WATER POLLUTION CONTROL RESEARCH SERIES • DAST-2
Sulfide Treatment
of
Acid Mine Drainage
J.8. DEPARTMENT OF THE INTERIOR • FEDERAL WATER POLLUTION CONTROL ADMINISTRATION
-------
WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Reports describe
the results and progress in the control and abatement
of pollution of our Nation's Waters. They provide a
central source of information on the research, develop-
ment and demonstration activities of the Federal Water
Pollution Control Administration, Department of the
Interior, through inhouse research and grants and
contracts with Federal, State, and local agencies,
research institutions, and industrial organizations.
Triplicate tear-out abstract cards are placed inside
the back cover to facilitate information retrieval.
Space is provided on the card for the user's accession
number and for additional keywords. The abstracts
utilize the WRSIC system.
Water Pollution Control Research Reports will be
distributed to requesters as supplies permit. Re-
quests should be sent to the Publications Office,
Dept. of the Interior, Federal Water Pollution
Control Administration, Washington, B.C. 202^2
-------
Sulfide Treatment of Acid Mine Drainage
FEDERAL WATER POLLUTION CONTROL ADMINISTRATION
DEPARTMENT OF THE INTERIOR
Bituminous Coal Research, Inc.
350 Hochberg Road
Monroeville, Pennsylvania
Program Number
FWPCA Grant No. 1^010 DLC
November, 1969
-------
This report has been reviewed by the Federal
Water Pollution Control Administration and
approved for publication. Approval does not
signify that the contents necessarily reflect
the views and policies of the Federal Water
Pollution Control Administration.
ii
-------
ABSTRACT
The studies reported herein were a continuation of a 12-month program
initiated in June 196? with funds from the Appalachian Regional
Commission. During this earlier program a process for the treatment
of coal mine drainage was conceived, involving the combined addition
of limestone and hydrogen sulfide to effect precipitation of iron
sulfides (LHS process).
Results of laboratory-scale continuous flow tests of the LHS process
during the current program indicated that hydrogen sulfide feed must
be accurately predetermined and controlled to effectively precipitate
iron, yet to avoid an excess of reagent in the process effluent.
The black sludge formed during treatment undergoes oxidation at a rate
depending on drying conditions, with formation of elemental sulfur.
X-ray diffraction analyses indicate the iron sulfide sludge component
is an amorphous material, although chemical analyses indicate it is
initially a stoichiometric compound whose empirical formula corresponds
to that of ferrous sulfide.
The unstable nature of the sulfide sludge, possibility of polysulfide
formation during treatment, instability of mine waters of the type
amenable to treatment, and inadequacies of available gas metering
equipment are among the factors which militate against the controlled
regulation of hydrogen sulfide feed necessary for successful operation
of the process. These factors, and additional disadvantages revealed
by an updated cost evaluation, lead to the conclusion that the LHS
process is less attractive than accepted methods of mine drainage
treatment.
The current program was cosponsored by the Federal Water Pollution
Control Administration (FWPCA) and the coal industry through its
research agency, Bituminous Coal Research, Inc.
This report was submitted in fulfillment of project (FWPCA Grant No.
lUOlO DLC) between the Federal Water Pollution Control Administration
and Bituminous Coal Research, Inc.
Key Words
Coal Mine Drainage
Iron Sulfides Sulfides
Pollution Abatement Waste Water Treatment
Limestone Stream Pollution
Hydrogen Sulfide
iii
-------
CONTENTS
Abstract
Section 1 Conclusions and Recommendations 1
Section 2 Introduction 5
Section 3 Experimental 13
Section k Discussion ^3
Section 5 References 6l
Abstract Cards 67
-------
FIGURES
Figure
1 Schematic Flow Diagram of Conceptual IBS
Treatment Process .............. ..... ............... 7
2 Continuous Flow Test Apparatus ....................... 8
3 General View of Continuous Flow System ............... 9
k Schedule of Work - Sulfide Treatment of Acid
Mine Drainage (Grant No. 1^010 DLC) ................ 12
5 Change in Residual Dissolved Iron Concentration
with Time; Enclosed Reactor 5-gallon Batch Tests... 20
6 Change in Residual Dissolved Iron Concentration
with Time after I^S Addition; Enclosed Reactor
5-gallon Batch Tests ............................... 21
7 Continuous Flow Runs with 60-gallon LHS Reactor;
Change of Fe8* Concentration in Process Effluent
with Time .......................................... 24
8 Effect of Various Coagulant Aids on Zeta Potential
of Sulfide Sludge .................................. 38
9 Changes in Iron Concentrations and pH with Time
During Stirring of a "Yellowboy" Sludge
Suspension Treated with HgS ........................ Ul
vi
-------
TABLES
Table Page
1 Representative Compositions of Four Mine Waters
Used in LHS Continuous Flow Experiments 15
2 Summary of Results of LHS Continuous Flow Tests
During July-September, 1968. 19
3 Summary of Results of LHS Continuous Flow Tests
Using the 60-gallon HgS Reactor 22
4 Results of Sulfide Sulfur and Iron Analyses on
Sulfide Sludges and Pure Ferrous Sulfide 28
5 Comparison of Settled Sludge Volumes and Solids
Contents of Sludges Obtained by Lime, Limestone, and
LHS Treatment of South Greensburg Mine Water 31
6 Costs of LHS Process at Different HgS Requirements
in Comparison With Costs of Lime and Limestone
Neutralization 59
vii
-------
Section 1
CONCLUSIONS & RECOMMENDATIONS
Conclusions
The following conclusions appear to be supported by the results
presented herein:
Regarding the process wherein limestone and hydrogen sulfide
are used in combination for the treatment of coal mine water (LHS
process), experience has indicated that the technique of utilizing
a closed vessel in which limestone, hydrogen sulfide, and mine
water are intimately mixed, is the most feasible method for
achieving the desired precipitation of iron.
hydrous ferric oxide, previously precipitated by natural oxida-
tion or limestone neutralization of mine water, is rapidly reduced
in suspension near pH 6 by hydrogen sulfide. Elemental sulfur is a
byproduct of this reduction reaction. Consequently, hydrogen
sulfide feed calculations should be based on concentrations of both
ferrous and ferric iron rather than on ferrous iron concentration
alone.
In the absence of sufficient hydrogen sulfide to precipitate
all iron as the sulfide, ferrous iron formed by hydrogen sulfide
reduction of ferric iron may be in equilibrium with hydroxyl ion
and ferrous hydroxide. This may account for the apparent resolubil-
ization of ferrous iron from the sulfide sludge observed during
some experiments.
Accurate determination of the hydrogen sulfide feed require-
ment is necessary for the successful operation of the LHS process
in terms of the amount of iron removed. Such accurate determina-
tion was found to be difficult because of the occurrence of iron
oxidation during the brief storage of mine water samples collected
for continuous flow treatment experiments.
Results of a refined analytical technique for sulfide sulfur
and iron indicate that the iron sulfide species in the freshly
formed sludge is a stoichiometric compound corresponding to the
empirical formula FeS.
The sulfide component of the sludge itself (presumably ferrous
sulfide initially) is unstable with respect to oxidation, and
undergoes decomposition in suspension or on exposure to air with
the formation of elemental sulfur and, probably, amorphous ferric
oxyhydroxide. If dried rapidly at room temperature by acetone
washing and pulverized to a fine powder, or if dried in air in an
-------
2.
oven held at 110 C, the sludge undergoes a strongly exothermic
reaction which is probably the result of rapid oxidation.
The ease of oxidation of the sludge, and the absence of lines
corresponding to those of known iron sulfide compounds in x-ray
diffraction patterns of freshly formed sludge samples, indicate that
the iron sulfide component of the sludge formed under the prevailing
experimental conditions is an amorphous, structurally disordered,
and (presumably) hydrated material. The sulfide sludge can
apparently be preserved temporarily in its original state by vacuum
drying at room temperature.
The elemental sulfur (probably of colloidal size) formed during
sludge oxidation or ferric iron reduction may undergo reaction with
the hydrosulfide ion (HS~) with the formation of soluble polysulfide
species. This may result in a significant loss of hydrogen sulfide
in the system, in terms of its availability for reaction with iron.
Freshly precipitated suspended sludge particles from the LHS
treatment process are negatively charged. Cationic flocculating
aids are effective in promoting coagulation and more rapid settling
of the LHS sludge; anionic flocculating aids, used alone, are not
effective in this regard.
The sludge obtained from LHS treatment of one mine water, after
gravity settling, had a smaller volume and a higher solids content
by weight than sludges obtained by either lime or limestone treat-
ment of the same mine water.
A current evaluation of the LHS process in terms of its techno-
logical soundness and economic feasibility indicates that the
process is less attractive than accepted methods of treatment
relative to its applicability to different types of mine waters,
its ease of control, its simplicity in Implementation, and its costs
of operation.
Recommendations
It is felt that future efforts should include activities in the
following areas:
Additional studies of ferric iron-hydrogen sulfide reactions
with the use of varying amounts of hydrogen sulfide; identification
of the solid reaction products, and determination of changes in the
oxidation states of iron species (both soluble and insoluble)with
time.
Investigation of material balances for sulfur in the sludge,
in an effort to determine the extent of sulfur formation during
oxidation of the iron sulfide component of the sludge.
-------
Attempts to promote crystal ordering of the iron sulfide
specie(s) in the sludge through heating to moderately elevated
temperatures (e.g., 100 C) under vacuum.
Further study of factors affecting the hydrogen sulfide stoi-
chiometric requirement, with particular attention to the formation
and stability of polysulf ide species.
Further tests, supplemented by zeta potential measurements, on
the use of cationic coagulant aids to promote sludge settling, for
the purpose of determining optimum dosages, methods of application,
and relative effectiveness.
Verification of the exact hydrogen sulfide stoichiometric
requirement for quantitative precipitation of iron during treatment
through the use of pure ferrous sulfate and ferric sulfate solutions,
and mixtures thereof.
In terms of the longer-range goals involving adaptation of the
LHS process for the continuous treatment of mine drainage, further
consideration should be given to:
The volume of mine water available for treatment with a high
ferrous:ferric iron molar ratio, relative to the total volume of
coal mine drainage.
The development of more rapid and more accurate methods for
the determination and control of hydrogen sulfide feed during
treatment. This could involve work in such areas as the development
of more rapid analytical techniques for ferrous and ferric iron,
automatic measurement and control of hydrogen sulfide in the process
effluent, and the development of more reliable gas metering equipment
The possible effects of changing chemical composition and
physical properties of the sludge after treatment, as they pertain
to the feasibility of a reagent recovery step in the process.
The actual economic feasibility of a reagent recovery-recycle
step in the overall process.
-------
5.
Section 2
INTRODUCTION
Water pumped from active coal mines and water draining from
abandoned coal mines are both potential sources of water pollution.
The composition of such mine water can generally be described as a
mixture of metallic sulfates in an acidic or alkaline solution together
with a varying amount of suspended solids. Of primary importance in
establishing the degree of pollution of any mine discharge is the
concentration of dissolved iron and, to a lesser degree, the concen-
trations of aluminum and manganese. The discharge of untreated mine
water into the stream can lead to the pollution of the receiving stream
water. Dilution of the mine discharge by the receiving stream
generally causes an increase in pH which subsequently causes hydrolysis
of the ferric iron present; this in turn provides an environment
conducive to rapid ferrous iron oxidation which then leads again to
hydrolysis and precipitation of the hydrous iron oxides familiarly
known as yellowboy.
To eliminate the source of pollution from coal mine discharges,
the most commonly accepted treatment process, lime neutralization, is
applied. Lime neutralization involves the addition of lime in a
suitable form, either slaked or unslaked, to mine water to raise the
pH of the discharge sufficiently to precipitate iron and aluminum
hydroxides and subsequently to oxidize ferrous to ferric iron in an
optimum environment. The cost of lime neutralization depends on many
factors, including the volume of water, the concentration of pollutants,
and the cost of available real estate for the construction of settling
lagoons and sludge disposal sites.
In June of 196?, BCR initiated a research program to investigate
the use of various sulfides for the treatment of mine water. The
results of this work, which was conducted over a 12-month period under
a grant from the Appalachian Regional Commission, indicated that the
reaction of sulfides in mine water was such as to produce insoluble
iron sulfides which could be recovered and utilized as a source of the
sulfide reagent itself. Preliminary cost estimates suggested that,
with the incorporation of sludge recovery and reagent recycling, the
process would enable the treatment of mine water at lower cost than
lime neutralization. Moreover, because of the small quantity of solid
residue expected to be produced, the difficult problem of sludge
disposal would be alleviated to a major extent. The results of this
earlier program have been reported by Zawadzki and Glenn. (16)
Subsequent to June, 1968, the research program was extended for a
12-month period under the current cosponsorship of the Federal Water
Pollution Control Administration (fWPCA) and Bituminous Coal Research,
Inc., with the objectives of continuing the development of the sulfide
process for treatment of coal mine drainage utilizing a laboratory-
-------
6.
scale continuous flow system, and, using results of these studies, to
make preliminary engineering evaluations and cost estimates on the pro-
cess projected to a full industrial scale.
As a result of the earlier studies conducted for the Appalachian
Regional Commission, a conceptual limestone-hydrogen sulfide (iflS)
process for the treatment of mine water was developed. A schematic
flow diagram of the process is shown in Figure 1. The process incor-
porated the use of limestone to increase the pH of the system to about
6, promoting the dissociation of hydrogen sulfide:
H,S(aq) = H* + HS"
HS~ = H1" + S8"
Subsequent addition of hydrogen sulfide resulted in the precipitation
of iron as an Insoluble iron sulfide:
Fe3* -f &~ = FeS
This conceptual sulfide treatment process involved the various
unit operations of pH adjustment, sulfide reaction, sludge separation,
and sludge utilization. The flow diagram presented in Figure 1 provides
for a system based on currently available technology; therefore,
specific sludge separation devices are not indicated. A system for
regenerating sulfide reagent from the sludge has not yet been developed
and the diagram, therefore, shows sludge disposal as a present
alternative.
Also as a result of the work sponsored by the Appalachian Regional
Commission, a continuous flow test apparatus was designed and constructed
at the BCR laboratory. This system, with a capacity of 1 to 5 gallons
per minute, provided a means for evaluating the sulfide treatment process
under dynamic conditions. A schematic drawing of the test apparatus is
shown in Figure 2, and a photograph of the flow system in its original
form is included as Figure 3. A detailed description of the continuous
flow system equipment follows. It should be noted that this description
applies to the equipment which evolved after several modifications (to
be discussed in a later section) had been incorporated during the course
of the work; consequently, there are some slight deviations from the
representations given by Figures 2 and 3.
Reservoir: The reservoir has a total capacity of 210 gallons
and consists of two Glascote glass-lined tanks1 each having bottom
discharge and lids. The tanks have 150 gallon and 60 gallon
capacities.
1 Mention of commercial products throughout this report does not imply
endorsement by the Federal Water Pollution Control Administration.
-------
Limestone
Storage
Tank
Treated Water
Bituminous Coal Research, Inc. 2030G6
Figure 1. Schematic Flow Diagram of Conceptual LHS
Treatment Process
-------
Legend
A - Stirred Tank Reactor
B — Reservoir
C — Reservoir
D — Vibn Screw Feeder
E — Limestone Separator
F-H,S Cylinder
G-H,S Reactor
H - Settling Tank
oo
Bituminous Coal R«»««rch, Inc. 2028O13
Figure 2. Continuous Flow Test Apparatus
-------
(2030P25)
Figure 3. General View of Continuous Flow System
0
-------
10.
Sample Cooling and Circulating System; A Forma "Forma-Temp"
portable cooler can be used to maintain the sample at temperatures
as low as k6 F. The cooler has a rated capacity of UOOO Btu/hr.
Circulation of the sample between the two reservoirs is accomplished
by use of a Gorman Rupp Model 11698, 5600 Series pump, with a Gorman
Rupp Model 12500-21 oscillating pump used for priming.
IflS Reactor; The LHS reactor consists of a 60-gallon stainless
steel tank equipped with four 2-inch vertical stainless steel
baffles welded to the inner wall of the tank. Stirring is ac-
complished by use of a "Lightnin" Model HD-1A, 1/U horsepower
portable mixer. A I/1*-inch thick Plexiglas lid, containing a
rectangular access port covered by a hinged Plexiglas panel, is
bolted to the top of the 60-gallon tank. Rubber foam weather
stripping (3/l6-inch thick) is used as gasket material to seal the
lid. The shaft of the "Lightnin" stirrer passes through a fitting
packed with teflon cord, forming a water-tight seal. Mine water
is pumped into the upper portion of the reactor from the reservoir
with the aid of a Gorman Rupp Model 11698, 5600 Series pump. The
flow rate is controlled by a 1/2-inch Whitey Model 1RS8-316 stain-
less steel needle valve, and is measured by a Brooks Model 1305
flowmeter. The hydrogen sulfide inlet, formed from 1/8-inch 3D
stainless steel tubing, enters near the top of the reactor and
is directed down the inside and to the center near the bottom
of the tank so that the gas is discharged just under the im-
peller blade. Gas flow rate is controlled and measured by means
of a Brooks Model V 1300-01-FlB flowmeter, equipped with a needle
valve.
Limestone Feeder; Pulverized limestone is fed to the reactor
by a Vibra Screw feeder equipped with a 1/U-inch diameter feed screw.
Settling Tanks; Two 90-gallon settling tanks in series are
used to collect the sulfide sludge. These tanks are constructed
of 316 stainless steel and are closed and vented to the outside air
by ducting to a high capacity blower. Plexiglas covers enable
visual observation of sludge settling behavior.
Piping; Stainless steel piping is in use throughout the system;
however, large diameter Tygon tubing is used for flexible connection
between the settling tanks.
Initial efforts under the present program were directed toward
utilization of the continuous flow system for sulfide treatment ex-
periments on mine waters of varying compositions. During the course
of these experiments, however, various unforeseen difficulties were
encountered which necessitated changes in the general direction of the
work. These difficulties, and the attempts made to overcome them,
will be described in subsequent sections of this report.
-------
11.
The work schedule covering the present program was revised by mutual
agreement between BCR and the FWPCA at the end of the first eight months
of the program, in consideration of certain findings during this period.
In essence, these revisions involved a de-emphasis of the process engi-
neering development work, and a return to more fundamental investigations
of certain aspects relating to the basic chemistry of the sulfide treat-
ment process for mine water. The revised work schedule resulting from
these changes is shown in Figure h.
Subsequent to the completion of the originally scheduled work pro-
gram as revised (Figure 4), limited additional work on sludge characteri-
zation was undertaken by mutual agreement between the project cosponsors.
The results of this work are discussed in later subsections entitled
Tendency Toward Sludge Oxidation.
As work on the present program progressed, it became apparent that
the reactions involved in the system were more complex than had been
originally anticipated. Consequently, a more detailed literature search
was initiated with the objective of acquiring an improved understanding
of the chemistry of sulfide-iron reactions in solution. A number of the
references obtained during this search would be of value in any future
theoretical studies; these have been included, therefore, in Section 5,
REFERENCES, under the subheading Additional References, Uncited.
-------
Aug I Sep Oct
MONTH, 1968-1969
Nov I Dec | Jan | Feb
Mar I Apr I May Jun
1. Continuous Flow Experiments
a. Optimization of Continuous
Flow Equipment
b. Mine Water Composition
c. Sludge Characterization
d. Sludge Separations
e. Literature Search
f. Control Systems
Development
g. lron(lll)-HjS Reactions
2. Engineering Evaluation and
Cost Estimates
3. Reporting
Continuous Effort
Revised 2/20/69
Intermittent Effort
Bituminous Coil R«i«irch, Inc. 2028O15R2
Figure 4. Schedule of Work-Sulfide Treatment of Acid Mine Drainage
(Grant No. 14010 DLC)
-------
13.
Section 3
EXPERIMENTAL
The major portion of the effort during the research program was
devoted to determining the applicability of the limestone-hydrogen
sulfide process for the continuous treatment of various types of mine
waters. Accordingly, numerous continuous flow treatment experiments
were conducted using mine waters of varying composition. At the same
time, modifications in both equipment and procedure were made in an
effort to refine the technique. These modifications were supplemented
by smaller scale batch and continuous flow tests with the aid of a 5-
gallon reactor.
During the development of the continuous flow treatment process,
however, certain discordant results were obtained which could not be
explained wholly on the basis of faulty technique or equipment design.
Consequently, during the last four months of the program efforts were
devoted essentially to studies of sludge formation, composition, and
behavior. These studies involved brief and somewhat interrelated in-
vestigations of sludge chemical composition, physical properties,
stability on exposure to air, electrophoretic mobility in the presence
of various coagulant aids, and formation in the presence of hydrous
ferric oxide.
This section deals with the experimental procedures and results
obtained in each of the above-mentioned work areas.
Continuous Flow and ^-Gallon Batch Experiments
Reagents and Analytical Techniques
One of the initial purposes of the continuous flow experiments was
to test the effectiveness of the sulfide treatment process on mine
waters of varying composition. In selecting mine waters for treatment
experiments, attention was given to parameters such as total iron con-
centration, ferric:ferrous iron ratio, and total acidity.
Four different mine waters were used in the earlier LHS continuous
flow treatment experiments. Three of these were selected based on data
presented in a report entitled "Sewickley Creek Area of Pennsylvania"
prepared for the Monongahela River Mine Drainage Remedial Project by
the FWPCA, June, 1966, cited by Zawadzki and Glenn. (16) The fourth
site is near Delmont, Pennsylvania, about ten miles east of the BCR
laboratory. The site locations, all in Westmoreland County, and the
nature of mine water samples taken from them, are described briefly
as follows:
South Greensburg (JWPCA No. 16-003); An inactive drift mine in
Hempfield Township located on the north side of Pennsylvania T-681
-------
northwest of its junction with Pennsylvania 6Ulll. Samples axe
characterized by the usual absence of ferric iron and a fairly
constant composition in spite of variations in flow rates.
Keystone (FWPCA No. 22-001); An inactive drift mine in
Sewickley Township located near the junction of Sewickley Creek with
the Youghiogheny River, approximately 200 feet west of Pennsylvania
6UlO^ and 0.6 mile south of the junction of this road with
Pennsylvania T-400. Samples are characterized by a relatively high
ferrous iron and sodium content and by a relatively high pH. The
ferric iron present tends to hydrolyze and precipitate fairly
rapidly during storage of the mine water.
Tarrs (fWPCA No. 9-H8); An abandoned mine site in East
Huntingdon Township about 0.5 mile southwest of the town of Tarrs,
Pennsylvania. Samples are characterized by a relatively high
aluminumttotal iron ratio and a relatively low pH, considering the
rather high ferrous:ferric iron ratio.
Thorn Run Reservoir; A mine water holding lagoon in Salem
Township about 0.3 mile east of Pennsylvania Route 66 and 3.3 miles
north of the junction of this road with U.S. Route 22. Samples are
characterized by a high ferric:ferrous iron ratio and low pH. The
composition of the mine water varies widely depending on flow rate,
suggesting a dilution effect from surface water run-off during
periods of rainfall.
Representative analytical data for these four sanples are shown in
Table 1.
In the later continuous flow tests, as well as batch tests and
sludge characterization studies, mine water from the South Greensburg
site was utilized almost exclusively. Because of the usual absence of
ferric iron in fresh samples and its relatively constant composition,
the South Greensburg mine water was employed in effect as a "control"
sample throughout the investigation.
The limestone used throughout the course of the investigation, des-
ignated as BCR No. 1809, was obtained from the West Winfield Lime and
Stone Company, West Winfield, Butler County, Pennsylvania. This
material is a deep-mined, high calcium limestone containing about 89
percent calcium oxide by weight of ignited (900 C) sample, and has been
found by tests conducted under another BCR project to be a relatively
effective neutralizing agent compared with other limestones of similar
composition (unpublished data). Preliminary continuous flow tests
were conducted using limestone which had been ground from lump size
to 88 percent by weight minus 200 mesh. In later work, the same lime-
stone was obtained as the agricultural grade, then ground and sieved to
minus 325 mesh. In all but a few 5-gaUon batch tests, the amount of
-------
TABLE 1. REIRESENTATIVE COMPOSITIONS OP FOUR MINE WATERS USED IN
LHS CONTINUOUS PLOW EXPERIMENTS (Concentrations in ppm)
Acidity,
CaCOj Fe3* /Fe3*
Source pH equiv. Fe8* Pea* Ratio Al Ca Mg Mil J& Na SO^
South Greensburg* k.86 2kk 112 0 0 8 l8l 7k U lb 70 1127
Keystone* 6.15 3^9 322 33 0.10 3 255 111 5 8 1125 37^6
Tarrs* 2.88 736 101 2k 0.2k 57 197 85 k Ul — ikjk
Thorn Run „, , . .
Reservoir** 2.69 90k 16 106 6.62 lk 189 66 Ik 2k 35 3*12
* Average of four samples collected in July-August, 1968. Notable characteristics;
Absence of Fe3* , fairly constant composition.
•*• Average of two samples collected on August 21 and 29, 1968. Notable characteristics;
High Fe3* and Na* content, relatively high pH, unstable during storage.
* Average of two samples collected on August 28 and September U, 1968. Notable
characteristics; Relatively high aluminum/total iron ratio (reflected in total acidity
value), relatively low pH considering low Fe^/Pe3* ratio.
** Average of three samples collected in July-August, 1968. Notable characteristics;
High Fe3*/Fe3* ratio, widely varying composition depending on flow conditions.
-------
16.
pulverized limestone employed, in treatment experiments was twice the
stoichiometric requirement based on mine water total acidity.
Reagent grade hydrogen sulfide used in all tests was obtained from
The Matheson Company in a pressurized cylinder. Gas flew rate was
measured with the aid of a Brooks "Sho-Rate" flowmeter calibrated to
read directly in standard liters per minute of gas with a specific
gravity of 1.1? at S.T.P. (i.e., hydrogen sulfide). The flowmeter
calibration was checked occasionally using air and a wet-test meter,
with appropriate density corrections. Ordinarily, hydrogen sulfide
was added in an amount equal to the exact stoichiometric requirement
based on the following reactions:
Fe3* + HgS = FeS + 2 if
2 Fe3* + IfeS = 2 Fe3* + 2 if + #
For reasons which will become apparent later, it was sometimes necessary
to back-calculate the hydrogen sulfide stoichiometry after a treatment
experiment due to changes in iron concentration and/or oxidation state
during storage of the mine water samples. In these instances, the
amount of hydrogen sulfide employed proved to be some multiple or
fraction of the exact stoichiometric requirement.
Ferrous iron concentration was determined by the colorimetric
method using ortho-phenanthroline. Sample absorbance was measured at
52Cto}i with the aid of a Beckman Model DB spectrophotometer. Unless
indicated otherwise, ferrous iron analyses were conducted on filtered
samples.
Total iron was determined by emission spectrographic analysis using
a solution technique. This same procedure was used for the determination
of other cations of interest. Concentrations of these cations (aluminum,
calcium, magnesium, manganese, silicon, and sodium) are expressed through-
out this report as milligrams of the elemental species per liter.
Total acidity was determined by titration to pH 8.2 with 0.1N
sodium hydroxide of the solution, previously treated with hydrogen
peroxide solution, boiled, and cooled. All acidity values reported
herein are expressed as milligrams per liter calcium carbonate
equivalents.
Experimental Procedure and Results
The experimental procedure for continuous flow tests was varied
considerably in the early stages of the investigation as changes in
equipment and technique were incorporated. In the first few flow ex-
periments, pulverized limestone was mixed with the mine water in a
separate vessel before the introduction of hydrogen sulfide to the
suspension. This approach proved to be unsatisfactory since there was
no provision for nijx-tng downstream from the point of hydrogen sulfide
-------
17.
introduction, and pH recovery after hydrogen sulfide addition was too
slow.
After various mechanical problems had been encountered (see below),
the 60-gallon mixing tank was converted to an enclosed stirred reactor
with the addition of a Plexiglas lid and appropriate inlet and outlet
lines. This innovation permitted simultaneous addition of limestone
and hydrogen sulfide to the mine water with high-speed stirring, and
provided for more efficient use of the reagents.
The procedure ultimately adopted for continuous flow treatment
experiments is as follows: Mine water samples were collected in the
field in four plastic-lined 55-gallon drums and, immediately upon re-
turn to the laboratory, were transferred to the holding tanks. During
storage, samples were maintained at a temperature in the range 8 to 12 C
(U6 to 5^ F) with a portable Immersion cooler to minimize iron oxidation
and precipitation. Samples of the raw mine water were taken immediately
for as-received determinations of pH, acidity, ferrous and total iron,
and in some cases, sulfate concentration. In addition to total iron,
concentrations of aluminum, calcium, magnesium, manganese, silicon, and
sodium were ordinarily determined by the emission spectrographic
technique on raw mine water samples.
At the start of a continuous flow treatment run, the 60-gallon
reactor was first charged with 50 gallons of raw mine water. The
entire amount of limestone required by the 50-gallon sample was weighed
and added to the contents of the reactor just before mixing and hydrogen
sulfide flow commenced. The system was then operated in this "batch
configuration" for a 50-minute period. At the end of this time period,
mine water influent and treated effluent flows were started at the rate
of 1 gallon per minute, continuous limestone feeding was commenced,
and the system was operated in a "flow configuration" for the duration
of the run. After the system was operating in the "flow configuration,"
samples were taken immediately from the reactor and at subsequent time
intervals from the first settling tank for pH measurement and iron
analyses. In addition, after all flow had ceased in the system, samples
were taken periodically from the settling tanks for the purpose of
determining changes in pH and dissolved iron content in the treated
effluent with time.
The results of the earlier continuous flow runs, in terms of the
degree of iron removal achieved, tended to be erratic and somewhat in-
conclusive. This was due in part to mechanical difficulties, as indi-
cated earlier, most of which gave rise to problems in maintaining
smooth flow and adequate hydrogen sulfide-liquid mixing within the
system. However, based on discoveries made later in the investigation,
it is likely that the unsatisfactory iron removals in earlier runs
were also influenced by changes in mine water composition prior to
treatment.
-------
18.
The results of these preliminary flow runs are typified by the data
in Table 2. As indicated therein, the iron removals effected were only
about 60 to 70 percent in all cases.
These preliminary continuous flow experiments were followed by a
series of 5-gallon batch tests, designed to provide closer insight into
the fate of dissolved iron during treatment through closer control of
experimental conditions. The batch tests were conducted in an enclosed,
stirred reactor built from a 5-gallon polyethylene screwcap bottle.
Stainless steel tubing was used to provide a mine water inlet near the
top and a hydrogen sulfide inlet at the bottom of the vessel.
In a typical batch treatment experiment, a 5-gallon sample of mine
water was mixed with the requisite amount of limestone for 15 minutes,
followed by introduction of hydrogen sulfide over a 5-minute period.
Samples were withdrawn from the tap at the bottom of the stirred reactor,
both before and for some time after hydrogen sulfide addition, for pH
and ferrous iron analyses.
Results of these 5-gallon batch tests are illustrated graphically
in Figures 5 and 6. The results shown by Figure 6 were obtained after
the hydrogen sulfide flowmeter was recalibrated and are, therefore,
probably more meaningful. The runs reflected by Figure 5 were conducted
using South Greensburg mine water with a pH of 5.00, 111 ppm ferrous
iron, 0 ppm ferric iron, and 267 ppm total acidity; those represented
by Figure 6 were conducted using mine water from the same source with
a pH of 5.0^, containing 123 PP* ferrous iron, 0 ppm ferric iron, and
260 ppm total acidity.
These batch tests were followed by conversion of the 60-gallon
mlv-ing tank to an enclosed, combination limestone-hydrogen sulfide
reactor and the conduct of five additional continuous flow experiments
by the procedure described earlier. During these experiments it was
discovered that significant amounts of ferrous iron were undergoing
oxidation even during brief storage periods of approximately two hours
prior to treatment. Evidently, this problem could not be avoided even
when the mine water samples were collected in the field, transported
to the laboratory, analyzed, and utilized in the flow system all within
an 8-hour period. Consequently, although preliminary hydrogen sulfide
feed requirements had been calculated based on the "as received"
analyses of the mine water, in each run it was necessary to back-
calculate the stoichiometric amount of hydrogen sulfide used after the
run had been completed, based on updated analytical results for ferrous
and total iron on samples taken during the run itself.
This situation is reflected in Table 3, which summarizes the per-
tinent analytical data for the five continuous flow experiments. These
data show that the original mine water composition was essentially the
same in each of the runs. Furthermore, the mine water was essentially
free of ferric iron at the time of collection. These results were con-
sistent with those obtained earlier for this particular mine water.
-------
TABLE 2. SUMMARY OF RESULTS OF LHS CONTINUOUS FLOW TESTS DURING JULY-SEPTEMBER, 1968*
Run No.
and Date
372-10
7/25
372-17
8/9
372-26
8/20
372-32
8/30
372-21
B/lU
372-31+
9/1+
AMD
Source
South
Greensburg
South
Greensburg
South
Greensburg
Keystone
Thorn Run
Reservoir
Tarrs
Sample
Raw AMD
LHS-Treated
flaw AMD
LHS-Treated+
Raw AMD
LHS-Treated
Raw AMD
LHS-Treated
Haw AMD
LHS-Treated
Raw AMD
LHS-Treated
pH
1+.78
6.1+0
U.88
5.28
1+.76
5.63
6.15
5.80
2.90
6.20
3.07
5.61
Acidity, ppm
CaCO, equiv.
21+5
—
21+0
266
"""
359
171
520
8
751
81+
Fe3*,
ppm
115
111
30
111
1+6
302
103
10
16
107
1+1
Spectrographic Analyses,
Fe
98
55
96
27
98
1+3
360
92
52
20
128
1+0
Al
9-5
6.5
7.8
<3
8
<3
3
<3
1+2
<3
59
<3
Ca
185
260
175
255
193
265
250
380
165
335
aoo
1+ltO
Mg
77
70
76
76
68
70
112
115
50
50
88
87
Mn
>+.5
3.5
3.5
3.6
3"4
3. '6
5
6
11
11
1+.2
1+.6
ppm
Si
12
8
15
12
15
12
7
8
16
11
1+2
35
Na
--
70
8s
73
81
1100
1080
30
..
--
Sulfate,
ppa
—
—
1127
—
371+6
3718
1090
1083
11+71+
11+20
Iron, Percent
Removal*
52
73
59
71
61
68
* Mine water flow rate was 1 gpm in all runs; limestone (BCR No. 1809) was added at 2 times
the stoichiometric requirement based on mine water acidity.
4- H9S feed was increased to 5 times the stoichiometric requirement in this run; in all other
runs, tigS feed was 1 times the stoichiometric requirement based on total iron.
* Where discrepancies in analytical results exist, the larger of the two values was used in
the calculations.
-------
120-1
100-
80
Z 60-
iu
o
8
i 40-
£
20-
Composition of Original Untreated Mine Water (South Greensburg):
111 ppm Fe*+, 0 ppm Fe3+,
267 ppm Acidity, pH 5.00
• Run 372-53-2XH; 2X Limestone, 2X H«S
o Run 372-56-4XH; 4X Limestone, 2X HjS
• Run 372-53-2X; 2X Limestone, IX HSS
» Run 372-56-4X, 4X Limestone, IX HjS
A: 15 Minute Limestone Mixing Time
B: 5 Minute H j S Reaction Time
10
30
50
TIME, minutes
70
~r
90
110
Bituminous Coal R»t»«rch, Inc. 2028G1C
Figure 5. Change in Residual Dissolved Iron Concentration with Time;
Enclosed Reactor 5-gallon Batch Tests
-------
ft
1
I
u
O
u
80-1
60-
40-
20-
Composition of Original Untreated Mine Water (South Greensburg):
123 ppm Fe*+, 0 ppm Fe3+,
260 ppm Acidity, pH 5.04
Run 372-62-2X; 2X Limestone, IX HjS
20
Run 372-64-2X; 2X Limestone, 1.5X HjS
40
T~
60
80
100
TIME AFTER HjS ADDITION, minutes
120
Bituminous Coal Research, Inc. 2028O19
Figure 6. Change in Residual Dissolved Iron Concentration with Time after H8S
Addition; Enclosed Reactor 5-gallon Batch Tests
-------
TABLE 3. SUMMARY OP RESULTS OP LHS CONTINUOUS PLOW TESTS
USING THE 60-GALLON Ha3 REACTOR!
Raw AMD Analysis
Raw AMD Analysis
LHS-Treated
Final Effluent
As Received
Run No.
372-78
372-98
372-8?
372-73
372-82
Fe~",
PPP
100
98*
100
107
93
Fe**,
ppm pH
3 5.55
0*
0
8
0
5.31
5.10
5.59
5. to
Storage
Period,
Hours
2k
2
2
1*8
2
After Storage Period
Fe**,
ppm
88
7k
75
79
68
Fe3*,
ppm
15
21*
25
36
25
Percent Fea+ Amount
Oxidized HgS
During Storage Used*
12.
21*.
25.
26.
26.
0
5
0
2
9
0.90
0.91
0.9^
1.21
1.31
Analysis**
Fe*T,
BPP
52
2k
15
0
8
_pH_
6.UO
6.1k
6.10
6.23
5.88
i South Greensburg mine water was used in all runs at a flow rate of 1 gpm; limestone (BCR No. 1809,
-325 mesh) was added at twice the stoichiometric requirement based on mine water acidity.
* Mole fraction of stoichiometric amount, calculated after the run and based on the reactions:
Fe3
2 Fea
FeS + 2 H*
2 Fe3* + 2 if + S°
** The "final" effluent sample was that taken from the first settling tank at the end of the
day on which the run was conducted. The treated effluent was then allowed to stand un-
disturbed overnight. Total and ferrous iron analyses on these samples agreed within 2 to
3 ppm of each other, essentially indicating the absence of Fea* in the treated effluent.
* Identical results were obtained on field samples acidified at the time of collection, indicating
essentially no Fea+ oxidation occurred while the samples were in transit to the laboratory.
-------
23.
With the exception of the results from Run 372-78, the data indi-
cate that about 25 to 2? percent of the ferrous iron in the mine water
underwent fairly rapid oxidation during storage of the sample prior to
treatment, and there is some evidence (Run 372-73) that no appreciable
further ferrous iron oxidation occurred during a subsequent prolonged
storage period (with cooling) up to two days. This seemingly anomalous
behavior may be explained by the fact that, as oxidation and hydrolysis
of iron occurs, the pH of the system decreases, thereby decreasing the
rate of ferrous ion oxidation to a point that the rate becomes negligible
and the system is in equilibrium. In addition, it should be recognized
that mine water samples were stored with cooling in covered tanks,
thereby limiting the access of air to the samples. If one assumes that
diffusion of oxygen through the air-liquid interface was the rate-
controlling step in the iron oxidation process, then it seems reason-
able that ferrous iron oxidation should either cease or proceed at a
negligible rate once dissolved oxygen and oxygen in the small air
spaces at the tops of the storage tanks was depleted. Based on the
analytical data (Table 2), one would expect the Keystone and, possibly,
the Tarrs discharges to behave in a similar fashion. Unfortunately,
however, usage of these two mine waters was limited in the present
study, and the data obtained therefrom are insufficient for the purpose
of comparison.
The data in Table 3 show a rough correlation between the amount
of hydrogen sulfide employed and the amount of ferrous iron remaining
in the treated effluent. This trend is illustrated more clearly by
Figure 7, which shows the change in ferrous iron concentration in the
effluent with time at the different hydrogen sulfide feed levels.
Concerning Figure 7, it should be pointed out that time zero indicated
thereon corresponds to the time at which mine water flow was begun
in the system. As explained earlier, this point in time was preceded
by a 50-minute mixing period during which the system was operating in a
"batch configuration" with only hydrogen sulfide flow into the mixing
tank containing the initial 50-gallon mine water-limestone charge. The
ferrous iron concentration at time zero was determined on a sample taken
from the reactor itself; all subsequent samples were taken from the
first settling tank. The solid-line portions of the curves in Figure 7
correspond to the time periods during which flow was occurring in the
system. The dashed-line portions correspond to the time periods after
flow ceased, during which the treated effluent was allowed to stand un-
disturbed in the settling tanks. The "final" treated effluent analyses,
shown in the extreme right-hand columns of Table 3, correspond to the
last points on the curves of Figure 7 before the breaks indicating the
overnight settling periods.
During the course of the continuous flow test work a number of
improvements were made in the design of the system. A few of these
were relatively minor, such as the installation of drains from all
holding tanks and the addition of a glass tube level indicator on
the mixing tank. The major modifications, resulting from recurring
-------
II
O4
£
60
50
40
30
20-
10
• Run 372-78
o Run 372-98
• Run 372-87
• Run 372-82
* Run 372-73
„ - 0.90 x H8S
0.91 x HsS
0.94 x H«S_ _ ^ — — *"
1.31 xHsS
i
60
120 180
TIME, minutes
240
Overnight
Bituminou* Coal R««««rch, Inc. 202B020
Figure 7. Continuous Flow Runs With 60-gal I on LHS Reactor; Change of
Fe*+ Concentration in Process Effluent with Time
-------
25.
mechanical difficulties encountered, are described in the following
paragraphs.
In the system as originally designed, the treated suspension con-
taining excess limestone particles was transferred by a small centrif-
ugal pump through a flovnneter into the first settling tank. This led
to frequent problems due to clogging in the flcwmeter and regulating
valve, resulting in erratic flow downstream from the mixing tank. This
difficulty was overcome by removing the pump and flowmeter, raising
the mixing tank to a higher elevation, and installing a length of 1/2-
inch ID stainless steel pipe directly from the bottom of the mixing
tank to the settling tank inlet to permit gravity flow of the process
effluent.
The limestone screw feeder installed in the continuous flow system
was originally purchased for another BCR project and was not designed
to provide directly the low feed rates (1.5 to 2.0 grams per minute)
required in the flow tests. It was found necessary to calibrate the
feed rate by a trial and error method prior to each run, and this
erratic response of the feeder led to uncertainties concerning control
of limestone addition. Various remedies were attempted, and other
types of feeders were evaluated during the course of the investigation.
Rather late in the experimental work, however, it was discovered that
replacement of the 1/2-inch feed auger with one of 1/4-inch size
(originally believed to be unavailable from the manufacturer) resulted
in satisfactory, controllable feed rates for the pulverized limestone.
Various methods of hydrogen sulfide introduction were attempted
early in the investigation. The original arrangement wherein hydrogen
sulfide was admitted through a tee ahead of a coiled section of tubing
(Figure 2) was wholly unsatisfactory due to insufficient reaction time,
inadequate gas-liquid mixing, and problems with back-pressure flow of
liquid into the hydrogen sulfide flowmeter. Admitting the hydrogen
sulfide through a fritted glass diffuser into a stirred compartment
at the head of the first settling tank resulted in only a slight im-
provement; difficulties due to inadequate gas-liquid contact time and
loss of hydrogen sulfide by diffusion were still encountered. The
technique which eventually proved satisfactory involved admitting the
gas through an open pipe (1/8-inch ID stainless steel tubing) directly
under the impeller of the high-speed mixer. Direct observation con-
firmed that the hydrogen sulfide, which is relatively soluble to the
extent of 2582 cc in 1000 cc of water at 20 C (10), apparently dis-
solved completely during stirring before it had a chance to escape
through the air-liquid interface.
Early in the research program, the following equipment manufac-
turers were contacted concerning hydrogen sulfide feed control in-
strumentation :
-------
26.
The Foxboro Company, Foxboro, Massachusetts
Honeywell, Minneapolis, Minnesota
API Instruments Company, Chester-land, Ohio
ITT Barton, Controls & Instrument Div., Monterey Park,
California
Among these, the only response during the course of the work was
from a representative of Honeywell who indicated that his company does
not manufacture analytical control devices for hydrogen sulfide. He
suggested, however, that it might be possible to couple a gas flow
measuring device such as Honeywell's integral orifice assembly with a
hydrogen sulfide-sensing device such as the "Toxgard," manufactured
by Mine Safety Appliances Company (MSA), Pittsburgh, Pennsylvania.
The latter instrument is an amperometric-type device capable of detect-
ing hydrogen sulfide in the range 0 to 50 ppm.
In a subsequent discussion, a representative of MSA advised that
the "Toxgard" would not be suitable as a hydrogen sulfide feed control
device, either alone or in conjunction with some type of flow regulator.
He explained that the "Toxgard" is designed primarily as a monitoring
instrument to warn of hazardous levels of toxic gases within a working
area (i.e., it is not designed for use on process streams). The MSA
representative suggested that a technique involving continuous chroma-
tography or ultraviolet spectroscopy might be the best approach for our
desired application; however, both analytical techniques require rather
expensive instrumentation.
In summary, the limited response to our inquiries concerning
hydrogen sulfide feed control instrumentation was rather indifferent,
and indicated that control instrumentation for use with our specific
application was not commercially available. In view of these findings,
therefore, it was felt that the development of such control instrumen-
tation would be an undertaking beyond the scope and resources of the
present project.
Sludge Characterization Studies
At the inception of the current research program, essentially no
precise information was available on the chemical composition or
physical properties of the freshly foamed, black precipitate produced
by the IBS treatment of mine water. Under the previous program (for
the Appalachian Regional Commission), some data were accumulated on
sludge samples dried under nitrogen at 300 C for 1 to 2 hours. For
example, sulfur;iron ratios on sludges treated in this manner were
found to be in the range l.k to 1.5; and among the iron sulfide
species identified by x-ray diffraction analyses of the pyrolyzed
samples were FeS, FeSg, and Fe3S4.
There was some doubt as to whether these results were indicative
of the actual composition of the freshly formed sludge. For this reason,
and in an effort to better understand the phenomena observed during the
continuous flow experiments, sludge characterization studies were initi-
ated on a limited scale during the later stages of the investigation.
-------
27.
Experimental Procedure and Results
Chemical Properties
A procedure was developed for the determination of iron and sulfide
sulfur on sludges formed during the MS treatment process. The proce-
dure was based on decomposition in 6N hydrochloric acid of the filtered,
washed sludge in a closed system. An all-glass apparatus connected with
standard taper joints and short lengths of Tygon tubing was used in
these studies.
A sample of the sludge, filtered on a 0.22ji pore size Millipore
filter disk and washed with deionized water, was wrapped in a small
piece of aluminum foil and immediately transferred to a 50-ml boiling
flask containing 25 ml 6N hydrochloric acid and equipped with a magnetic
stirring bar. The reaction flask was connected with two gas bubbler
tubes in series each containing a solution of 1M cadmium chloride.
The entire system was purged with prepurified nitrogen gas for at least
1/2 hour before the sample was introduced.
As the sludge sample decomposed in the stirred, acidic solution,
hydrogen sulfide liberated during the reaction was swept with nitrogen
into the bubbler tubes wherein cadmium sulfide precipitated. The de-
composition was driven to completion by heating the acidified solution
to boiling, after which the contents of the flask were allowed to cool,
diluted to 250 ml, and analyzed for total iron by the emission spectro-
graphic technique. The cadmium sulfide precipitate was filtered in a
fine-porosity fritted glass crucible, washed, dried at kOO C for 2 hours,
and weighed.
During the development of the procedure and analyses of actual
sludge samples, control runs with pure ferrous sulfide were conducted
for comparison of the results. In addition, a blank experiment with
aluminum foil alone in the system showed that no measurable iron was
being contributed to the solution by the aluminum itself.
The results of sulfide sulfur and iron analyses conducted on
sulfide sludges and pure ferrous sulfide by the above-described pro-
cedure are given in Table k. The sulfide sludge samples were all
prepared by treatment of South Greensburg mine water in batch experi-
ments using the exact stoichiometric requirement of hydrogen sulfide.
Each of these batch tests involved the use of twice the stoichiometric
requirement of limestone except for Test No. 385-^8, in which IN
sodium hydroxide was used for reasons to be discussed later. In
addition, in all tests involving sulfide sludge samples except Test No.
385-37, the original mine water sample was filtered through 3.0^ pore
size Millipore filter disks immediately prior to treatment. This was
done to reduce interferences due to suspended hydrous ferric oxides
in the mine water (also to be discussed later).
-------
TABLE 1*. RESULTS OF SULFIDE AND IRON ANALYSES ON SULFIDE SLUDGES
AND PURE FERROUS SULFIDE
Test No.
385- 7
385-10
385-18
385-23
385-30
385-31
385-37
385-38
385-^2
385-^
Sample
Description
FeS
FeS
FeS
FeS
Sulfide Sludge
FeS
Sulfide Sludge
FeS
Sulfide Sludge
Sulfide Sludge
Amount
Fe
0.918
0.980
0.953
0.985
0.255
0.962
0.219
Q.Qkl
0.139
0.309
Found by Analysis,
mmole
s—
0.996
1.011
1.022
0.989
0.255
0.961
0.238
1.01*0
0.162
0.306
Sa":Fe
Mole
Ratio
1.08
1.03
1.07
1.00
1.00
1.00
1.08
1.2U
1.16
0.99
-------
29.
X-ray diffraction analyses were conducted using the powder camera
technique and a scanning diffractometer, each employing copper IQy
radiation. Outside services were required for the use of the latter
instrument. It was found that sludge samples, which ordinarily under-
went oxidation rather rapidly after exposure to air, could be preserved
in their original state for at least several hours by drying them in a
vacuum oven at room temperature.
X-ray diffraction analysis was conducted on the sulfide sludge
from Test No. 385-^2. The sludge was prepared specifically for this
purpose, and x-ray analysis was accomplished immediately after pre-
paration while the vacuum-dried sludge was still black in color. A
sample of powdered, pure ferrous sulfide was also submitted as a
reference material.
The peaks of the x-ray pattern for the ferrous sulfide reference
sample corresponded closely in position and intensity to those for
ordered ferrous sulfide (Troilite; Powder Diffraction File card No.
l<-0832). For the sulfide sludge sample, all peaks in both the powder
photograph and the recorded diffractograin were assignable to either
calcite (CaCGj) or cr-quartz (Si03), both compounds presumably orig-
inating from unreacted limestone in the sludge sample. Although it
was possible that the 100 percent peak of ferrous sulfide at d=2.09 A
was masked by a relatively strong 18 percent peak of CaCO^ appearing
in the diffractogram at d=2.08 A (reference value d=2.095 A), a
numerical analysis of the relative peak intensities indicated that
this was not so.
To avoid this possible interference due to the presence of calcite
in the sample, the experiment was repeated using IN sodium hydroxide
during sludge preparation. The mine water was maintained at pH 6 by
incremental additions of the base during hydrogen sulfide addition.
X-ray diffraction of the resulting black, vacuum-dried solids failed
to produce any definitive pattern.
Physical Properties
Again, only limited data were available on the physical properties
of the freshly formed sludges at the inception of the current research
program. As sludge chemical properties began to take on added impor-
tance, it was decided to investigate briefly two selected physical
properties of the sludge, namely, solids content and settled sludge
volume.
Two experiments on comparison of sludge volumes and solids contents
were conducted based on lime, limestone, and LHS treatment of South
Greensburg mine water, according to the following procedure.
-------
30.
Concurrent with the continuous flow LHS treatment experiment, two
batch samples of the raw mine water were taken for lime and limestone
neutralization tests. Sufficient 30 percent hydrogen peroxide solution
was added to each sample to oxidize ferrous iron, then one sample was
neutralized with twice the stoichiometric requirement of limestone (-325
mesh BOB No. 1809), and the second sample was neutralized with an amount
of 10 percent hydrated lime slurry sufficient to raise the pH to 7. The
limestone-treated suspension was stirred for °X) minutes, and the lime-
treated suspension was stirred just long enough to effect the desired
neutralization and precipitation of iron. The final pH of both suspen-
sions was about 7. These samples, as well as a sample taken during
the IHS treatment experiment, were transferred to 1-liter Imhoff cones
and sludge volumes were measured after overnight settling. The super-
natant liquid was then decanted from each sample by careful siphoning
to avoid disturbing the settled solids, and portions of the gravity-
settled solid material were transferred to small, tared beakers.
The solids were weighed, dried at 110 C for 1 hour, allowed to cool,
and weighed again. In addition, similar weighed samples of the settled
sludge were washed on No. h2 Whatman filter paper, and the residue ob-
tained after evaporation of the filtrate was weighed. In this manner,
the contribution due to dissolved solids in the original solids content
determination could be ascertained.
The results of sludge volume comparison tests and solids content
analyses, conducted during continuous flow LHS treatment experiments
372-82 and 372-98, are shown in Table 5. As indicated therein, fairly
good agreement between the results of the two tests was obtained in
most cases.
Tendency Toward Sludge Oxidation
Rather early in the experimental work, it was observed that damp
sludge filter residues resulting from the LHS treatment of mine waters
tended to change coj.or from black to rust-orange within minutes after
exposure to air. Scheduling arrangements involving the process develop-
ment and continuous flow testing phases of the project did not permit a
fuller investigation of this phenomenon until much later in the experi-
mental studies, however. This aspect of sludge behavior was investigated
briefly in the following experiment:
A sulfide sludge was prepared by the LHS treatment of South Greens-
burg mine water in the 5-gallon reactor. Two samples of the suspension
were withdrawn 45 minutes after the reaction and filtered on 0.22jji. Milli-
pore paper.
The first filter cake was washed with dilute hydrochloric acid,
leaving a brown residue which was found to be a mixture of sulfur and
quartz. The sulfur presumably arose from hydrogen sulfide reduction of
ferric iron, known to be present in the mine water prior to treatment,
and/or partial sludge oxidation before filtration.
-------
TABLE 5. COMPARISON OF SETTLED SLUDGE VOLUMES AND SOLIDS CONTENTS OP SLUDGES
OBTAINED BY LIME, LIMESTONE, AND LHS TREATMENT OF SOUTH GREENSBURG MINE WATER
Sludge Volume Percent
after Overnight Gravity Settling
Run No.
372-82
372-98
Lime Sludge
U. 00
5.00
Limestone Sludge
1.25
1.20
LHS Sludge
0.50
0.55
Weight Percent Solids
in Unwashed Sludge
372-82
372-98
0.77
2.28
7.80
5.U9
Weight Percent Dissolved
Solids in Unwashed Sludge
372-82
372-98
0.16
0.16
0.06
0.21*
Weight Percent Solids
in Washed Sludge*
372-82
372-98
0.63
0.78
2.12
2.W
7.56
5.35
* Weight percent solids in washed sludge calculated as weight
percent solids in unwashed sludge minus weight percent dissolved
solids.
U)
H
-------
32-
The second filter cake was washed with water alone and allowed to
air dry at room temperature, during which time it changed from black to
a rust-brown color. X-ray analysis of a portion of the dried solids re-
vealed sulfur and calcium carbonate; no lines due to iron compounds
could be identified. The remaining rust-brown solids were washed with
dilute hydrochloric acid, leaving a substantial amount of a grey residue
identified as sulfur. The acidified washings were found to contain 1
ppm ferrous iron, 36 ppm ferric iron, and no sulfate.
The ease of sludge oxidation was demonstrated dramatically
early in the investigation, when attempts were made to prepare x-ray
specimens by washing filtered sludge samples with a water-miscible,
low-boiling solvent (i.e., acetone) to effect rapid room temperature
drying. For example, the sludge resulting from a continuous flow LHS
treatment experiment using South Greensburg mine water was allowed to
stand overnight in the settling tanks. On the following day the rela-
tively clear supernatant liquid was decanted from one of the tanks and
a portion of the sludge, still jet black in color, was filtered in a
medium frit glass filter funnel. The damp filter cake was washed
several times with acetone and then pulverized inside the funnel to a
dry, black powder by grinding with a glass rod.
The bulk of the powder was then poured out onto a clean sheet of
paper for subsequent transfer to a plastic vial. Within 10 seconds,
however, a strongly exothermic reaction occurred which caused the paper
to smolder and char. The powder changed color from black to dark brown
during this reaction, and a pungent odor, possibly that of sulfur
dioxide, was detected.
The experiment was repeated later on sludge from the IBS treatment
of a different (Keystone) mine water. A portion of the filtered sludge
washed only once with acetone and ground to a granular powder failed to
produce the expected reaction. However, a second portion washed four
times with acetone and ground to a fine, dry powder underwent the exo-
thermic reaction about two minutes after being placed in contact with
the paper. The powder was Immediately placed in a small beaker, and
the beaker became quite hot to the touch. There was no evidence of
a sudden or violent decomposition reaction, however.
In both cases, small amounts of the acetone-washed filter residues
(which may or may not have undergone the exothermic reaction exhibited
by the bulk of the powder) were recovered and submitted for x-ray dif-
fraction analyses. Only limestone residues (calcite, silica) and
possibly calcium sulfate were identified from the x-ray patterns; no
lines assignable to iron compounds were present. Also, in both cases,
the powders continued to lighten in color from dark brown to a rust-
or chocolate-brown during exposure to air for several days.
The pyrophoric behavior of the IBS sludge was demonstrated again,
somewhat unexpectedly, during later attempts at solids content deter-
minations. A sample of the sulfide sludge, filtered on Whatman No. h2
-------
33.
paper, was placed in an oven for drying at 110 C. Shortly afterward,
smoke was observed coining from the oven. Inspection of the oven con-
tents revealed that the portion of the filter paper which had been
covered by the sludge was completely disintegrated; only a black ash
remained. Thus, there is evidence that the sulfide sludge, wet with
water alone, will undergo the exothermic reaction when dried under
conditions of moderately elevated temperature. This finding tends to
preclude the possibility, considered when the phenomenon was first
observed, that residual acetone may have initiated the exothermic
reactions of sludges dried at room temperature.
One possible explanation for the pyrophoric properties of the
sludge, observed under accelerated drying conditions, involves the
occurrence of finely divided elemental iron (formed by reduction of
ferrous ion in the presence of hydrogen sulfide) and its reaction
with sulfur residues in the sludge. Such a reaction is known to be
strongly exothermic, at least when initiated by heating. To explore
this possibility further, limited additional work was undertaken with
the objective of confirming the presence or absence of metallic iron
in the pyrophoric sulfide sludge.
The overall procedure involved the precipitation and characteri-
zation of sulfide sludges, formed by the addition of hydrogen sulfide
to synthetic and actual mine waters in the presence of an alkali.
Tendencies toward pyrophoric behavior were observed after sludge
samples were filtered, washed with acetone, and dried at room tempera-
ture. Efforts to separate and identify iron in the precipitates in-
volved employment of the following techniques: X-ray diffraction
analyses of sludges and sludge-iron powder admixtures, centrifugation
tests, and tests for magnetic behavior of gravity-settled and centrifuged
solids. In addition to the experimental work, a brief review of
pertinent literature was conducted to determine whether reduction of
iron to the elemental state is thermodynamically possible in aqueous
systems, and to determine if other researchers have encountered similar
pyrophoric behavior with iron sulfide mixtures.
A synthetic ferrous sulfate solution (1000 ppm ferrous ion)
acidified to pH 3.0 with sulfuric acid was employed for most experiments
in this series. All experiments were conducted at room temperature on
2-liter aliquots of the synthetic or natural mine water. The reaction
vessel consisted of an open k-liter beaker equipped with a magnetic
stirring bar, a fritted gas dispersion tube for hydrogen sulfide intro-
duction, and electrodes for pH measurement. In every case, hydrogen
sulfide was added at twice the stoichiometric requirement based on
total iron concentration.
In the initial experiment, 0.5 N sodium hydroxide was used to
maintain the suspension at pH 6 during hydrogen sulfide addition. The
resulting suspension was almost colloidal in nature, however, in that
sedimentation of the bulk of the black solids did not occur after an
overnight settling period. The solids passed easily through a fine
-------
porosity flitted glass filter, and could not be separated by centrifuga-
tion. A small amount of black, ferromagnetic material was recovered
during centrifugation tests with the aged (24 hours) sludge. This
material was washed successively with deionized water and acetone by
repeated resuspension, recentrifugation, and decantation of the super-
natant liquid. X-ray diffraction analysis of a portion of the dried
solid revealed the compounds Fe^ (greigite), tetragonal FeS
(mackinawite), and sulfur. Both the dried solid and a portion of the
centrifugate retained under acetone were still magnetically active two
weeks after preparation.
This initial experiment was followed by three subsequent experiments
involving the use of limestone (BCE No. 1809, minus 325 mesh) at twice
the stoichiometric requirement based on mine water acidity. The result-
ing sludges settled on standing and could be filtered quantitatively on
a medium porosity fritted glass filter.
In the first of the three additional experiments with the synthetic
mine water- limestone-hydrogen sulfide system, the sludge was allowed to
settle overnight and the settled black precipitate was weakly magnetic.
This precipitate, after being filtered, washed with acetone, and dried
at room temperature, was identified by x-ray diffraction analysis as
tetragonal FeS (the weak magnetic properties were probably due to the
presence of a small amount of Fe^ which was below the detectability
limits of the instrument). In addition, a portion of the aged sludge
was centrifuged, washed with acetone, and recentrifuged. This solid was
also identified as tetragonal FeS.
The bulk of the black solid was filtered, washed with acetone, and
ground to a fine powder. After being deposited in a pile on a piece of
paper, the material underwent a strongly exothermic reaction within
seconds. During this reaction the solid changed color from black to
reddish-brown, became strongly ferromagnetic, and sulfur dioxide (identi-
fied by odor) was evolved. The residue was found by x-ray analysis to be
a mixture of Fe,j04 (or, possibly, v-FegQg), cr-FCgOg, and possibly
Finally, a residue of the solid remaining on the glass filter was
observed to become warm on exposure to air. The solid was ground to a
Jet-black powder which exhibited strongly magnetic properties. X-ray
analysis of this powder revealed lines due to Fe^ as well as those of
tetragonal FeS.
In the second experiment with the synthetic mine water- limestone-
hydrogen sulfide system, a portion of the sludge was filtered, washed
with acetone, and subjected to x-ray analysis one-half hour after prep-
aration. Lines due to tetragonal FeS and CaCCfe (calcite) were identified,
and an extraneous line at d = 2.08 A was assigned to hexagonal FeS
(troilite). The relatively fresh, acetone-washed sludge was nonmagnetic
and failed to produce the exothermic reaction, although a second portion,
which was also filtered, washed, and ground to a powder about 2 hours
-------
35.
after sludge preparation, underwent the reaction readily with the
formation of the rust-brown, strongly magnetic residue and the
evolution of sulfur dioxide.
The bulk of the nonmagnetic black solid was ground to a coarse,
granular powder and stored in a plastic vial for two days. During
this storage period it became ferromagnetic but remained black in
color. This material was used in the preparation of admixtures with
iron powder containing 5 percent and 1 percent iron by weight,
respectively. X-ray analyses of the admixtures showed FegS4, tetrag-
onal FeS, and calcite; lines due to elemental iron were clearly
evident in the 5 percent iron sample, but were not discernible for the
1 percent iron mixture.
In the third experiment with the synthetic mine water-limestone-
hydrogen sulfide system, a portion of the sludge was filtered one-
half hour after preparation, washed with deionized water, and extracted
twice with about 20-ml portions of carbon disulfide. The filter cake
was stored under water for about 1 hour, then the solids were dewatered
again and washed with acetone. The acetone washings were collected
separately and the solvent was evaporated at room temperature. The
remaining portion of the sludge was stored under water for 21* hours
(it remained black and nonmagnetic), then it was filtered and the
solids extracted with carbon disulfide as before. During these
operations, a brassy-yellow scum was present which tended to creep
up and coat the filter funnel walls. This yellow material, as well
as the solids obtained after evaporation of the acetone extract and the
two (fresh and aged sludge) carbon disulfide extracts, were all identi-
fied as sulfur by x-ray analyses. The actual amounts of sulfur obtained
in each extraction were quite small, probably only a few milligrams.
Furthermore, the amount of sulfur obtained by the carbon disulfide
extraction of the aged sample did not seem to be significantly greater
than that obtained in the same manner from the freshly prepared sample.
The solid residue remaining from the carbon disulfide extraction
of the aged sample was ground with a glass rod to a coarse, black, non-
magnetic powder. It was transferred quickly to a plastic vial, but the
vial began to get very warm indicating that the exothermic reaction had
been initiated. The vial was capped quickly and the contents were
shaken to disperse them over the inner walls of the container. This
technique seemed to quench the reaction, since the powder remained
black, although it had become strongly ferromagnetic. X-ray analysis
showed that the powder was essentially Fe3S4.
For the studies utilizing natural coal mine water, freshly obtained
field samples of the South Greensburg discharge were employed. As
before with the synthetic mine water, limestone and hydrogen sulfide
were used in the sludge preparation reaction at twice the stoichiometric
requirement based on acidity and total iron content, respectively.
-------
36.
Freshly prepared acetone-washed sludges underwent the strongly
exothermic reaction within seconds after being ground to a fine powder.
X-ray analysis of the light brown, ferromagnetic residue from the exo-
thermic reaction showed the presence of Fe304 (or, possibly, v-FegOs)
as well as a number of lines assignable to limestone residues (calcite
and silica). With the exception of the residue from the exothermic
reaction, however, ferromagnetic properties were completely absent in
sludges from the natural mine water system, whether fresh or aged, and
regardless of whether the solids had been dried or retained in aqueous
suspension. Furthermore, it was not possible to recover a magnetic
fraction by centrifugation of the sludge samples.
In contrast with the results of studies utilizing synthetic mine
water systems, no crystalline iron sulfide species (Fe^, tetragonal
FeS) were identified in x-ray patterns of sludges prepared from the
natural mine water system. This was the case in spite of the fact that
considerable care was taken to minimize air exposure time by loading the
x-ray specimen capillary tube and initiating x-ray .diffraction analysis
Immediately after filtering, acetone washing, drying, and grinding the
sludge to a black powder. Moreover, the sulfide sludges formed by LHS
treatment of South Greensburg mine water were converted by air exposure
to nonmagnetic, x-ray amorphous tan solids within a few hours when dry,
and somewhat more slowly when covered by water or acetone.
Sludge Separation Studies
As the work progressed, it became apparent that the chemical com-
position and physical behavior of the sulfide sludge were neither well
defined nor consistent in nature. For this reason, the originally
scheduled solid-fluid separation tests were considered premature until
more was known concerning the nature of the sludge itself. Instead,
work under the above subheading was limited to a brief study of sludge
settling behavior, with and without the use of coagulant aids, supple-
mented by zeta potential measurements.
Experimental Procedure and Results
Electrophoretic mobility measurements were made with the aid of
a Zeta-Meter, Zeta-Meter Inc., New York, N. Y. Flocculating aids were
evaluated in beaker-scale batch tests by their addition at various
concentrations to 100 ml samples of the sulfide sludge suspensions,
selected at random from both continuous flow and batch LHS treatment
tests.
Electrophoretic mobility measurements, conducted on the fresh
sulfide sludge obtained during continuous flow test No. 372-82, re-
vealed that the corresponding zeta potential value was -12 mv. Thus,
unlike sludges from direct lime or limestone neutralization, which
are characteristically positive in charge, the sludge particles from
-------
37.
the LHS treatment process are negatively charged. This result in-
dicated that cationic coagulant aids could be employed effectively
to enhance sludge settling and separation.
This expectation was confirmed in later experiments using the
freshly formed sludge obtained during the continuous flow LHS treatment
test No. 372-98. One cationic flocculant (Primafloc C-7, Rohm and
Haas Co.) and two anionic flocculants (Poly-Floe 1130, Betz Laboratories,
Inc. and Calgon C-55, Calgon Corp.) were added at various concentra-
tions to 100 ml samples of the sulfide sludge suspensions, and zeta
potentials were determined at each concentration level. Subsequent
to this work, two additional cationic coagulant aids (Calgon 225,
Calgon Corp. and Arquad 2C-75, Armour Industrial Chemical Co.) were
evaluated relative to their effect on the zeta potential of a freshly
formed sulfide sludge, obtained during a 5-gallon batch LHS treatment
test. The curves relating zeta potential to concentration of additive
are shown in Figure 8.
Further tests were conducted on 2000 ml samples of sulfide sludge
suspensions, prepared by LHS batch treatment of South Greensburg mine
water, to compare the relative efficacy of the three cationic coagulant
aids. The coagulant aids were added as 1000 ppm stock solutions to the
sulfide sludge suspensions, contained in 2-liter graduated cylinders,
in dosages necessary to achieve the following concentrations (required
to attain the iso-electric point based on the data in Figure 8): Prima-
floc C-7, 1.5 ppm; Arquad 2C-75, 2 ppm; CaLgon 225, 3 ppm. A fourth
sample, to which no coagulant aid was added, was included as a control.
The samples were stirred gently for 30 seconds during addition of the
coagulants, then allowed to stand undisturbed.
After 90 minutes, there was no appreciable difference in appear-
ance among the four samples; all remained opaque. When viewed with a
strong light behind the cylinders, however, there was evidence of floe
formation in each of the treated samples.
At this point in time, addition of the flocculating aids was re-
peated using the same dosages as before. This time, floe formation
was quite evident when observed by transmitted light. The floe size
and settling rate appeared to be about the same in each of the three
treated samples, with perhaps slightly better clarity in the sample
treated with Arquad 2C-75. After about 1/2 hour settling time, the
three treated samples were of definitely better clarity than the
control, although a certain amount of very fine particulate material
tended to remain in suspension in all samples for several hours.
Ferric Iron - Hydrogen Sulfide Reactions
A brief experiment was conducted early in the present investiga-
tion in which hydrogen sulfide was bubbled through a "yellowboy" sludge
suspension prepared by limestone neutralization and aeration of Thorn
Run mine water. The suspended solids readily changed color from orange-
brown to black, although there was no change in settled sludge volume.
-------
15-1
+ 10-
+ 5-
0-
2
-10-
•15-
Arquad 2C-75 (Cationic)
Primafloc C-7 (Cationic)
/ Calgon 225 (Cationic)
Calgon C-55 (Anionic)
Poly-Floe 1130 (Anionic)
1 1 1 F
COAGULANT AID, PARTS PER MILLION
10
Bituminous Coal R«»t«rch, Inc. 2028022
Figure 8. Effect of Various Coagulant Aids on Zeta Potential
of Sulfide Sludge
-------
39.
On the basis of these observations, the following reactions were
assumed
2 Fe3* + EgS = 2 Fe2+ + S° + 2 if
2 Fe8* + 2 HS = 2 FeS + k if
and hydrogen sulfide feed requirements were adjusted accordingly during
LHS treatment experiments when ferric iron was known to be present.
As in other cases, however, this aspect involving the basic
chemistry of the process was not explored further until considerably
later in the overall research program.
Experimental Procedure and Results
The procedure used for experiments in this series was similar to
that employed for the 5-gallon batch IBS treatment experiments, except
that no limestone was added during the reaction. In all cases, sludge
suspensions were prepared by direct limestone neutralization (twice the
stoichiometric requirement based on acidity) and prolonged aeration of
South Greensburg mine water.
A preliminary test involved the addition of hydrogen sulfide to 5
gallons of a "yellowboy" sludge suspension prepared as indicated above.
During hydrogen sulfide addition the suspension changed color from orange
to black, and the pH decreased slightly. Analytical results on samples
taken subsequent to hydrogen sulfide addition, with continued stirring,
revealed the following:
Dissolved ferrous iron content increased slightly during the
first 60 minutes, then decreased gradually thereafter.
The pH of the suspension increased during the first two hours
to a value greater than that of the original suspension (before
hydrogen sulfide addition). Thereafter, the pH decreased to near its
original value.
Insoluble material was present in the unfiltered samples taken
for analysis after acidification with hydrochloric acid. The amount
of this insoluble material increased with time after hydrogen sul-
fide addition. The material was identified by x-ray analysis as
predominantly elemental sulfur, with some silica present (presumably
from unreacted limestone).
The above observations were substantially confirmed in a later 5-
gallon batch experiment. To a hydrous ferric oxide suspension, prepared
by limestone neutralization and aeration of South Greensburg mine water,
hydrogen sulfide was added at a calculated rate sufficient to reduce all
of the ferric iron and precipitate it as ferrous sulfide according to
the above equations.
-------
Samples were taken from the reactor before hydrogen sulfide addition
and, thereafter, over a five-hour period for dissolved ferrous iron (in
filtrate), total ferrous iron, and total iron (unfiltered samples), as
well as pH measurement. The results of this experiment are shown in
Figure 9, and may be summarized as follows:
Total iron concentration in the suspension remained essentially
unchanged throughout the experiment, as expected.
Total ferrous iron concentration increased rapidly during the
addition of hydrogen sulfide. After hydrogen sulfide flow ceased,
total ferrous iron concentration in the suspension gradually de-
creased with time. This decrease, which became more rapid after
about two hours as indicated by the abrupt change in slope of the
line, was accompanied by the formation of increasing amounts of
elemental sulfur (confirmed by x-ray analysis of acid-insoluble
residues).
The pH of the suspension increased gradually to a maximum value
after hydrogen sulfide addition, then decreased slightly to a reason-
ably constant level about 3 hours after hydrogen sulfide flow was
ceased.
It should be pointed out that the results shown by Figure 9 were
obtained for the case where sufficient hydrogen sulfide was added to
both reduce ferric iron and precipitate the resulting ferrous iron as
the sulfide. It was anticipated that somewhat different results might
be obtained for the case wherein a deficient amount of hydrogen sulfide
was added. To explore this possibility further, an experiment was con-
ducted in which hydrogen sulfide was added to a "yellowboy" suspension,
prepared by limestone neutralization and aeration of South Greensburg
mine water, in the five-gallon reactor. The amount of hydrogen sulfide
added was only enough to reduce the ferric iron.
The results of this experiment were qualitatively similar to those
shown in Figure 9 in terms of changes in dissolved ferrous iron, total
ferrous iron, total iron, and pH with time, as well as regarding the
formation of sulfur during the course of the reaction. In addition,
however, the following phenomena were observed:
The suspension, which was black immediately following hydrogen
sulfide addition, gradually returned to its original orange-brown
color after kO minutes with continuous stirring in the enclosed
reactor.
Immediately after hydrogen sulfide addition a portion of the
suspension was filtered on 0.22^ Millipore paper. The filter cake
appeared to consist of three layers:
-------
a.
7.0-i
6.8-
6.6-
6.4-
6.2-
6.0-
I
2
U
O
u
O
at
130-,
110-
90-
70-
5 Minute HjS
Flow Period
120 160 200 240
TIME AFTER HSS ADDITION, minutes
280
320
Bituminous Co*I Research, Inc. 2028G23
Figure 9. Changes in Iron Concentration and pH with Time During
Stirring of a "Yellowboy" Sludge Suspension Treated with HZS
H
-------
Black solids nearest to the filter disk; a middle layer of
blue-green material which turned a rust-brown color on exposure
to air; and a rust-brown surface coating. A portion of the filter
cake was analyzed immediately for iron and sulfide sulfur by the
usual method. However, essentially no measurable cadmium sulfide
was formed in the bubbler tube, although both ferrous and ferric
iron were present in the acidified residue in a molar ratio of
about 1:1 (about 60 ppm each after dilution of the acidified
solution to 250 ml).
Sludges obtained by the hydrogen sulfide reduction of hydrous
ferric oxide (i.e., "yellowboy") were also analyzed by x-ray diffrac-
tion techniques after vacuum drying. The only substances identified
by x-ray analyses were sulfur and limestone residues (calcite and
silica). No extraneous peaks corresponding to those of known iron
sulfides or oxides were present in the x-ray patterns.
Of further interest is the observation, made during these x-ray
studies, that sludges obtained by hydrogen sulfide reduction of
"yellowboy" sludge tended to undergo oxidation much more rapidly than
those obtained by LHS treatment of filtered mine water. For example,
two samples were prepared on different occasions by the foimer method
and vacuum-dried for 2 hours at room temperature. Each developed a
definite olive-green surface coating within minutes after being re-
moved from the vacuum oven. One of the samples continued to lighten
in color to an ocher shade during grinding to a fine powder, while
the second remained essentially black after grinding. By contrast,
vacuum-dried sludges resulting from LHS treatment of filtered South
Greensburg mine water remained black for several hours after grinding.
-------
Section, k
DISCUSSION
Continuous Flow and 5-Gallon Batch Experiments
In addition to the low iron removals experienced in the earlier
continuous flow treatment experiments, certain trends regarding the
concentrations of other species present were found. These trends, in-
dicated by the data in Table 2, were observed consistently throughout
the investigation, and are as follows:
The aluminum concentration almost invariably decreased to less
than 3 ppm (i.e., below the detectable limit of the analytical
method) during treatment. This is undoubtedly due to the precipita-
tion of hydrous aluminum oxide during neutralization by limestone
(aluminum sulfide hydrolyzes immediately on contact with water, and
can therefore not exist in an aqueous environment).
As expected, the calcium content increased during the solution
of limestone in the system.
The concentrations of magnesium, manganese, silicon, sodium,
and sulfate remained essentially unchanged throughout the treatment
process.
The data in Table 2 also reveal some rather significant discrepan-
cies between ferrous and total iron concentration in some cases. Al-
though this problem was overcome later in the investigation through the
use of common calibration standards, the uncertainties involved in
hydrogen sulfide feed calculations based on these analytical data un-
doubtedly affected the overall results of the earlier treatment
experiments.
The data from subsequent 5-gallon batch treatment experiments, in-
dicated in Figures 5 and 6, yielded the somewhat unexpected result that
when the amount of hydrogen sulfide employed was no more than the exact
stoichiometric requirement based on total iron, there was a tendency
for the ferrous iron concentration to decrease to some minimum value
and then to increase thereafter with time.
In terms of overall iron removals achieved, Figure 5 shows that
slight improvements were effected when the amounts of limestone em-
ployed were in excess of twice the stoichiometric requirement.
The results of later continuous flow tests, shown by Table 3 and
Figure 7, agree substantially with those obtained during the batch
tests and may be summarized as follows:
-------
With amounts of hydrogen sulfide only slightly less than the
exact stoichiometric requirement, the desired degree of iron removal
from South Greensburg mine water could not be achieved under the ex-
perimental conditions employed. Moreover, based on ferrous iron
analyses of treated effluent samples after overnight settling, there
was confirmatory evidence that the ferrous iron concentration in the
process effluent tended to increase with time after attaining some
minimum value when the amount of hydrogen sulfide during treatment
was only slightly less than the exact stoichiometric requirement.
With amounts of hydrogen sulfide slightly greater than the exact
stoichiometric requirement, the desired degree of iron removal
from South Greensburg mine water apparently could be achieved under
the experimental conditions employed. Furthermore, after decreasing
to some minimum value, the ferrous iron concentration did not appear
to change appreciably with time when the amount of hydrogen sulfide
during treatment was greater than the exact stoichiometric
requirement.
The results shown by Figure 7 emphasize the need for accurate de-
termination and control of hydrogen sulfide feed in the LHS treatment
process so as to achieve and maintain the desired degree of iron re-
moval from those mine waters amenable to treatment. The attainment of
such accurate hydrogen sulfide feed control, however, would seem dif-
ficult for the following reasons:
Control of hydrogen sulfide addition by measurement of rate of
gas flow into a strongly agitated suspension is an operation which
cannot be carried out with a high degree of confidence, at least
with the apparatus used for these studies.
There is strong evidence that the stoichiometric hydrogen sul-
fide requirement varies depending not only on changes in iron con-
centration but also on changes in the oxidation state of the iron in
the mine water. Thus, a practical LHS treatment process would be
likely to necessitate fairly rapid determinations of changes in iron
concentration and oxidation state occurring in the process influent,
if these changes did in fact occur. To our knowledge, such determi-
nations cannot be made rapidly enough to provide the necessary proc-
ess control with the instrumentation or techniques currently
available.
It is reasonable to expect that a small amount of the hydrogen
sulfide added will become unavailable for reaction with the iron as a
result of losses such as those due to escape of the gas into the va-
por phase, involvement in salvation (hydration) equilibria, and reac-
tion with other reducible species such as dissolved oxygen. The
extent of these losses will depend on variables such as mine water
composition and temperature, and would therefore be considered diffi-
cult to predetermine or control in a practical mine water treatment
process.
-------
Evidently, the only alternative to accurate determination and con-
trol of hydrogen sulfide feed would be the use of an excess of hydrogen
sulfide in the process, with the concomitant disadvantages of increased
cost and probable undesirable effects on effluent quality.
Sludge Characterization Studies
Chemical and Physical Properties
The results of chemical analyses shown by Table k reveal that,
with only two exceptions, sulfide sulfur:iron mole ratios were within
10 percent of the value 1.00, and that the experimentally determined
values of this ratio tend to be high. Correlations with the amounts of
ferrous sulfide used in control tests indicated a positive error in the
sulfide sulfur determination. The good agreement among sulfide sulfur:
iron ratios for the sulfide sludges and ferrous sulfide control samples
provides strong evidence that the iron sulfide species, present in the
freshly formed sulfide sludge, does in fact correspond to the empirical
formula FeS.
It is significant that no peaks corresponding to any known iron
sulfide compound were identified during x-ray diffraction analyses of
freshly formed, vacuum-dried sulfide sludge samples. This result indi-
cates that the iron sulfide component of the sludge is either of ex-
tremely small particle size or that it is a structurally disordered,
amorphous material. The ease of oxidation and, under certain condi-
tions, pyrophoric behavior of the sludge (discussed below) tend to sup-
port the latter possibility.
The absence of a recognizable iron sulfide pattern in the x-ray
diffractograms of the sulfide sludges takes on added significance when
one realizes that appreciable amounts of iron and sulfide sulfur were
present in these samples. For example, material balance calculations
based on the sulfide sulfur and iron analyses for the sludge from Test
No. 385-48 revealed the following percentages by weight of the dried
sample: Iron, 1*2.6 percent; sulfide sulfur, 2U.2 percent. In addi-
tion, the sample contained 9.9 weight percent of a hydrochloric acid-
insoluble residue, identified as elemental sulfur by x-ray analysis.
If one assumes that this sulfur was formed as a result of oxidation of
ferrous sulfide before or during sludge chemical analysis, then the
amount found corresponds to an additional 1?.2 weight percent iron and
9.9 weight percent sulfide sulfur, leaving only 6.1 weight percent
(corresponding to 2.5 mg of the original sample taken for analysis) un-
accountable in the material balance. Interestingly, sulfur (as well as
silica) was also found in a hydrochloric acid-insoluble residue during
Test No. 385-30. In both of these cases, since the mine water samples
were filtered to remove hydrous ferric oxide immediately prior to
treatment, the sulfur present in the sludge was presumably formed from
either oxidation of hydrogen sulfide by oxygen (air) or other species
during sludge preparation, oxidation of the iron sulfide species
-------
present in the sludge daring its decomposition in 6N hydrochloric acid,
or oxidation of the iron sulfide species in the sludge during its ex-
posure to air prior to chemical analysis. Available experimental evi-
dence tends to favor the last of these three possibilities.
With regard to sludge volumes and solids contents, the data in
Table 5 show that, based on averaged values from the two experiments,
the relative settled sludge volume percentages are roughly in the ratio
1.0:2.3:8.5 for the LHS, limestone, and lime sludges, respectively; the
corresponding weight percent solids values are in the ratio 9.4:3.3:1.0,
respectively. These results indicate that the LHS sludge is appreci-
ably denser than sludges formed from either lime or limestone neutrali-
zation alone. However, it should be realized that the sludges from
limestone and LHS treatment presumably contained excess unreacted lime-
stone, the latter probably containing a larger proportion thereof.
This fact should be considered in any direct comparison of sludge
densities.
A few other differences in sludge properties were observed during
these tests. For example, the sludge from lime treatment was orange in
color, and gelatinous, but settled fairly rapidly leaving a clear super-
natant liquid. The sludge from limestone treatment was yellowish and
more granular, but the solids tended to cling to the wall of the Imhoff
cone during settling, giving the appearance of a yellowish turbidity to
the supernatant liquid. The sludge from LHS treatment was black, and
the bulk of the solids tended to settle fairly readily. A small amount
of finely divided solid material tended to remain in suspension, how-
ever, making the supernatant liquid completely opaque for several hours.
(No flocculating aids were employed in any of these tests.) These sus-
pended solids also tended to deposit on the glass wall of the cone dur-
ing settling, and changed to a rust-orange color during the overnight
settling period. The resulting effect was the appearance of a grayish-
orange turbidity in the supernatant liquid.
Tendency Toward Sludge Oxidation
The absence of detectable amounts of sulfate ion in the acidified
washings of oxidized LHS sludges indicates that the initially formed
iron sulfide does not oxidize to a sulfate on contact with air. Bather,
free sulfur appears to be the only oxidation product of this element
during exposure of the filtered sludge to air. This conclusion is
supported by the occurrence of sulfur in freshly precipitated sludges,
produced by LHS treatment of South Greensburg mine water (with an initial
pH near 5) which was filtered immediately prior to treatment (see above
subsection, Chemical and Physical Properties).
-------
With the knowledge that the sludge undergoes gradual oxidation
under mild conditions (i.e., exposure of the damp material to air at
room temperature), one may reasonably assume that the pyrophoric be-
havior described earlier is a result of sudden and rapid oxidation under
conditions of forced dehydration. This reaction may also be accelerated
by the existence of sulfur, dispersed in very small particles throughout
the sludge, and formed either before or as a result of the reaction.
The results of the additional experimental studies on the nature of
sludges, formed by IBS treatment of synthetic mine water, are remarkably
similar to those reported previously by Berner (2) with regard to the
iron sulfide forms observed. Berner found that tetragonal ferrous sul-
fide (corresponding to the mineral mackinawite) was formed over a wide
range of pH and temperature during the reaction of hydrogen sulfide with
a wide variety of iron-bearing reactants. In particular, the reaction
of hydrogen sulfide with acidic (pH 3) ferrous sulfate solutions at
room temperature gave either an amorphous ferrous sulfide or mixtures of
the tetragonal ferrous sulfide, cubic iron sulfide (Fe3S4), and sulfur,
depending on the availablility of air to the suspension and the duration
of precipitate aging. (Berner did not report the iron concentrations
of ferrous sulfate solutions employed, nor was there any mention of
pyrophoric tendencies on the part of the reaction products.) The re-
sults indicated that the magnetic phase, Feg S4, was formed primarily
during air oxidation of suspensions containing amorphous or tetragonal
FeS initially. The aging of the initially amorphous ferrous sulfide
precipitate to crystalline phases was enhanced by increasing tempera-
ture and decreasing pH.
X-ray data from the present study are in accord with Berner's
findings, and support the following conclusions:
The room temperature reaction of hydrogen sulfide with pure
ferrous sulfate solutions in the pH range k to 6 results initially in
a precipitate of tetragonal FeS (the precipitate may be amorphous
originally, although crystalline forms were found to be present 30
minutes after the reaction). The tetragonal FeS converts to the cubic
Fe3S4 on air oxidation, rather slowly in aqueous suspension, and quite
rapidly when the dried solid is pulverized to a powder. This rapid
oxidation reaction is strongly exothermic, and if permitted to go to
completion, the ultimate iron-bearing reaction product is apparently
hematite, a-Fea03. The overall conversion of the sulfide to the oxide
probably involves the formation of the cubic magnetic iron oxides Fe304
(black) and Y-Feg03 (brown) as intermediate products, evidenced by
their appearance in x-ray patterns of products from the exothermic
reaction and the color and magnetic properties of these products.
The overall process may be summarized by the following reaction
sequence:
aging Os
-f Fe8*—+• amorph. FeS —*• tetrag. FeS —-*Fe3S4 + Fe304(?)
-------
°s
_» Fes04 + S .^v-FegOg + SOS
The above reactions are not meant to Imply that the oxidation proceeds
by discrete steps. To the contrary, the experimental observations
suggest that the sulfur- oxygen anion substitutions and crystal Lattice
rearrangements indicated are probably proceeding simultaneously, at
least when the reaction reaches a high velocity (i.e., becomes exo-
thermic). For example, the presence of sulfur in relatively fresh
sludges, confirmed during carbon disulfide and acetone extraction ex-
periments, indicates that conversion of the iron sulfide to an oxide
occurs at an appreciable rate very soon after precipitation of the
sulfide.
The scheme suggested above is entirely analogous to that in-
volving interconversion of iron oxide crystalline forms. Some of
these structural transformations are considered to be classical
examples of a phenomenon known as topotaxy. In the particular case
of the iron oxides, topotactic reactions are usually manifested by
the preservation of an oxygen atom framework within the crystal
lattice; changes take place by the regrouping of cations within the
lattice, but the number of oxide ions per unit volume is not very
much changed. These reactions, some of which are shown below, have
been reviewed by Mackay and others, (l, U, 11)
Fe(OH)s — * FeO — * Fe304 — » Y-Fea03 .-*. a-FeaOs
(amorphous (cubic) (cubic) (cubic) (hexagonal)
to hexagonal)
The lattice systems, shown in parentheses, indicate the arrangement of
the oxygen atom framework. Each transformation shown is a topotactic
reaction except for the Y-FegQ3 to a-FegCXj transformation, which is better
regarded as epitaxial. (11) These reactions are included merely to show
the possibility of comparison with those involving the sulfide species,
presented earlier.
Redox reactions pertinent to the present discussion are listed
below, with the corresponding standard oxidation potentials in acidic
solution as given by Moeller: (13)
IP, Volts
Fe = Fe3* + 2 e~ +0.^0
Fe = Fe3* + 3 er +0.036
= &+€- 0.000
- S + 2 if +2e" -O.lUl
-------
Fe3* = Fe3* + e~ -0.771
1/2 Ofl -»- 2 H" -i- 2 e- -1.229
This listing reveals that among the species considered, oxygen
is the strongest oxidizing agent and metallic iron is the strongest
reducing agent. Thus, theoretically, metallic iron will be oxidized
to Fe3* in aqueous media by If , & , Fe3* , or Oa . More importantly,
these data show that hydrogen ion should be reduced to hydrogen gas
by elemental iron (this reaction vill in fact occur, but only at
elevated temperatures is the rate appreciable in slightly acidic
solution) . Expressed in a different way, redox potential data shew that
any reducing agent strong enough to reduce ferrous ion to elemental iron
will first reduce water to hydrogen. Consequently, theoretical consider-
ations show that elemental iron cannot be formed from its oxidation
products in an aqueous system. This conclusion can be derived in a
similar manner by a consideration of standard free energies of reaction
for various combinations of the half -react ions shown above.
The foregoing considerations are of course qualified by the
assumption of the usual conventions regarding standard states, such
as unit activities for ionic species, and are therefore essentially
approximate in character.
The tendency for freshly prepared ferrous sulfide to undergo
spontaneous self -decomposition was apparently confirmed by very early
workers in the field. Thus, Mellor reports that as early as 1826,
J. J. Berzelius observed that "moist ferrous sulfide readily oxidizes
in air with the separation of sulfur, and the heat evolved may be
great enough for incandescence." (12) Later workers observed a similar
reaction when samples were rubbed between the fingers or ground in a
mortar. In the 100- year period following Berzelius' original discovery,
at least 17 other workers studied the phenomenon under various condi-
tions. (12) The magnetic Fe304 was mentioned more recently as one of
the reaction products. (8)
Some German workers have reported that "highly pyrophoric mixtures"
of iron mercaptides, iron oxide hydrates, iron sulfides, and sulfur were
formed by the reaction of mercaptans on alcoholic solutions of ferric
chloride. (15) Other workers have reported the preparation of an ammonia-
containing pyrophoric ferric sulfide by the addition of bydrosulfide ions
to ammoniacal ferric iron solutions. (3)
Also of interest is a recent disclosure (6) that Freke and Tate,
during their studies on the formation of magnetic iron sulfides by
bacterial reduction of iron sulfate solutions (5), occasionally ob-
served rapid, exothermic reactions of dried sulfide sludges. These
workers reported that the magnetic sulfide, dried "under anaerobic
conditions," appeared to be perfectly stable and retained its magnetic
-------
50.
properties even after repeated wetting and drying. Air drying or re-
tention of the settled sludge in water produced apparent surface oxida-
tion with an overall decrease in magnetic properties, while continuous
agitation of the material in suspension resulted in complete oxidation
and loss of magnetic properties.
In summary, the results of this investigation have failed to indi-
cate the presence of metallic iron in sludges prepared by the reaction
of hydrogen sulfide with natural or synthetic mine water in the presence
of an alkali (limestone or sodium hydroxide). The results of x-ray dif-
fraction analyses showed that any iron, if present, must be below the
detectability limit of the instrument for this element, namely, about 5
percent by weight. These findings are consistent with theoretical con-
siderations regarding the thermodynamic stability of elemental iron in
slightly acidic, aqueous media. Moreover, the apparent tendency of
freshly prepared ferrous sulfide to undergo autoxidation seems to be
adequately documented in the chemical literature. All things considered,
therefore, it would seem reasonable to conclude that metallic iron is
not involved in the exothermic reaction.
Magnetic sludge fractions, when obtained by centrifugation or
filtration of sludges from LHS treatment of synthetic mine water, were
invariably found to contain F^S^ the magnetic sulfur analog of
magnetite.
Besults with acidified ferrous sulfate solutions indicated that
the predominant iron sulfide species in the freshly formed precipitate
is tetragonal PeS. This is in accord with findings concerning sulfur:
iron mole ratios in sulfide sludges from natural mine water. The
tendency of ferrous sulfide to convert to Fe3S4 during aging may explain
why sulfur:iron mole ratios found by experiment were occasionally greater
than one (see Table !+).
Still puzzling is the fact that at least two distinct iron sulfide
crystalline forms were obtained from limestone-hydrogen sulfide treat-
ment of an acidified ferrous sulfate solution, while treatment of a
natural (South Greensburg) mine water, under identical conditions, con-
sistently yielded an x-ray amorphous iron sulfide. It is possible that
certain foreign impurities, formed by coprecipitation during treatment
of natural mine water (e.g., aluminum hydroxide), interfere in orderly
crystal nucleation and growth processes. Until further experimental
efforts are brought to bear on this question, however, it must remain
open to speculation.
Sludge Separation Studies
One objective in flocculating aid addition is to decrease the ab-
solute value of the electrostatic charge to or near zero (the iso-
electric point), thereby minimizing the forces of mutual repulsion
between the particles. Figure 8 shows that the only flocculants
-------
51.
effective in this regard were the cationic agents. Qualitative ob-
servations of sludge coagulation and settling behavior did in fact
confirm this finding.
The best results, in terms of rapid coagulation and settling rates,
were observed during electrophoretic mobility measurements at the 2 ppm
level with the cationic Primafloc C-7. Settling behavior at the higher
Primafloc C-7 concentrations was still considerably improved over that
observed in the absence of any flocculating aid. Sludge settling be-
havior was Judged qualitatively to be fair-to-good near the iso-electric
point with Calgon 225 and Arquad 2C-75. Both of these coagulant aids,
however, were judged to be inferior to Primafloc C-7 in their ability to
promote more rapid settling of the sludge. $y contrast, coagulation and
settling behavior was judged poor for those samples treated with the
anionic flocculating aids, which is in agreement with the fact that the
iso-electric point was never approached closely when the anionic floc-
culating aids were employed.
As has been noted earlier, observations made during sedimentation
tests revealed that the bulk, of the black solid (presumably including
excess unreacted limestone) tended to settle readily, leaving a small
amount of very fine material suspended in the supernatant liquid. The
amount of this residual suspended material was sufficient to render the
sample completely opaque for periods of several hours. This behavior
was observed consistently with both flocculated and unflocculated sludge
suspensions, and suggests the possibility of a two-stage flocculant
addition process in which the almost colloidal suspension, resulting
after initial settling, might be treated in a different manner or with
different reagents than the bulk of the material.
Ferric Iron - Hydrogen Sulfide Reactions
From a consideration of the following standard oxidation potential
data (13)
HaS(aq) = ^ + 2 H1" + 2er E° = -O.lUl volt
Fea+ = Fea+ + e" E° = -0.771 volt
one would predict that hydrogen sulfide would reduce ferric iron in
aqueous solution. It is also known (7) that the precipitation of ferric
sulfide free from sulfur is not possible since hydrogen sulfide, hydro-
sulfide ion, and sulfide ion all reduce ferric iron.
It is not surprising, therefore, that the results of this study
indicate fairly rapid reduction of dissolved or suspended ferric iron
present in mine water during the LHS treatment. The reduction and sub-
sequent reprecipitation of ferric iron presumably proceed according to
the equations:
-------
52.
2 Fe3* + HeS = 2 Fe3* + S° + 2 H* (l)
2 Fe3* + 2 BgS = 2 FeS + 1* H* (2)
In this regard, Kumai has studied reactions between hydrous ferric oxide
and hydrogen sulfide by extracting the resulting sulfur with carbon
tetrachloride and determining its amount. (9) He claims that above pH
10, a stable ferric sulfide is formed according to the reaction
2 Fe3* + 3 %S = FeaSg + 6 H* (3)
ferric sulfide is reportedly unstable at lower pH, however, and de-
composes into ferrous sulfide and sulfur (almost completely at pH 3),
presumably according to the reaction
FeaSa = 2 FeS + S° (k)
It is apparent that whether the actual reduction mechanism involves re-
actions 1 and 2 or reactions 3 and k, the overall reaction stoichio-
metry is identical in both cases.
As shown by Figure 9> the total ferrous iron concentration in the
suspension decreased with time after hydrogen sulfide addition (with a
concomitant increase in the amount of sulfur present). These phenom-
ena probably occurred as a result of air oxidation of the sulfide
sludge, possibly according to the reaction
2 FeS + 3/2 GS + 3 HjjO = 2 Fe(QH)a + 2 S° (5)
(Although the suspension was maintained in the five -gallon enclosed re-
actor during the run, there was probably enough air in the space at the
top of the reactor to permit gradual air oxidation to occur.)
The actual processes occurring during sulfide sludge oxidation are
probably more complicated than indicated by equation 5> For example,
since the oxidation may involve ionic species in solution, it is con-
ceivable that the ferrous sulfide precipitate undergoes some
dissociation:
FeS = Fe3* + S3" (6)
Above pH 6, hydroxyl ions may compete for dissolved ferrous iron,
leading to the following reactions:
Fea* + 2 OH" m Fe(OH)a (?)
2 Fe3* +1/2 Oa + 5 HB<) » 2 Fe(OH)3 + k H* (8)
Sulfide ion resulting from reaction 6, may undergo hydrolysis as well
as oxidation.
-------
53.
S3" + HaO = HS~ + OH" (9)
Sa" + 1/2 Oa + HgO = S° -»• 2 OH" (10)
In addition, in the presence of free sulfur, and since over 90 percent
of the hydrogen sulfide is present as the species HS~ at or near pH 6,
the formation of soluble polysulfide species is quite likely:
m S° + HS~ a HS^ (2 < n < 8) (ll)
According to a very recent report, reaction 11 can occur easily and
under a wide variety of conditions in aqueous systems (1*0, and i* should
be noted that this reaction represents a potentially significant source
of hydrogen sulfide consumption in the system.
Concerning the changes in pH, indicated by the dashed line in
Figure 9> several of the above reactions can contribute to pH fluctua-
tions. This fact, and the presumed presence of excess limestone in the
system, make a straightforward explanation of pH variations somewhat
difficult. It should be mentioned, however, that the same pH trends as
shown in Figure 9 were observed during the preliminary tests described
earlier. Thus, the initial decrease in pH during addition of hydrogen
sulfide is probably due to reactions 1 and 2 above, as well as to the
dissociation of hydrogen sulfide.
HsS = HS" + H* (12)
After hydrogen sulfide flow is stopped, neutralization reactions
involving limestone probably account for the gradual increase in pH,
although this increase may be augmented by reactions such as 9 and 10.
Finally, as the rates of ferrous iron and sulfide oxidation in-
crease in the later stages of the reaction, there is a tendency for the
pH to decrease with time. This may be due to reactions involving the
hydroxides, such as 7 and 8, as well as to consumption of hydrosulfide
ion via reaction 11 driving the dissociation reaction 12 to the right-
hand side.
The foregoing considerations are admittedly speculative, and sug-
gest the need for more refined work with pure systems before a clearer
understanding of the actual mechanisms can be obtained.
In the experiment where a deficiency of hydrogen sulfide was em-
ployed, the absence of measurable sulfide sulfur and the known presence
of ferrous iron in the filtered solids would seem to indicate that un-
der these conditions (i.e., in the absence of sufficient hydrogen sul-
fide to precipitate all iron as the sulfide), ferrous iron may be
present in the sludge largely as some species other than ferrous sul-
fide. It would appear that this other species is ferrous hydroxide,
which supports the earlier comment regarding competition between
hydroxyl and sulfide ion for ferrous ion. (The pH of the suspension
was within the range 7.0 to 7.k during this experiment.)
-------
Process Evaluation and Cost Estimates
In view of the experimental results obtained during the past year's
work, it seems clear that the IBS process for mine water treatment most
be re-evaluated in terms of its technological feasibility. In ad-
dition, certain findings in the present investigation have impli cations
which bear on the economi.cs of the overall treatment process. These
matters will be considered more folly in this section.
Process Evaluation
There is no question that hydrogen sulfide, in combination with
pulverized limestone to raise the pH of the system to near or above 6,
is effective in precipitating iron from mine water. The major problem
which arises, based on these studies, appears to be related to determi-
nation and control of hydrogen sulfide feed. It is evident that suf-
ficient hydrogen sulfide must be added during treatment to adequately
precipitate ferrous iron, and to adequately reduce and precipitate
ferric iron, whether present as soluble ferric ion or insoluble hydrous
ferric oxide.3 Moreover, sufficient additional hydrogen sulfide pre-
sumably must be added to compensate for any possible physical losses,
such as escape from the reactor by diffusion, and/or any possible
chemical losses, such as oxidation by oxygen (air), involvement in
solvation or hydrogen bonding reactions with water, or formation of
polysulfide species with by-product sulfur.
The extent of additional hydrogen sulfide consumption through any
of these reactions not involving iron is unknown at the present time,
and a study thereof was not within the scope of the present investiga-
tion. Undoubtedly, however, such extraneous losses of hydrogen sulfide
would be variable depending on mine water composition, temperature, and
other factors.
As stated earlier, the alternative to accurate determination and
control of hydrogen sulfide feed would be to introduce a known excess
of the gas during treatment. Such practice would, of course, require
additional provision for the removal of unreacted hydrogen sulfide from
the process effluent; the presence of residual dissolved hydrogen sul-
fide in the treated effluent would create a serious problem obvious to
anyone familiar with the noxious properties of this chemical or cogni-
zant of water quality criteria. The potential hazards involved in hand-
ling hydrogen sulfide during the actual treatment operation are serious
enough, without compounding the problem by the introduction of addi-
tional risk factors.
3 This statement involves the implicit assumption that the product
water should meet quality standards for residual total iron con-
centration (7 ppm or less) established by the Commonwealth of
Pennsylvania.
-------
55.
The more recent experimental evidence concerning sludge chemical
composition, tendency toward oxidation, and ferric iron-hydrogen sulfide
reactions has brought to light two additional uncertainties regarding
the technological feasibility of the process.
The first of these involves the abundantly clear fact that the
sulfide sludge, formed at or near room temperature under the conditions
prevailing during the treatment process, is unstable and undergoes
oxidative decomposition. This decomposition is fairly rapid on expo-
sure of the damp, filtered sludge to air, and although slower while the
sludge is still in suspension, the decomposition nevertheless occurs
gradually as oxygen diffuses through the air-liquid interface. The
products of this oxidation are sulfur and, presumably, amorphous
hydrous ferric oxide, although this statement must be qualified as
follows:
Although sulfur was invariably present in those oxidized sludge
samples subjected to analysis, it was never clear whether this material
arose entirely from oxidation of a ferrous sulfide component of the
sludge, or whether it was formed at least in part due to hydrogen sul-
fide reduction of ferric iron species present in the mine water during
treatment. Neither has the possibility of direct air oxidation of
hydrogen sulfide, via the reaction
HsS + 1/2 Oa = HgO + S°
been completely ruled out. This reaction is known to be catalyzed in
the gaseous phase by ferric oxide. However, the appearance of sulfur
in oxidized sludges derived from mine waters filtered immediately prior
to treatment, is taken as tentative evidence that oxidation of ferrous
sulfide in the sludge must occur to some extent.
The ease of decomposition exhibited by the sludge is consistent
with the known behavior of solids possessing a highly disordered
crystalline structure. Probably, therefore, the black precipitate pro-
duced under the prevailing experimental conditions would best be con-
sidered as an amorphous, disordered and presumably hydrated material,
even though chemical analyses indicate that it is a stoichiometric com-
pound whose empirical formula corresponds to that of ferrous sulfide.
Indications of structural disordering in the iron sulfide component of
the freshly prepared sludge are further confirmed by the complete ab-
sence of peaks assignable to any such species in the x-ray patterns.
The ease of sludge oxidation is considered detrimental to the ex-
tent that changes occurring in the chemical composition and, presumably,
physical properties of the sludge after treatment would unnecessarily
complicate sludge handling and disposal operations. Certainly, any
scheme for sludge processing, including that involving recycling or re-
generation of a sulfide reagent conceived prior to the inception of the
present program, must take into account the unstable nature of the
sludge itself.
-------
56.
The second uncertainty regarding the applicability of the LBS
treatment process for mine water concerns the presence of ferric iron
species in the mine water during treatment and their effect on hydrogen
sulfide consumption, the latter as a consequence of chemical reduction
and due to potential polysulfide fonnation with byproduct sulfur. AH
things considered, it is not unreasonable to state that the presence of
ferric iron species creates a nusiance factor which militates against
the efficacy of the LHS treatment process.
One obvious way to avoid interferences due to ferric iron would be
to limit the use of the LHS process to those mine waters in which it is
completely absent, either as ionic species or in the form of suspended
solids. Coal mine water discharges of this type are presumably quite
limited in number. The South Greensburg discharge, used extensively in
these studies, probably approaches the idealized specimen; it is a bi-
carbonate-buffered mine water of relatively high pH and containing
essentially no ferric iron at the point of discharge. Paradoxically,
however, it is precisely this type of mine water which is the least
stable with respect to oxidation of iron, a fact which is brought out
forcibly by the data in Table 3.
In retrospect, the presence of ferric iron species in amounts
greater than those indicated by initial analytical data probably ac-
counts in large part for the anomalous results of earlier continuous
flow runs and the subsequent uncertainties regarding the stoichiometric
hydrogen sulfide requirement. The results of later continuous flow
runs, shown by Figure 7, in which the stoichiometric hydrogen sulfide
requirement was back-calculated based on ferrous and ferric iron analy-
ses conducted on raw mine water samples taken during the run, tend to
affirm that nearly all iron in either oxidation state can be removed
with the exact stoichiometric amount of hydrogen sulfide (disregarding
extraneous losses, discussed earlier). The apparent resolubilization
phenomenon (i.e., increase in ferrous iron concentration with time
after treatment with < 1 X hydrogen sulfide) may very well be due to
hydrogen sulfide reduction of ferric iron species and, in the absence
of sufficient hydrogen sulfide to quantitatively precipitate ferrous
sulfide, the gradual establishment of equilibria involving the more sol-
uble (at pH 6) species, ferrous hydroxide. This possibility is enhanced
by the finding, disclosed earlier, that with a deficiency of hydrogen
sulfide, ferrous hydroxide apparently is formed as a stable product of
the hydrogen sulfide reduction of hydrous ferric oxide.
The data indicate that the sludge formed by the LHS process is
denser than those produced by lime or limestone neutralization alone.
Offsetting this, however, is the fact that the comparatively poor set-
tling characteristics of the LHS sludge dictate the use of flocculating
aids if a solid-fluid separation is to be effected by the conventional
methods of sedimentation.
-------
57.
In terns of the development of the process to full industrial
scale, much still remains to be done. For example, the development of
a low-cost method for on-site production of the sulfide reagent from
recycled sludge or coal refuse material is an aspect of the overall LHS
treatment process which has received only cursory attention to date.
It is felt that the reactions of hydrogen sulfide with ferrous and
ferric iron, as well as reactions of the freshly formed sludge itself,
are just beginning to be fully appreciated relative to their effect on
hydrogen sulfide consumption in the system (i.e., as concerns hydrogen
sulfide feed control). It seems clear that further studies of these
reactions will be imperative before the LHS treatment process can be
considered technologically feasible. It is suggested, therefore, that
any future effort be organized to include more detailed studies of the
basic chemistry involved. Such studies would reasonably involve a re-
turn to beaker-scale batch tests, the use of chemically pure reagents,
and more precise control of experimental conditions. Incidentally,
this type of approach based, on more fundamental studies was not included
in the scope of the present research program, nor was there time after
the adopted schedule revisions to carry out extensive, systematic inves-
tigations into the more interesting aspects of sludge formation, compo-
sition, and behavior.
Cost Estimates
Preliminary cost data, based on a comparison of the LHS process and
lime neutralization, were developed in an earlier publication. (16) For
both processes, capital investment costs were derived for the design,
purchase, and installation of the necessary equipment. The systems re-
quirements were assumed to be similar for both processes, and included
mine water holding ponds, lime (or limestone) storage bins and feeders,
reactors (mixing tanks), instrumentation, and sludge settling ponds,
together with auxiliary equipment such as pumps, piping, and electrical
supplies. Operating costs were calculated to include the costs of labor
($3.50 an hour), electrical power (1.66 cents per kilowatt hour), and
raw materials. The cost data indicated that the capital investment and
operating costs for the two processes were generally of the same order
of magnitude for plants of similar size. The major cost differentials
between lime neutralization and IBS treatment arose primarily from the
differences in chemical reagent costs for the two processes.
For the IBS process, computations were based on three different
prices of hydrogen sulfide: $200 and $1,000 per ton, quoted by suppliers,
and $15^.13 per ton, estimated to be the cost of producing hydrogen
sulfide at the treatment site by the Girdler process. In addition, cost
data were developed to include three different mine water flow rates:
Case A. 1 x 106 gallons per day (gpd); Case B, 1 x 10* gpd; and Case C,
3 x 1CP gpd. Although not indicated in this preliminary cost study,
calculations based on the data presented show that the concentrations
of iron and total acidity in the mine water considered for treatment
by the IBS process were as follows: 110 ppm ferrous iron, 0 ppm ferric
-------
58.
iron, and 200 ppm acidity. On this basis, the South Greensburg dis-
charge with a normal flow rate of approximately 1 x Iff gpd would
correspond closely to Case B.
Because of the findings in the present investigation, it was deemed
appropriate to augment these preliminary data with cost data reflecting
the effects of increased hydrogen sulfide requirements. As indicated
above, the preliminary cost analysis was based on the use of the exact
stoichiometric requirement (l.O X) of hydrogen sulfide assuming *n iron
to be in the ferrous state. Table 6 shows the results of computations
based on two additional hydrogen sulfide feed requirements, namely
1.125 X hydrogen sulfide for the case of 25 percent oxidation of initial
ferrous iron, and 1.250 X hydrogen sulfide for the case of 50 percent
oxidation of initial ferrous iron. The additional hydrogen sulfide
costs were incorporated into the overall operating costs per year, which
were then converted to costs in terms of dollars per 1,000 gallons, based
on the three flow rates cited above, over depreciation periods of 5, 10,
and 20 years. Also included in Table 6 are comparative cost data for
lime neutralization, derived during the preliminary cost study, as well
as cost data for limestone neutralization, developed recently on the
same basis3 under another BCR research program (unpublished data).
The data in Table 6 show that, in terms of cents per thousand gal-
lons of mine water treated, the costs apparently do not increase apprec-
iably at the higher hydrogen sulfide feed requirements. However, al-
though not shown by Table 6, the actual increase in overall operating
costs per year ranges between $200 (Case A, 1.125 X hydrogen sulfide @
$15^.13/ton) and $77,250 (Case C, 1.250 X hydrogen sulfide @ $l,000/ton).
As would be expected, the increase in cost becomes more obvious as the
cost of hydrogen sulfide per unit increases.
The data in Table 6 also indicate that limestone neutralization is
significantly less expensive than either lime neutralization or the LHS
process. In this regard, it is felt that the aspect of initial iron
concentration has not received sufficient attention in past cost com-
parison studies. For example, from the following hypothetical reactions
for the oxidation of pyrite, the hydrolysis of ferrous iron, and the
precipitation of ferrous sulfide it is apparent that doubling the amount
of ferrous iron in solution (i.e., doubling the amount of pyrite oxi-
dized per given volume of solution) will also double the requirement of
limestone for direct limestone neutralization and double the hydrogen
sulfide requirement in the LHS process.
3 The limestone requirement was assumed to be twice the stoicniometric
requirement based on mine water total acidity.
-------
59.
TABLE 6. COSTS OF LHS PROCESS AT DIFFERENT IfeS REQUIREMENTS
IN COMPARISON WITH COSTS OF LIME AND LIMESTONE NEUTRALIZATION
Operating Costs, Dollars per 1000 gallons*
Case A, 1 x 105 gpd flag
5-year depreciation:
HgS @ $15l+.13/ton
H..S @ $200. 00/ton
H-,3 @ $l,OOO.OO/ton
10-year depreciation:
Hj,S @ $15l+.13/ton
H-,3 @ $200. 00/ton
HgS @ $1,000. 00/ton
20-year depreciation:
ILjS @ $15l+.13/ ton
HgS @ $200. 00/ton
HgS @ $1,000. 00/ton
Case B, 1 x 10s gpd flow
5-year depreciation:
HeS @ $15l+.13/ton
H;,S @ $200. 00/ton
HgS © $1,000. 00/ton
10-year depreciation:
HsjS @ $15^.13/ton
HS @ $200. 00/ton
@ $1,000. 00/ton
20-year depreciation:
H-,3 @ $15l*.13/ton
@ $200. 00/ton
@ $1,000. 00/ton
Case C, 3 x 10s gpd flow
5-year depreciation:
IL,S @ $15l+.13/ton
H..S @ $200. 00/ton
^S @ $1,000. 00/ton
10-year depreciation:
HjS @ $15l+.13/ton
H-,3 @ $200. 00/ton
HgS @ $l,000.00/ton
20-year depreciation:
H,,S @ $15^.13/ton
HaS @ $200. 00/ton
HjS @ $1,000. 00/ton
1.000 x H,S4.
0.71
0.73
0.96
0.62
0.62
0.85
0.55
0.57
0.8o
0.19
0.20
0.1+3
0.16
0.17
0.1+0
0.11+
0.15
0.38
0.15
0.16
0.39
0.12
0.13
0.36
o.n
0.12
0.35
LHS Process
1.125 x HaS
0.72
1.00
0.63
0.63
0.89
0.56
0.58
0.81+
0.20
0.21
0.1+7
0.17
0.18
0.1+1+
0.15
0.16
0.1+2
0.16
0.17
0.1+3
0.13
0.11+
0.1+0
0.12
0.13
0.39
1.250 x H,S
0.72
1.03
0.63
0.63
0.92
0.56
0.58
0.87
0.20
0.22
0.50
0.17
0.19
0.1+8
0.15
0.17
0.1+6
0.16
0.18
0.1+7
0.13
0.15
0.1+1+
0.12
0.1+3
Lime
neutralization
0.78
0.63
0.55
O.lU
0.12
0.10
0.11
0.09
0.08
Limestone
Neutral i zat ion
0.73
0.58
0.51
0.11
0.08
0.07
0.07
0.05
0.01+
* Rounded to the nearest $0.01
4- Previously presented in Table B-6, Reference 16
* The limestone requirement is assumed to be twice the stoichiomatric
requirement based on mine water total acidity
-------
60.
FeSa + 7/2 Oa + HgO = FeS04 + HgS04
FeS04 + 1/k 02 + 5/2 HaO = Fe(OH)a
FeS04 + HgS = FeS +
Confutations on this basis for Case B, 5-year depreciation period, with
1.0 X hydrogen sulfide @ $15U.13/ton show that yearly treatment costs
would increase by $15,900, corresponding to an increase from 19$f to 2k j
per 1,000 gallons (26 percent increase) for the IBS process. In contrast,
doubling the limestone requirement for Case B, 5-year depreciation period
with 2.0 X limestone @ $3.00/ton results in an increase in yearly treat-
ment costs of $3»000, corresponding to an increase from 11^ to 12^ per
1,000 gallons (9 percent increase) for direct limestone neutralization
(the actual increase is to 11.6^/1,000 gallons).
Finally, in the preliminary cost evaluation study it was suggested
that recovering the sulfide reagent from the sludge for reuse would re-
duce the cost differential between the LHS process and lime neutraliza-
tion sufficiently to make the former process economically competitive.
On the basis of the preliminary cost data, this suggestion appeared to
be reasonable and warranted. However, no consideration was given to
the fact that perhaps the most promising approach to such a reagent re-
covery process, based on experimental results, involved heating the
sludge (under nitrogen) to 750 C (1382 F). Although final judgment
should be withheld pending further investigation of this facet of the
overall process, it would seem that the energy requirements in such a
thermal treatment of the sludge would militate strongly against any
cost advantage to be gained thereby.
In summary, in view of the considerations presented throughout
this section, it seems that a critical re-appraisal of the limestone-
hydrogen sulfide treatment process for mine drainage at this point in
time would justifiably cast some doubt as to its technological and
economic feasibility. It is our opinion, and available experimental
evidence indicates, that in terms of its applicability to different
types of mine waters, its ease of control, and its simplicity in imple-
mentation, the LHS process suffers by comparison with accepted methods
of treatment.
-------
61.
Section 5
REFERENCES
Cited References
1. Bernal, J. D., Dasgupta, D. R., and Mackay, A. L., "The
oxides and hydroxides of iron and their structural inter- relationships,"
Clay Minerals Bull, k (21), 15-30 (1959).
2. Berner, R. A., "Iron sulfides formed from aqueous solution
at low temperatures and atmospheric pressure," J. Geol. 72 (3), 293-306
~~
3. Boehm, H. P. and Flaig, E., "Iron(lll) sulfide," Angew.
Chem., Intern. Ed. Engl. 5 (11), 963 (1966).
k. Burkin, A. R. , "The Chemistry of Hydrometallurgieal
Processes," Princeton: D. Van Nostrand, Inc., 1966. pp 87-89.
5. Freke, A. M. and Tate, D., "The formation of magnetic iron
sulphide by bacterial reduction of iron solutions," J. Biochem.
Microbiol. Technol. Eng. _3 (l), 29-39 (1961).
6. Glover, H. Gordon, National Coal Board, Yorkshire, England,
Mar. 1969. Private communication.
7. Heslop, R. B. and Robinson, P. L., "Inorganic Chemistry,"
Amsterdam: Elsevier, I960, p 466.
8. Kirk, R. E. and Othmer, D. F., "Encyclopedia of Chemical
Technology," 2nd Ed., Vol. 12, New York: J. Wiley & Sons, Inc., 1967.
p U2.
9. Kumai, T. , "Reactions between ferric hydroxide and hydrogen
sulfide. Part 2. Formation of ferrous sulfide," Nippon Kagaku Zasshi
70^ 7^9-52 (1958); Chem. Abstr. 53_, 6860c (1959).
10. Lange, N. A., "Handbook of Chemistry," Eighth ed., Sandusky,
Ohio: Handbook Publishers, Inc., 1952. p 1083.
11. Mackay, A. L. , "Some aspects of the topochemistry of the
iron oxides and hydroxides," in "Proceedings of the ktb. International
Symposium on the Reactivity of Solids," J. H. DeBoer, ed., Amsterdam:
Elsevier, 1961. pp 571-83.
12. Mellor, J. W., "A Comprehensive Treatise on Inorganic and
Theoretical Chemistry," Vol. XIV, London: Longmans, Green & Co., Ltd.,
1935. PP 157-8.
-------
62.
13. Moeller, T., "Inorganic Chemistry," New York: J. Wiley &
Sons, Inc., 1952. PP 286-7*
lU. Monscvltz, J. T. and Ainsworth, L. D., "Hydrogen polysulfide
in water systems," ACS 157th National Meeting, Div. Water, Air, and
Waste Chemistry, Minneapolis, Minn., Apr. 14-lfi, 1969. See also,
Monscritz, J. T. and Ainsworth, L. D., "Unique polysulflde taste and
odor problem at Santa Barbara, California," Taste Odor Control J. 34
(2), 1-4 (1968).
15. Schultze, R., Schellhorn, J., and Boberg, F., "Pyrophors.
Iron-sulfur compounds," Arbeitsschutz 1964, 194-6 (1964); Chem. Abstr.
65, I48o8g (1966).
16. Zavadzki, £. A. and Glenn, R. A., "Sulfide treatment of acid
mine drainage," Bituminous Coal Research, Inc., Final Kept. L-290 to
Appalachian Regional Comm. and Federal Water Pollution Control Admin.
(1968).
Additional References, Uncited
17. Aksel'rud, N. V. and Spivakovskii, V. B., "Solubility product
and solubility," Ukrain. Khim. Zhur. 25, l4-7 (1959); Chem. Abstr. 54,
46d (I960).
Calculation of dissociation constant and solubility of
ferrous sulfide.
18. Bouet, J. and Brenet, J. P., "Potential/pH diagram for iron
in sulfide-bearing media," Corrosion Sci. 3 (l), 51-63 (1963); Chem.
Abstr. 59_, H033b (1963).
19. Bowers, J. W., Fuller, M. J. A., and Packer, J. E.,
"Autoxidation of aqueous sulfide solutions," Chem. Ind. (London) 1966
(2), 65-6 (1966).
Concentration-dependence of E^S oxidation products,
including sulfur and polysulf ides .
20. Boznanski, A., Kowalowa, S., and Falecki, M., "Suitability of
mixtures of bog iron ore and pickling sludges for removal of hydrogen
sulfide from gases," Koks, Smola, Gaz 7 (2), 63-6 (1962); Chem. Abstr.
63, 8074f (1965).
21. Czamanske, G. K., "Sulfide solubility in aqueous solutions,"
Econ. Geol. 54, 57-63 (1959).
Solubility products and solubilities at various pH values
and temperatures for several metal sulfides, including FeS.
-------
63.
22. Egorov, A. M. and Vol'skii, A. N., "Hydrolysis of
difficultly soluble or slightly dissociated electrolytes," Zh.
Neorgan. Khlm. 5, 2677-80 (I960); Chem. Abstr. 56, 1^99^a (1962).
Calculated hydrolysis constants for various metal sulfides,
including FeS and FeaS3.
23. Hem, J. D., "Some chemical relationships among sulfur
species and dissolved ferrous iron," U.S. Geol. Surv., Water Supply
Paper No. 1^59-C, 57-73 (i960).
24. Jaulmes, P. and Brun, S., "Solubility and precipitation
of slightly soluble salts of weak or moderately strong acids," Trav.
Soc. Pharm. Montpellier 25 (2), 98-110 (1965); Chem. Abstr. 65, 8o68c
(1966).
25. Kaczorek, M., Leszczynska, H., Sobiesiak, R., and Hoffman,
P. M., "Absorption of post-aeration gases from deposit waters on
ferric hydroxide," Przemysl Chem. tjl, 513-7 (1962); Chem. Abstr. 58,
351d (1963).
Absorption of HgS in Fe(OH)a suspensions.
26. Khodakovskii, I. L., "The hydrosulfide form of the heavy
metal transportation in hydrothermal solutions," Geokhimiya 1966 (8),
960-71 (1966); Chem. Abstr. 65, l4502e (1966).
Stability of hydrosulfide (HS~) complexes of heavy metals,
including iron.
27. Korolev, D. F. and Kozerenko, S. V., "The formation of
iron sulfides from solutions," Dokl. Akad. Nauk SSSR l6_5 (6), 1^402-^
(1965); Chem. Abstr. 64, 7721g (1966).
28. Kremer, V. A. and Vail, E. I., "Device for continuous
potentiometric determination of ionic concentration in solutions, e.g.,
sulfides," USSR Pat. 180,397 and 180,398 (Mar. 21, 1966); Chem. Abstr.
65, 9738a,h (1966).
29. Leszczynska, H., Kaczorek, M., Sobiesiak, R., Witkowska,
B., Hoffman, B., and Pfeffer, A., "Development of a method for the
removal of IfeS from mine waters," Freiberger Forschungsh. 350A,
183-94 (1965); Chem. Abstr. 6U, 4783f (1966).
-------
Absorption of IfeS by an alkaline ferric hydroxide sus-
pension.
30. Loy, H. L. and Hinmelblau, D. M., "The first ionization
constant of hydrogen sulfide in water," J. Phys. Chem. 65, 26U-7 (196!)
31. Meuly, W. C. and Seldner, A., "Method for abating stream
pollution," U.S. Pat. 3,226,320 (Dec. 28, 1965).
Use of iron chelate compounds to remove IfeS from polluted
streams.
32. Nelson, N. H. and Jepson, C. A., "I^S removal from water
without air pollution," Public Works 9j± (l) , 97-8 (1963).
Use of FeS04 in removing EfeS from city water supplies.
33. Petrucci, R. H. and Moews, P. C., Jr., "HjjS equilibriums;
the precipitation and solubilities of metal sulfides," J. Chem. Educ.
32 (8), 391-1* (1962).
34. Pohl, H. A., "Solubility of iron sulfides," J. Chem. Bag.
Data 7, 295-306 (1962) .
Changes in solubility and stoichiometry of iron sulfides
with temperature.
35. Simon, A. and Reichelt, D., "Iron sulfides and their
oxidation products," Chem. Zvesti 13, 731-2 (1959); Chem. Abstr. 54,
(1960).
Study of the composition and oxidation of iron sulfides
formed by the action of HjS on Fe(OH)3.
36. Simon, A. and Reichelt, D., "Dry gas purification," Part 4,
"Chemical and x-ray investigations of the oxidation of iron sulfide
as a model for the regeneration of sulfided iron oxide masses,"
Z. Anorg. Allgem. Chem. 305, 108-15 (I960); Chem. Abstr. 55,
(1961) . —
37. Simon, A. and Reichelt, D., "Dry gas purification," Part 5,
"The sulfidation of iron(lll) hydroxide and the regeneration of iron
sulfide," Z. Anorg. Allgem. Chem. 319, 24-36 (1962); Part 6, "The
-------
65.
influence of the pH and the temperature on the activity and capacity of
iron hydroxides toward hydrogen sulfide," Ibid, 37-*A (1962); Simon, A.
and Scheibitz, M., Part 7, "The sulfidability of differently prepared,
definite iron hydroxides and oxides as a demonstration of the generality
of the •skeleton-theory,1" Ibid, 1*5-51 (1962); Chem. Abstr. 58,
(1963).
The composition and reactivity of iron sulfide compounds
formed by the action of HJaS on various ferric oxyhydroxides, dried
and in aqueous suspension, at various pH values and temperatures.
38. Uspenskii, F. and Diev, N. P., "Some particular reactions
of sulfates with sulfides," Zhur. Neorg. Khim. 5, 1022-7 (I960); Chem.
Abstr. 56, 4359d (1962).
The reaction of CaS04 and FeS is discussed as an example.
39. Vasil'ev, V. and Rodicheva, N. A., "Preparation of
solutions for the detection of anions," Vestn. Leningr. Univ. l6 (10),
Ser. Fiz. ± Khim. (2), 1^5-7 (1961); Chem. Abstr. 55, 196lle (1961).
Mole percentage compositions of weak acids, including
and their dissociation products at pH 5, 6, 7, 8, 9> 10, and 11.
IK). Yamada, M., "Mine water treatment by I^S at the Akita
mine," Nippon Kogyo Kaishi 8l (6), 30-8 (1965); Chem. Abstr. 65,
(1966) . — ~~
Use of H^S to precipitate Cu3* from a mine water containing
Cu3*, Zna+, Fe3* and Fe3*, after preliminary limestone neutralization.
Ifl. Yamada, M., "Processing of mine waters from the Akita mine
with hydrogen sulfide," Nippon Kogyo Kaishi &L, 55^-62 (1965); Chem.
Abstr. 66, 108077g (1967).
-------
ABSTRACT: The studies reported herein were a continuation of a
12-month program initiated in June 1967 with funds from
the Appalachian Regional Commission. During this earlier
program a process for the treatment of coal mine drainage
was conceived, involving the combined addition of lime-
stone and hydrogen sulfide to effect precipitation of iron
sulfides (LHS process).
Results of laboratory-scale continuous flow tests of the
LHS process during the current program indicated that
hydrogen sulfide feed must be accurately predetermined
and controlled to effectively precipitate iron, yet to avoid
an excess of reagent in the process effluent.
The black sludge formed during treatment undergoes oxi-
dation at a rate depending on drying conditions, with
formation of elemental sulfur. X-ray diffraction analyses
indicate the iron sulfide sludge component is an amorphous
material, although chemical analyses indicate it is initially a
stoichiometric compound whose empirical formula cor-
responds to that of ferrous sulfide.
ABSTRACT: The studies reported herein were a continuation of a
12-month program initiated in June 1967 with funds from
the Appalachian Regional Commission. During this earlier
program a process for the treatment of coal mine drainage
was conceived, involving the combined addition of lime-
stone and hydrogen sulfide to effect precipitation of iron
sulfides (LHS process).
Results of laboratory-scale continuous flow tests of the
LHS process during the current program indicated that
hydrogen sulfide feed must be accurately predetermined
and controlled to effectively precipitate iron, yet to avoid
an excess of reagent in the process effluent.
The black sludge formed during treatment undergoes oxi-
dation at a rate depending on drying conditions, with
formation of elemental sulfur. X-ray diffraction analyses
indicate the iron sulfide sludge component is an amorphous
material, although chemical analyses indicate it is initially a
stoichiometric compound whose empirical formula cor-
responds to that of ferrous sulfide.
ABSTRACT: The studies reported herein were a continuation of a
12-month program initiated in June 1967 with funds from
the Appalachian Regional Commission. During this earlier
program a process for the treatment of coal mine drainage
was conceived, involving the combined addition of lime-
stone and hydrogen Sulfide to effect precipitation of iron
sulfides (LHS process).
Results of laboratory-scale continuous flow tests of the
LHS process during the current program indicated that
hydrogen sulfide feed must be accurately predetermined
and controlled to effectively precipitate iron, yet to avoid
an excess of reagent in the process effluent.
The black sludge formed during treatment undergoes oxi-
dation at a rate depending on drying conditions, with
formation of elemental sulfur. X-ray diffraction analyses
indicate the iron sulfide sludge component is an amorphous
material, although chemical analyses indicate it is initially a
stoichiometric compound whose empirical formula cor-
responds to that of ferrous sulfide.
ACCESSION NO:
KEY WORDS:
Coal Mine Drainage
Iron Sulfides
Pollution Abatement
Limestone
ACCESSION NO:
KEY WORDS:
Coal Mine Drainage
Iron Sulfides
Pollution Abatement
Limestone
ACCESSION NO:
KEY WORDS:
Coal Mine Drainage
Iron Sulfides
Pollution Abatement
Limestone
-------
The unstable nature of the sulfide sludge, possibility of
polysulfide formation during treatment, instability of mine
waters of the type amenable to treatment, and inadequacies
of available gas metering equipment are among the factors
which militate against the controlled regulation of hydro-
gen sulfide feed necessary for successful operation of the
process. These factors, and additional disadvantages re-
vealed by an updated cost evaluation, lead to the conclu-
sion that the LHS process is less attractive than accepted
methods of mine drainage treatment.
The current program was cosponsored by the FWPCA and
the coal industry through its research agency, Bituminous
Coal Research, Inc.
This report was submitted in fulfillment of FWPCA Grant
No. 14010 DLC between the Federal Water Pollution
Control Administration and Bituminous Coal Research,
Inc.
The unstable nature of the sulfide _sludge, possibility of
polysulfide formation during treatment, instability of mine
waters of the type amenable to treatment, and inadequacies
of available gas metering equipment are among the factors
which militate against the controlled regulation of hydro-
gen sulfide feed necessary for successful operation of the
process. These factors, and additional disadvantages re-
vealed by an updated cost evaluation, lead to the conclu-
sion that the LHS process is less attractive than accepted
methods of mine drainage treatment.
The current program was cosponsored by the FWPCA and
the coal industry through its research agency, Bituminous
Coal Research, Inc.
This report was submitted in fulfillment of FWPCA Grant
No. 14010 DLC between the Federal Water Pollution
Control Administration and Bituminous Coal Research,
Inc.
The unstable nature of the sulfide sludge, possibility of
polysulfide formation during treatment, instability of mine
waters of the type amenable to treatment, and inadequacies
of available gas metering equipment are among the factors
which militate against the controlled regulation of hydro-
gen sulfide feed necessary for successful operation of the
process. These factors, and additional disadvantages re-
vealed by an updated cost evaluation, lead to the conclu-
sion that the LHS process is less attractive than accepted
methods of mine drainage treatment.
The current program was cosponsored by the FWPCA and
the coal industry through its research agency, Bituminous
Coal Research, Inc.
This report was submitted in fulfillment of FWPCA Grant
No. 14010 DLC between the Federal Water Pollution
Control Administration and Bituminous Coal Research,
Inc.
Hydrogen Sulfide
Sulfides
Waste Water Treatment
Stream Pollution
Hydrogen Sulfide
Sulfides
Waste Water Treatment
Stream Pollution
Hydrogen Sulfide
Sulfides
Waste Water Treatment
Stream Pollution
------- |