bituminous
Coal Research, Inc., the Pennsylvania Coal Research Board, and the
United Mine Workers of America, and expanded February 7, 1968 with
additional financial support for the program through Grant WPRD 63-01-68
to the Commonwealth of Pennsylvania by the Federal Water Pollution
Control Administration.(l)
Work on the project was conducted according to the BCR Research Program
Proposal RPP-171R, dated November 19, 1970, and the Grant Application,
dated May 18, 1971 > both of which were submitted to the Environmental
Protection Agency. A schedule of work detailing the plan of operation
used on the project was also submitted to EPA on July 7» 1971' The
experimental work was conducted during the period June 1, 1971 to
May 31, 1972.
Objectives
The objectives of the studies conducted in the period covered by this
report were to delineate the process variables influencing the catalytic
oxidation of ferrous iron with activated carbon and to determine the
number and types of coal mine water to which this process may be
successfully applied.
This work is a part of a long-range program being conducted at BCR the
objective of which is to design and develop an improved process, based
on treatment with limestone, for the control and prevention of pollution
of waters by drainage from coal mines.
Nature and Scope of the Problem
The occurrence of acid drainage associated with coal mining has been
well documented. (2) By 1970, over 200 mine water treatment plants were
operating and an additional 100 plants were in various stages of design
and construction in the state of Pennsylvania alone as part of the coal
industry's vigorous abatement program of constructing water treatment
plants wherever polluting waters might be discharged from active mines
into open streams.(3)
-------
Because of certain conditions, treatment of coal mine water by neutrali-
zation does not always proceed as rapidly as desired, ^ecifically,
treatment of coal mine water containing iron in the ferrous, Fe , state
involves oxidation as well as neutralization, and this oxidation is
responsible for generating more acid. The oxidation and hydrolysis of
each mole of ferrous iron, Fe2*, results in a net gain of two moles or
acid, if, to the system as can be seen in the following equations:
Fea+ + lACb + H+-»Fea+ + 1/2 KfeO
Fe8* + 3 !%0 -»Fe(OH)3 + 3H
The oxidation and hydrolysis of the resultant product of oxidation to a
more insoluble ferric compound, the familiar yellowboy, is responsible
for slowing up the treatment process since the acid produced requires
further neutralization. The problem is particularly acute when lime-
stone is the neutralizing agent (4) since limestone is comparatively
insoluble and a weakly basic material. In a continuous limestone treat-
ment process, the pH rarely increases to as high as 7. Since the rate
of oxidation of ferrous iron increases a hundredfold for every increas-
ing unit of pH (5, 6), that rate cannot be accelerated by pH control
with limestone, as is possible during treatment with lime, soda ash, or
sodium hydroxide. In addition, other processes which might be used for
treatment of coal mine water, e.g., reverse osmosis (7), have exhibited
problems associated with the oxidation of ferrous iron.
The present program of laboratory investigation was designed to aid in
solving these problems.
Approach to the Problem and Research Procedure
The approach taken to oxidation of ferrous iron in coal mine water with
activated carbon was conceived, for the most part, as a result of two
publications. The first, a 1$W patent by Schumacher and Heise (8),
described the preparation and use of activated carbons to oxidize
"aqueous solutions of salts of metals of variable valency." The second,
a 1965 publication by Thomas and Ingraham (9), described studies of the'
rate of oxidation of aqueous ferrous sulfate solutions using activated
carbon as a catalyst. Both were concerned with catalytic oxidation of
leach liquors which routinely consisted of concentrated amounts of iron
in contrast to the more dilute concentrations of iron typically found in
coal mine drainage.
-------
As early as 1931, the concept had been tested by Lamb and Elder (10)
using a commercial, steam-activated, coconut charcoal and later in 1953
by Posner (11) using a catalyst prepared from sugar, FeClg, and urea.
The concept was tested at BCR in 1967 (12) and found to be feasible as
part of the limestone treatment research program (l, 13) and later under
a separate research program, BCR Project No. 20UO. Early in 1968, as
part of the limestone treatment program (l), the procedure of preparing
activated carbons for testing was abandoned and commercially available
materials were obtained and tested (l^, 15) with favorable results. The
Bureau of Mines, in a limited study (16) conducted in 1969 in coopera-
tion with the Pittsburgh Activated Carbon Division of Calgon Corpora-
tion, confirmed our earlier results.
To delineate the process variables influencing the catalytic oxidation
of ferrous iron with activated carbon and thereby achieve one of the
stated objectives of this program, both batch and continuous flow
experiments were conducted. The first part of the program involved the
use of laboratory-prepared synthetic coal mine waters. The latter
portion of the program involved actual coal mine waters to aid in
achieving the other stated objective of the program which was to deter-
mine the number and types of coal mine waters to which this process
might be successfully applied.
Consideration was given early in this project to statistical design
(17, 18, 19) and, more specifically, to sequential analysis (20, 21,
22, 23), and these techniques were used where appropriate to obtain the
maximum amount of useful information from the experimentation.
Process variables explored during this study included (a) the surface
area, particle size, and origin of the activated carbons; (b) the pH
and concentrations of iron, aluminum, manganese, and sulfate as well as
the temperature of the mine water; (c) the aeration rate and flow rate
of water through the carbon columns; and (d) the configuration of the
carbon column. The role of bacteria in this process was examined by
(a) analyzing the waters for bacteria which might influence the process
and (b) inoculating some waters or carbons with Thiobacillus
ferrooxidans and/or Thiobacillus thiooxidans.
Procedures used, process variables explored, and conduct of the experi-
ments on this project were selected by mutual agreement between BCR
personnel and the sponsors (EPA) prior to and during the experimental
program*
-------
SECTION IV
General pyocedixres, including apparatus, materials, and analytical
procedures for conducting the experiments and for evaluating the effect
of process variables, are reported here.
Analytical Procedures
Both yaw »ine watejr and treated water samples were analyzed routinely
for pH and for ferrous and total iron eolorimetrically (2*0 using
o-phenwttopoliiie and a, Bach Chemical Co. Model AC-DR colorimeter.
Iron, aluminum, and manganese were determined by emission spectro-
graphic techniques using a Jarrell Ash Model 78-000 1.5 meter Wadsworth
grating or "by atomic absorption techniques using an instrumentation
Laboratory Model 3&-153 atomic absorption speetrophotometer. Sulfates
were determined gravianetrically. Dissolved oxygen was determined with
a Yellow Springs Model 5^ oxygen meter equipped with a Clark-type
» covered polarographic probe.
General procedures
The general procedure for conducting the batch experiments was as
follows:
A 1$ g portion of activated carbon was placed in a gas washing bottle
wttfe 150 ml of synthetic coal mine water. Air was bubbled at a rate
of IfO ml/min through the carbon-water mixture and the water analyzed
at specified periods Of time. For most tests, ferrous iron, Fe , and
total iron, Fe«, act well as pH were determined. A sketch of the appa-
ratus used for these experiments is shown in Figure 1. The gas washing
bottles are a standard, off-the-shelf item (Fisher Scientific, Catalog
3^0. 3-037) consisting of a lyrex bottle approximately 270 mm overall
height, 350 ml capacity with a side inlet and fritted disc 60 mm in
diameter at the bottom apd a standard taper (S. T. Qk/kO) ground joint
at the top.
general procedure for conducting the continuous flow experiments
was as follows;
A glass column, 19 in* long x 3 in* inside diameter, was packed with
BOO g of Buchar W-W, 12 x ^0 mesh, activated carbon (Westvaco). A
rubber stopper and a plastic screen were used to contain the carbon. A
reservoir of coal mine water wai? placed approximately 1,5 ft above the
top of the column. The coal miae water flowed through the column at a
rate of 30 ml/min. Air, at a rate of 150 ml/min, was introduced by
means of a gas dispersion tube inserted down through the activated
-------
Inlet Air
Synthetic Coal
Mine Water
Activated Carbon
Frit
Bituminous Cool Research, Inc. 2042G33
Figure 1. Apparatus for Batch Experiments on
Catalytic Oxidation of Ferrous Iron
10
-------
carbon to the plastic screen at the bottom of the column. The water was
analyzed at specified periods of time, usually for ferrous iron, Fe ,
and total iron, Fe_, and pH.
A sketch of the apparatus used for most of these continuous flow tests *
is shown in Figure 2. Preliminary tests were conducted using a similar
glass column, 22 in. long x 1 in. inside diameter, packed with 80 g of
activated carbon. Tests were also conducted with molded acrylic
columns. Other changes in the procedure and apparatus are discussed
for the individual tests.
To control the flow of compressed air for the experiments, flowmeter-
regulator combinations were mounted on panels. The flowmeters were
Dwyer Instruments, Inc., Visi-Float Model VFA-SSV-PF-21. Each was con-
nected to a matching Dwyer Model EKA constant differential pressure
regulator behind the panel. The regulators were preset at 3 Psi- Tb-e
flowmeters were calibrated and covered the range of 0.1 to 0.5 liters
per minute.
Materials
Activated carbons and corresponding technical information were received
from a number of suppliers. Those carbons used in most of the experi-
ments include Westvaco's Nuchar WV-W, 12 x ^0 mesh, and Pittsburgh
Activated Carbon's Filtrasorb 100 and 300, both 8 x ho mesh. Other
activated carbons will be described as they are used in the experiments.
Synthetic coal mine water was prepared on demand in 15 liter batches.
The "standard".synthetic used in most tests consisted of a solution of
approximately 250 mg/1 of Fea* added as FeS04*7HgO and adjusted to pH
3.0 with sulfuric acid. For other tests, the pH of the synthetic was
changed by addition of different amounts of acid to the desired pH.
The concentration of iron also was changed in other tests and other
constituents such as aluminum, manganese, and sulfate were added to it.
Two actual coal mine waters were used. One was a discharge from an
abandoned coal mine in Westmoreland County, Pennsylvania, which con-
tained 13 mg/1 of Fes* and 1+5 mg/1 of Fe_, with a pH of 3.0. The other
was a stream sample from a tributary of Little Plum Creek in Allegheny
County, Pennsylvania. The concentration of iron varied with the par-
ticular spot chosen for sampling. All sampling points of this stream
were within 3 miles of each other. The concentration of Fe8* varied
from 55 to 390 mg/1, total iron from 71 to 535 mg/1 and the pH from
2.8 to k.6 as the stream was diluted by run-off from various sources.
11
-------
Inlet Air
I
••^
i.
'I
1 ,-
ff
II
t i '
J I
M
M
i
i
i;-'
;
H
s
-
-
s
'
, ,
n
Ml '
II
n
' '*-. *'
* i •
^
/
/
bbe
Inlet Water
Activated Carbon
Plastic Screen
bber Stopper
Outlet Water
Bituminous Cool Research, Inc. 2042G34
•••••••^•••••••^^•••^•••••••••^•••^^•^•••^^^^""'*"^fc^~~~
Figure 2. Apparatus for Continuous Flow Experiments
on Catalytic Oxidation of Ferrous Iron
12
-------
Effect of Variables
Column Design
Preliminary batch and continuous flow tests were conducted to obtain
data for design of activated carbon columns for use on this project.
Carbons of different particle size and three different glass columns,
22 in. x 1 in., 53.5 in. x 1.5 in., and 19 in. x 3 in. (length x inside
diameter) were used in a series of experiments with standard synthetic
coal mine water to establish flow rates of the water through the columns.
The effects of dry versus wet packing, of omitting aeration or aerating
the water exterior to the column, and of aerating by suction on the
effectiveness of iron oxidation were determined.
Role of Adsorption
To determine the adsorptive capacity of the carbon, 25 ml aliquots of
standard synthetic coal mine water were placed in 500 ml Erlenmeyer
flasks with the following amounts of Nuchar WV-W carbon, 12 x Uo mesh:
0.05, 0.10, 0.15, 0.20, 0.25, 0.50, 0.75, 1.00, 1.25, 1.50, 1.75, 2.00,
2.25, 2.50, 5.0, 7-5, 10.0, 12.5, 15-0, 17-5, and 20.0 g. The flasks
were each agitated for 60 minutes. At the end of this tine, the samples
were filtered and analyzed for ferrous iron, total iron, and dissolved
oxygen. The experiment was repeated with Nuchar WV-W carbon, minus 325
mesh. In the duplicate experiment with the powdered carbon, only the
range 0.05 to 2.00 g of carbon was included.
To determine the oxidation state of the iron adsorbed on the carbon, two
10 g samples of Nuchar WV-W carbon, 12 x ^0 mesh, were rinsed with 50 ml
of deionized water. One of the samples was placed in 50 ml of standard
synthetic coal mine water; the other, which was used as a blank, in
deionized water. Each was aerated for 30 minutes, stirred occasionally,
and samples taken for Fe * and Fe^,. Each was removed from the water,
rinsed with 50 ml of deionized water (samples taken of the deionized
water), removed from the water, extracted with 25 ml of hot (80-90C)
6N HC1, and the acid analyzed for Fe3* and
Batch tests were then conducted according to the general procedure to
determine any differences in the rate of adsorption of ferrous iron,
Fes+, and ferric iron, Fe3*, on the surface of the activated carbon.
Solutions of 300 ml each of (a) standard synthetic coal mine water and
(b) ferric nitrate containing approximately 250 mg/1 of Fe3+ were
added to 32 g portions each of Nuchar WV-W carbon, 12 x kO mesh, in two
600 ml beakers. The carbon samples had each been soaked overnight in
100 ml of deionized water. Each was stirred at a rate of 51 revolutions
per minute (rpm) with identical mixers, and pH, Fe3+, and FeT determined
at selected intervals for a 2k hour period.
13
-------
Since no attempt was made to control pH in the previously described
batch tests, the tests were repeated and greater pH control was main-
tained during the repeat tests. Nuchar WV-W carbon, 12 x hO mesh, 32 g
portions for each test, was soaked overnight in 100 ml of deionized
water prior to each test. A new portion of carbon was used for each
test. The pH of each solution was adjusted to the desired value by
adding HgS04 at the beginning of each test. Each solution was then
added to the carbon portions which were each in separate 600 ml beakers;
the mixture was stirred at a rate of 51 rpm, and pH, Fest, and FeT were
monitored at selected intervals for 300 minutes. Tests were conducted
at pH 1.0, 2.0, 2.5, 2.7, 3.0, U.O, and 5-0.
Results of initial tests showed that the pH of the mixture changed more
rapidly if the carbon was agitated slightly; therefore, the mixer was
placed into the carbon at the bottom of the beaker. This resulted in
gentle agitation of the carbon without aerating the mixture. Results
of initial tests also demonstrated that the pH was not maintained at
the desired value throughout the test; therefore, the mixture was ti-
trated with JfeS04 during the 300-minute test to maintain the desired
value-throughout the test. A concentration of 3K EfeS04 was selected for
the titration since the addition of this did not appreciably change the
volume of the initial solution and, therefore, change the concentration
of iron, yet controlled the pH satisfactorily. It was also discovered
in the initial tests that iron precipitated rapidly from the solution of
ferric nitrate even at relatively low pH. Fresh solutions were then
prepared on the day of the test.
Reduction of Iron
Both batch and continuous flow experiments were conducted with Nuchar
WV-W carbon, 12 x kO mesh, and a synthetic coal mine water containing
250 mg/1 of ferric iron, Fe3* , added as Fe(NQ3 )3 '91^0. The pH of this
water was 2.5- The reduction of iron from the Fe3+ state to the Fe2*
state in the aerated carbon-water systems was determined by analyzing
the water for ferrous and total iron at specified periods of time.
Types of Carbon
Six different carbons having various origins were each used in batch
and continuous flow tests. These include, with their origin in paren-
theses: (a) Barnebey Cheney AE, k x 6 mesh (coconut shell); (b) Witco
k x 10 mesh (petroleum); (c) Westvaco Nuchar WV-W, 8 x 30 mesh(bitu- '
minous coal); (d) Pittsburgh Activated Carbon BPL, 6 x 16 mesh (bitu-
minous coal); (e) Atlas Darco, U x 12 mesh (lignite); and Barnebev
Cheney 2?^, 3x8 mesh (mineral).
The batch tests were conducted with standard synthetic coal mine water
according to the general procedure in the gas washing bottles. The
carbon-water mixture was aerated at a rate of 150 ml/min for 120 minutes
Duplicate tests were conducted with the carbon being rinsed with 50 ml
of deionized water between tests.
Ik
-------
For the continuous flow tests, the general procedure vas followed but
400 g each of the same six carbons as were employed in the batch tests
were used. Sufficient quantities of Witco carbon, k x 10 mesh, and
Atlas Darco carbon, k x 12 mesh, were not available; therefore, a
k x 16 mesh and a 4 x *K) mesh fraction, respectively, were used. Syn-
thetic coal mine water flow rate was maintained at 30 ml/min for the
first 120 minutes of each test; at 200 ml/min for an additional 30
minutes; and then at l600 ml/min until the water sample was exhausted.
In some cases, l6oo ml/min could not be attained due to the nature of
the particular carbon; flow was then set at the maximum attainable.
Statistical Design I
This phase of the research program involved a study of the effects of
particle size and surface area of the activated carbon, as well as
aeration rate and concentration of ferrous iron in the synthetic coal
mine water on removal of iron. To this end, a 2* factorial experiment
was designed to include these variables. The methodology involved in
this factorial experiment was based on a recent publication (18) on
statistically designed experiments. The 16 tests required that two car-
bons be available of the same type or origin with different surface
areas and which could be pulverized to different particle size fractions.
The carbons selected were Filtrasorb 100 and Filtrasorb 30° (Calgon
Corp.)
From the results of preliminary tests, 3 ml of 3N I^S04 was selected as
the approximate amount necessary to neutralize 16 g of carbon in 150 ml
of deionized water. All carbons used in the 16 tests were first neutral-
ized by standing overnight in this amount and concentration of acid.
Separate batch and continuous flow tests were conducted. In all cases,
the pH of the two synthetic coal mine waters used for these tests was
adjusted to 3.0 with sulfuric acid.
For the batch tests, the general procedure was essentially followed. A
16 g portion of the selected activated carbon was placed in a gas washing
bottle with 150 ml of synthetic coal mine water. The concentration of
Fea* in the water was either 100 or 1,000 mg/1 depending on the parti-
cular test. The water was analyzed at specified periods of time during
a 180-minute aeration period. Duplicate tests were performed immediately
following the first test for each one of the 16. The carbon was rinsed
with 50 ml of deionized water between duplicate tests.
For the column experiments, the general procedure was essentially fol-
lowed. An 80 g portion of the selected activated carbon was placed in
the 22 in. x 1 in. column. Air, at a rate of either 50 or 500 ml/min
depending on the particular test, was introduced by means of a gas dis-
persion tube at the bottom of the column. Synthetic coal mine water,
containing either 100 or 1,000 mg/1 of Fe * depending on the particular
test, flowed through the column at a rate of approximately 30 ml/min.
Duplicate tests were performed immediately following the first test for
each one of the 16. The carbon was rinsed with 250 ml of deionized water
15
-------
between duplicate tests.
Statistical Design II
This phase of the research program involved a study of the effects of
changes in the concentration of several components of the synthetic coal
mine water other than ferrous iron on removal of iron. These additional
components included pH and concentrations of aluminum, jnanganese, and
sulfate in the synthetic coal mine water^ and another & factorial ex-
periment was designed. The methodology involved in this factorial exper-
iment was identical to that used in the Statistical Design I experiments.
For each of the 16 tests, a fresh quantity of Nuchar WV-W carbon,
12 x hO mesh, was soaked in deionized water overnight prior to use. The
synthetic coal mine water contained approximately 250 mg/1 of Fe »
added as ferrous sulfate. The pH was eithei- 1.5 or 3*5 depending on the
particular test and was adjusted with BC1, instead of H^SO* as In pre-
vious tests, so that this adjustment would not change the sulfate conceit-
tration. The concentration of sulfate in the water was either approxi-
mately 1*50 mg/1 from the ferrous sulfate or approximately 8,000 fflg/1
from the ferrous sulfate and the additional amount added as magnesium
sulfate, MgS04'7IfeO, depending on the particular test. The concentra-
tion of manganese was either approximately 25 mg/1 or approximately 100
mg/1, added as manganous chloride, MhClg*4ffeO, depending on the parti-
cular test. The concentration of aluminum in the synthetic coal mine
water was either approximately 50 mg/1 or approximately 200 mg/1, added
as aluminum nitrate, Al(NOa)8 "9^0, depending 6n the particular test.
Initially, aluminum hydroxide was added to the synthetic coal mine water
as the source of aluminum but it was observed that this material would
not solublize readily even at the low pH (1.5) of the test. Aluminum
nitrate was then used. No special precautions such as drying, etc.,,
were taken with the chemicals which were added as hydrates, it was
assumed that each chemical contained the exact amount of water of hydra*
tion as specified on the label, although some samples of the same
chemical, e.g. MgS04'7IfeO, appeared to be more hygroscopic than others.
Again, separate series of batch and continuous flow tests were conducted.
Each series consisted of 16 tests.
For the batch tests, the general procedure was essentially followed. A
16 g portion of carbon was placed in a gas hashing bottle with 150 ml of
the particular synthetic coal mine water and the mixture aerated at a
rate of 150 ml/min fot 120 minutes. Duplicate, tests were performed
immediately following the first test for each of the series of 16 tests
The carbon was rinsed with 50 ml of deionized water between duplicate
tests. Inmost cases, the concentrations of mangahese, aluminum and
sulfate were also determined. *
For the column experiments, the general procedure was followed and the
19 in. x 3 in. column and 800 g of activated carbon were used. The
carbon-water mixture was aerated at a rate of 150 ml/min and synthetic
16
-------
coal mine water was permitted to flow through the column at a rate of
approximately 30 ml/min for a total of 360 minutes. Duplicate tests
were performed on the next day for each test of the series of 16. The
carbon was soaked overnight in the column with deionized water, which
was drained prior to the duplicate test. In most eases, the concentra^-
tions of manganese j aluminum, and sulf ate were also determined.
Additional 5 tati-sticaF Design Tests
The resultso of*. the two factorial experiments labeled Statistical Design
I and :II indicated'that certain*variables affected or did not affect the
removal of iron with activated carbon*? Tests were conducted only at
high and-"low levels of the selected variables. : To more precisely define
the effectsof these variables, additional continuous flow tests were
conducted.^ The; procedure, apparatus, and carbons used were similar to
those of Test HO.<12 of the"continuous flow Statistical Design II
experiment/" .The results of each'of these additional tests were Compared
with those of. Test No. 12.-' The synthetic coal mine water for Test No. 12
consisted of approximately 250 mg/1 of Fe2+, a high concentration of
aluminum,: 200 mg/1, a Iciw concentration of manganese, 25 mg/1, a high
concentration of sulf ate, 8,000 mg/1, and a high pH, 3.5. Test No/ 12
was selectedifor^comparison purposes because the iron removal was rela-
tively constant throughout the duration of that test and because its re-
sults were reproducible.
In the first of this series, the effect of pH over-the range 1.5 to 3-5,
the low and high values in the Statistical Design II experiment, was
examined by conducting tests at pH 1.5, 2.0, 2.5, 3.0, and 3.5. The
first ..test at pH -I.-5- was not used in the evaluation of pH effect. The
second test at pH>1.5 was used in the evaluation since past tests (BCR
Project-20^0) had shown a first test with a fresh sample of carbon to be
different from each of the succeeding ones, and that the second and suc-
ceeding tests gave similar results until the carbon was spent. This was
probably;due to the elation of soluble material present in the carbon
during the first test. The same 6*00 g portion of Nuchar WV-W carbon,
12 x kO mesh, was used in each test of this series. The carbon was kept
overnight between tests in the column filled with deionized water.
Tests were conducted in the following sequence: pH 1.5, 1.5, 3-5, 2.0,
3.0, 2.5, and, finally 1.5 again. The last test at pH 1.5 demonstrated
that the carbon performed similarly to the second test at pH 1.5 and was
not rendered inactive during these tests.
Test No. 12 was repeated but this time with no-manganese, no aluminum,
and no sulfate in the synthetic coal' mine water7. In another series,
sodium chloride replaced the sulfate in the:
-------
Effect of Bed Depth
The effect of depth of the bed of activated carbon on removal of iron
was examined by conducting continuous flow tests using the 19 in. x 3 in-
column according to the general procedure with the following exception:
in five tests, the amount of Nuchar WV-W carbon, 12 x kO mesh, used was
80, 260, kkOt 620, and 800 g for a bed depth of 3.5, 9-2, 15-6, 21.9,
and 27.9 cm, respectively. A new portion of carbon was used for each
test. Two tests were conducted with each portion of carbon. The first
of each of the two tests was conducted with a ratio of carbon to total
volume of mine water of 1 g/13-5 ml. In other words, with 80 g of car-
bon used, the first test consisted of flowing 1080 ml of standard syn-
thetic coal mine water through the column. This took 36 minutes at 30'
ml/min. The first test with 800 g of carbon was conducted with 10,800
ml of water for a total test of 360 minutes duration. Each of the
second tests was conducted for 360 minutes.
Effect of Column Diameter
The effect of diameter of the column on removal of iron with activated
carbon was examined by conducting a series of continuous flow tests
according to the general procedure with Nuchar WV-W carbon, 12 x kO
mesh, standard synthetic coal mine water, and three molded acrylic tubes
of 3 in., k in., and 5 in. diameter. The tests were repeated using
actual mine water.
Effect of Temperature
To examine the effect of temperature on removal of iron, three continu-
ous flow tests were conducted according to the general procedure with
the 19 in. x 3 in. column, Nuchar WV-W carbon, 12 x kO mesh, and stan-
dard synthetic coal mine water at temperatures of 12, 21, and 30 C.
Use of Materials Other Than Carbon
In a series of continuous flow tests with the 19 in. x 3 in. column con-
ducted according to the general procedure, activated alumina (Aluminum
Company of America) was used in place of activated carbon with standard
synthetic and actual coal mine waters, and charcoal briquettes (Cumber-
land Charcoal Corporation) were used with standard synthetic coal mine
water.
Tests with Actual Coal Mine Waters
Two series of continuous flow tests with the 19 in. x 3 in. column were
conducted, with Nuchar WV-W carbon, 12 x ko mesh, essentially according
to the general procedure. In the first series, the Tarrs coal mine
water was used. In one test, the concentration of Fe3* was supplemented
in the raw water by adding ferrous sulfate to it. In the second series
a stream sample, collected at various points along a tributary of Little
18
-------
Plum Creek, was used.
Effect of Bacteria
Samples of water collected from selected catalytic oxidation experiments
were analyzed for the presence or absence of Thiobacillus ferrooxidans
and Thiobacillus thiooxidans by the Most Probable Number Method. Dilu-
tions of the reaction systems were prepared in Silverman and Lundgren
9K basal salts plus ferrous sulfate or elemental sulfur as the energy
source. Each dilution series was monitored for at least several days
for bacterial activity. The bacterial analyses were conducted under the
supervision of Dr = Francis W. Liegey, Chairman, Biology Department,
Indiana University of Pennsylvania.
Cultures of Thiobacillus ferrooxidans and Thiobaeillus thiooxidans were
also obtained from Dr. Liegey and used in combination with the activated
carbon.
- <
Tests with Bacteria-inoculated Water
Active cultures of Thiobacillus ferrooxidans and Thiobacillus thiooxidans
were obtained from Indiana University of Pennsylvania. Both batch and
continuous flow experiments with the 22 in. x 1 in. column were conducted
according to the general procedure with Nuchar WV-W carbon, 12 x kO
mesh, with (a) standard synthetic coal mine water which was inoculated
with Thiobacillus ferrooxidans, (b) standard synthetic coal mine water
which was inoculated with Thiobacillus thiooxidans, and (c) standard
synthetic coal mine water which was inoculated with a mixture of
Thiobacillus thiooxidans and Thiobacillus ferrooxidans. In each case,
the cultures were mixed with the standard synthetic coal mine water in a
10:1 water to suspension of culture ratio. The culture was used at the
48-hour growth period, upon recommendation of Dr. Liegey, who stated
that this is the time at which the bacteria are at the peak of their
growth curve.
Tests with Bacteria-inoculated Carbon
For the batch tests, 500 ml of an aqueous suspension containing an active
culture of Thiobacillus ferrooxidans was added to 80 g of Nuchar WV-W
carbon, 12 x ^0 mesh, in a gas washing bottle, and the mixture aerated
for 2k hours at a rate of about 50 ml/min. At the end of this time, the
water was decanted from the mixture and replaced with standard synthetic
coal mine water containing ferrous iron as an energy source for the
bacteria. The procedure was repeated three times per week with continuous
aeration for a total of 18 days. Another 80 g portion of Nuchar WV-W
carbon was placed in a separate gas washing bottle. No bacteria were
added to this carbon. The contents were treated in a manner identical to
the carbon-bacteria, i.e., standard synthetic coal mine water was added
at the same intervals and aeration continued; this carbon system was used
19
-------
as a blank for comparison of the results with the carbon-bacteria sys-
tem. After 18 days of this treatment, the synthetic coal mine water
was decanted from each of the carbons and replaced with 150 ml of fresh
standard synthetic mine water. Aeration at 150 ml/min was continued
for 120 minutes. Two 120-minute tests were conducted with each system.
For the first of the column tests, 500 ml of the same aqueous suspension
of Thiobacillus ferrooxidans used in the batch tests was added to 80 g
of Nuchar W-W carbon, 12 x kO mesh, in the 22 x 1 in. column. This
mixture was aerated at about 75 ml/min for 2k hours. The water was
drained from the bottom of the column and replaced with' standard syn-
thetic coal mine water as an energy source for the bacteria. As in the
batch tests, the water was "changed" three times per week and aeration
continued for 18 days. After this time, fresh synthetic mine water was
placed in the column and a test conducted according to the general pro-
cedure for a total of 360 minutes.
To examine the long term effects of the iron-oxidizing bacteria-
activated carbon combination, the carbon samples which had previously
been treated with Thiobacillus ferrooxidans culture on September 30,
1971> and used in a batch test (80 g) and a continuous flow test (80 g)
on October 15, 1971, which were just described, were combined in the
19 in. x 3 in. column, "fed" with ferrous sulfate solutions, additional
Nuchar WV-W carbon added to this column in batches of 80 g to a total of
800 g of carbon, and the column aerated continually. Continuous flow
tests were conducted with this column, which still contained iron-
oxidizing bacteria, and with standard synthetic coal mine water. These
tests were conducted on January 5, 1972, approximately two months after
receipt of the original culture. A sample of the feed water drained from
the column prior to this series of tests was sent to Dr. Liegey at
Indiana University for evidence of survival of the bacteria. Seven tests
were conducted according to the general procedure. Ferrous sulfate
solution was added and the column aerated between tests to maintain the
activity of the bacteria.
Effect of Long Aeration Periods Prior to Testing
A series of continuous flow tests in the 19 in. x 3 in. column were con-
ducted with Nuchar WV-W carbon, 12 x Uo mesh, and with standard synthetic
and actual coal mine waters using a bacteria-inoculated carbon and a
carbon without bacteria to examine the effect on iron removal of aerating
these carbons continually for long periods of time, in this case longer
than one month, prior to their use. Similar tests were conducted using
activated alumina instead of activated carbon.
Tests with Bacteria-inoculated Coal
Continuous flow tests were conducted with the 19 in. x 3 in. column
according to the general procedure with actual coal mine water and coal
being used in place of the activated carbon. The coal was inoculated
20
-------
with TMobacillus ferrooxidans approximately 15 days prior to the
tests and treated in the same manner as was the activated carbon.
Statistical Design III
This phase of the research program involved a study of the effects on
removal of iron from an actual coal mine water of water flow rate, amount
of carbon, bacteria versus no bacteria, and aeration rate. A 2* factori-
al experiment was designed. The methodology involved in this factorial
experiment was identical to that used in Statistical Design I and II
experiments. For eight of the l6 tests, a fresh quantity of Nuchar WV-W
carbon, 12 x kO mesh, was placed in a 1§ in. x 3 In. column with deion-
ized water and aerated for 2^ hours prior to the test. The other eight
of the 16 tests were conducted with a fresh supply of Nuchar WV-W carbon,
12 x lj-0 mesh, which had been placed in the column and inoculated with a
suspension of Thiobacillus ferrooxidans. A new solution of standard
synthetic coal mine water was added to the column every other day. The
column was then used in a test approximately three weeks after the
original inoculation.
The water flow rate was either 20 ml/min or 80 ml/min, the aeration rate
was either 50 ml/min or 500 ml/min, and the amount of carbon was :either
500 g or 1,000 g, depending on the particular test. For the l6 individ-
ual continuous flow tests, the general procedure was followed with the
changes as indicated in the individual tests.
Effect of Water Flow Rate
In the first of this series, a sample of Nuchar WV-W carbon, 12 x 40
mesh, was used in a continuous flow test with a 22 in. x 1 in. column and
standard synthetic coal mine water according to the general procedure.
Then this same carbon was used in a series of tests essentially according
to the general procedure but with water flows of 30, 100, 150, 200, 300,
and 600 ml/min, respectively, for 90 minutes each.
In the second of this series, stream samples, collected at various points
along a tributary of Little Plum Creek, were used. The effect of flow
rate was examined by conducting continuous flow tests with a 19 in. x
3 in. column and Nuchar WV-W carbon, 12 x Uo mesh and at water flows of
30, 60, and 120 ml/min.
In the last of this series, the activated carbons, which were used in the
Statistical Design III experiment and which contained Thiobacillus ferro-
oxidans, were combined for a total of approximately 6500 g of carbon.
This carbon was placed in a 52 in. long x 5 in. inside diameter molded
acrylic column similar to the other columns used on this project. The
carbon filled this column to a depth of 34 in. The column was fed and
aerated continually, similar to other columns containing bacteria, and
then used in five tests with actual coal mine water according to the
general procedure but with water flow rates of 20, 200, ^00, 800, and
1600 ml/min.
21
-------
SECTION V
RESULTS AW DISCUSSION
Effect of Variables
Past studies (l, 12, 13, 1^, 15, 16) established that aeration of low
pH acid mine drainage containing ferrous iron in the presence of acti-
vated carbon resulted in a significant decrease in concentration of
ferrous iron, Fe8*, in the water. Aeration of the same quality water
in the absence of activated carbon resulted in no significant change
in concentration of Fe3+ (See, particularly, Figure 16, p. 6^ of Ref-
erence l). To utilize this information in a practical mine drainage
treatment process, a study of the variables influencing the oxidation
of Fe3* in acid mine drainage was undertaken.
Column Design
Initial tests were conducted with gas washing bottles and with a num-
ber of glass tubes used as columns to obtain information which was
used to design a column or columns for this project. A variety of car-
bons were used in the glass tubes to establish flow rates of the water
through the columns under various conditions. Then, the smallest-
diameter column was used in a series of tests with synthetic coal mine
water and activated carbon, Nuchar WV-W, to note the effect of aera-
tion versus no aeration on the effectiveness of the column to oxidize
(remove) iron. These effects were double checked by conducting similar
batch tests with the gas washing bottles. The effects of both the
method of packing the carbon and aeration by suction through the column
were also examined.
Water Flow Rates through Column—A 22 in. x 1 in. glass column was
charged with 100 g of BPL carbon, h x 10 mesh, and then filled with the
standard synthetic coal mine water. With no aeration, flow by gravity
into the top and out of the bottom of the column was ioo ml/min. With
the 6 x l6mesh BPL carbon, the flow rate was 300 ml/min; with the 12 x
30 mesh BPL carbon, 200 ml/min. There was a direct relationship be-
tween the particle size of the carbon and the flow rate. As expected,
the smaller particles restricted the flow.
Next, a 19 in. x 3 in. glass column was charged with approximately
800 g of Nuchar WV-W carbon, 12 x kO mesh, and then filled with the
standard synthetic coal mine water. A maximum flow of 1030 ml/min was
obtained with gravity flow. This corresponds to 390 gal/day. With
larger columns and pumped water, it seemed probable that any desired
flow could be obtained.
Effect of Aeration—Three sets of conditions were used in column experi-
ments: (a) the synthetic coal mine water-carbon mixture was not aer-
23
-------
ated, (b) the water was aerated prior to entering the column, and (c)
the water-carbon mixture was aerated in the column.
In the first series, involving no aeration, three separate tests were
conducted with the 22 in. x 1 in. column, Nuchar WV-W carbon, 12 x kO
mesh, and the following flow-rates through the column,: 150, 75, and
30 ml/min. On the basis of Fe2*, as monitored during the test, flow
rates of 150 and 75 ml/min were too fast for effective iron removal.
Even at 30 ml/min, the effectiveness of removal falls off drastically
after 30 minutes.
Batch tests were conducted to check the results of th|+column tests.
As in the column tests, the oxidation (removal) of Fe in the non-
aerated systems was unsatisfactory in all cases.
In the next series, the standard synthetic coal mine water was aerated
by means of a gas dispersion tube before it was passed through the
column. If aeration of the mine water prior to its passing through
the column would result in satisfactory removal of Fe *, then a sim-
pler column could be designed and utilized for this program. For
these1 tests, Darco carbon, k x 12 mesh, and the 22 in. x 1 in. column
were used. The water was held and aerated for 3 min and then flow
started through the column at 30 ml/min. Aeration of the water be-
fore it entered the column was continued for the duration of the test.
A duplicate test was conducted using Darco carbon, 4 x 12 mesh, and
an additional one using Nuchar WV-W carbon, 12 x kO mesh. As in the
case of the tests with no aeration, the effectiveness of these car-
bons decreased after a very short period of time.
Batch tests with Darco and Nuchar WV-W were also conducted to check
the results of the column tests. As in the column tests, aerating
the water prior to contact with the carbon produced unsatisfactory re-
sults in terms of the removal of Fe3*.
In the final series, aeration was carried out by means of a gas dif-
fuser directly in the column (with appropriate venting of the column).
In duplicate experiments, Darco carbon, h x 12 mesh, and the 22 in. x
1 in. column were used. The flow rate of the synthetic coal mine
water was 30 ml/min and air flow rate was 150 ml/min. Effectiveness
of iron removal with Darco was no better than in the tests with no
aeration and aeration outside of the column. But in similar tests
with Nuchar WV-W carbon, 12 x ko mesh, the effectiveness was improved.
In these same tests with Nuchar WV-W carbon, either the water was in-
troduced at the top of the column and air introduced at the bottom
opposite to the flow of the water, or air was introduced at the top
of the column with the water. In the latter case, with no vent at the
top of the column, the air then flowed down through the carbon-water
mixture and out of the bottom of the column. Aerating through the
top was less effective in removing iron than was aerating through the
bottom.
-------
In the next test, the 53.5 in. x 1.5 in. column was charged with
approximately kOO g of Nuchar W-W carbon, 12 x kO mesh. The syn-
thetic coal mine water was introduced at the top of the column at a
rate of 35 ml/min and air introduced at the bottom at a rate of 150
ml/min.
More effective removal of iron was demonstrated with the longer column
containing more carbon, partially due to the higher pH as a result of
the greater amount of carbon. During the test, the iron removal de-
creased with decreasing pH as the carbon contacted more and more water.
This implies oxidation of Fe2+ as the factor controlling iron removal
since more iron would be oxidized at the higher pH values. The rate
of oxidation increases a hundredfold with each increasing pH unit
(53 6), but adsorption, which might also be influenced by pH, could
not be excluded.
Again, to check these results, similar batch tests were performed with
the gas washing bottle. The results of one of these tests and a con-
.tinuous flow test with Nuchar WV-W are presented in Table 1. In the
batch test, the concentration of Fe2+ decreased to about 3 mg/1 after
10 minutes. The time of contact between water and carbon in the col-
umn was much shorter; therefore, the decrease in Fea
-------
TABLE 1. PRELIMINARY TESTS FOR EFFECT OF AERATION
ON FERROUS IRON REMOVAL
Continuous Flow Test Batch Test
Time,
min
0
5
10
15
30
^5
60
75
90
105
120
135
150
165
180
Experiment No. ^33-30
Fe~ , mg/1
251
10
23
39
6k
89
no
125
Ito
-
-
-
-
-
-
Experiment No. 1*3^-21
Fe8*, mg/1
21*5
^5
3
3
2
2
2
0
0
0
0
0
0
0
0
26
-------
would present some problems in adaptation to a continuous flow system
as long as some form of aeration is necessary.
The results of a BCR continuous flow test are presented in Table 2 and
compared with the Bureau's results. At the end of their test of 2.6
minutes duration, 98.2 percent of the original Fe2* has been removed.
At thesend of 5 minutes in the BCR tests, 96.6 percent of the origi-
nal Fe had been removed. But in the BCR tests, which were of a
much longer duration, there was a gradual decrease in effectiveness of
iron removal with time. There was no difference in effectiveness of
removal of iron when aeration was accomplished by suction.
Role of Adsorption
During the tests on column design which were described in the previous
section, there was also a decrease in concentration of total iron,
Ferp, which includes both ferrous, Fea+, and ferric, Fe3*, iron. This
was an indication that adsorption of iron may play a role in the mech-
anism for iron removal since removal by oxidation would result in a
decrease in concentration of Fe3* with no change in concentration of
Ferp. Adsorption of Fes* onto the surface of the activated carbon
without oxidation to the Fe3* state in an actual treatment operation
might be detrimental to the effectiveness of this method of iron re-
moval since back-washing of the carbon would redeposit Fe back to
the stream. Any advantage gained would be in concentrating the iron
and thereby treating a smaller volume of water. Assuming then that
only adsorption and no oxidation had taken place, experiments were
conducted to measure the adsorptive capacity of Nuchar WV-W carbon.
Determination of the Freundlich Adsorption Isotherm -- Aliquots of
standard synthetic coal mine water and samples of Nuchar WV-W carbon,
12 x kO mesh, were agitated as described in the Experimental section
and the water analyzed for Fe8* and Feij. The weight of carbon, M,
and the concentration of Fe * remaining in solution at the end of 60
minutes, C, were used in the Freundlich adsorption equation (25):
X i
I-KC"
where X = units of Fe3* removed
M = weight of carbon required to remove X
C = concentration of Fe2* remaining in solution
K,n = constants
27
-------
TABLE 2. COMPARISON OF BCR AND BUREAU OF MINES RESULTS
OF CONTINUOUS OXIDATION OF FERROUS IRON
Time,
min
0
2.6
5
10
15
30
60
75
90
105
120
135
150
BCR Bur. of Mines
Experiment Experiment
No. U33-37 No. 65 (Ref. 16)
Fes+, mg/1 Fe3*, mg/1
250 830
15
6
50
83
122
153
1W
163
163
166
167
168
28
-------
The equation can also be stated in logarithmic form
log | = log K + i log C
M n
which is the equation of a straight line. The data were plotted, but
a straight line was not obtained, probably because some oxidation
occurred concurrently with adsorption.
The experiment was repeated with the minus 325 mesh fraction of Nuchar
WV-W carbon since the Freundlich equation more properly applies to
powdered carbons. The results were essentially identical.
Adsorption-Acid Extractions--One sample of carbon was treated with 50
ml of standard synthetic coal mine water and a second sample with de-
ionized water as described in the Experimental section. Originally,
50 ml of the synthetic contained 24l mg/1 of Fea+ or an input to the
system of 12 mg of Fe24". Carbon containing the adsorbed iron was
washed with hot HC1. Analysis of the acid wash accounted for 5.7 mg
of Fes+, or less than 50 percent of the Fes* originally used. The
same type of calculation was done with Ferpj in that case, only 60 per-
cent of the total iron was accounted for. It could not be established
from this test whether any oxidation of Fe2"1" to the Fe3+ state accom-
panied the adsorption since, surprisingly, some of the iron remained
on the carbon, even after extraction with hot acid.
The test was repeated and this time the carbon was washed (a) with 25
ml of hot (80-90 C) 6N HC1, (b) again with 25 ml of hot 6N HC1 and
(c) a third time with 25 ml of hot 6N HaS04. The total amount re-
covered by these acid extractions was surprisingly similar to that from
the first experiment. The carbon apparently clings tenaciously to
part of the adsorbed iron.
Comparison of Adsorption of Fe24> and Fe3*— To determine the rates of
adsorption of Fe2* and Fe34" on activated carbon, identical batch tests
were conducted with activated carbon and with solutions of (a) Fe24" -
ferrous sulfate adjusted to pH 3.0 with sulfuric acid and (b) Fe34" -
ferric nitrate with pH 2=5 without adjustment. A comparison was then
made of the differences between the rate of adsorption of each species.
A plot of the log of concentration of iron versus time from the data
obtained is shown in Figure 3- The differences in the rate of ad-
sorption of each species are evident. In addition, the reasonable
"fit" of the data by the method of least squares to a straight line
indicates that the rate of adsorption in each case is first-order
with respect to the concentration of that particular species,
-------
IRON ADSORBED,
mg /I
1000-,
100-
10-
3+
50
100
150 200 250 300 350
DURATION OF TESTS AT CONSTANT
FLOW OF WATER, MINUTES
Bituminous Cool Research, Inc. 2042G3
O -4-
Figure 3. Comparison of Rates of Adsorption of Fe
and Fe^ on Activated Carbon
30
-------
Effect of pH on Adsorption of Fea+ and Fe3* — The studies just de-
scribed have shown that there was a difference between the rate of
adsorption of Fe and Fe3* on the surface of activated carbon. In
those studies, no effort was made to measure the rates of adsorption
from solutions having the same pH or to control the pH throughout the
duration of the tests. In this series of batch tests, pH was con-
trolled throughout the tests. The results of the tests to measure
•the adsorption of Fe on activated carbon over the range of pH values
most encountered in those actual coal mine waters to be neutralized
are listed in Table 3« It can be seen from the data in this table
that iron was adsorbed more rapidly at the higher pH values.
These data were then used in developing rate curves for adsorption of
Fea* by plotting the log of the concentration of Fe2* remaining in
solution versus time for each of the tests at the pH values listed.
These are shown in Figure h. It can be seen from Figure 4 that there
was not much difference in the rate of adsorption of Fe2* from pH 1.0
to 2.5. There was a substantial increase, though, in this rate be-
tween 2.5 and 3.0. In fact, the test at pH 2.7 was conducted after
this graph was originally drawn and the rate curve for this pH value
was then drawn as anticipated. From pH 3.0 to 5.0 there is a less
rapid but still substantial increase in this rate. From these data,
a process for removal of iron based solely on adsorption of Fe3* might
be applicable only to those actual coal mine waters having a pH greater
than 2.5. In reality, very few coal mine waters have a pH below 2.5.
The data from Table 3 and Figure k were used in the calculation of the
amount of Fe * adsorbed per hour with pH and these data are plotted in
Figure 5. From this figure the amount of iron adsorbed per hour can
be determined for any pH value over the range 1.0 to 5»0.
The results from tests to measure the adsorption of Fea+ on activated
carbon over the pH range of 1.0 to 5.0 are presented in TabJ.e k. The
concentration of Fe3* was then calculated by subtracting Fe + and Fei
and these concentrations are listed in Table 5- The data in Table 5
seem to indicate a more rapid adsorption of Fe3* on activated carbon
at pH 1.0, 2.0, and 2.5 than at the higher pH values. Closer exami-
nation of the data in Table U, however, reveals a substantial re-
duction of Fe8* to Fe3* occurring in the solutions of ferric nitrate
at these low pH values. A plot of the data in Table 5 showing the log
of the concentration of Fe3 remaining in the solution versus time, as
seen in Figure 6, indicates little differences in the rates of adsorp-
tion of Fea+ at the pH values listed, at least as compared to the
differences shown in Figure k for rates of adsorption of Fe2 as af-
fected by pH. The adsorption rates of Fe3* are complicated by the
appearance of Fe2* with time in the ferric nitrate as shown in Table
k. A plot of the log of the concentration of Fe2* in solution versus
time (from Table U) is shown in Figure ?• There is a strong resem-
blance of this data to that in Figure h. This indicates that the Fe2*
is produced by reduction of Fe3* at the start of the test and that
31
-------
TABLE 3. ADSORPTION OF Fe3 ON ACTIVATED CARBON
^__ „ Fe3* Remaining in Solution, mg/1
ExperimentExperimentExperimentExperimentExperimentExperimentExperiment
Number Number Number Number Number Number Number
Time, 440-93 44-0-85 440-83 440-97 440-91 440-89 440-95
min pH 1.0 pH 2.0 pH 2.5 pH 2.7 pH 3.0 pH 4.0 pH 5.0
u, 0 205 234 231 221 244 237 227
10 5 179 179 156 91 117 8l 94
10 182 195 159 104 91 68 75
15 188 211 195 120 84 65 55
30 188 201 179 117 62 36 19
45 195 208 172 120 39 16 10
60 182 172 156 114 36 6 0
90 172 143 156 71 23 60
120 166 127 114 55 . 3 0 0
150 156 110 91-39 0 o 0
180 i4o 110 78 26 o oo
210 84 104 8l 19 o oo
24o 75 101 81 13 o o o
270 68 97 78 6 0 o 0
300 62 78 75 0 0 0 0
-------
2.50
UJ
2.00
1.50
LOG Fe2+
REMAINING
IN SOLUTION
pH 2.0
pH 1.0
Straight Lines by Method of Least Squares
120
180
240
300
DURATION OF TESTS AT CONSTANT FLOW OF WATER, MINUTES
Bituminous Coal Research, Inc. 2042G36
Figure 4. Effect of pH on Rate of Adsorption of Fe* on Activated Carbon
-------
mg/l of Fe2 +
ADSORBED PER HOUR
200-i
175-
150-
125-
100-
75-
50-
25-
pH
1.0
2.0
2.5
2.1
3.0
4.0
5.0
Time to
Reduce
Fe2+to
O mg/l
22.04 hr
27.49 hr
25.06 hr
6.72 hr
2.73 hr
2.06 hr
1.11 hr
Initial
Fe2 +
mg/l
205
234
231
221
244
237
227
mg/l
Fe2 +
Adsorbed
per Hour
8.9
8.5
9.2
32.9
89.4
115.1
204.5
1.0
2.0
3.0
PH
4.0
5.0
6.0
Bituminous Coal Research, Inc. 2042G37
Figure 5. Effect of pH on Fe2+ Adsorbed
per Hour on Activated Carbon
-------
TABLE 1*. ADSORPTION OF Fe8* ON ACTIVATED CARBON. I,
VI
Experiment
Number
1*1*0-91*
pH 1.0
Time,
min
0
5
10
15
30
1*5
60
90
120
150
180
210
2UO
270
300
2/1
6
211
185
172
182
172
172
ll*6
107
9U
75
71
62
62
55
FeT
mg/1
2l*7
21*1*
237
21*7
256
21*7
256
2l*0
21*1*
21*7
2l*7
23!*
2l*0
2l*l*
21*1*
Experiment
Number
1*1*0-86
pH 2.0
mg/1
0
172
179
182
195
192
11*0
120
101*
101*
91
91
91*
81*
68
FeT
mg/1
250
205
218
227
227
211*
208
198
188
185
166
166
159
162
156
Experiment
Number
PH2.5
s+
3
123
120
133
ii*o
li*6
ll*0
ll*3
117
9k
81
65
71
71
68
2k
21*1*
185
188
188
185
185
185
185
175
166
11*6
ll*6
ll*9
ll*6
li*3
Experiment
Number
1*1*0-98
pH 2.7
2/1
6
16
36
1*2
52
1*9
32
1*2
32
23
19
19
16
16
16
FeT
237
201
188
198
185
192
195
182
175
169
159
162
162
159
156
Experiment
Number
1*1*0-92
pH 3.0
g^
0
16
13
13
13
13
10
3
0
O
0
0
0
0
0
mg/1
21*1*
2i*0
227
221
211
221*
227
23^
227
221*
22k
211
208
179
133
Experiment
Number
1*1*0-90
pH 1*.0
mg/1
0
10
10
0
0
0
0
0
0
0
0
0
0
0
0
ipg/X
Q|I ji
2l*l*
221
22k
231*
227
221
231*
211
211
211
221
231*
22k
211*
Experiment
Number
1*1*0-96
pH 5.0
mg/1
0
19
3
0
0
0
0
0
0
0
0
0
0
0
0
FeT
256
2l*0
221
231
221
211
221*
237
227
227
221
227
231
231
237
-------
TABLE 5. ADSORPTION OF Fe3+ ON ACTIVATED CARBON II
CO
Time,
min
0
5
10
15
30
1*5
60
90
120
150
180
210
21*0
270
300
Fe3+ Remaining in Solution, mg/1
pH 1.0
21*1
33
52
75
71*
75
81*
9!*
137
153
172
163
178
182
189
- pH 2.0
250
33
39
1*5
32
22
68
78
81*
81
75
75
65
78
88
pH 2.5
21*1
62
68
55
1*5
39
**5
1*2
58
72
65
81
78
75
75
pH 2;7
231
185
152
156
133
11*3
163
ll*3
ll*6
ll*0
11*3
11*6
1 |iO
1^4-0
pH 3.0
21*1*
221*
211*
208
198
211
217
231
227
221*
221*
211
208
179
133
pH I*.0
2l+i*
231*
211
221*
231*
227
221
231*
211
211
211
. 221
. 231*
221*
213
pH 5.0
256
221
218
231
221
211
221*
237
227
227
221
227
231
231
237
-------
tjO
LOG Fe3 +
REMAINING
IN SOLUTION
2.50-1
2.00-
1.50-
1.00-
.50-
pH 2.5
pH 2.0
Straight lines by Method of Lea?) Squares
60
120
180
240
300
DURATION OF TESTS AT CONSTANT FLOW OF WATER, MINUTES
360
Bituminous Coal Research, Inc. 2042G38
Figure 6. Effect of pH on Rate of Adsorption of Fe^+ on Activated Carbon
-------
2.50
2.00
LOG Fe2 +
REMAINING
IN SOLUTION
1.50
CO
1.00
pH 2.5
Straight Lines by Method of Least Squares
180 240 300 360
DURATION OF TESTS AT CONSTANT FLOW OF WATER, MINUTES
Bituminous Coal Research, Inc. 2042G39
Figure 7. Concentration of Fe* from Reduction of Fe during Adsorption Tests
-------
is plotted in Figure 7 represents the adsorption of this Fe2*
with time on the activated carbon.
From Table 4, the concentration of Fes+ produced in 5 minutes by the
reduction of Fe3 was plotted versus the initial pH in Figure 8. It
can be seen from this figure that more Fe2"*" was produced from the re-
duction of Fe3 at the lower pH values. Again, as in the adsorption
of Fe , the pH value of 2.5 is important, since, below that value,
the reduction of iron to the ferrous, Fe , state proceeds at a rela-
tively rapid rate.
Reduction of Iron
Because of theaunexpected reduction of ferric iron, Fe3+, to the fer^
rous state, Fe , which occurred during the previously described ex-
periments, batch and continuous flow experiments were conducted with
a synthetic coal mine water containing approximately 250 mg/1 of ferric
iron, Fe3 , added as ferric nitrate and having a pH of 2.5.
Batch Tests -- Tests were conducted with l6 g of Nuchar WV-W carbon,"
12 x M) mesh, which had previously been washed with two 20 ml portions
of 6N HC1 and two 20 ml portions of deionized water. Other tests were
conducted with carbon which was not acid-washed and carbon which was
washed with IfeS04 instead of HC1. The results of typical tests are
presented in Table 6. Duplicate tests gave essentially identical re-
sults . In every instance, Fe3+ was reduced to the Fea state even
though the experimental conditions, including aeration, would seem
more conducive to oxidation. The acid-washed carbons were more effec-
tive reducing agents in these tests than the carbon that was not acid-
washed. Little difference in effectiveness was noted between carbons
washed with HC1 and IfeS04, although the HCl-washed carbons were slight-
ly more effective reducing agents.
Continuous Flow Tests -- In a manner similar to the batch tests, tests
were conducted with the column and "ferric" synthetic coal mine water
according to the general procedure. The first test was conducted with
carbon which had been washed with HC1. The results are listed in
Table 7. As in the batch tests, reduction of Fe3 to the Fe state
was observed. The test was repeated with carbon which had not been
acid-washed. The results are also presented in Table 7. This time no
reduction occurred until the relatively basic, soluble portion of the
carbon was removed by elution. Below pH 6, reduction of iron was ap-
parent and continued for the duration of the test. Both the batch and
continuous flow tests resulted in the reduction of Fe3 to Fe and the
extent of this reduction was dependent on the pH of the water.
Types of Carbon
Six different types of carbon having different origins were evaluated
for this process by using each in both batch and continuous flow tests
to relate the type of carbon to effective iron removal.
39
-------
Fe2+, mg/l
IN 5 MINUTES
BY REDUCTION OF F03 +
250i
200-
150-
100-
50-
1.0
2.6
3.0
4.0
5.0
PH
Bituminous Coal Research, Inc. 2042G40
Figure 8. Effect of pH on Concentration of Fe2 + from
Reduction of Fe^+ during Adsorption Tests
-------
TABLE 6. REDUCTION OP IRON. BATCH TESTS.
Experiment No.
Time,
min
0
5
10
15
30
60
PH
2.5
i.h
Carbon Washed
Fes+ , mg/1
0
158
163
163
205
201
with HC1
Fe«p , mg/1
2^7
231
227,
- 219
221
Experiment No. ^3^-58;
Time,
min
0
5
10
15
30
£5
60
pH
2.5
h.l
Carbon Not Acid
Fe3* , mg/1
0
127
116
103
85
Washed
Fe
-------
0
5
10
15
30
45
60
90
120
150
0
5
10
15
30
45
60
90
120
150
7. REDUCTION OF IRON.
CONTINUOUS FLOW TESTS.
Experiment No. 433-54;
Acid Washed Carbon
gH
2,5
2.6
2.9
3.0
3.2
3.2
3.1
2.7
2.6
2.5
Fe* ', mg/1
1
185
212
215
217
222
228
231
232
230
Ferp, mg/1
258
250
233
223
228
233
235
256
236
233
Experiment No. 433-56;
Carbon Not Acid Washed
ES
2.5
8.5
7.3
6.3
5.7
5.4
M
4.6
4.6
3.0
FeiT, rog/1
0
0
0
0
95
118
140
200
209
220
FeT, mg/1
251
8
5
26
100
158
173
220
233
259
42
-------
Batch Tests — The carbons used, their origin, and the results of the
batch tests are summarized in Table 8.
From the results on both Fes+ and Ferp remaining after. 60 and after
90 minutes in the second test, it can be seen that the bituminous coal-
derived carbons were most effective in removing iron in this series.
Lignite-, petroleum-, and coconut-derived materials were less effec-
tive. The mineral-derived carbon (manufacturer's description) was
completely ineffective.
Continuous flow tests -- The same six carbons used in the batch tests
were also used in continuous flow tests although some particle size
fractions were slightly different than those used in the batch tests.
The data from the continuous flow tests are summarized in Table 9 along
with the carbons and their origin. The carbon which removed iron most
effectively during the first 120 minutes of the test, while the flow
rate was 30 ml/min, was a lignite-derived material. The two bitumi-
nous coal-derived carbons were next in effectiveness. These results
are similar to those from the batch tests with the same carbons. In
the batch tests, the lignite-derived material showed a substantial
decrease in effectiveness when the same portion of that carbon was used
in a second test.
After the initial 120 minutes of these tests, the flow was increased
from 30 ml/min to 200 ml/min for an additional 30 minutes; then the
flow was increased to the maximum attainable for each carbon with this
system. Only Darco, k x kO mesh, the lignite-based carbon, and Witco
Carbon, k x l6 mesh, restricted the flow through the column. At the
increased flow rates, no carbons removed iron effectively.
From the relatively small differences between Fe2* and Per concentration
for most samples taken during both the batch and continuous flow tests
(See Tables 8 and 9), adsorption of iron seems to be the mechanism in-
dicated for removal of iron during these tests. Removal by oxidation
alone would result in low Fe2* values with little or no change in the
FeT concentration from the original synthetic coal mine water — in
other words, relatively large differences between Fe and Fej. A com-
bination of adsorption-oxidation, though, cannot be eliminated as a
possibility since that portion of the original Fe concentration which
might have been oxidized to the Fea+ state may possibly also undergo
rapid hydrolysis. The ferric iron compound thus formed could be pre-
cipitated even at relatively low pH. The end effect would again be
relatively small differences between Fe and Fe,p concentration, as were
observed during these tests.
In summary, the results of both batch tests and continuous flow tests,
indicate that the bituminous coal-derived carbons were the most effec-
tive carbons for removing iron. The lignite-derived material, though
initially most effective, decreased in effectiveness with time and
restricted the flow through the column.
-------
TABLE 8. COMPARISON OF TYPES OF CARBON. BATCH TESTS
First Test**
Second Test**
After
Carbon
Barnebey-
Cheney AE
Witco
Chemical
Westvaco
Nuchar WV-W
Pittsburgh
Activated
Carbon BPL
Atlas
Darco
Barnebey-
Cheney 274
Origin
coconut
shell
petroleum
bituminous
coal
bituminous
coal
lignite
mineral
Mesh
Size
4x6
4 x 10
8 x 30
6 x 16
4 x 12
3x8
5 min
Fe8**
229
217
189
190
178
234
FeT*
230
222
198
192
185
24l
After
60
Fe2+*
90
73
9
1
1
2kO
min
FeT*
101
91
23
17
27
242
After
120
Fes+*
28
12
0
0
1
24l
min
FeT*
33
32
6
8
2
245
After
5 min
Fe2"1"* FeT*
240 241
222 23^
202 208
19? 198
191 209
244 2*8
After
60
Fe3**
156
134
33
32
70
238
min
FeT*
173
141
36
39
96
246
After
90
Fe8**
123
82
2
0
29
241,
min
FeT*
134
89
4
9
34
248
* All Fe8+ and FeT concentrations are expressed as milligrams per liter (mg/l).
** With synthetic coal mine water: pH, 3.0; Fe3"1", 244 mg/l; FeT 257 mg/l
-------
TABLE 9. COMPARISON OF TYPES OF CARBON. CONTINUOUS FLOW TESTS
Carbon
Barnebey-
Cheney AE
Witco
Chemical
Westvaco
Nuchar WV-W
Pittsburgh
Activated
Carbon BPL
Atlas
Darco
Barnebey-
Cheney 27^
Origin
coconut
shell
petroleum
bituminous
coal
bituminous
coal
lignite
mineral
Mesh
Size
It x 6
k x 16
8 x 30
6 x 16
k x Uo
3x8
Flow
After
5 min**
Fe3+* Fet*
50 88
20 33
25 ^5
75 95
30 60
222 250
Rate, 30 ml/min
After
60 min
Fe2"1" * Fej*
88 115
88 117
k3 68
80 103
^5 63
238 265
After
120 min
Fe2+*
138
170
115
123
90
238
Pejr*
150
202
138
170
110
262
Flow Rate,
After
135 min
Fe2**
228
239
228
225
198
238
Per*
252
288
250
258
212
262
200 ml/min
After
IgO min
Fe2**
2^0
252
232
2^5
215
238
Fer*
253
278
2J+5
262
227
262
Maximum
Flow
Attainable
After
155 min
Fe2**
228
262
238
2^7
21*1
233
Fer*
262
268
2^8
265
262
252
* All Fe24" and Fej concentrations are expres-sed as milligrams per liter (mg/l).
** With synthetic coal mine water: pH, 3-Oj Fes*, 262 mg/lj FeT, 280 mg/l.
-------
Statistical Design I
Both batch and continuous flow tests were conducted to examine the
effects of particle size and surface area of the carbon as well as
aeration rate and concentration of ferrous iron in the synthetic coal
mine water on removal of iron, whether by oxidation, adsorption, or a
combination of the two. A 2* factorial design was selected. The l6
tests required that an adequate supply of one type of carbon be avail-
able in two distinct and separate surface area ranges and two distinct
and separate particle size ranges. After rejecting the use of many of
the carbons solicited for this project, either because they did not meet
the criteria established for these particular tests or because the
properties were not always as represented on the labels, Piltrasorb 100
and Filtrasorb 300 were selected since (a) they are both coal-based
carbons, (b) the surface area of Filtrasorb 100 is reported to be
800 ms/g, of Filtrasorb 300, 950-1100 m2/g, and (c) they were received
in sufficient quantity in 8 x Uo mesh particle size so that an 8 x 10
and a 20 x kO mesh particle size fraction could be obtained by screening.
The aeration rates selected were 50 and 500 ml/min; the concentrations
£%^U 0 %
of Fe in the synthetic coal mine water (at pH 3.0) were 100 and
1000 mg/1.
A preliminary batch test with Filtrasorb 100 conducted according to the
conditions of one of the l6 factorial tests resulted in the synthetic
coal mine water having a pH of 6.9 after one 2-hour test and 7.8 after a
duplicate test. In each case there was no Fes* remaining in solution.
Similar results were obtained with Filtrasorb 300. Since oxidation is
dependent on pH (5, 6) and this process as envisioned involves oxidation
at low pH, a high pH caused by the relatively basic, soluble portion of
these carbons would result in precipitation of the iron and fouling of
the carbon and could not be tolerated in this system. In addition, con-
tinuous treatment would soon result in the soluble portion of the carbon
being removed. The resultant carbon would be of a different nature and
react differently toward this process.
To alleviate this problem, it was decided to neutralize the basic
portion of the carbon by an acid wash before use. This was done reluc-
tantly since past experience (BCR Project No. aAo) had shown that the
adsorption-oxidation effects from an acid-washed carbon can be quite
different from those effects from a carbon which has not been acid
washed. Thus, on the day prior to each test, the carbon to be used was
immersed in deionized water containing 3 ml of 3N !feSC4. For each test,
a fresh sample of carbon was used.
Since past tests (BCR Project 20*1-0) had shown a first test with a fresh
sample of carbon to be different from each of the succeeding ones, and
that the second and succeeding ones gave similar results due to the
elution of soluble material present in the carbon during the first
test, two successive tests were performed with each carbon in both the
batch and continuous flow series. In each case only the second test was
-------
evaluated, since equilibrium had probably been reached by that time and,
therefore, this second test would be more representative of actual per-
formance of that particular carbon.
Batch Tests — The data representing the batch tests are summarized in
Table 10. For ease of interpretation, the four separate variables are
identified, with the high and low level of each variable being repre-
sented by + and - signs respectively.
The,data in Table 10,were employed more fully in the calculation of
main effects and variable interactions for each of the four responses.
For this,purpose, a codified design and calculation matrix, shown in
Table 11, was used. The derivation of this matrix is described by
Davies.(^9) The positive or negative signs for variable interaction
were obtained simply by algebraic multiplication for the signs of each
single variable involved in the interaction. Main effects and inter-
actions were calculated by multiplying the response values from Table 10
by the corresponding effect elements from Table 11, summing the results,
and dividing by 8. The resulting value for single variables is also
sometimes designated as an average effect, since, for two levels of the
variable only, it is simply the difference between the average response
to all tests conducted at the first level of the variable and that of
all tests at the second level. If the effect of one variable changes
at different levels of another variable, the two variables are said to
interact. The relative importance of the variable or variables in
question is reflected by the absolute magnitude of the calculated main
effect or interaction. First-order (two variable) interactions are
sometimes easy to interpret in terms of the physical behavior of the
system. Second- and higher-order interactions are often less readily
understood, and are often of minor significance in any case.
The main effects and interactions obtained in the manner described
above are listed in Table 12. The calculations were less time-consuming
since a computer program was employed. This enabled the evaluation of
additional responses in this and in subsequent experiments. As stated
before, only the second test of a series was evaluated since the results
of these second tests are felt to be representative of the performance
of the carbon after the soluble material present with the carbon has
been removed.
As shown in Table 12, the main effects and interactions for each
response are grouped according to the number of variables involved and
are listed in order of decreasing absolute magnitude in each group.
For the batch tests, the data indicated that the most significant
variable affecting percent Fe2* removed was Variable k, concentration
of Fes* in the original water. The negative sign for that main effect
simply meant that the Fe2* was removed more effectively at the lower
level of Fe2* concentration (100 mg/1 in this test). This effect was
particularly apparent when one compares the results of Test 1 through
Test 8 with the results of Test 9 through Test 16 in Table 10 under the
-------
Test
1
2
3
k
6
7
8
9
10
n
12
13
1^
15
16
TABLE 10. RESULTS OF STATISTICAL DESIGN I EXPERIMENT
ON VARIABLES AFFECTING IRON REMOVAL. BATCH TESTS
Variables* Responses
123 4
Surface Particle Aeration Fe8* Concen-
Area Size Rate tration
; : :. :
+
- 4.
•f" ™ •• *4"
••• -4^ •» •!•
: ; : :
^Variable Identification
Variable +
Surface Area 950-1100 ma/g
Particle Size 8 x 10 mesh
Aeration Rate 500 ml/min
Fe2+ Concentration 1000 mg/1
Fes* Removed, Fej Removed,
^percent percent
81 77
89 86
62 6l
6k 59
82 75
9k 93
7k 70
82 75
25 18
33 30
19 20
17 18
7^ 60
ko 36
39 3^
30 25
Levels
-
800 ms/g
20 x kO mesh
50 ml/min
100 mg/1
-------
TABLE 11. CODIFIED DESIGN AND CALCULATION MATRIX
Variable
Interactions
vo
est 1 2 3 U (,1^2) (1,3)
1 + +
2 +
3 - •*• - - - ' +
k + + - - -f
5 - - + - +
6 + - + - - 4
Y _ 4 4. _
8 + + + - -f +
9 - - - -f + +
10 + - - +
11 _ 4- - + - +
12 4- + - + +
13 - - 4 + +
Ik + - +. + - -f .
15 - + + -K -
16 -f + -f + + +
(1,U) (2,t3) (2,i|) (3,1*) (1,2,3) (2,3,^) (1,2,^) (1,3^) (1,2^3^4)
4 + + +- - - - +
.4444 _ 4 4
----f- + - + +
__4- - -f-f - -f
4 + -- - - + + +
4- + - - + + 4-
__4- + - - - + +
4 - + - - _—:•;+
4 I I 4 * I * ' 1 " 4 r
4 + H- - 4 _
-------
TABLE 12. MAIN EFFECTS AND INTERACTIONS
BASED ON DATA FROM BATCH TESTS
Responses
Single Variable:
,2+
Fe" Removed, percent
Main Effects
and
Interactions
Variable(s)
4
2
3
1
•43.88
-16.38
15.62
- 0.88
FeT Removed, percent
Variable(s)
4
2
3
1
Main Effects
and
Interactions
-44.38
-14.12
12.38
0.88
Two Variable
Interactions:
Three Variable
Interactions:
Four Variable
Interactions:
1,4
3,4
1,3
1,2
2,4
2,3
2,3,4
1,2,3
1,2,4
1,3,4
1,2,3,4
8.38
6.62
4,
o,
o,
62
38
0.12
5.88
4.62
3.12
0.?4
4.12
2,3
1,4
3,4
1,3
1,2
2,4
1,3,4
2,3,4
1,2,3
1,2,4
1,2,3,4
8.75
6.62
4.88
3.38
2.88
2.38
7.38
5.88
3.38
3.12
3.f
Designation of Variables:
1
2
3
4
= Surface area
= Particle size
= Aeration rate
= Fes+ concentration
50
-------
heading "Responses". With the single exception of Test 13, the percent
Fe^ removed was greater in Test 1 through Test 8, at the lower level of
Fe in the original water, than in Test 9 through Test 16, at the higher
level of Fe in the water. This indicated contact of the water (con-
taining Fe ) -carbon mixture to be extremely important to removal of
Fe . The same can be said of the results with percent Fe™ removed.
(See Table 10.) *•
''* \
Particle size and aeration rate affected both removal of Fe2* and of
total iron, FeT (which"includes 'both Fea+ and Fe3*). The negative value
for particle size (Table 12) indicated more effective iron removal at
the lower level of that particular variable while the positive value for
aeration rate indicated that a higher aeration rate was desirable. The
absolute magnitudefof both of these indicated the influence of particle
size- and aeration rate to be much less than that of Fe2* concentration.
Each, though, again implied that more efficient contact would have re-
sulted in more effective iron removal.
Surface area did not seem to influence the results of the test since the
magnitude of that particular main effect was less than 1. It should not
be surprising that the relatively large iron molecule did not seem to be
affected by a change in the number of relatively small micropores of the
carbon as measured by surface area. There was no significant interaction
between the variables, since the absolute magnitude of the first-order
interactions from these tests did not even begin to approach the •magnitude
of the single variable main effects.
f
In summary of the batch tests, the concentration of Fe2* in the synthetic
coal: mine water was the single most important variable. The most effec-
tive iron removal occurred at the lower concentrations of Fes*. Of
lesser importance were particle size (finer particles for more effective
iron removal) and aeration rate (higher rate for more effective removal).
Surface area was not a significantly affecting variable nor was there .-' ;
significant interaction between variables. All results of the factorial
experiment for the batch tests indicated that more efficient contact of
the carbon-water system would have resulted in more effective iron
removal.
Continuous Flow Tests — As with the batch tests, continuous flow
(column) tests were conducted with an identical factorial experiment and
evaluated in the same manner. The data representing these continuous
flow tests are summarized in Table 13i Again, two successive tests were
conducted for each of the l6. For the reasons stated in the discussion
of the batch tests, only the second test was evaluated. The data in
Table 13 were again employed in the calculation of main effects and
variable interactions for each of the 12 responses using the codified
design and calculation matrix already presented in Table 11 and the pro-
cedure as already outlined.
-------
TABLE 13. RESULTS OF STATISTICAL DESIGN I EXPERIMENT ON VARIABLES AFFECTING
IRON REMOVAL. CONTINUOUS FLOW TESTS.
VJl
ro
1
Surface
Test Area
1
2 +
3
k +
5
6 +
7
8 +
9
10 +
11
12 +
13
Ik +
15
16 +
Variables*
2 34
Particle Aeration Fe2* Coneen-
Size Rate tratlon
.
_ - —
•f
+
+
+
+ +
+ +
+
+
+ - +
+ - +
+ +
- -f +
•f + +
+ + +
After
Fe2*
Removed,
percent
68
82
27
3^
55
25
11
32
ii^S
71).
6
22
62
80
32
16
5 min
FeT
Removed,
percent
62
80
16
23
k8
26
Ik
22
36
68
5
llj-
55
70
25
12
After
Fes+
Removed,
percent
16
61
18
11
>5
^5
Ik
15
22
23
15
10
37
38
5
12
120 min
FeT
Removed,
percent
5
k
5 .
k
k3
9
6
16
16
5
5
12
18
9
9
After
Fe3*
Removed
percent
17
6k
21
11
2
61
17
17
18
28
15
8
39
ko
2
10
180 min
Fef
, Removed,
percent
10
kl
6
9
13
k3
12
12
11
10
5
5
20
12
9
6
* Variable Identification
Levels
Variable
Surface Area
Particle Size
Aeration Rate
Fe3* Concentration
+
950-1100 ma/g
8 x 10
mesh
500 ml/ain
1000 mg/1
-
800 ma/j
20 x kO
I
mesh
50 ml/min
100 mg/1
-------
These main effects are listed in Table Ik. The absolute magnitude of the
values for the variable interactions were small; therefore, the inter-
actions were judged to be insignificant and the values were not presented
here. For these tests, the data indicated that the most significant
variable affecting percent Fes+ and FeT removed was Variable 2, particle
size of the carbon. The negative sign for that main effect simply meant
that iron was removed more effectively at the lower level of particle
size (20 x kO mesh in this experiment). It is doubtful whether this
effect from particle size would have been particularly apparent had it
not been for the calculation of this main effect as described. In any
event, the absolute magnitude of this, the most significant variable, is
only about one-half of the absolute magnitude of the most significant
variable^in the batch tests (Table 12) indicating a degree of importance
for particle size in the continuous flow tests less than that for aera-
tion rate, the most significant variable in the batch tests. ;
a
Variables 1 and k were similar in their effect on the continuous flow
tests (except for the first 5 min of a test). The negative sign on
Variable h simply meant that more effective .iron removal occurred at the
lower level of Fe2* (100 mg/1 in this experiment). The positive sign for
Variable 1 simply meant that more effective iron removal occurred at the
higher level;of surface area (950-1100 m2/g). In the evaluation of this
test, it must be pointed out that there was, at times during the test,
not much difference in the absolute magnitude of the main effects and
that the relative importance of the variables changed as the test
progressed.
It must- also be pointed out that considerable difficulty was experienced
in maintaining (a) a constant flow and (b) any flow at all during some
of the l6 tests. In fact, these tests were completed only by (a) stop-
ping the flow and aerating for 1 minute followed by (b) stopping the
aeration and draining the tube for two minutes. This was necessary in
Tests 2, 5, 6, 10, 13, and 14. Tests 5, 6, 13, and Ik involved the small
particle size carbon and high aeration rate. Tests 2 and 10 also in-
volved small particle size carbon, but low aeration rate. The use of the
carbon having high surface area in combination with the aforementioned
conditions made this unusual procedure necessary. The particular set of
conditions resulting in erratic or no flow through the column was not
discovered until well into the tests for this factorial experiment. The
experiment was completed but with reservations concerning the evaluation
of the data. The similarity of the signs (positive or negative) on the
variables from this experiment with those from the batch tests again
indicated that more efficient contact should have resulted in more
effective iron removal. In addition, the mechanical difficulties ex-
perienced served to indicate that the column used in these tests was not
adequate. It had already been established that the same type of flow
problems would not be experienced with a column of 3 in. diameter (as
compared to the column of 1 in. diameter used for the tests reported
here). A column having at least a 3 in. diameter was used in most of the
tests after this experiment.
53
-------
TABLE Ih. MAIN EFFECTS EASED ON DATA FROM
CONTINUOUS FLOW TESTS.
Response
Fe2* Removed,
percent
FeT Removed,
percent
Variable Main Effects Variable Main Effects
After 5 Minutes
gle
iable : 2
1
3
k
-38.88
7.38
- 5.62
0.38
2
1
3
k
-39.25
6.75
- k. 00
- 0.75
After 120 Minutes
gle
iable : 2
k
1
3
-23.38
- 7.88
5.38
k.38
2
1
k
3
-13.12
10.12
- 3.62
1.38
After 180 Minutes
gle
iable : 2
1
h
3
Designation of
-21.00
13.50
- 6.25
0.75
Variables : 1
2
1*
2
1
3
= Surface Area
= Particle Size
-12.00
- 8.50
6.50
3.75
= Aeration Rate
= Fe3* Concentration
-------
Statistical Design II
The second set of factorial experiments covered the effect of pH, and
concentrations of aluminum, manganese, and sulfate in the synthetic
coal mine water on removal of iron. Both batch and continuous flow
tests were conducted. Each series consisted of 16 tests. For reasons
discussed previously, only the second of two successive tests was
evaluated.
Batch tests — The data representing the batch tests are summarized in
Table 15. For ease of interpretation, again the four separate variables
are identified, with the high and low level of each variable being repre-
sented by + and - signs respectively. The procedure for evaluating this
factorial experiment was identical to that described previously. The
data from Table 15 were employed in the calculation of both main effects
and variable interactions for the eight responses using the codified
design and calculation matrix already presented in Table 11 and the pro-
cedure as already outlined. These main effects are listed in Table l6.
For these tests, the data indicated that the most significant variable
affecting percent Fe and Fem removed was Variable 1, the pH of the
synthetic coal mine water. The positive sign for that main effect
simply meant that the iron was removed more effectively at the higher
level of pH (3.5 in this experiment). This effect was particularly
apparent when one compares the results of the odd-numbered tests with
the results of the even-numbered tests in Table 15. In almost all cases,
the percent Fe3+ and Fe^ removed was greater in the even-numbered tests,
at the higher level of pH in the water, than in the odd-numbered tests,
at the lower level of pE in the water.
The magnitude of Variable k indicated that this variable also affected
iron removal. The positive sign on this variable indicated that iron was
removed more effectively at the higher concentration of this variable,
concentration of sulfate in the synthetic coal mine water. From the
batch tests, no other variable or interaction between variables was
judged to have an effect on iron removal.
Continuous Flow Tests — As with the batch tests, continuous flow tests
were conducted with a 2* factorial experiment and evaluated in the same
manner as before. The data representing these continuous flow tests are
summarized in Table 17. Two successive tests of 6-hour duration each
were conducted for each of the 2.6, and only the second test was evaluat-
ed for reasons already discussed. The data in Table 17 were again em-
ployed in the calculation of main effects and variable interactions for
each of the 12 responses using the codified design and calculation
matrix presented in Table 11 and the procedure as already outlined.
These main effects are listed in Table 18. For these tests, the data
indicated that the most significant variable affecting percent Fe and
Fe-r removed was Variable 1, the pH of the synthetic coal mine water.
The positive sign for that main effect meant that iron wa§ removed more
55
-------
TABLE 15. RESULTS OF STATISTICAL DESIGN II EXPERIMENT ON VARIABLES
AFFECTING IRON REMOVAL. BATCH TESTS.
Variables'*
After 5 min
After 30 min
After 60 min
After 120 min
123
Test pH Al Mn
1 -
2 + - -
3 - + -
4 + + -
5 - - +
6 + - +
i 7 - + +
8 + + +
9 - - -
10 + -
11 - + -
12 + +
13 - - +
14 + - -H
15 - + +
16 + + + -
Fe"**
4 Removed,
S04te~ _P_e_rcent
26
39
21
15
0
24
8
21
+ 16
+ 28
+ 14
+ 28
4- 20
+ 24
+ 26
+ 27
Fey
Removed,
percent
0
24
2
13
0
19
0
16
2
27
3
20
2
20
2
20
Pd3*
Removed ,
percent
26
53
20
31
8
52
5
34
20
56
14
51
27
53
32
36
FeT
Removed,
percent
0
48
0
30
0
51
0
33
2
51
3
43
2
48
2
33
Fed*
Removed,
percent
26
70
16
50
17
72
8
57
35
76
25
73
31
77
43
61
FeT
Removed,
percent
0
64
2
49
2
70
0
55
7
73
3
67
5
75
10
58
Fe*+
Removed,
percent
26
89
18
78
25
94
15
85
53
95
36
91
48
96
57
90
FeT
Removed,
percent
0
86
2
71
4
91
4
82
12
92
5
89
10
93
12
85
* Variable Identification
Levels
Variable +
pH
Aluminum
Manganese
Sulfate
3-5
200 mg/1
100 mg/1
8,000 mg/1
1.5
50 mg/1
25 Mg/1
450 mg/1
-------
TABLE 16. MAIN EFFECTS BASED ON DATA FROM BATCH TESTS
Single
Variable:
Single
Variable:
Single
Variable:
Single
Variable:
F6"
Removed,
percent
Variable Main Effects
1
3
If
2
1
2
If
3
1
if
2
3
1
2
3
After 5 Minutes
9.38
-4.62
3.62
-2.12
After 30 Minutes
26.75
-9.00
7-50
-3.00
After 60 Minutes
If 1.88
13.12
-8.88
-0.62
After 120 Minutes
55.00
17.00
-7.00
3.00
Designation of Variables: 1
2
3
U
Fey Removed,
percent
Variable Main Effects
1
if
2
3
1
2
If
3
1
if
2
3
1
If •
2
3
= Aluminum
= Manganese
= Sulfate
18.50
2.75
-2.25
-1.50
Ul.OO
-7.25
2.75
-1.00
60.25
7.00
-6.50
1.25
80.00
7.25
-if. 75
3.00
57
-------
TABLE 17. RESULTS OF STATISTICAL DESIGN II EXPERIMENT ON VARIABLES AFFECTING IRON REMOVAL.
CONTINUOUS FLOW TESTS,
Vari ables*
VJ1
00
2 3
Test pjK Al Mn S042'
1
2
3
5
6
7
8
9
10
11
12
13
lU
15
16
After 5 min
Fe2*
Removed,
percent
95
100
100
99
74
100
66
78
81
100
61
84
100
100
95
86
FeT
Removed,
percent
66
100
75
91
38
100
35
67
48
91
49
85
77
100
54
78
After
re2*
Removed,
percent
14
76
10
72
20
71
9
28
16
92
10
75
4i
97
24
68
60 min
FeT
Removed,
percent
0
62
0
68
0
56
5
17
1
83
9
70
16
83
7
55
After
Fe3*
Removed,
percent
13
55
18
62
12
58
13
25
21
91
19
85
16
88
38
67
120 min
FeT
Removed,
percent
4
45
0
53
0
43
0
18
0
76
4
70
7
87
4
66
* Variable Identification
Variable
PH
Aluminum
Manganese
Sulfate
,.v
3.5
200
100
8,000
Levels
+: ' ' .
mg/1
mg/1
rag/1
-
1-5
50 mg/1
; 35 mg/1
250 mg/1
After 180 min
Removed,
percent
6
48
16
60
15
50
8
21
19
80
19
85
19
79
28
71
Removed,
percent
1
40
1
52
4
43
0
19
6
70
2
72
4
72
6
62
-------
TABLE 17. RESULTS OF STATISTICAL DESIGN II EXPERIMENT ON VARIABLES AFFECTING
IRON REMOVAL. CONTINUOUS FLOW TESTS. (Continued)
Variables*
Test
1
2
4
6
7
8
9
10
11
12
13
14
15
16
1 2 3 ^4
JH Al Mn SO.2"
-t-
+
+
After ;
Fe2"1-
Removed,
percent
4
46
10
58
12
4o
0
21
9
7S
9
84
10
73
21
72
24o min
Fe^
Removed,
percent
3
35
1
51
5
0
19
4
70
if
76
5
65
4
68
After 300 min
Fe3*
Removed,
percent
13
33
0
64
10
50
0
20
7
80
l4
84
23
73
15
74
Fey
Removed,
percent
4
32
0
57
4
44
0
19
4
68
1
76
4
66
2
70
After
Fe3*
Removed,
percent
6
30
2
58
6
48
0
25
11
76
14
82
25
69
12
68
360 min
Fej
Removed,
percent
4
0
3
56
0
44
7
24
4
70
2
76
5
65
6
64
* Variable Identification
Variable
PH
Aluminum
Manganese
Sulfate
3.5
200
100
8,000
Levels
mg/1
mg/1
mg/1
-
1.5
50 mg/1
25 mg/1
250 mg/1
~
-------
TABLE 18. MAIN EFFECTS BASED ON DATA FROM
CONTINUOUS FLOW TESTS
Fe Removed, Fe^ Removedj
Single
Variable:
Single
Variable:
Single
Variable:
Single
Variable:
percent
percent
o>
H
aj
o
W
After 5 Minutes
2
1
3
-10.13
9.38
-2.63
-0.63
1
2
3
4
33.75
-10.75
-7.00
1.25
After 120 Minutes
3
2
47.63
21.13
-5.88
-3-38
1
4
2
3
54.88
18.88
-5.88
-3.38
After
Minutes
1
4
3
2
49.63
20.63
-6.13
0.38
1
4
3
2
50.25
17.25
-4.25
-1.00
After 360 Minutes
1
4
3
2
47.50
22.75
-3.25
-1.25
1
4
2
3
46.00
19.25
5.75
0.00
Fes* Removed, Fei Removed,
percent percent
Variable
1
2
1*
3
1
1*
3
2
1
1*
3
2
,
After
54
-16
15
-0
After
45
22
-5
-1
After
49
22
-3
-2
Effects
Ci
fi
•H
i
Effects
60 Minutes
.38
.38
.38
.88
180
.50
.00
.25
.00
300
.50
.50
.75
.25
1
4
2
3
57
14
-8
-6
.00
.50
•75
.75
Minutes
1
4
3
2
50
16
-4
-3
.75
.75
.25
.25
Minutes
1
4
3
2
51
16
-4
-0
.63
.38
.13
.13
Designation of Variables:
1
2
3
k
pH
Aluminum
Manganese
Sulfate
60
-------
effectively at the higher level of pH (3.5 in this experiment). The
importance of this variable remained apparent throughout the duration
of each 6-hour test as indicated by the relatively large value for that
main effect (See Table 18) .
Following approximately the first hour of the tests, the absolute magni-
tude of ^ Variable 4, concentration of sulfate in the synthetic coal mine
water, indicated that this variable influenced percent Fes+ and Fem re-
moved. The positive sign meant that more effective iron removal occurred
at the higher level of sulfate concentration. A comparison of the magni-
tude of this effect and the magnitude of the pH effect from Table 18 in-
dicated that pH had a much greater effect on iron removal than did
sulfate concentration. In addition, after about an hour, interaction
between Variable 1 and Variable k, pH and sulfate concentration, respec-
tively, was evident. From these continuous flow tests, no other variable
or interaction between variables was judged to have an effect on iron
removal. ;
In summary, both from the batch tests and continuous flow tests, the pH
of the synthetic coal mine water was the most important variable affect-
ing percent Fe * and Fer removed. Most effective iron removal was accom-
plished at the higher pH (3.5) in this experiment. This fact is consis-
tent with oxidation as the mechanism for iron removal since the rate of
oxidation of Fe2* is known to be greater at the higher pH. A greater
degree of adsorption at the higher pH, though, was also possible. Again,
both from the batch and continuous flow tests, the concentration of
sulfate in the synthetic coal mine water also affected percent Fe2* and
removed. More effective iron removal was accomplished at the higher
concentration of sulfate. This effect, though, was of lesser importance
than the effect from pH as evidenced by the relative magnitude of
Variables 1 and k from Table 18. The interaction between those two
variables might be explained in a number of ways . The presence of
(a) large amounts of sulfate and (b) the relatively high concentration
of H* at the pH range of this experiment might have interacted to form
the HS04~ species. The high concentration of sulfate, 8000 mg/1, might
also have simply saturated the system, and, therefore, forced the iron
out of the solution by precipitation onto the activated carbon resulting
in more effective iron removal. More extensive precipitation would have
occurred at higher pH.
An examination of the analyses for aluminum, manganese, and sulfate in
the synthetic coal mine water prior to and after the tests showed that no
significant removal of any of these three constituents was obtained with
activated carbon.
Additional Statistical Design Tests
A series of continuous flow experiments was designed to fill in gaps in
the data left by the Statistical Design I and II experiments.
61
-------
Effect of pH — Previous experiments had demonstrated more effective
iron removal at pH 3.5 than at pH 1.5. Experiments were then conducted
with synthetic coal mine waters containing approximately 250 mg/1 of
Fes* and pH values of 1.5, 2.0, 2.5, 3.0, and 3.5. The results are
listed in Table 19. Only the results of the second of two tests at
each pH level are presented for reasons discussed previously. The data
from Table 19 on removal of Fe3* were plotted in Figure 9. From these
data, tests at low pH, 1.5 and 2.0, did not result in effective removal
of Fe3+ for any reasonable period of time. In fact, at these pH values,
a level of removal of 50 percent was achieved only for a very few min-
utes. At the higher pH values of 2.5, 3.0, and 3.5, a 50 percent level
of removal was achieved for over 2 hours.
The data from Table 19 on removal of total iron, Fe^, including both
Fe2* and Fe3* are presented in Figure 10. As in the evaluation of Fe
removal, more effective removal of Fe^ occurred at the higher level of
pH.
Effect of No Manganese — Previous tests have demonstrated that a high
or a low level of manganese in the synthetic coal mine water had no
effect on removal of iron. This was carried one step further by con-
ducting a test with no manganese in the synthetic mine water. The re-
sults of this test are presented in Table 20 and compared in the same
table with the results of a past experiment from the Statistical Design
II set (See Table 17, Test No. 12). The results of the two tests, as
listed in Table 20, were judged to be similar, thus reenforcing the
conclusion that the concentration of manganese in the water had no
effect on iron removal. The data from Table 20 are compared in
Figure 11. The differences in iron removal during this test with no
manganese and during Test No. 12 were judged, from this figure, to be
small.
Effect of No Aluminum — Previous experiments have demonstrated that a
high or low level of aluminum in the synthetic coal mine water had no
effect on removal of iron. In a manner similar to the test with no
manganese, a test was conducted with no aluminum in the synthetic coal
mine water. The results of this test are presented in the first part of
Table 21 and these data are also plotted in Figure 11. The results of
the test with no aluminum were judged to be similar to the results of
the standard test, Test No. 12, as described in Table 20 and Figure 11,
reenforcing the conclusion that the concentration of aluminum in the
water had no effect on iron removal.
Effect of No Aeration — Previous experiments have demonstrated that a
high rate of aeration favorably affects the removal of iron. A test
was conducted in which the water-carbon mixture in the column was not
aerated for the duration of the 6-hour test. The results are presented,
also, in Table 21 and Figure 11. From the data, the effectiveness of
iron removal was judged to have decreased drastically in comparison with
the standard test, Test No. 12.
62
-------
TABLE 19. EFFECT OF pH ON REMOVAL OF IKON
Experiment
No. UUH-7
pH 1.5
Time,
min
0
30
60
120
180
2to
300
Percent
Feiv
100
5U
25
17
1U
3
U
removed
FeT
100
1*
20
11
7
5
11
Experiment
No, Wf-11
pH 2.0
Percent
Fe*"p
100
1*
32
20
10
21
28
removed
FeT
100
29
16
0
0
0
1
Experiment
No. MA-lU
pH 2.5
Percent
Fea+
100
70
63
61
52
50
ho
removed
FeT
100
53
U3
28
18
13
11
Experiment
No. Wt-12
pH 3.0
Percent
Fe3*
100
6ki
65
63
62
60
^9
removed
FeT
100
h6
U8
38
31
26
19
Experiment
No. W±-9
PH 3.5
Percent
Fei+
100
71
62
56
UU
Ul
31
removed
FeT
100
6k
55
U7
35
35
26
-------
10
REMOVED,
PERCENT 40
PH 1.5^ ^^-_
1 60 120 180
1 III HIIMM.HI1I III.MI.III •
pH 2.
— —
240 30
DURATION OF TESTS AT CONSTANT FLOW OF WATER, MINUTES
Bituminous Cool Research, Inc. 2042G41
Figure 9. Effect of pH on Removal of
-------
VJI
FeT
REMOVED,
PERCENT
180
240
300
DURATION OF TESTS AT CONSTANT FLOW OF WATER, MINUTES
Biruminous Coal Research, Inc. 2042G42
Figure 10. Effect of pH on Removal of Fey
-------
TABLE 20. EFFECT OF NO MANGANESE ON REMOVAL OF IRON
Test with No Manganese
Experiment No. 444-15
o\
0
5
10
15
30
45
60
90
120
150
180
210
240
270
300
330
360
_ 24* _
Fe Fe
pH
3.5
4.3
4.3
4.3
4.3
4.2
4.2
4.2
4.2
4.2
4.1
4.1
4.1
4.1
4.1
4.1
4.0
mg/1
256
6
0
6
16
26
29
13
23
29
37
45
45
55
58
62
58
Percent
removed
—
__
--
94
»
89
._
91
•• •»
86
--
82
• •»
77
__
77
mg/1
269
16
13
19
26
32
39
29
29
49
57
65
65
68
68
71
71
Percent
removed
--
--
__
90
«•
86
_-
86
• •»
T9
__
76
«*»
75
__
74
Standard Test for Comparison Ex-
periment No. 443-81 (Test No. 12)
.PL
3.5
4.5
4.4
4.3
4.3
4.3
4.3
4.2
4.2
4.2
4.2
4.2
4.2
4.1
4.1
4.1
4.1
mg/J
248
40
23
33
45
53
63
33
38
38
35
33
40
45
40
40
45
Fea+
Percent
L removed
--
__
on *•
82
••^
75
__
85
«
-------
Fe2+ REMOVED,
PERCENT
120-,
100
80-
60-
40-
20-
No Al
Straight Lines by Method of Least Squares
Test No. 12
(High Sulfate)
Low Sulfate, High NaCI
No Sulfate
Low Sulfate, Low NaCI
60 120 180 240 300 360
DURATION OF TESTS AT CONSTANT FLOW OF WATER, MINUTES
Bituminous Coal Research, Inc. 2042G43
Figure 11. Comparison of Effects of Various Process
Variables on Removal of Iron
67
-------
TABLE 21. EFFECT OF NO ALUMINUM AND NO AERATION ON REMOVAL OF IRON
O\
CD
Test -with No Aluminum
Experiment No. 444-30
Test with No Aeration
Experiment No. 444-22
Fe2* FeT
Time,
min
0
5
10
15
30
45
60
90
120
150
180
210
240
270
300
330
360
PH
3.5
6.5
6.4
6.4
6.2
6.3
5.9
4.*9
4.8
4.8
4.7
4.6
4.6
4.5
4.5
4.5
mg/1
244
0
0
0
0
0
0
0
0
13
26
31
36
55
62
65
68
Percent
removed
*•«•
--
100
_ _
100
--
100
MM '
89
— -
85
MM
75
--
72
mg/1
263
0
0
0
0
0
0
0
6
25
32
37
42
62
68
75
81
Percent
removed
—
__
100
„ „„
100
—
98
_ _
88
_—
84
__
74
--
69
.PH.
3.5
4.7
4.7
4.6
4.5
4.4
4.4
4.4
4.3
4.3
4.3
4.2
4.2
4.2
4.2
4.2
4.2
Fe2+
i
mg/1
273
62
71
68
107
175
205
227
224
218
224
221
218
227
250
263
256
FeT
Percent
removed mg/1
—
_-
61 -
MM
25
<•••*
18
M««
18
MW
20
M-M
8
M M
6
286
88
97
97
123
188
211
240
237
250
256
256
256
,244
256
266
276
Percent
removed
—
__
57
M**
26
__
17
M»%
10
«.*,
10
M«ta
10
«.*»
3
-------
Removal of iron by adsorption should not have required aeration yet
this test demonstrated aeration to be necessary. This is an addi-
tional indication that iron removal was accomplished by an oxidation
mechanism in this process, or, at least, by a combination of oxidation
and adsorption.
To further explore this aeration effect, continuous flow tests were also
conducted according to the general procedure with actual coal mine water
and with air, nitrogen, and oxygen, each at flow rates of 150 ml/min and
with no air. The results of these tests are presented in Table 22 and
the data concerning Fes+ plotted in Figure 12. From these data, tests
with nitrogen and no air resulted in ineffective removal of iron. As in
past tests, aeration aided in removing iron more effectively. The most
effective removal occurred in the test where: oxygen was bubbled through
the carbon-water mixture.
During this test and the test with nitrogen sparging, the dissolved oxygen
content of the mine water was k to 6 mg/1, having decreased from an
original value of 10 mg/1 in the raw water; therefore, the improved
effectiveness cannot be attributed to an increase in dissolved oxygen
content of the water from the use of oxygen. The mechanism could involve
adsorption of oxygen by the carbon and subsequent reaction by the dirad-
ical -0-0- thus formed (26) to extract electrons from the adsorbed
ferrous ions.
Effect of No Sulfate — Previous tests have indicated that a high concen-
tration of sulfate favorably affects the removal of iron. A test was
conducted in which sulfate was omitted from the synthetic coal mine water
by adding the Fe3* as ferrous chloride instead of as ferrous sulfate as
was normally done. The results are presented in Table 23 and the data
also plotted in Figure 11. Also included in Table 23 and Figure 11 are
results of a test with low concentrations of both sulfate and NaCl and a
test with a low concentration of sulfate and a high concentration of
NaCl. The test with no sulfate present in the synthetic coal .mine water
resulted in significantly poorer iron removal as judged by comparison
with data from the standard test, Test No. 12, in Table 20 and Figure 11.
This seemed to reenforce the original conclusion that a high concentra-
tion of sulfate favorably affected iron removal.
At this point, there was concern on the part of the BCR investigators as
to whether, in fact, sulfate itself was affecting iron removal. The con-
centration selected as the high level of sulfate for the Statistical
Design II experiment was 8,000 mg/1 since concentrations of sulfate of
this level are not too uncommon in actual coal mine waters. But this was
a higher concentration than that of any other species chosen for these
tests. It was possible that the high total concentration of material in
the water was resulting in a somewhat saturated solution and that this
saturation was, itself, responsible for driving the iron out of solution.
A test was planned, then, with a high concentration of ions other than
sulfate.
69
-------
TABLE 22. EFFECT OF AERATION, NO AERATION, NITROGEN SPARGING,
AM) OXYGEN SPARGING ON REMOVAL OF IRON
Aeration
Experiment No. 444-88
No Aeration
Experiment No. 444-90
Fes* FBT Fes*
Time,
min
0
30
60
120
180
240
300
360
EIL
2.8
4.3
4.3
4.2
4.1
4.1
4.1
4.2
mg/l
305
175
198
231
234
224
224
218
% Re-
moved
—
43
35
24
23
27
27
29
% Re-
mg/1 moved
429
188
205
240
253
234
244
234
—
56
52
44
41
45
43
45
PJL
3.0
4.3
4.2
4.2
4.1
4.1
4.1
4.1
Nitrogen Sparging
Experiment No. 444-92
Fes*
Time,
min
0
30
60
120
180
240
300
360
£!_
2.9
4.4
4.2
4.3
4.3
4.2
4.1
4.1
ng/1
175
62
133
169
198
198
185
188
% Re-
moved
—
65
24
3
0
0
0
0
FeT
% Re-
mg/1 moved
279
78
159
192
208
205
192
205
__
72
43
31
25
27
31
27
pH_
2.9
4.3
4.3
4.2
4.1
4.0
3.9
3.9
mg/l
240
94
175
224
234
263
263
263
^Re-
moved
—
61
27
7
3
0
0
0
FeT
mg/l
334
127
198
253
261
273
273
273
% Re-
moved
—
62
41
24
22
18
18
18
Oxygen Sparging
Experiment No. 444-94
Fe
s£
224
26
71
84
73
75
97
78
3+
% Re-
moved
—
88
68
63
67
67
57
65
FeT
mg/l
237
39
94
110
94
101
120
94
% Re-
moved
—
84
60
54
60
57
49
60
70
-------
Fe2+ REMOVED, PERCENT
100H
90-
80-
60
120
180
240
300
360
DURATION OF TESTS AT CONSTANT
FLOW OF WATER, MINUTES
Bituminous Coal Research, Inc. 2042G44
Figure 12. Effect of Aeration, No Aeration, Nitrogen, Sparging
and Oxygen Sparging on Removal of Iron
71
-------
TABLE 23. .EFFECT OF NO SULFATE AND VARIOUS CONCENTRATIONS OP NaCl ON REMOVAL OF IRON
-3
ro
Test with No Sulfate
Experiment No. 444-36
Fed
FeT
Test with Low Sulfate, Low
NaCl Experiment No. 444-40
FeT
Test with Low Sulfate, High
NaCl Experiment No. 4U4-46
FeT
Time, Percent Percent
min pH mg/1 Removed tng/1 Removed
Percent Percent
pH mg/1 Removed mg/1 Removed
Percent Percent
pH mg/1 Removed mg/1 Removed
0
5
10
15
30
45
60
90
120
150
180
210
240
270
300
330
360
• i
4.3
4.5
4-3
4.3
4.2
4.2
4.2
4.2
4.2
4.1
4.1
4.1
4.1
4.1
4.1
4.1
•••^^••••aiM
218
13
0
13
45
91
110
127
123
159
133
130
127
127
110
114
120
—
—
79
50
—
44
—
39
42
__
50
__
45
•™ ^Hd^*MH^B
240
23
23
29
65
104
133
140
146
175
159
151
143
143
136
143
133
—
—
73
45
—
39
34
40
—
43
—
45
•tr-tru) 1
. . . 1
U) U) VJ1 |
4.3
4.2
4.1
4.1
4.0
4.2
i»-.o
4.0
4.0
4.0
3.9
3-9
4.0
4.0
^••dVofe^vw
253
13
16
26
45
29
101
110
127
143
140
143
146
133
127
127
133
—
—
73
—
49
—
45
—
34
--
36
—
45
—
47
A HfflWBIBH^M
273
36
36
45
68
104
130
130
140
166
166
164
162
150
140
143
133
—
—
75
—
52
--
49
—
39
--
41
—
49
—
51
••W-HI^
3.5
4.3
4.5
4.3
4.2
4.2
4.2
4.2
4.1
4.1
4.0
4.0
4.0
4.0
4.0
3-9
, 3-9
IVi^VB^BV-v
273
36
45
55
52
49
52
58
68
71
81
73
65
78
45
45
42
--
81
—
81
—
75
—
70
--
76
--
84
—
85
A IMi^^hMMBV^
279
68
78
81
68
81
81
120
114
10k
107
108
110
91
58
58
58
—
--
76
—
71
—
59
—
62
—
61
• --
79
--
79
-------
Sodium chloride, NaCl, seemed to be a logical choice as a replacement
for (magnesium) sulfate. A test was first conducted with low, concentra-
tions of both WaCl and sulfate to observe any slight effect of NaCl on
removal of iron. The results of this test are included in Table 23
and Figure 11. The effect of the aforementioned was judged to be
similar to the effect when no sulfate was used for the test; in other
words, the overall result was ineffective removal of iron. The NaCl
itself did not seem to affect the iron removal.
For the test with a high level of NaCl, the concentration of NaCl was
selected so that the overall concentration of ions, the total ionic
strength, in the synthetic coal mine water would be equivalent to that
of the water used in the standard test, Test No. 12. The results of
this test are also presented in Table 23 and Figure 11. From the data,
the results of this test were judged to be the same as the results of
the standard test. Either a high concentration of sulfate or a similarly
high concentration of NaCl resulted in effective removal of iron. What
was observed was the effect of a high concentration of ions regardless of
their nature, and not specifically the effect of sulfate concentration.
Effect of Bed Depth
The results of a series of tests with various amounts of carbon in the
19 in. x 3 in. column, resulting in different depths of the bed of
carbon, are presented in Table 2k. From the results in Table 2k, the
greater the depth of the bed of carbon, the more effective was the iron
removal. This can be seen by a plot of the percent Fes* removed as
shown in Figure 13 , and a plot of the percent Ferp removed as shown in
Figure 1^. A comparison of the time during which 50 percent of the iron
was removed from each of these figures demonstrates this particular
point.
That even 25 percent of the Fes* was not removed with the 3.5 cm-deep
column can be seen from Figure 13. Again from Figure 13, the 9.2 cm
deep column effected 50 percent iron removal for only a very short+
period of time. Using the time during which 50 percent of the Fe is
removed for each of the other three columns, 15.6, 21.9, and 27.9 cm
deep, from Figure 13, the depth in centimeters necessary to achieve 50
percent removal for a time period of 1 minute was calculated to be 0.71,
0.72, and 0.72, respectively, for an average of 0.72 cm/min of 50 percent
removal.
A similar calculation for total iron removal, Fej, from Figure iH yields
values of 0.72, 0.72, and 0.73, respectively, for the 15.6, 21.9, and
27.9 cm deep columns for an average for 0.72 cm/min of 50 percent removal
of Pap.
Since the water used in these tests contained approximately 250 mg/1 of
iron and this water flowed through the column at a rate of 30 ml/min,
the value of 0.72 cm/min was also calculated to be 0.72 cm/ 30 ml
73
-------
TABIE 2k. EFFECT OF BED DEPTH ON REMOVAL OF IRON
80g Carton
3.5cm Depth
Experiment
Number
W*-38
Time,
min
0
5
30
60
120
180
2kO
300
360
Percent
Fe"""
100
^3
22
18 •>
12
19
19
Ik
19
removed
Ferj
100
35
22
. 13
13
13
12
9
7
260g Carbon
9.2cm Depth
Experiment
Number
W*l*-37
Percent
Pea+
100
82
U5
3k
k2
29
21*
31
28
removed
FeT
100
76
k7
38
39
26
21
32
28
W*0g Carbon
15.6cm Depth
Experiment
Number
-1*1*1*-31
Percent
Fea*
100
100
100
62
36
30
23
21
23
removed
Fej
100
100
96
59
3^
26
21
20
23
620g Carbon
21.9cm Depth
Experiment
Number
1^-27-
Percent
Fea+
100
100
100
85
5k
ko
36
31
19.
removed
FeT
100
100
100
79
U9
36
33
29
J2O
800g Carbon
27.9cm Depth
Experiment
Number
kkk-23
Percent
Fei+
100
100
100
100
75
5^
k2
36
27 .
removed
Fem
100
100
100
100
73
U8
33
33
28
-------
Fe*"1"
REMOVED,
PERCENT
50 Percent Fe^"1" Removed
\
60 120 180 240 300
DURATION OF TESTS AT CONSTANT FLOW OF WATER, MINUTES
360
Bituminous Coal Research, Inc. 2042G45
Figure 13. Effect of Bed Depth on Removal of Fe2 +
-------
FeT
REMOVED,
PERCENT
50 Percent Fey Removed
60
120
180
240
300
360
DURATION OF TESTS AT CONSTANT FLOW OF WATER, MINUTES
Bituminous Coal Research, Inc. 2042G46
Figure 14. Effect of Bed Depth on Removal of-Fey
-------
(0.021* da/mi) or 0.72 aft/7;5 mg of iron (0.0096 cm/mg of iron) for 50
percent removal of iron.
Effect of Column Diameter
Three molded acrylic tubes of 3 in., k in., and 5 in., diameter were
used in a series of tests, each with standard synthetic coal mine water
and each with a separate 800 g sample of Nuchar WV-W carbon, to examine
the effect of diameter of the column on iron removal. The results of
these three tests are summarized in Table 25. Examination of the data
in Table 25 reveals that the rate of removal of iron was similar regard-
less of the diameter of the column.
The same tubes were also used in similar tests with an actual coal mine
water. The results\of these tests are summarized in Table 26. The data
showed little influence of column diameter on rates of removal of iron.
Effect of Temperature
Three continuous flow tests were conducted according to the general pro-
cedure and with standard synthetic coal mine water at temperatures of 12,
21, and 30C, to examine the effect of temperature on removal of iron.
The results are summarized in Table 27, and the data plotted in Figures
15 and 16. The data showed little difference in iron removal between
the tests at 12C and 21C, but a substantial difference in the test at
30C. More effective iron removal at the higher temperature can be seen
from these figures.
Use of Materials Other Than Carbon
In the first test of this series, activated alumina was used in place of
the carbon. The results of this test are presented in Table 28. The
iron could be seen as a yellow precipitate on the top two inches of
alumina in the column. The data in Table 28 also indicated ineffective
iron removal, at least in this single test of relatively short duration*
In the second test of this series, charcoal briquettes were used in place
of the carbon. The results of this test are presented in Table 29. The
extremely effective removal of iron was attributed to the high pH result-
ing from the briquettes. An attempt was made after this test to lower
the pH by passing IN IfeS04 through the briquettes in the column. This
only resulted in clogging the column, a fact which precluded the use of
this material for the process.
The use of activated alumina and coal were examined in connection with
bacteria and the results of these tests are described later in this
report.
77
-------
00
TABLE 25. EFFECT OF DIAMETER OF COLUMN ON REMOVAL OF IRON
FROM SYNTHETIC COAL MINE WATER
3 in. Diameter U in. Diameter 5 in. Diameter
Experiment No. UUU-68 Experiment No. Wt-57 Experiment No. UUU-66
Fe3+
Time,
min
0
30
60
120
180
240
300
360
mg/1
268
0
45
68
120
169
153
175
Percent
removed
-
100
83
75
55
37
43
35
FeT Fea*
mg/1
276
0
62
127
136
192
172
192
Percent
removed
-
100
78
54
51
30
38
30
mg/1
253
0
32
97
127
133
153
172
Percent
removed
-
100
87
62
50
47
40
32
Tjflrffc T^*^^^^^ ^71 MK^^
reip re rey
mg/1
263
16
58
136
136
156
179
188
Percent
removed
-
94
78
48
48
41
32
29
mg/1
263
0
13
78
133
146
179
195
Percent
removed
-
100
95
70
49
44
32
26
mg/1
292
0
36
117
169
188
1B5
208
Percent
removed
-
100
88
60
42
36
37
29
-------
TABLE 26. EFFECT OF COLUMN DIAMETER ON REMOVAL OF IRON
FROM ACTUAL COAL MINE WATER
r\J_
FBT
mg/1 Percent Removed mg/1 Percent Removed
3 Inch Column
28 cm Carbon Depth
Experiment No. U51-U2
0 3.0 2UO ~ 33U
30 U.3 107 55 127 62
6t> U.2 123 U9i iU3 57
120 U.2 156 35 179 U6
l6p U.I - 162 33 182 U6
2UO U.I 172 28 188 UU
300 U.O 192 20 211 37
360 U.O 192 20 208 38
U Inch Column
18 cm Carbon Depth
Experiment No.
0
30
60
120
180
2UO
300
360
2.9
U.5
U.U
U.3
U.2
U.2
U.2
U.I
175
65
130
153
169
169
172
169
W«B
63
26
13
3
3
2
3
68
153 U5
172 38
179 36
175 37
192 31
185 3U
5 Inch Column
11 cm Carbon Depth
Experiment No. U51-U6
26 195 ^
120 U.3 175 22 205 M
180 U.2
2UO U.2 192 - , -
300 U.2 221 1 2UO 31
360 U.I 237 0 24° ^
185 17 211 3
79
-------
TABLE 27. EFFECT OF TEMPERATURE ON REMOVAL OF IRON
Low Temper attire (12 C)
Experiment No.
High Temperature (30 C)
Experiment No. U51-15
Ambient Temperature (21 C)
Experiment No. U51-10
34" 3H" 2Hh
Time,
min
0
30
60
120
180
2^0
300
360
mg/1
273
6
U2
117
153
162
175
185
Percent
removed
98
85
57
Ul
36
32
mg/1
292
13
78
179
188
201
Percent
removed
96
73
39
36
31
27
3fflg/X
Pilll
0
0
16
39
78
So
Percent
removed
100
100
93
Qh
68
53
mg/1
282
0
0
23
62
130
156
Percent
removed
100
100
92
78
63
*5li
|rCJ
253
10
75
Uh
136
153
169
192
Percent
removed
96
70
55
1*6
ko
33
2k
mg/1
263
39
81
166
175
188
208
Percent
removed
85
69
37
33
21
-------
Fe2+ REMOVED,
PERCENT
High Temperature
Low Temperature
Ambient Temperature
60 120 180 240 300 360
DURATION OF TESTS AT CONSTANT FLOW OF WATER, MINUTES
Bituminous Coal Research, Inc. 2042G47
Figure 15. Effect of Temperature on Removal of
81
-------
FeT REMOVED,
PERCENT
lOO-i
80-
60-
40-
20-
High Temperature
Ambient Temperature ^Sl"»^
—T~
60
T
T
T
0 60 120 180 240 300 360
DURATION OF TESTS AT CONSTANT FLOW OF WATER, MINUTES
Bituminous Coal Research, Inc. 2042G48
Figure 16. Effect of Temperature on Removal of FeT
82
-------
TABLE 28. EFFECT OF AIAMI1& ON REMOVAL OF IRON
Fe3* FeT
Percent Percent
pH nag/1 removed mg/1 removed
.. •_&MH>^ ••!• l/«l»lll Ilium* ^«W«^M»B • n II in AIM .1.1 III IB* • I il i»m II • Will
t
3.0 256 - 289
k.k 133 ^8 179 38
k.k 185 28 198 31
k.k 198 23 221 2k
k.b 198 23 221 2k
k.k 192 25 201 30
k.k 198 23 221 2k
k.k 192 25 201 30
k.k 185 28 198 31
k.k 192 25 198 31
k.k 182 29 198 31
k.k 182 29 192 3^
k.k 175 32 185 36
U.U 166 35 175 39
k.k 169 3^ 175 39
k.k 166 35 175 39
k.k 167 35 175 39
83
-------
TABLE 29. EFFECT OF CHARCOAL BRIQUETTES ON REMOVAL OF IRON
Fe3*
Time,
min
0
5
10
15
30
^5
60
90
120
150
180
210
2kO
270
300
330
360
pH
3.0
8.0
8.0
8.0
7.8
7.7
7.6
7.*
7.3
7.2
7.0
7.0
7.1
7.0
6.9
6.8
6.8
mg/1
263
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Percent
removed
...
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
mg/1
292
0
0
0
0
0
0
0
0
0
0
0
0
0
10
10
23
Percent
removed
100
100
100
100
100
100
100
100
100
100
100
100
100
96
96
92
-------
Tests with Actual Coal Mine Waters
Studies on this project with synthetic coal mine waters have demonstrat-
ed that the rate of removal of iron from the waters tested decreased
drastically when the pH of that water was 2.5 or lower. When the pH of
the water to be treated by this process is higher than h.O to k-.5, then
significant precipitation of the iron compounds would occur, thereby
blocking the activated sites on the carbon to further reaction and
physically clogging the carbon column. Therefore, the coal mine waters
containing Pe which are most amenable to treatment by this process
should have -a pH between 2.5 and k.O. In fact, coal mine waters having
a pH below 2.5 are relatively rare, and those having a pH above k.O
usually present little difficulty in treatment. Difficulties in treat-
ment usually occur in those coal mine waters in the pH 2.5 to k.O range,
and this process was designed to overcome some of the difficulties.
Two coal mine waters having the above characteristics desirable for this
process were found and used in continuous flow tests. For the first
series of tests, water was collected from the Tarrs discharge, which is
near Tarrs, Westmoreland County, Pennsylvania and which has been used in
past projects at BCR. In the past, the water typically contained ko to
100 mg/1 of iron as Fe2* , an additional 50 to 80 mg/1 of iron as Fes* ,
50 to 100 mg/1 of aluminum, and had a pH of approximately 3. As sampled
for these tests, the water contained only 13 mg/1 of Fe , V? mg/1 of
Ferp, and had a pH of 3.0. Although the Fes+ was considered lower than
desired for these studies, this water was used in one test, stored
overnight at a temperature of 1 C, and used the next day in a second
test. The results are listed in Table 30. Throughout the duration of
the first two tests, «n iron was removed. The removal was more
effective than with the synthetic coal mine water, but the low Initial
concentration of iron may have influenced the effective removal.
After the water had been stored at a temperature of 1 C for a weekend,
iron (FeS04) was added to the water to increase the concentration of
Fe2* to the values shown in the last two columns in Table 30, and a third
test was conducted with this water. The results of this test showed
that a 50 percent level of iron removal, either Fe or Fe,p, was attained
throughout the duration of the 6-hour test. No further tests were con-
ducted using this coal mine water.
For the second series of tests, water was collected at various points
along a tributary of Little Plum Creek in Allegheny County, Pennsylvania,
about five miles from BCR. The analyses of the first samples obtained at
five locations along the stream are presented in Table 31. Since sampl-
ing points No. 3, 4, and 5 were progressively further downstream, they
were more affected by weather conditions and by dilution from run-off.
Tests were not conducted with water from points h and 5 both for this
reason and because of difficulties in accessibility to the stream at
these points.
85
-------
TABLE 30. PRELIMINARY CONTINUOUS FLOW TESTS WITH TARRS COAL MINE WATER
Experiment Number Experiment Number Experiment Number
UUU-U2 UUU-U3 UUU-U7
Fe3* Fe-p Fe3* Fey Fe3* Fej
pH mg/1 mg/1 pH
0
5
10
15
30
U5
60
90
120
150
180
210
2UO
270
300
330
360
3.0
6.9
7.2
7.5
7.9
7.8
7.5
7.U
7.U
7.2
7.1
7.0
6.7
5.3
5.0
—
—
13
0
0
0
0
0
0
0
0
0
0
0
0
0
0
—
—
U5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
—
—
3.0
6.2
6.U
6.3
6.0
5.2
U.8
U.6
U.6
U.5
U.5
U.6
U.8
li IL
]i C
IL Ji
Ji O
13
0
0
0
0
0
0
0
0
0
0
0
0
O
0
0
0
U5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
3.1
U.6
u.u
u.u
4.5
U.5
U.5
U.3
U.2
U.I
U.o
U.O
U.2
U.o
u.o
U.o
3.9
266
0
O
0
0
0
0
55
68
nu
120
130
127
136
133
136
1U3
3U7
0
0
0
0
0
13
8U
10U
133
136
1U3
1U9
153
159
159
159
-------
TABLE 31. ANALYSES AT VARIOUS POINTS ALONG
TRIBUTARY OF LITTLE PLUM CREEK
a+ Dissolved
Sampling Fe PeT . Oxygen,
Point No. pH mg/1 mg/1 mg/1
1 3.0 390 516 8.5
2 3.^ H7 153 9,1
3 3.3 156 198 11.6
k 3.U 107 123 12.0
5 If .6 62 71 11.5,
87
-------
Continuous flow tests were conducted according to the general procedure
with water from sampling points 1, 2, and 3. The results of three
tests each with these waters are presented in Tables 32, 33, and 34.
Under the conditions specified by the general procedure, satisfactory
removal of either Fe2* or Fe^ was not achieved from the water contain-
ing relatively high concentrations of iron from sampling point No. 1,
as indicated by the data in Table 32. A 50 percent level of iron re-
moval was attained only for a very short duration of time during these
tests. Similarly, the data from tests with water from sampling point
No. 2, Table 33, show that a 50 percent or greater level of iron re-
moval for any significant time was attained only in the first of the
three tests. In fact, the results of the last of the three tests shown
in Table 33, indicated elution of some of the iron which had been ad-
sorbed and/or oxidized during the first two tests with this water. For
these tests, a fresh sample of carbon was used for the waters from each
sampling point and that carbon was then used for all three tests with
the same water.
The results in Table 3^ of tests with water from sampling point No. 3
in the last experiment of this series, indicate effective removal
throughout the duration of the three tests and even slightly improved
results with each of the three succeeding tests. The more effective
removal of iron from water at sampling point No. 3, which is farther
downstream, could be attributed to the dissolved oxygen content of the
water, since this water had a concentration of 11.6 mg/1 as compared to
8.5 mg/1 for water from sampling point No. 1 and 9.1 mg/1 for water
from sampling point No. 2 (See Table 31)• In addition, each water also
contained a different initial concentration of iron.
Effect of Bacteria
Active cultures of Thiobacillus ferrooxidans, iron-oxidizing bacteria,
and Thiobacillus thiooxidans, sulfur-oxidizing bacteria, were obtained
from Indiana University of Pennsylvania at the 48-hour period of growth,
the time at which they were said to be at the peak of their activity
(27). Standard synthetic coal mine water was inoculated with the sus-
pension of Thiobacillus ferrooxidans in a 10/1 ratio of water/suspen-
sion. A 500 ml portion of the inoculated synthetic coal mine water was
aerated and the sample analyzed for Fes
-------
TABLE 32. RESULTS OF PRELIMINARY CONTINUOUS PLOW TESTS WITH
TRIBUTARY OF LITTLE PLUM GREEK. SAMPLING POINT NO. 1.
Experiment No. 1*51-3**
Experiment No. 1*51-35
Experiment Ho. 1*51-36
CO
VQ
Fes+ FeT
Time,
min
0
30
60
120
180
21*0
300
360
PL
2.9
7.0
6.6
M
U.8
l*.l*
1*.3
«*.3
mg/1
279
0
0
91
172
188
192
211*
Percent
removed
—
100
100
67
38
33
31
23
mg/1
U03
0
10
111*
201
2n
221*
2l*0
Percent
removed
—
100
98
72
50
1*8
1*1*
1*0
2.8
li O
1|, O
1*.2
1*.2
1*.2
U.2
l*.l
Fe3* Ferp
mg/1
289
71
175
231
263
263
276
286
Percent
removed
—
75
39
20
9
9
1*
1
mg/1
91
205
253
286
289
295
299
Percent
removed
—
77
*9
37
29
28
27
26
Fes+
pH mg/1
2.9 276
1*.2 10i*
1*.2 169
l*.l 198
1*.2 229
l*.l 21*1*
i*.o 256
1*.0 273
Percent
removed
—
62
39
28
17
12
7
1
mg/1
377
130
192
227
21*8
276
273
276
FeT
Percent
removed
—
66
U9
1*0
3*
27
28
27
-------
TABLE 33- RESULTS OF PRELIMINARY COKTIHUOUS FLOW TESTS WITH
TRIBUTARY OF LITTLE PLUM CREEK. SAMPLIBG POINT NO. 2.
Experiment No. Uj?l-21
Experiment No. U51-26
Experiment No. 1*51-27
Fea* FeT Fe2* FeT
Time,
min
0
30
60
120
180
2UO
300
360
.PJL
8.1
7.9
7.5
5.3
U.6
U.5
u.u
mg/1
117
0
0
0
0
3
3
3
Percent
removed
—
100
100
100
100
97
97
97
mg/1
153
0
0
0
0
29
23
19
Percent
removed
—
100
100
100
100
81
85
88
pH mg/1
3.0 11U
U.6 13
U.3 32
U.3 91
U.3 99
U.2 101
U.I 101
U.l 101
Percent
removed mg/1
—
89
72
20
13
11
11
11
282
U2
91
120
13U
133
133
129
Percent
removed
—
85
68
57
52
53
53
5U
PL
2.9
U.2
U.O
U.O
U.o
3.6
Fe8*
mg/1
123
55
107
166
178
192
208
2lU
Percent
removed
—
55
13
0
0
0
0
0
FeT
mg/1
383
78
123
179
20U
218
22U
231
Percent
removed
—
80
68
53
U7
U3
U2
Uo
-------
TABLE 3k. RESULTS OF HRELBHNARY CONTINUOUS FLOW TESTS WITH
TRIBUTARY OF LITTLE PLUM CREEK. SAMPLING POINT NO. 3-
Experiment No. UUU-8U
Experiment No. UUU-85
Experiment No. UUU-86
Time,
min
0
30
60
120
180
2UO
300
360
Pe
pH ng/1
3.0 55
7.8
7.0
5.0
U.8
U.5
U.U
U.3
0
0
0
13
26
23
29
3+
Percent
removed
—
100
100
100
76
53
58
U7
Pei
ng/1
198
0
0
10
32
39
U2
68
Percent
removed
—
100
100
95
8U
80
79
66
J2L !
3.2
U.5
U.3
U.3
U.2
U.2
U.I
U.I
_ a+ .*»
Fe PeT
«g/l
62
0
0
3
0
6
3
6
Percent
removed
—
100
100
95
100
90
95
90
192
0
0
23
23
29
23
23
Percent
removed
—
100
100
88
88
85
88
88
Oi
Pe PeT
3.U 55
U.3 0
U.3 0
U.2 0
U.2 0
U.2 0
U.2 0
U.i o
Percent
removed
—
100
100
100
100
100
100
100
mg/1
97
0
0
0
0
0
0
0
Percent
removed
—
100
100
100
100
100
100
100
-------
Synthetic coal mine water which had not been inoculated was subjected
to the same treatment, which resulted in no measurable change in Fe
even after having been aerated for 7 days. In neither experiment was
carbon present. The bacteria were alive and active during these tests.
Tests with Bacteria-inoculated Water
Batch tests and continuous flow tests were conducted with synthetic
coal mine water and Nuchar W-W carbon, 12 x ho mesh, according ^to the
general procedure. These "standard" tests were compared with Mists
with synthetic coal mine water which had been inoculated with Thio-
bacillus ferrooxidans and/or Thiobacillus thiooxidans.
Batch tests — Active cultures of Thiobacillus ferrooxidans -were, ob*
tained from Indiana University of Pennsylvania in Silverman and
Lundgren 9K basal salts + ferrous sulfate as the energy source. As
received, the culture media contained kl mg/1 of Fe2*, 6lO mg/1 of Fej,
and had a pH of 1.9 by our analyses. A 10:1 mixture of synthetic to
culture media was prepared and used in these batch tests and the contin-
uous flow tests. The results of a "standard" batch test, with no bacte-
ria present, and a test with Thiobacillus ferrooxidans are compared in
Table 35.
It is apparent from the data in Table 35 that, in spite of the presence
of iron oxidizing bacteria and aeration, the Fe3+ present with the
culture as an energy source for the bacteria had been reduced to the
Fea* state. Two samples of water from these two tests were analyzed
for bacteria. No bacteria were found, indicating rather complete
adsorption of the bacteria on to the activated carbon.
Another test was conducted with the synthetic coal mine water which had
been inoculated with Thiobacillus thiooxidans. The results of this test
are also presented in Table 35. This particular culture, by itself, was
not expected to and, in fact, did not affect iron removal. It was
desirable, though, to determine the effect of this culture on the pro-
cess before combining it with the iron-oxidizing bacteria. The combina-
tion of iron-oxidizing and sulfur-oxidizing bacteria was said to
accomplish oxidation of iron at a faster rate than if only the iron-
oxidizing bacteria were present.
Finally, a test was conducted with the combination of iron-oxidizing
and sulfur-oxidizing bacteria in the synthetic coal mine water. One
liter of each culture was added to 10 liters of synthetic. The results
are also presented in Table 35. As in the past tests with Thiobacillus
ferrooxidans, the presence of bacteria did not enhance the oxidation of
Fe and, in fact, again resulted in reduction of the Fe3+ to the Fe8*
state.
92
-------
35. BATCH TESTS WITH BACTERIA- INOCULATED WATER
."Standard" Test-No Bacteria Thiobacillus ferrooxidans Added
fime, Experiment No. 1*1*0-26 Experiment No. 1*33-1*8
Fes+, mg/1 FeT> mg/1 Pe8*, mg/1 FeT> mg/1
0 2l*5 21*8 227 615
5 208 215 1*11 595
10 195 199 *K>7 575
15 175 178 1»01 555
30 158 166 39? 1*55
1*5 109 130 3^5 390
60 76 93 321* 375
Thiobacillus ferrooxidans
THiobacillus thiooxidans Added Thiobacillus thiooxidans Added
Time, Experiment No. lfUO-25 Experiment No. 1*1*0-27
,m^ Fe3*, mg/1 FeT, mg/1 Fe3*, mg/1 FeT, mg/1
0 219 231 188 560
5 223 233 337 5^0
1,0 212 226 331 520
15, 199 225 333 5^0
30 175 193 310 510
1*5 163 180 21*7 500
60 159 177 272 530
93
-------
Continuous Flow Tests — As in the case of the batch tests, a "standard"
continuous flow test with no bacteria in the synthetic coal mine water
was conducted. The results of this test are compared in Table 36 with
the results of a test with synthetic containing Thiobacillus ferrooxi-
dans and also with the results of a test with synthetic containing a
combination of Thiobacillus ferrooxidans and Thiobacillus thiooxidans.
£?^L
Again, as in the batch tests, reduction of iron back to the Fe state
was seen in the results of tests with bacteria present.
A further difficulty was experienced in the test with the combination
iron-oxidizing and sulfur-oxidizing bacteria. The cultures, as receiv-
ed, contained iron as an energy source for the bacteria. This iron had
been converted to the Fea+ state by the bacteria during their growth
period. The precipitated ferric Iron caused the column to be blocked
at 150 minutes into this test. Flow could then be maintained only by
eliminating the aeration. Flow was continued for an additional 60 min-
utes with no aeration, for a total of 210 minutes.
In summary, the results of neither the batch nor the continuous flow
tests have demonstrated effective removal of iron from synthetic coal
mine water which had been inoculated with bacteria. In addition, re-
duction of iron from the Fes+ back to the Fe3* state occurred under
these conditions, and this is undesirable to this process.
Tests with Bacteria-inoculated Carbon
Based on results of tests with bacteria, described above, the time
element in oxidation of ferrous iron with carbon-bacteria systems was
considered by conducting tests at least two weeks after suspensions of
bacteria had been mixed with carbon. Both batch tests and continuous
flow tests were conducted with standard synthetic coal mine water and
Nuchar WV-W carbon, 12 x ho mesh, which had been inoculated with iron-
oxidizing bacteria, Thiobacillus ferrooxidans, 18 days before these
tests so that the bacteria could first adapt to the new environment
before being called upon to oxidize the iron. The bacteria were obtain-
ed from Indiana University of Pennsylvania and were the same as used in
previous tests.
Batch Tests — The results of "standard" batch tests with no bacteria
were compared with the results of tests with the bacteria-inoculated
carbon. The results of the "standard" batch tests are presented in
Table 37. Two tests of 120 minute duration each were conducted with
the no-bacteria system and two, also of 120 minute duration, with the
bacteria-carbon system. The results of these are also shown in
Table 37. From these data, it is apparent that iron removal was not
enhanced by the presence of bacteria and that there was no evidence of
the reduction of iron back to the Fe3+ state as in some past tests with
bacteria.
-------
TABLE 36. CONTINUOUS FLOW TESTS WITH BACTERIA-INOCULATED WATER
Thiobacillus ferrooxidans and
"Standard" Test-No Bacteria Thiobacillus ferrooxidans Added Thiobacillus thiooxidans Added
Experiment No. 4U3-5 g:xpgr;ynent No. 4^^-4Q Exoeriment No.
s»«**" m/» /l TP»n «M /n m^,2+
Fe3 , mg/1 Fe?, mg/1 Fe3*, mg/1 Fef, mg/1 Fe8'*', mg/1 Fef, mg/1
235 2kO 228 590 188 560
28 39 166 - 182 iQk
28 29 326 338 208 266
38 39 ^01 - 220 292
133 150 328 483 2?8 525
156 173 197 588 170 500
170 175 18U 588 163 5^0
176 182 169 584 150 585
182 190 157 590 138* 575*
190 196 153 660 120* 565*
* Stopped aeration but maintained flow
-------
TABIE 37. BATCH TESTS WITH BACTERIA-INOCULATED CARBON
"Standard" Tests - Ho Bacteria
Bacteria Added
Time,
min
0
5
10
15
30
^
60
90
120
First
Experiment
Fes% mg/1
251
15
7
2
1
1
1
1
1
Test
No. kkO-32
FeT, mg/1
259
73
^5
30
17
lU
12
11
8
Second Test
Experiment No. Ml-0-35
Fe2+, mg/1
251
10
6
2
1
1
1
1
1
FeT, mg/1
259
65
^7
37
25
18
17
13
11
First
Experiment
Fe2+, mg/1
251
58
U5
25
2k
20
20
Ik
12
Test
No. W-0-3^
FeT, mg/1
259
670
63k
395
370
360
320
250
205
Second Test
Experiment No. kkO-36
Fes+, mg/1
251
30
21
19
15
11
11
11
11
FeT, mg/1
259
580
570
570
570
570
5to
505
500
-------
Continuous Flow Tests — The results of the test with bacteria-
inoculated carbon and the 22 in. x 1 in. column are presented in
Table 38 and compared in the same table with the results of a past
"standard" test and a past test with bacteria which was added to the
water, and not to the carbon. In this test, the carbon in the column
had been inoculated with the bacteria 18 days prior to the test and
growth of the bacteria was encouraged while they were given tme to
adapt to their new environment.
From the data in Table 38, there is no evidence of the reduction of
Fe3* during the test with the inoculated carbon, in contrast to the
reduction evident during the test with the inoculated water. In the
test with the inoculated carbon, there was specific evidence of a sub-
stantial amount of oxidation since (a) greater than 50 percent of the
Fe3* was removed throughout the 6-hour duration of the test, and (b)
during this 6-hour period, there was a substantial difference between
the Fej analyses and the Fe3* analyses at any single point in time.
This was the first evidence from this project of any substantial enhance-
ment of iron removal with iron-oxidizing bacteria.
To examine the long term effects of the iron-oxidizing bacteria-
activated carbon combination, continuous flow tests were conducted
according to the general procedure with the carbon which had been inocu-
lated with Thiobacillus ferrooxidans two months prior to these tests.
Analysis of the water drained from the carbon prior to the tests did not
show the presence of the bacteria. This did not necessarily indicate
that the bacteria had not survived; those which survived could cling
tenaciously to the carbon. The results of these tests are presented in
Table 39. The differences between Fe3* and 'Fey concentrations at any
point in time during the tests indicated that a significant amount of the
iron had been removed by an oxidation process due to the iron-oxidizing
bacteria which might still have been present. The amount of Fe3* removed
was at the 100 percent level for 3^ of the total k& hours of the seven
6-hour tests. For the entire duration of the test, at least a 50 percent
level of Fe3* removal was attained.
A high level of removal of Ferp was also demonstrated during these tests.
Adsorption is the probable mechanism for the removal of Fe
(Fem = Fe3* + Fe3*). The 50 percent removal level for this species of
iron was attained for the same total time period as in the removal of
Fe3*, 3^ of the total ^2 hours of the tests. At a flow rate of
30 ml/min, this amount of the bacteria-impregnated carbon could be used
to treat 6l,200 ml of water of this quality and attain 50 percent removal
of the iron. Stated differently, one pound of this carbon could treat
over 9 gallons of water of this quality based on these tests. For this
process to be practical, the removal of iron from actual mine waters must
clearly be more effective than demonstrated in these tests with synthetic
coal mine waters.
97
-------
TABLE 38. CONTINUOUS FLOW TESTS WITH BACTERIA-INOCULATED CARBON
"Standard" Test - No Bacteria
Experiment No.
CD
150
180
210
270
300
330
360
Fes+, mg/1
235
28
28
38
133
145
156
170
176
182
190*
FeT, mg/1
240
39
29
39
150
162
173
175
182
190
196*
Bacteria Added to Water
Experiment No. W+3-7
Fes+, mg/1
188
182
208
220
278
180
170
163
150
138
120
145*
FeT, mg/1
560
184
266
292
525
550
500
54o
185
575
565
600*
Bacteria Added to Carbon
Experiment No. 443-31
Fe2*, mg/1
224
37
19
20
12
10
24
12
42
70
83
100
no
106
n4
n4
108
FeT, mg/1
260
336
223
214
202
211
2l4
213
216
204
194
200
195
191
211
191
183
Test discontinued after 180 min.
Test discontinued after 210 min.
-------
TABLE 39. LONG TERM EFFECTS OF THZOBACILLUS FERROOXIDANS
Time,
min
0
5
10
15
30
45
60
90
120
150
180
210
240
270
300
330
360
Experiment
No. 440-99
.mg/1
Fe"T
250
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
6
FeT
266
13
29
23
16
13
13
23
36
58
81
89
107
123
127
120
127
Experiment
No. 451-2
mg/1
Fe*T
263
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
FeT
286
42
52
55
71
91
97
114
120
133
143
143
143
143
156
159
156
Experiment
No. 451-3
mg/1
Fe*T
250
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
FeT
295
75
78
78
75
84
94
117
156
149
153
159
169
166
169
169
172
Experiment
No. 451-4
mg/1
Pe»-r
253
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
3
FeT
266
42
45
49
49
55
75
94
107
117
123
131
136
143
156
153
Experiment
No. 451-5
mg/1
Pe"*
260
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
FeT
279
62
71
94
101
104
io4
127
127
127'
130
140
149
153
149
156
159
Experiment
No. 451-6
mg/1
Fe--r
263
0
0
0
0
0
0
-
0
H
3
32
78
88
FeT
269
88
107
117
120
156
159
-
175
188
-
208
_
201
—
208
Experiment
No. 451-7
mg/1
Fe'T
250
0
13
13
16
26
19
-
23
26
-
42
_
104
^
127
FeT
266
58
71
103
117
130
136
-
143
_
195
221
_
224
—
234
-------
Effect of Long Aeration Periods Prior to Testing
While the differences between the effectiveness of iron removal using
this bacteria-inoculated carbon and carbon which had not been inoculat-
ed with bacteria were attributed to the bacteria, there may have been
other reasons for differences in effectiveness. Before using the column
containing the carbon and bacteria, the column was aerated continually
and the ferrous sulfate solution "changed" approximately every other
day (on the weekend, the solution was changed on the third day). The
continual aeration of the carbon for the expressed purpose of maintain-
ing bacterial activity might also have resulted in changing the nature
of the carbon. Lamb and Elder (10) reported (a) that charcoal when
prepared fresh and dry showed a positive test for a peroxide (b) that
no further test for peroxide was seen after extraction of the charcoal
with acid and (c) that the sample of charcoal once extracted could again
be made to give an equally positive test for peroxide by shaking the
residual charcoal with more water and air. The effect of air (oxygen)
on charcoal was also recognized by others (29-3*0.
To demonstrate the effect of a long aeration period, a comparison was
made between a bacteria-inoculated carbon and a carbon without bacteria.
Each carbon had been aerated continually for at least one month under a
solution of ferrous sulfate which was "changed" every other day. The
first series of continuous flow tests according to the general procedure
was conducted with synthetic coal mine water. The results are presented
in Table ^0 and were similar whether bacteria were present or not.
The second series of continuous flow tests was conducted with an actual
coal mine water. The results of these tests are presented in Table Ul
and again are similar whether bacteria were present or not.
In both series of tests, aerating the carbons for long periods of time
prior to using them in the tests resulted in very effective removal of
Fe2* accompanied by less effective removal of Fej. An additional con-
tinuous flow test was conducted with activated alumina which was aerat-
ed continually for 18 days prior to the test and a solution of standard
synthetic coal mine water added as in past tests. The results of this
test with actual coal mine water showed virtually no removal of Fe8*
and only a slight removal of Fej, indicating little or no oxidation
with the activated alumina and possibly a slight amount of adsorption
or filtration of precipitated Fe3* compounds by the alumina.
From these tests of relatively short duration, the long aeration of the
carbon prior to use resulted in effective removal of Fe8* and less
effective removal of Ferj., which is most desirable for this process since
the iron would be oxidized but not adsorbed by the activated carbon.
This favorable removal of iron occurred whether bacteria were present
or not.
100
-------
TABLE 40. COMPARISON OF THE EFFECT OF LONG AERATION PERIODS AND BACTERIA
VERSUS NO BACTERIA ON REMOVAL OF IRON FROM SYNTHETIC COAL MINE WATER
No Bacteria Added
Experiment No. 444-95
Bacteria Added
Experiment No. 444-96
Fe3* FeT
H
3
Time,
min
0
30
60
120
180
240
300
360
3.1
2.8
3.7
2.5
2.4
2.4
2.4
2.4
240
0
0
o
0
0
0
0
Percent
Removed
—
100
100
100
100
100
100
100
mg/1
260
0
0
49
78
94
no
123
Percent
Removed
—
100
100
81
70
64
58
53
J2S-
3.0
2.3
2.3
2.2
r—
2.2
2.2
2.2
2.2
Fe3* FeT
Percent
mg/1 Removed mg/1
253
0
0
0
0
0
0
0
—
100
100
100
100
100
100
100
266
52
68
81
101
io4
no
123
Percent
Removed
—
80
74
70
62
61
59
54
-------
TABLE iH. COMPARISON OF THE EFFECT OF LONG AERATION PERIODS AND BACTERIA
VERSUS NO BACTERIA ON REMOVAL OF IRON FROM ACTUAL COAL MINE WATER
No Bacteria
Experiment No.
Bacteria
Experiment No. Wi-98
2^* S4*
Time,
min
0
30
60
120
180
21*0
300
360
pH mg/1
3.1 127
2.8
2.7
2.7
2.8
2.7
2.7
2.7
0
0
0
0
0
0
0
Percent
Removed
—
100
100
100
100
100
100
100
mg/1
159
te
^
62
62
62
65
68
Percent
Removed
—
7k
72
61
61
61
59
57
JBS. i
3.1
2.7
2.7
2.7
2.8
2.7
2.7
2.7
mg/1
127
0
0
0
0
0
0
0
Percent
Removed
—
100
100
100
100
100
100
100
Percent
mg/1 Removed
159-
62
^5
62
62
62
65
68
--
61
72
61
61
61
59
57
-------
Tests with Bacteria-inoculated Coal
Tests were conducted using bacteria alone, carbon alone, and combinations
of bacteria and carbon. To examine the effect of the combination of
bacteria and a solid surface other than activated carbon, nine tests
vere conducted according to the general procedure but with coal which had
been inoculated with Thiobacillus ferrooxidans 15 days prior to the first
test with this material. The results of these tests seemed to indicate
effective removal of iron with the use of material other than activated
carbon. In spite of the fact, though, that the pH of the effluent from
the column remained low throughout these tests, iron could be seen pre-
cipitating onto the coal in increasing amounts as these tests progressed.
This precipitation, by itself, may have been responsible for the appar-
ently effective removal of iron. Since these tests were conducted near
the end of the project, time was not available for further study in this
area.
Statistical Design III
To further explore the matter of the need for bacteria or long aeration
of the carbon prior to use, a third set of factorial experiments was
designed to cover the effects of water flow rate, amount of carbon,
bacteria versus no bacteria, and aeration rate on continuous removal of
iron from an actual coal mine water. Each series consisted of 16 tests.
In eight of the tests, separate samples of carbon were inoculated with
Thiobacillus ferrooxidans at least 18 days prior to the test, the carbon
columns aerated continually, and a ferrous sulfate solution added period-
ically as in past tests. The other eight tests were conducted with
carbons which had not been inoculated with bacteria. As in past experi-
ments, only the second of two successive tests with each column was
evaluated. The data representing the 16 tests are presented in Table U2.
For ease of interpretation, again the four separate variables are identi-
fied, with the high and low level of each variable being represented by
•f and - signs respectively. The procedure for evaluating this factorial
experiment was identical to that described previously in the Statistical
Design I and II experiments. The data in Table h2 were again employed in
the calculation of both main effects and variable interactions for the
seven responses for this test using the codified design and calculation
matrix previously presented in Table 11 and the procedure previously out-
lined. The main effects are listed in Table lj-3.
From the data in Table ^3, the most significant variable affecting Fes+
and FeT removal was Variable 1, the flow rate of the water through the
column. The negative sign for that main effect simply meant that iron
was removed more effectively at the lower flow rate of raw water
(20 ral/min in this experiment). The importance of this variable remained
apparent throughout the duration of each test as indicated by the rela-
tively large value for that main effect (See Table ^3)• In the first few
hours of each test this effect of flow rate was greater on the removal of
103
-------
TABLE 42.
Variables*
RESULTS OF STATISTICAL DESIGN III EXPERIMENT
ON VARIABLES AFFECTING IRON REMOVAL
1
Water
Flow
Test Rate
1
2 +
3
4 +
5
6 +
7
8 +
9
10 +
11
12 +
13
14 +
15
16 +
2 . 3 „ 4 After
Amount Aera- Fe**
of Bacte- tion % Re-
Carbon ria Rate moved
- - 100
59
+ 100
+ 100
-I- - 100
+ - 75
+ + - 100
+ + - 100
+ 100
+ 83
+ + 100
+ + 100
+ + 100
+ + 69
+ + + 100
+ + + 100
30 min
FeT
% Re-
moved
""W
52
100
100
89
32
100
87
100
71
100
99
8l
4l
100
88
After
Fe
% Re-
moved
"1 /"%^\
J»vVi/
36
100
100
100
71
100
100
100
79
100
95
100
63
100
100
60 min
FeT
% Re-
moved
~100~
44
100
97
82
22
100
73
100
54
100
91
66
25
100
44
After
Fes
-------
TABLE 42. RESULTS OP STATISTICAL DESIGN III EXPERIMENT
ON VARIABLES AFFECTING IRON REMOVAL (continued)
Test
1
2
3
4
5
6
7
8 v
9
10
11
12
13
14
15
16
1
Water
Flow
Rate
4-
_
4-
—
4-
-
4-
_
4-
-
+
«*
4-
-
4-
Variables*
234
Amount Aera-
of Bacte- tion
Carbon ria Rate
^ —
_ _ _
+
4-
4-
4-
4- 4-
4- 4-
4-
+
4- - 4-
4- - 4-
-4-4-
4- 4-
4- 4- 4-
4- 4- 4-
After 240 min
Fe3*
% Re-
moved
100
4o
100
65
100
39
100
100
100
64
100
59
100
43
100
100
Fey
% Re-
moved
100
24
100
67
48
7
100
45
100
38
100
69
43
10
100
29
After 300 min
Fe3*
% Re-
moved
100
41
100
26
92
37
100
100
100
64
100
26
99
38
100
90
FeT
% Re-
moved
100
25
100
58
41
8
100
42
100
37
100
58
42
22
100
29
After 360 min
Fe34"
% Re-
moved
100
44
100
20
84
37
100
100
100
68
100
20
72
43
100
76
FeT
io Re-
moved
100
29
100
57
33
8
100
4o
100
34
100
57
39
24
100
29
* Variable Identification
Variable
Water Flow Rate
Amount of Carbon
Bacteria
Aeration Rate
Levels
4- • -
80 ml/min 20 ml/min
1,000 g 500 g
YES NO
500 ml/min 50 ml/min
-------
TABLE U3. MAIN EFFECTS BASED ON DATA FROM
: CONTINUOUS FLOW TESTS
Fe3* Removed,
percent
Fej Removed,
percent
Single
Variable:
Single
Variable:
Single
Variable:
Single
Variable:
Variable
1
2
1*
3
1
2
3
1*
1
2
3
1*
1
2
3
1*
Main
Effects
After
-ll*.
li*.
2.
•
After
-27.
18.
5.
1.
After
-36.
17.
6.
2.
After
-*3.
8.
7.
*" •
1 Variable
Main
Effects
30 Minutes
25
25
25
25
120
25
50
50
00
21*0
25
25
75
75
360
50
50
50
75
2
1
3
*
26.
-25.
-13.
2.
00
00
00
50
Minutes
1
2
3
U
-1*2.
29.
-25.
-2.
75
00
00
25
Minutes
1
2
3
-50.
30.
-27.
-.
25
00
00
25
Minutes
1
2
3
I*
-1*9.
27-
-25.
2.
25
00
50
00
Fe Removed, Feg; Removed,
percent percent
0)
H
,0
<3
•H
&
>
CO
•p
0
C! fl)
^_J ^^J
53 W
0)
H ra
oJ o
•H GO)
pi *rt yH'
cjj c3 *H
p> J2 |3c)
After 60 Minutes
1
2
1*
3
1
2
3
1*
1
3
2
1*
-19.50
18.25
3.75
3.00
After 180
-33.50
20.00
5.50
3.25
After 300
-46. 33
12! 38
8.88
2.63
1 -37.25
2 26. 50
3 -21.75
k ~k'75
Minutes
i -¥*.oo
2 28!?5
3 -28.50
k 0.00
Minutes
1 -50.50
2 26. 50
3 -2U.25
^ 1.75
Designation of Variables:
1 = Water Flow Rate
2 = Amount of Carbon
3 = Bacteria
k = Aeration Rate
106
-------
than it was on the removal of Fes+; but by the end of the 6-hour
tests, the effect was equally pronounced on both Fej and Fes*.
Of lesser importance was the effect of Variable 2, the amount of carbon
used. The value of the effect of this variable on Fej was approximately
one-half the value of Variable 1. This variable had a practically
insignificant effect on removal of Fes*. The positive value for this
effect meant that greater removal was attained with the greater amount of
carbon used (1,000 g in this experiment).
Similarly, Variable 3 affected the removal of Fes* and FeT to a degree
similar to that of Variable 2. The negative sign for the effect on Fei
indicated that greater removal of iron was demonstrated when bacteria
were not added to the carbons prior to use. The positive sign for the
effect of this variable on Fe2* indicated that the bacteria aided in the
removal of Fe34>. The effect on Fe8"1", though, was much less than the
effect on Fej as indicated by the absolute magnitude of this variable.
These results are consistent with the probabilities that (a) the presence
of iron-oxidizing bacteria or the use of long aeration periods of the
carbon prior to use should aid in the oxidation of Fe3+, and (b) adsorp-
tion of iron onto the surface of the activated carbon as measured by Fe^
would be inhibited by the presence of bacteria blocking the active carbon
sites for adsorption.
Variable U, aeration rate, at the levels of this experiment, had no
affect on removal of iron as demonstrated by the relatively small value
for this effect from Table k-3,
The values for many two-variable interactions were as great as those for
Variables 2 and 3, but these relatively high values did not persist for
single two-variable interactions throughout the duration of the tests.
Effect o£ Water Flow Sate
In the first of this series, standard synthetic coal mine water was used.
The results of continuous flow tests conducted according to the general
procedure but with water flows of 30, 100, 150, 200, 300, and 600 ml/min
are presented in Table kk. From the data in Table Mf, the amount of iron
removed decreased for both Fe and FeT with increasing flow rate.
In the second of this series, an actual coal mine water was used. Water
from a tributary of Little Plum Creek, sampling point No. 2 (See Table
31), was chosen. The results of these tests at water flow rates of 30,
60, and 120 ml/min are presented in Table 45. The flow rate of 30 ml/min
is included in the general procedure. Only the results of the second of
two successive tests with water at the specified flow rate are included
in Table 45, for reasons previously discussed. From these data, little
direct relationship could be found between water flow rate and removal of
Fe8*. For this series of tests, more effective Fe removal was seen at
the higher flow rate, 60 ml/min, than at a lower flow rate, 30 ml/min.
10?
-------
TABLE hk. EFFECT OF FLOW RATE ON REMOVAL OF IRON FROM SYNTHETIC COAL MINE WATER
o
00
30 ml/min
100 ml/min
150 ml/min
200 ml/min
300 ml/min
600 ml/min
Experiment
No. 1M-53
Time,
min
0
5
10
15
30
U5
60
90
BH.
3-0
2.9
2.9
3-0
2.9
2.9
2.9
2.8
Fe8*
mg/1
253
166
162
156
133
133
130
1*3
Fej
mg/1
266
211
201
162
179
166
159
172
Experiment
jjo 9 lil^lj..- 5^4-
J2H-
3.0
2.8
2.9
2.9
2.8
2.8
2.8
2.8
Fe3*
mg/1
253
159
159
153
153
159
199
159
mg/1
266
208
208
201
208
205
208
205
Experiment
No. W<-5^
J*L
3.0
2.9
2.9
2.9
3.0
2.9
2.8
2.8
Fe3*
mg/1
253
101
101
110
11U
1*9
156
159
mg/1
266
137
ll+O
159
159
£01
22^
218
Experiment
No. Wf-53
SS
3-0
3.0
3-3
3-3
3-1
3.0
2.9
2.9
mg/1
253
9^
130
156
166
175
179
175
mg/1
266
101
169
205
205
201
22k
208
Experiment
No. W*-l*9
pH
3-0
3-9
3-9
3.7
3.6
3-3
3-2
3.1
Fe3*
mg/1
263
192
201
237
21*
250
231
22^
mg/1
286
208
si*
253
256
260
266
2*10
Experiment
Uo. ***~*9
pH
3.0
3-1
3-1
3-1
3-0
3-0
3-0
3-0
Fea*
mg/1
263
218
208
208
208
20^
201
207
Fey
mg/1
286
OllA
OliO
237
250
2kj
2^0
2^0
-------
TABLE U-5. EFFECT OF WATER FLOW RATE ON REMOVAL OF IRON FROM
TRIBUTARY OF LITTLE PLUM CREEK. SAMPLING POINT NO. 2.
Plow Rate = 30 ml/min
Experiment No. U51-26
Flow Rate = 60 ml/rain
Experiment No. 1*51-29
Flow Rate = 120 ml/min
Experiment No. U51-32
tn 3+ 3*
Time,
min
H 0
o
VO
30
60
120
180
2UO
300
360
3.0
U.6
\t O
H ^
U.3
U.2
U.I
U.I
llU
13
32
91
99
101
101
101
Percent
removed
89
72
20
13
11
11
11
I mg/1
282
U2
91
120
13U
133
133
129
Percent
removed pH mg/1
85
68
57
52
53
53
5U
3.3 172
U.6 29
U.5 71
U.U 9U
U.3 97
U.2 101
U.I 10U
3.9 136
Percent Percent
removed mg/1 removed
83
59
^
UU
Ul
Uo
21
227
71 69
10U 5U
120 U7
117 U8
llU 50
117 U8
1U6 36
i
I pH
3.2
U.U
U.U
U.I
3.7
3.2
3.2
3.1
Fe3+ FeT
182
156
159
156
188
175
169
166
Percen
remove1
lU
13
lU
0
U
7
9
t
d mg/1
256
185
182
205
22U
221
198
192
Percent
removed
28
29
20
13
lU
23
25
-------
However, the least effective Fe2* removal was found at the highest flow
rate of this series, at 120 ml/min. ;
Also from these data, the expected decrease in Fej removal with increas-
ing flow rate can be seen.
It must be pointed out here that with this particular actual coal mine
water sample, the rate of removal of either FeS4> or Fem was not good at
any flow rate, nor as good as was obtained in earlier tests with syn-
thetic coal mine water. The reasons for this have not yet been clearly
defined, but the low dissolved oxygen content of this water as sampled
has already been discussed.
In the last of this series, the combined activated carbons from the
Statistical Design III experiment, a total of 6500 g of carbon, were
placed in a 52 in. x 5 in. column. The carbon was aerated continually
for about 3 weeks with a ferrous sulfate solution added periodically as
in past tests. This carbon was then used in tests with actual coal
mine water with water flow rates of 20, 200, hOO, 800, and 1600 ml/min.
Water from a tributary of Little ELum Creek, sampling point No. 1, was
seclected because of the high concentration of iron in this water. The
results of these tests are presented in Table U6. Generally speaking,
the results show a decrease in iron removal with increasing flow rate.
These data will be employed more fully in the following section.
110
-------
TABLE 1+6. EFFECT OF FLOW RATE ON REMOVAL OF IRON FROM TRIBUTARY
OF LITTLE PLUM CREEK. SAMPLING POINT NO. 1
Fe:
.a*
0
60
120
180
21+0
300
360
a
60
120
180
21+0
300
360
0
60
120
180
21+CX
300
360
0
60
120
180
21+0
300
360
0
60
120
180
21+0
300
360
pH mg/1 Percent Removed mg/1 Percent Removed
Flow
2.8
2.7
2.7
2.7
2.8
2.8
2.8
Flow
2.9
2.7
2.7
2.7
2.6
2.6
2.6
Flow
2.8
2.6
2.6
2.6
2.7
2.7
2.7
Flow
2.8
2.5
** * X
2.5
fc • „/
2.5
*•• * x
2.6
2.7
*•• V |
2.7
Flow
2 8
fc» • W
2.6
2.7
*"" • i
2.7
2.7
2.7
2.7
Rate 20 ml/min
279
0
0
0
0
0
0
Rate 200
282
**5
55
81+
101+
117
136
Rate 1+00
292
1+9
75
127
172
179
201
Rate 800
318
0
0
62
133
188
205
Rate 1600
286
172
201
218
231
231
231+
__
100
100
100
100
100
100
ml/min
__
81+
80
70
63
59
52
ml/min
__
83
57
1+1
39
31
ml/min
MM IM
100
100
81
58
1+1
36
ml/min
*•<»
1+0
30
21+
19
19
18
Experiment No. 1+51-75
1+19
13
13
13
13
16
16
Experiment No. 1+51-78
1+22
156
182
21+1+
286
325
331*
Experiment No. 1+51-77
1*35
237
31+1
370
373
373
373
Experiment No. 1+51-76
1+1+8
295
35**
383
390
373
386
Experiment No. 1+51-79
1+68
1+19
1+19
1+19
1+16
1+19
1+19
_-
97
97
97
97
96
96
_„
63
57
1+2
32
23
21
m J*
1+6
22
15
ll+
ll+
ll+
«•*•
^li-
21
15
13
17
ll+
—
10
10
10
11
10
10
in
-------
SECTION VI
SUMMARY EVALUATION OF IRON REMOVAL WITH ACTIVATED CARBON
Based on the results of the laboratory studies which were presented in
the preceding section, the following is a summary evaluation of the
process for removal of iron from acid mine drainage with activated
carbon.
Aeration with No Carbon Present
To demonstrate the effect of aerating acid mine drainage containing
ferrous iron at low pH, the following experiment was performed. A
19 in. x 3 in. glass column was packed with approximately 800 g of
3/8 in.-diameter glass beads and a continuous flow test conducted
according to the general procedure with actual coal mine water contain-
ing Fes* . The results are presented in Table ^7. From these data,
aeration of the acid mine drainage at low pH in the absence of activated
carbon had no effect whatsoever on removal of iron.
Aeration with Activated Carbon Present
From the results of the laboratory studies described in the preceding
section, aeration of acid mine drainage in the presence of activated
carbon resulted in removal of iron from the water at a rate which was
affected greatly by the experimental conditions. In addition, it was
also demonstrated that the relative rates of removal of Fe * and Fe3*
were influenced by the experimental conditions. The removal of Fe3*
proceeded by a combination of oxidation and adsorption. The removal of
Fe3* proceeded by adsorption. The most desirable mechanism for removal
of iron from acid mine drainage is removal of only Fe3* by oxidation.
This oxidation would result in conversion of Fe3* to Fe3* which would
then remain in solution. The activated carbon would not eventually be
rendered inactive and would not need to be replaced with fresh activated
carbon. Removal solely by an oxidation process, however, has not been
experienced on this project.
In the last series of tests on this project, the experimetnal conditions
selected were those found from all previous tests to result in most
effective removal of iron. The results of these tests with 6500 g of
carbon and actual coal mine water flow rates of 20, 200, ii-00, 800, and
1600 ml/min were listed in Table U6. For the purposes of this evalua-
tion, the data from Table h6 were plotted in Figure 17 and consideration
given to the time during which 50 percent of the Fe2* was removed in
these tests. The 50 percent level of iron removal was somewhat arbitrar-
ily chosen as an attainable, or, at least, a measurable goal.
113
-------
TABLE 4?. EFFECT OF AERATION ON REMOVAL OF IRON. NO CARBON
Time,
mln
0
5
10
15
30
60
120
180
QkO
300
360
2.7
2.7
2.6
2.6
2.6
2.6
2.6
2.6
2.6
2.6
2.6
mg/1
350
350
350
350
350
350
350
350
350
350
350
Fe
Percent Removed
0
0
0
0
0
0
0
0
0
0
0
mg/1
535
535
535
535
535
535
535
535
535
535
535
Fex
Percent Removed
0
0
0
0
0
0
0
0
0
0
0
llU
-------
Fe2+ REMOVED, PERCENT
100
90-
80-
70-
60-
50-
40-
30-
20-
10-
20 ml/min
Straight Line by Method
of Least Squares
T
T
T
] 60 120 180 240 300 360
DURATION OF TESTS AT WATER FLOW RATES LISTED, MINUTES
Bituminous Coal Research, Inc. 2042G49
Figure 17. Effect of Water Flow Rate on Removal of
115
-------
As can be seen from Figure 1?, at a flow rate of 20 ml/min, 100 percent
of the Fe2* was removed for the duration of the test. This flow repre-
sents only a few gallons of water treated per day, a rate too small to
be treated practically. At that low water flow rate, no indication was
given of the deterioration of the carbon with time during this short
test. Conversely, from Figure 17, at the highest water.flow rate,
1600 ml/min, the rate of removal was very low and the 50 percent level
of removal was never attained.
For this evaluation, consideration was given first to the tests at a
water flow rate of 200 ml/min. A 50 percent level of Fe3+ removal was
attained for slightly better than 6 hours in this test. At 200 ml/min,
this corresponds to a rate of approximately 19 gallons treated to the
50 percent Fes+ removal stage in a 6-hour period using 6500 g of carbon
before the carbon was spent. At about 22 cents per pound for the par-
ticular carbon used, this represents 0.75 lb of carbon necessary to
treat one gallon of water, or a materials cost of $165 per thousand
gallons of water treated.
Similarly, at a flow rate of 800 ml/min, 6500 g of carbon was used to
treat approximately 51 gallons of water in a 4-hour period to the
50 percent level of Fe3* removal. At the same price per pound of
carbon, this corresponds to a materials cost of approximately $62 per
thousand gallons of water treated. (From Figure 17, the apparently
more effective removal of Fe3* at 800 ml/min as compared to UOO ml/min
can be attributed simply to the fact that the 800 ml/min test was con-
ducted prior to the 400 ml/min test and the carbon had not yet been
rendered quite so inactive.)
Since total costs for treatment of acid mine drainage including chemi-
cals, labor, capital cost, etc., are reported to be from about 10 cents
to slightly more than $1.00 per thousand gallons of water treated, it
is clear that cost of treatment with activated carbon based on these
studies would be prohibitive to the use of this process for acid mine
drainage.
It is felt that further studies in this area might result in some in-
creases in the rate of removal of iron with activated carbon by changes
in the configuration of the reactor or by using many columns in series.
Increases in the rate of removal of a sufficiently large magnitude to
make this process practical are not anticipated.
The key to a practical process seems to be in finding a method of
easily removing the adsorbed iron from the carbon. These studies have
shown that this adsorbed iron clings tenaciously to the carbon thereby
rendering it inactive. "Regeneration" of the spent carbon should not
be considered in the classical sense; i.e., reactivation and removal of
adsorbed contaminants by a heat treatment step. Neither heat treatment
nor an acid wash, as found from these studies, will remove adsorbed
iron from the activated carbon.
116
-------
Another possibility is that a replacement for the activated carbon might
be found which might react similarly and yet not adsorb iron so strongly
or else be sufficiently inexpensive that once having adsorbed the iron,
the spent material might be discarded.
A third possibility is that some level of iron removal less than 50 per-
cent might be tolerated in a continuous treatment system, such as in a
limestone neutralization system. The effectiveness of this combination
on an activated carbon-limestone treatment process can only be ascer-
tained by actually testing such a system.
In summary, a method has been found for removing iron from acid mine
drainage containing ferrous iron by using activated carbon. The method,
though technically feasible, is prohibitively expensive.
117
-------
SECTION VII
ACKNOWLEIX»ffiNTS
Work on this project was supervised by C. T. Ford, Project Scientist,
J. F. Boyer, Jr., Principal Investigator, and R. A. Glenn, Project
Director.
Technicians involved in the conduct of the experimental work were
J. R. Allender and D. A. Bowser. The assistance of R. R. Care in the
spectrographic analyses is also acknowledged.
The helpful suggestions and comments from E. Harris, EPA Project Officer,
during the course of the work, from J. M. Shackelford, EPA Project
Manager, in shaping the course of the work, and F. W. Liegey, Chairman
Biology Department, Indiana University of Pennsylvania during the
bacteria studies, were sincerely appreciated. Dr. Liegey deserves
special thanks for repeatedly adjusting his busy schedule to fit the
needs of this project and for tolerating our unfamiliarity with bacteria.
A significant objective of this project was to investigate practical
means of abating mine drainage pollution. Such research projects, in«i
tended to assist in the prevention of water pollution by industry, are
required by Section 6b of the Water Pollution Control Act, as amended.
This project of EPA was conducted under the direction of the Pollution
Control Analysis Section, Ernst P. Hall, Chief, Dr. James M. Shackelford,
Project Manager, and Eugene Harris, Project Officer.
119
-------
SECTION VIII
REFERENCES
1. Bituminous Coal Research, Inc., "Studies on limestone treatment of
acid mine drainage" Water Pollution Control Res. Series DAST-33,
14010 EIZ, 01/70 Water Quality Office, Environmental Protection'
Agency, Washington, D. C. (1970).
2. "Acid mine drainage in Appalachia," Appalachian Regional Comm. ,
Washington, D. C. (1969).
«
3. Maneval, D. R., "Mine drainage pollution abatement — theory and
practice," Coal Mining Inst. Am. 6%th Annual Meet., Pittsburgh,
Pa., Dec. 10, 1970. 5pp.
k. Bituminous Coal Research, Inc., "Studies on limestone treatment of
acid mine drainage Part II," Water Pollution Control Res. Series,
1^010 EEZ, 12/71, Environmental Protection Agency, Washington,
D. C. (1971).
5. Stumm, W. and Lee, G. F., "Oxygenatlon of ferrous iron," Ind. Eng.
Chem. 53 (2), 1^3-6 (l9°"l).
6. Harvard University, "Oxygenation of ferrous iron," Water Pollution
Control Res. Series 1^010, 06/69 Water Quality Office, Environmental
Protection Agency, Washington, D. C. (1970).
7. Mason, D. G., "Treatment of acid mine drainage by reverse osmosis,"
3rd Symp. Coal Mine Drainage Res. Preprints, Pittsburgh, Pa., 1970
pp. 227-1*0.
8. Schumacher, E. A. and Heise, G. W. (to National Carbon Co., Inc.),
"Activated carbon catalyst bodies and their preparation and use",
U. S. Pat. 2,365,729 (Dec. 26,
9. Thomas, G. and Ingraham, T. R., "Kinetics of the carbon-catalyst
air oxidation of ferrous ion and sulfuric acid solutions," Unit
Process Hydromet. 1, 67-79
10. Lamb, A. B, and Elder, L. W., Jr., "The electromotive activation of
oxygen," J. Am., Chem. Soc. 53, 137-63 (1931).
11. Posner, A. K., "The kinetics of the charcoal catalyzed autoxidation
of Fe ion in dilute HC1 solutions," Trans. Faraday Soc.
389-95 (1953).
121
-------
12. Zawadzki, E. A., "Acid mine drainage research at Bituminous Coal
Research, Inc.," AIME Natl. Meet., Las Vegas, Nev., 1967. 12 pp.
13. Bituminous Coal Research, Inc., "BCR research planning reflects
coal industry needs." Coal Research, No. 32, Winter 1968-69. 7 pp.
lU. Calgon Corp., Pittsburgh, Pa., 1967. Private communication. 1 p.
15. Pittsburgh Activated Carbon Co., Pittsburgh, Pa., 1968 Private
communication. 1 p.
16. Mihok, E. A., "Mine water research. Catalytic oxidation of ferrous
iron in acid mine water by activated carbon," U. S. Bur. Mines,
Rept . Invest . 7337 (19^9) . 7 PP .
17. Hunter, W. G. and Hoff, M. E., "Planning experiments to increase
research efficiency," Ind. Eng. Chem. 59 (3), ^3-8 (19^7).
18. Pavelic, V. and Saxena, U., "Basics of statistical-experiment
design," Chem. Eng. 76 (21), 175-80 (1969).
19. Davies, 0. !>., ed., "The Design and Analysis of Industrial
Experiments," New York:Hafner Publishing Company, 1956, Chapter 7.
20, Armitage, P., "Restricted sequential procedures," Biometrika kk,
9-26 (1957).
21. Snell, E. S. and Armitage, P., "Clinical comparison of diamorphine
and pholcoline as cough suppressants by a new method of sequential
analysis," Lancet 272, 860-62 (1957).
22. Armitage, P., "Sequential Medical Trials," Springfield, Ill.:Thomas
Publishing Co., 1960. 105 PP.
23. Wald, A. "Sequential Analysis," New York: John Wiley & Sons, Inc.,
212 pp.
2k. "Standard Methods for the Examination of Waste and Wastewater,"
Am. Pub. Health Ass., Inc., 13th ed., New YorkrAm. Pub. Health
Ass., Inc., 1971. p. 189.
25. Barrow, G. M., "Physical Chemistry," New York:McGraw-Hill Book Co.,
Inc., 1961 pp. 631-2.
26. Laidler, K.J., "Catalyses," Vol. 1, New York:Reinhold Publishing
Corp., 195^. p. 152.
27. Liegey, F. W'. , Indiana Univ. of Pa., Indiana, Pa., 1971. Private
communication. 1 p.
122
-------
28. Beafore, P. J., "Field trip Consolidation Coal Co.-tour of
Mountaineer Division Operations," Acid Mine Drainage Workshop,
Athens, Ohio, by Ohio Univ., 1971. 7 pp.
29. Stenhouse, J., "Charcoal respirator for purifying the air by
filtration," Pharmaceutical Journal 1£, 1^-7 (1851*).
30. Stenhouse, J., "Charcoal as a disinfectant," Pharmaceutical
Journal 1^, 328-30 (185*0.
31. Calvert, P. C., "Experiments on oxidation by means of charcoal,"
5 J. Chem. Soc. (London) 20, 293 (1867).
32. Degener, P. and Lach, J., "Verfahren zur Behandlung von Knochkohle,"
Dingierfs Po3yteehnisch.es, J. 25j$, 519 (1885).
33. Lowry, H. H., and Hulett, G. A., "Studies in the absorption by
charcoal—Part 2, relation of oxygen to charcoal," J. Am. Chem.
Soc. Ug, 11*08 (1920).
3lf. Swinarski, A. and Siedlewski, J., "The effect of adsorbed oxygen on
the catalytic properties of activated carbon," Rocznlki Chemii 35»
999-1008 (1961).
OU.S. GOVERNMENT PRINTING OFFICE: 1973 514-15E/165 1-3
123
-------
SELECTED WATER i. Report No.
RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
3. Accession No.
Tlt'e 5. Report Date
TREATMENT OF FERROUS ACID MINE DRAINAGE WITH ACTIVATED 6-
S. Performing Organization
7. Author(s) -p^^n ro, i „„ m Report No.
10. Project No.
Boyer, James F.
l^KXLOGYH
9. Organization
11. Contract/Grant No.
Bituminous Coal Research, Inc. (Contractor)
13. Type of Report and
Period Covered
12. Sponsoring Organization
IS. Supplementary Notes
Environmental Protection Agency report
number, EPA-R2-73-150. January 1973.
16. Abstract
Laboratory studies were conducted with activated carbon as a catalyst
for oxidation of ferrous iron in coal mine water. Batch tests and
continuous flow tests were conducted to delineate tjfe process variables
influencing the catalytic oxidation and to determine the number and
types of coal mine water to which this process may be successfully
applied.
The following variables influence the removal of iron with activated
carbon: (a) amount and particle size of the carbon; (b) pH, flow rate,
concentration of iron, temperature, and total ionic strength of the
water; and (c) aeration rate. Adsorption as well as oxidation are the
mechanisms involved in iron removal by this process.
An evaluation of this process indicated technical feasibility which
would permit acid mine drainage neutralization using an inexpensive
reagent, such as limestone. The major disadvantage is the cost of the
activated carbons since they are rendered inactive after relatively short
use by apparently irreversible adsorption of iron. This cost appears to
be sufficiently high to prohibit the use of this process for treating
i7a.Descriptorscoal Jaine drainage. (Ford—Bituminous Coal Research, inc.)
*Acid Mine Water, *Waste Water Treatment, *Activated Carbon,
Iron Compounds, Oxidation, Adsorption, Ferrobacillus
17b. Identifiers
*Iron Removal, Oxidation Catalyst
17c.COWRR Field & Group 05D
IS. Availability 19. Security Class.
(Report)
20. Security Class.
(Page)
21. No. of Send To:
Pages
WATER RESOURCES SCIENTIFIC INFORMATION CENTER
U.S. DEPARTMENT OF THE INTERIOR
WASHINGTON. D. C. 20240
Abstractor Charles T. Ford \iustitution Bituminous Coal Research, Inc.
WRSIC 102 (REV. JUNE 1971) 3"26'
------- |