&EPA
United States
Environmental Protection
Agency
Industrial Environmental Research EPA-600/7-80-006
Laboratory January 1980
Cincinnati OH 45268
Research and Development
Rotating Disc
Biological
Treatment of
Acid Mine Drainage
Interagency
Energy/Environment
R&D Program
Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4 Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-80-006
January 1980
ROTATING DISC BIOLOGICAL TREATMENT
OF ACID MINE DRAINAGE
by
Harvey Olem and Richard F. Unz
The Pennsylvania State University
University Park, Pennsylvania 16802
Grant No. R-805132
Project Officer
Roger C. Wilmoth
Resource Extraction and Handling Division
Industrial Environmental Research Laboratory
Cincinnati, Ohio 45268
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Industrial Environmental Research
Laboratory-Cincinnati, U.S. Environmental Protection Agency, and approved for
publication. Approval does not signify that the contents necessarily reflect
the views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
ii
-------
FOREWORD
When energy and material resources are extracted, processed, converted,
and used, the related pollutional impacts on our environment and even on our
health often require that new and increasingly more efficient pollution
control methods be used. The Industrial Environmental Research Laboratory -
Cincinnati (lERL-Ci) assists in developing and demonstrating new and improved
methodologies that will meet these needs both efficiently and economically.
This report describes studies that successfully utilized the rotating
biological contactor (RBC) as a unit process in the treatment of acid mine
drainage. The RBC, using indigenous bacteria, oxidized ferrous iron to the
ferric state, thus making the mine drainage amenable to limestone neutraliza-
tion and widening the mine operator's choice of treatment options. The
information contained in this report will be of interest to mine operators,
regulatory agencies, and academia. For further information, please contact
the Resource Extraction and Handling Division.
David G. Stephan
Director
Industrial Environmental Research Laboratory
Cincinnati
iii
-------
ABSTRACT
The rotating biological contactor (RBC) was investigated under field
and laboratory conditions for application to Fe(II) oxidation in acid mine
drainage. At three coal mining locations in Pennsylvania and West Virginia,
treatment of six heterogeneous mine waters was investigated in experiments
with pilot-scale (0.5-m diameter) and prototype (2.0-m) RBC units. In the
field and laboratory, synthetic and supplemented natural mine drainage were
used with bench-scale RBC units to study certain potentially influential
factors on Fe(II) oxidation under controlled conditions.
Continuous biological oxidation of Fe(II) to less soluble Fe(III) was
accomplished at natural mine water temperatures as low as 0.4°C at Hawk Run,
Pa. and as high as 29°C at Crown, W. Va. Reduction of Fe(II) oxidation
efficiency at 0.4°C amounted to 10 to 20 percent of that achieved at 10°C at
the same site. Oxidation efficiency was above 80 percent at each location
for both RBC field units at mine water temperatures of 10 to 29°C. Micro-
biological oxidation with the 0.5-m RBC was unaffected at influent mine water
pH values in the range of 2.18 to 5.50 (Crown, W. Va.). Lower influent
Fe(II) in Hollywood, Pa. mine drainage (71.6 mg/£) resulted in higher oxida-
tion efficiency at equilibrium for the 0.5-m and 2.0-m RBC units as compared
to treatment efficiency when mine water contained about twice the influent
Fe(II) content (5.2- and 2.9-percent change, respectively).
Fe(II) oxidation efficiency was an average 10 percent lower with the
2.0-m than with the 0.5-m RBC under equivalent conditions with homologous
mine drainages. The observed decrease in oxidation was due, in large part,
to nonmicrobiological factors such as increased short-circuiting, lower
residence time, and a smaller effective surface area which may be increased
through proper design.
In experiments with synthetic mine drainage, Fe(II) oxidation in the
bench-scale RBC was improved by supplementation with natural mine drainage.
Examination of solids samples removed from disc surfaces of the 0.5-m RBC
operating at Hollywood, Pa. revealed the presence of iron-oxidizing and
heterotrophic bacteria in a gelatinous, iron-containing matrix. A gelatinous
surface covering was not seen on disc surfaces in field experiments at Hawk
Run, Pa., where Fe(II) oxidation efficiency was 10 to 20 percent less than at
other locations. Heterotrophic bacteria recovered from mine water and disc
solids may produce the gelatinous film.
Costs for Fe(II) oxidation with the RBC were estimated to be about
twice the amortized capital costs and one-half the operating costs compared
iv
-------
to a conventional chemical oxidation process. Neutralization of RBC effluent
and separation of precipitated iron solids must be applied in a complete
treatment scheme to produce water of a suitable quality for stream-release.
This report was submitted'in fulfillment of Grant No. 805132 by The
Pennsylvania State University under the sponsorship of the U.S. Environmental
Protection Agency. This report covers the period May 23, 1977 to October 31,
1978.
-------
CONTENTS
Foreword
Abstract iv
Figures ix
Tables xii
Abbreviations and Symbols xiv
Acknowledgments xv
1. Introduction 1
Background 1
Study objectives 1
2. Summary and Conclusions 3
3. Recommendations 6
4. Acid Mine Drainage Pollution 7
Introduction 7
Treatment of acid mine drainage 8
Chemistry of iron in mine drainage 8
Oxidation processes for Fe(II) 9
Biological Fe(II) oxidation 16
Biological treatment of acid mine drainage 16
5. Rotating Biological Contactor Treatment System 19
Principles 19
History 19
Process operation 21
Applications ,23
Inorganic reactions 23
Fe(II) oxidation - Studies at Hollywood, Pa 24
6. Field Studies: Pennsylvania and West Virginia 26
Location of field sites 26
Crown, W. Va 26
Hollywood, Pa 26
Hawk Run, Pa 26
Mine water characteristics 27
Crown, W. Va 27
Hollywood, Pa 32
Hawk Run, Pa 32
Experimental apparatus and procedures 33
Rotating biological contactors 33
Other equipment 35
Experimental procedures 35
Sampling 36
vii
-------
Results 36
Start-up 36
Equilibrium conditions 38
Effect of temperature 43
Effect of pH 49
Effect of influent Fe(II) 49
Kinetics of Fe(II) oxidation 49
Effect of hydraulic loading 54
Effect of unanticipated shutdown 54
Characteristics of biological film 56
Hydraulic Characteristics of RBC 66
Cost analysis 66
Estimated costs 66
Cost comparison 70
7. Bench-Scale Studies 73
Introduction 73
Mine water characteristics 73
Experimental apparatus and procedures 77
Crown, W. Va 77
Laboratory 77
Results 79
Adjustment of pH of natural mine water 79
Supplementation of synthetic mine water
with natural mine drainage 79
8. Discussion 84
Role of RBC in mine drainage treatment 84
Relationship of temperature to Fe(II) oxidation 86
Relationship of pH to Fe(II) oxidation 86
Synthetic versus natural mine drainage 87
Development of solids on RBC disc surfaces 87
Comparison of different size RBC units 89
References 91
Appendices
A. Chemical equilibria of iron 97
B. Analytical procedures 100
C. Mixing characteristics of mine drainage streams
at Hawk Run, Pa 106
D. Characteristics of experimental RBC units 109
E. Formulas for statistical analysis 112
viii
-------
FIGURES
Number
1 Solubility of pure Fe(OH>3 in the presence of free
Fe3+, Fe(III) - OH and Fe(III) - 804 complexes 10
2 Solubility of pure Fe(OH)2 in the presence of free
Fe2+, Fe(II) - OH and Fe(II) - S04 complexes 11
3 Chemical oxidation rate of Fe(II) as a function of pH 13
4 Location of alternative Fe(II) oxidation process and
neutralization options in two acid mine drainage
treatment schemes 15
5 Schematic of rotating biological contactor package
unit showing disc assembly suspended in contoured
tank 20
6 Large-scale rotating biological contactor installation
showing disc assemblies suspended in below-ground
concrete basins 20
7 Schematic of mine drainage streams available at
Hawk Run, Pa. showing locations of mine water
feed lines 28
8 RBC units with disc diameters of 2.0 m (top) and
0.5 m (bottom) employed in field experiments 34
9 Start-up and time required to obtain equilibrium
Fe(II) oxidation with the 0.5-m RBC in treatment
of Crown mine drainage, Hollywood Proctor No. 2
mine water and combination Hawk Run + MH 1
discharge 37
10 Iron-oxidizing bacteria at surface of 0.5-m RBC
discs for each mine drainage location 39
11 Iron-oxidizing and heterotrophic bacteria at surface
of 0.5-m RBC discs at Hollywood, Pa 40
ix
-------
12 Effect of mine drainage temperature fluctuations over
one 24-hr period on Fe(II) oxidation with 2.0-m RBC
in treatment of MH 1 discharge (top) and combination
Hawk Run and MH 1 drainage (bottom) 45
13 Effect of temperature fluctuations over one 24-hr period
on Fe(II) oxidation with 2.0-m RBC in treatment of
Hawk Run mine water 46
14 Effect of mine drainage temperature on Fe(II) oxidation
for the 0.5-m RBC in treatment of mine waters at
Hawk Run, Pa 47
15 Effect of mine drainage temperature on Fe(II) oxidation
for the 2.0-m RBC in treatment of mine waters at
Hawk Run, Fa 48
16 Changes in mine water pH with retention time in the
0.5-m RBC 50
17 Effect of influent Fe(II) concentration on Fe(II)
oxidation efficiency in treatment of Hollywood
mine drainage 51
18 Kinetics of Fe(II) oxidation with the 0.5-m RBC 52
19 Kinetics of Fe(II) oxidation with the 2.0-m RBC 53
20 Solids formation on stage 1 discs of 0.5-m RBC in
treatment of mine drainage at Hollywood, Pa. (top)
and Hawk Run, Pa. (bottom) 57
21 Phase-contrast photomicrograph showing bacteria
embedded in gelatinous matrix from stage 1 of
0.5-m RBC in treatment of mine drainage at
Hollywood, Pa 58
22 Scanning electron micrographs. Top: Glass cover
slips removed from stage 1 disc of 2.0-m RBC after
54 days of continuous treatment at Hawk Run. Bottom:
Microorganisms filtered from Hawk Run mine water onto
a 0.40-ym diameter filter 60
23 Recovery of tracer in each stage of RBC field units .. 67
24 Bench-scale RBC units. Top: Experimental 15-cm
diameter units used in pH studies at Crown, W. Va.
Bottom: 10-cm units used in laboratory studies on
the supplementation of natural mine drainage to
synthetic mine water feed 74
-------
25 Apparatus and concentrated solutions employed in
continuous metering of synthetic mine drainage
to 10-cm bench-scale RBC units 78
26 Fe(II) oxidation in 15-cm RBC units operated on Crown
mine drainage with a natural pH range of 4.05 to 5.74
(RBC A) and acidified to pH 3 (RBC B) 80
27 Iron-oxidizing bacterial densities at surface of
discs and total disc solids dry weight. Bench-scale
(15-cm) RBC units in treatment of Crown drainage
with a natural pH range of 4.05 to 5.74 (RBC A) and
an adjusted pH of 3 (RBC B) 81
28 Fe(II) oxidation in bench-scale (10-cm) RBC units in
treatment of synthetic mine water at 10°C and
initial pH of 3 82
29 Proposed treatment process for acid mine drainage
85
30 Functions of RBC disc surface, disc solids, liquid
film, and atmosphere in mine water treatment 88
C-l Schematic of mine drainage streams at Hawk Run, Pa.,
showing locations of sampling stations and mine
water feed lines 107
D-l Plan and side view showing flow pattern and stage
sampling locations for trough of 0.5-m RBC (left)
and 2.0-m RBC (right) HI
xi
-------
TABLES
Number Page
1 Examples of Forms of Dissolved Iron in Mine Drainage 9
2 Potential Methods for Fe(II) Oxidation in Acid Mine
Drainage Treatment 12
3 Characteristics of Acid Mine Drainages Studied at the
U.S. EPA Crown Field Site Near Morgantown, W. Va 29
4 Characteristics of Acid Mine Drainages Studied at the
Experimental Mine Drainage Treatment Facility at
Hollywood, Pa 30
5 Characteristics of Acid Mine Drainages Studied at the
Hawk Run Mine Water Treatment Plant Near Philllpsburg, Pa. . . 31
6 Comparative Performance of 0.5- and 2.0-m RBC Units at
Equilibrium Fe(II) Oxidation for Each Mine Drainage
Treated 41
7 Dissolved Oxygen in Mine Drainage Influent and Individual
Stage Compartments of 0.5- and 2.0-m RBC Units at
Equilibrium Fe(II) Oxidation Efficiency 42
8 Iron-oxidizing Bacteria in Mine Drainage Influent and
Effluent of 0.5- and 2.0-m RBC Units at Equilibrium
Fe(II) Oxidation Efficiency 44
9 Effect of Hydraulic Loading on Fe(II) Oxidation
Efficiency with the 0.5- and 2.0-m RBC Units in
Treatment of Mine Hole 1 Discharge 55
10 Characteristics of Solids at Surface of Discs in
0.5-m RBC 61
11 Characteristics of Solids at Surface of Discs in
2.0-m RBC 61
12 Iron-oxidizing Bacterial Densities of Disc Solids at
Each Stage of the 0.5- and 2.0-m RBC 62
13 Total Solids Content of Disc Solids at Each Stage of
the 0.5- and 2.0-m RBC 63
xii
-------
14 Total Iron Content of Disc Solids at Each Stage of
the 0.5- and 2.0-m RBC 64
15 Aluminum Content of Disc Solids at Each Stage of
the 0.5- and 2.0-m RBC 65
16 Hydraulic Characteristics of 0.5- and 2.0-m RBC Units
68
17 Estimated Costs for Fe(II) Oxidation with Rotating
Biological Contactors 69
18 Estimated Costs for Fe(II) Oxidation by Conventional
Chemical Oxygenation with Mechanical Aerators 71
19 Characteristics of Acid Mine Drainage Employed in
Bench-scale Experiments at the U.S. EPA Crown Field
Site Near Morgantown, W. Va 75
20 Composition of Synthetic Mine Drainage Feed Solution
Employed in Laboratory Studies 76
21 Iron-oxidizing Bacteria in Influent and Effluent Mine
Water and Disc Solids of 10-cm RBC units in Treatment
of Synthetic Mine Water Supplemented with Hollywood
Mine Drainage (RBC 1) and Unsupplemented (RBC 2) 83
22 Comparison of Available Surface Area for Different
Size RBC Units 90
A-l Chemical Equilibrium Describing Solubility and
Complexation of Fe(II) and Fe(III) 98
B-l Culture Medium Used to Recover Iron-oxidizing
Bacteria from Mine Water and Disc Solids 103
B-2 Culture Medium Used to Recover Heterotrophic Bacteria
from Mine Water and Disc Solids 104
C-l Mine Water Fe(II) Concentrations at Stations on
Streams at Hawk Run, Pa 108
D-l Design and Normal Operational Parameters of RBC
Experimental Units 110
xiii
-------
ABBREVIATIONS AND SYMBOLS
Al
cm
co2
DO
Fe(II)
Fe(III)
H+
HC1
H2S04
kw
kwh
In
log
Li
LiCl
m
m/mln
ra3/d
mg/cm^
mg/£
mi
MPN
N
NH3-N
(N02 + N03)-N
Org-N
PH
rpm
R
RBC
RO
SEM
Total Fe
Total P
aluminum
centimeter
carbon dioxide
dissolved oxygen
ferrous iron
ferric iron
hydrogen ion
hydrochloric acid
sulfuric acid
kilowatt
kilowatt-hour
natural logarithm (base e)
common logarithm to base 10
lithium
lithium chloride
meter
meters per minute
cubic meters per day
cubic meters per day per square meter
milligrams per square centimeter
milligrams per liter
milliliter
most probable number
normality
ammonia nitrogen
nitrite plus nitrate nitrogen
organic nitrogen
negative log hydrogen ion activity
revolutions per minute
correlation coefficient
rotating biological contactor
reverse osmosis
scanning electron microscope
ferrous iron plus ferric iron
total phosphorus
cents per cubic meter
degrees Celsius
xiv
-------
ACKNOWLEDGMENTS
Appreciation is expressed to Howard A. Jacobs'on for his valuable
assistance toward the successful completion of the investigation. The
contributions of Russell E. Harms, Dr. Donald Langmuir, Dr. David A. Long,
Dr. Harold L. Lovell, Dr. Archie J. McDonnell, Dr. John B. Nesbitt,
M. Elizabeth Olem, Dr. Raymond W. Regan, Loren C. Trick, Paul A. Wichlacz,
and Dama L. Wirries are greatfully acknowledged.
Appreciation is expressed to C. H. McConnell and other personnel of
the Pennsylvania Department of Environmental Resources for use of the
Hollywood and Hawk Run facilities, and to Robert B. Scott and other personnel
of the U.S. Environmental Protection Agency for use of the Crown field site.
Laboratory experiments were conducted with funds provided by the Pennsylvania
Science and Engineering Foundation under Project No. 416.
xv
-------
SECTION 1
INTRODUCTION
BACKGROUND
Acid mine drainage may occur as a consequence of mining activities.
Iron and sulfur-bearing minerals present in coal seams and surrounding
strata become exposed to air and water; all of which enter into the formation
of mine drainage. Source abatement and the treatment of mine discharges are
the two current means available for mine drainage pollution control.
Due to recent legislative requirements (67), acid and ferruginous
discharges must be treated before stream-release to reduce the net acidity
and metal content of drainage. A typical mine drainage treatment scheme
includes a neutralization process to remove acidity followed by oxidation of
reduced iron and subsequent removal of iron hydroxide solids from the water
before effluent discharge. Oxidation of soluble ferrous iron [Fe(II)] is
usually performed by chemical oxidation or addition of a commercial oxidant.
Biological processes have been examined for treatment of Fe(II) in mine
drainage although uncertainties exist regarding their overall applicability,
cost, and process effectiveness.
The rotating biological contactor (RBC) process has been shown to
provide appreciable reduction in the Fe(II) content of coal mine drainage
(54). Biological oxidation of Fe(II) to Fe(III) precedes precipitation of
the ferric species as insoluble ferric oxyhydroxides which are formed at the
near neutral pH values required for discharge (see Section 4).
The emphasis of this project was to determine if naturally-occurring
bacteria could be utilized to effectively and efficiently oxidize the
ferrous ion to the ferric form. By visiting different acid mine drainage
sources, a wider variety of real world situations could be evaluated.
STUDY OBJECTIVES
Both field and laboratory studies were conducted to examine the Fe(II)
oxidizing capacity of the RBC in the treatment of acid mine drainage. Field
studies were performed at several mine drainage locations under various
conditions. A prototype RBC unit was evaluated under conditions simulating
actual treatment of mine drainage. Unaltered, unsupplemented mine drainage
was studied in field experiments.
-------
Bench-scale investigations were performed with synthetic and
supplemented natural mine drainage in order to study certain factors under
controlled conditions. Previous RBC studies were performed under field
conditions using only a small pilot unit and a single source of mine water
(54).
Specific objectives of the present study included:
1. Treatment of mine drainages of contrasting physical, chemical,
and microbiological nature.
2. Comparative performance evaluation of different size RBC units
for determination of potential scale-up problems in design and
operation of the process.
3. Pattern of microorganism development on the rotating disc
surfaces and their possible role in the treatment system.
4. Requirements for treatment of specific drainages and necessary
modifications to accomplish desired effluent quality.
5. Estimated costs for use of the RBC process in mining pollution
control and comparison with costs of other Fe(II) oxidation
systems.
6. Required processes to be coupled with the RBC to provide a total
package for treatment of mine drainage to a suitable quality for
stream-release.
-------
SECTION 2
SUMMARY AND CONCLUSIONS
Application of the rotating biological contactor (RBC) wastewater
treatment system to Fe(II) oxidation in acid mine drainage was examined in
the field and laboratory. Two sites in Pennsylvania and one in West Virginia
provided six heterogeneous Fe(II)-containing acid waters for use in
experiments with pilot scale and prototype RBC units. Specific character-
istics examined in detail for their effect on biological treatment were pH,
temperature, and Fe(II) concentration. Bench-scale investigations were
performed with synthetic and supplemented natural mine drainage in order to
study pH and the effect of constituents in mine waters other than Fe(II) and
basal salts. Costs of RBC treatment were estimated and compared to costs of
other Fe(II) oxidation systems. Neither the RBC nor other oxidation systems
were considered to provide complete treatment of acid mine drainage.
Neutralization of RBC effluent and separation of precipitated iron solids
would be necessary in order to produce a water suitable for stream-release.
Within the limits of the experimental conditions set forth in this
study, the following conclusions were drawn:
1. Under all conditions, continuous microbiological oxidation of
Fe(II) to less soluble Fe(III) was accomplished.
2. Mine water temperatures as low as 0.4°C did not inhibit Fe(II)
oxidation in field RBC units; however, oxidation efficiency was
reduced 10 to 20 percent in comparison to performance at 10°C.
The effect of temperature was less pronounced for the 2.0-m RBC
than for the smaller (0.5-m) pilot unit.
3. The larger field RBC unit produced an average 10 percent lower
Fe(II) oxidation efficiency than the 0.5-m unit under equivalent
operating conditions. The observed decrease in oxidation for
scale-up to the 2.0-m RBC was due, in part, to nonmicrobiological
factors, which included the effect of an increased effective
surface area for the 0.5-m RBC due to biological activity on the
trough surfaces exposed to mine water, the longer residence time
of mine water in the 0.5-m RBC trough (54.4 min) as compared to
the larger unit (48.0 min), and an increased short-circuiting of
flow through stages of the 2.0-m unit as determined by addition
of a tracer (lithium) to RBC influents.
-------
4. No relationship was observed between Fe(II) oxidation efficiency
and influent pH over the ranges examined in field experiments.
Acid adjustments of influent mine waters of pH 4.05-5.74 to pH 3
did not result in improvement in maximum Fe(II) oxidation in
bench-scale RBC units. The time to obtain an equilibrium
oxidation efficiency increased by approximately five days for the
RBC which received mine water adjusted to pH 3 as compared to the
response on unadjusted mine water.
5. Lower influent Fe(II) in Hollywood mine drainage (71.6 mg/£)
resulted in a higher oxidation efficiency at equilibrium for
the 2.0-m prototype RBC (95.7 percent) as compared to treatment
efficiency when mine water contained about twice the influent
Fe(II) content (90.5 percent). A similar but less pronounced
observation was made for the smaller field unit (2.9 percent
change). Oxidation efficiency exceeded 96 percent when the
0.5-m RBC treated the brine from the Crown reverse osmosis unit
that contained between 356 and 453 mg/£ of ferrous iron. These
higher influent Fe(II) concentrations did not correlate (5-percent
level) with oxidation efficiency.
6. Amendment of synthetic mine water feed with natural mine drainage
resulted in a net incfease in rate of Fe(II) oxidation in the
10-cm bench-scale RBC.
7. Under similar operating conditions, Fe(II) oxidation efficiencies
in treatment of Hawk Run mine waters with the RBC field units
were 10 to 15 percent lower than treatment efficiencies observed
with Hollywood mine drainage. A gelatinous outer layer on RBC
discs seen during treatment of Hollywood mine drainage was not
observed during experiments at Hawk Run. The oxidation efficiency
for treatment of Hawk Run Mine Hole 1 discharge was increased to
those levels observed at other locations by lowering of hydraulic
loading to both RBC units from 0.16 to 0.08 m3/d-m2. The
gelatinous layer did not form on RBC discs following the modifi-
cation.
8. Disc solids, in treatment of mine waters at all field locations,
contained iron (25 to 50 percent dry weight), a small proportion
of aluminum (0.05 to 0.1 percent dry weight), and viable iron-
oxidizing bacteria. Heterotrophic bacteria were recovered from
disc surfaces of the 0.5-m RBC during treatment of Hollywood mine
drainage.
9. Costs for Fe(II) oxidation with the RBC in treatment of 500 and
5,000 m^/d of mine drainage were estimated to be 5.6 and 4.8
cents per cubic meter treated, respectively. For both design
flows, amortized capital costs for the RBC method were estimated
to be about twice those of the conventional oxidation process.
Operating costs for the RBC, however, were almost one-half those
estimated for chemical oxidation due to,lower electrical,
maintenance, and personnel requirements.
-------
10. Sizing the RBC for mine drainage treatment may be accomplished
after determination of (1) mine water Fe(II), pH, and temperature,
(2) desired effluent Fe(II)", and (3) expected mine water flow.
The oxidation rate constant (k, min~l) for a mine drainage
closest in quality to one of the waters examined in this study
would be inserted into equation 6.1 (see Section 6, page ^9) to
determine the required detention time (t, min). The required
surface area would then be determined based on flow and RBC
surface-to-volume ratio (available from manufacturer).
-------
SECTION 3
RECOMMENDATIONS
The biological oxidation of Fe(II) in acid mine drainage treatment was
accomplished using a commercial rotating biological contactor. Other
devices are available for biological treatment which rely on the development
of microbial films. Any of these devices may potentially perform equally as
well as the RBC and could be more cost-effective for a particular application.
Regardless of the method used to complete biological oxidation of
Fe(II), neutralization and solids separation of the effluent must follow.
The choice of a neutralization system in conjunction with the treatment of
the effluent discharged from the biological oxidation process may be expected
to bear technically and economically on the total mine drainage treatment
system.
Suggested general areas warranting further investigation are:
1. Evaluation of other fixed-film biological treatment devices which
operate on the same principle as the RBC and may be more cost-
effective for a particular situation. An example of such a unit
is the biological tower (packed bed column).
2. Examination of a full-scale mine drainage treatment plant
comprising rotating discs for biological oxidation of Fe(II)
and alternative methods of acid neutralization.
Specific aspects of the research which did not provide detailed
information and which deserve future study are:
1. Kinetic characterization of the RBC at extremely high influent
Fe(II) concentrations (above 500 mg/fi.).
2. Interrelationships between iron-oxidizing and heterotrophic
microorganisms in the treatment process.
3. Additional detail on the physical and chemical make-up of solids
formed on RBC discs.
4. Possible inhibition in Fe(II) oxidation by various types and
concentrations of heavy metals and organic molecules.
-------
SECTION 4
ACID MINE DRAINAGE POLLUTION
INTRODUCTION
The acid mine drainage problem is not new to the coal industry and
is not restricted to the mining of coal. Acid discharges to streams have
occurred long before the mining of coal. In fact, one method of locating
coal seams during the early days of mining (1800* s) involved the discovery
of acidic, iron-bearing streams (2) .
The production of acidity and dissolved contaminants in mine waters
occurs through a complex chemical-biological process in which sulfide
minerals, particularly pyrites (FeS2) , undergo an oxidation-reduction
reaction as illustrated in Equation (4.1). Pyrite occurs naturally in
association with many coal seams. A second group of reactions may occur
spontaneously if the pH approaches neutrality [Equations (4.2) and (4.3)].
At distinctly acid pH, iron-oxidizing bacteria may mediate Equation (4.2).
The Fe3+ generated may then chemically oxidize pyrite according to the
stoichiometric reaction illustrated in Equation (4.4). Additional Fe2+ is
formed which may re-enter the cycle via Equation (4.2).
2+ 2~ +
FeS2 (s) + 02 + H20 - Fe + 2S03~ + 2H (4.1)
Fe2+ + -| 02 + H+ = Fe3+ + | H20 (4.2)
Fe3+ + 3 H20 = Fe(OH)3 (s) + 3H+ (4.3)
(s) + 14 Fe3+ + 8H0 = 15 Fe2+ + 2S02~ + 16H+ (4.4)
These mechanisms, acting independently or in combination, assist in
formation of a highly mineralized, acid water. Mining enhances the
production of mine drainage by exposing the sulfide materials to oxygen and
water,' thereby, permitting chemical reactions to proceed much more rapidly.
Several publications may be consulted for a more complete discussion of the
complex reactions which may occur in acid mine drainage, mechanisms believed
to be important in mine drainage formation, nature and extent of the problems
in mining regions across the nation and the world, and various control
measures which may be employed to aid in prevention of mine drainage forma-
tion (2, 10, 14, 32, 33, 39, 40, 49, 58, 61).
-------
TREATMENT OF ACID MINE DRAINAGE
Variations in mine drainage characteristics often exist in relation to
geology of mining regions. Consequently, several treatment methods may be
required. The method chosen for a particular discharge will depend on the
quality and quantity of the mine drainage and the ultimate use of the water.
Treatment in certain instances may be practiced for potable water production
where no other suitable municipal or industrial supply exists (36, 71, 75).
In order to._mandate improvement and maintenance of stream water quality,
Water Pollution Control Act Amendments of 1972, P.L. 92-500, provides for
regulation of the concentration of certain elements in mine discharges (55).
Current discharge guidelines were published in the Federal Register on April
26, 1977 (67). A maximum daily effluent concentration of 7.0 mg/£ total iron
was adopted with the restriction that a 30-continuous day average iron
concentration may not exceed 3.5 mg/£. The term "acid or ferruginous mine
drainage", as defined in the Federal Register, means "any water drained,
pumped or siphoned from a coal mine which before any treatment either has a
pH of less than 6.0 or a total iron concentration of more than 10 mg/£." In
addition to regulations on total and dissolved iron content, neutralization
of the acid is required to meet pH discharge requirements. This regulation
may be met by addition of any of several neutralizing chemicals (47). Other
limitations currently imposed include those for manganese and total suspended
solids. The reader is referred to a report by the U.S. Bureau of Mines (66)
for a more complete discussion of current regulations under P.L. 92-500 and
their impact on the mining industry.
CHEMISTRY OF IRON IN MINE DRAINAGE
There are several forms of dissolved iron and these may be divided into
two general categories based on oxidation state, namely, Fe(II) and Fe(III).
Dissolved iron may exist in the free, dissociated state (Fe2+, Fe3+) or may
be complexed with other elements (Table 1). Iron content in mine drainage is
usually removed from solution by oxidation of soluble Fe (II) to the ferric
state [Fe(III)] with corresponding neutralization of the water wherein Fe(III)
becomes highly insoluble and will precipitate from solution as pure or impure
Fe(III) oxyhydroxides (Figure 1). This precipitation of ferric hydroxide
will occur above pH A. Insoluble Fe(II) hydroxides, on the other hand, do
not form to any significant degree unless the pH is elevated above 9
(Figure 2) which, together with the low density of the precipitates,
contributes to the difficulty in removing the Fe(II) hydroxides from
effluents by plain sedimentation (33).
Soluble Fe(II) - OH complexes are insignificant in ferruginous acid
waters (Figure 2); however, Fe(III) - OH complexes are important in acid
mine drainage and their presence accounts for the higher solubility of
Fe(III) than would be expected when only Fe^+ is considered (Figure 1).
Equilibrium relationships and calculations employed in constructing
solubility diagrams (Figures 1 and 2) are given in Appendix A. Polynuclear
complexes of iron are not considered in calculations and may be important in
iron solubility. Most acid mine drainage waters contain appreciable concen-
tration of sulfate (804). Chemical equilibria calculations indicate that
8
-------
TABLE 1. EXAMPLES OF FORMS OF DISSOLVED IRON IN MINE DRAINAGE
Oxidation state Free and complexed forms
Fe(II) Fe2+
FeOH+
Fe(OH)2 (aq)
Fe(OH)~
J
FeSO° (aq)
Fe(III) Fe3+
Fe(OH)3
Fe(OH)~
FeSO+
iron-sulfate complexes (FeSO* and FeSO* ) are also important in acid mine
drainage. These forms contribute to further increases in soluble Fe(II) and
Fe(III) above the level expected with free and hydroxy forms of iron
(Figures 1 and 2). Therefore, large concentrations of soluble Fe(II) and
Fe(III) may occur in acid mine drainage and their removal is facilitated by
Fe(II) oxidation and subsequent hydrolysis to Fe(III) oxyhydroxides at the
near neutral pH values required for stream-release.
OXIDATION PROCESSES FOR Fe(II)
Potential methods for oxidation of Fe(II) in mine drainage treatment
may be broadly divided into chemical and biochemical classes with further
subdivision of the chemical category to include processing using atmospheric
oxygen, commercial oxidants, and electrochemical oxidation (Table 2).
Purely chemical Fe(II) oxidation with atmospheric oxygen as the sole
oxidant is generally employed in contemporary mine drainage treatment plants.
This method relies on the more rapid oxidation rates which occur in neutral
and alkaline pH ranges (Figure 3). At lower pH values, the rate is
-------
Q)
O
6
§
o>
-10-
1
e
10
Figure 1. Solubility of pure Fe(OH).j in the presence of free
Fe3+, Fe(III) - OH, and Fe(III) - S04 complexes.
10
-------
Figure 2. Solubility of pure Fe(OH)o in the presence of free
Fe2+, Fe(II) - OH, and Fe(II) - 804 complexes.
11
-------
TABLE 2. POTENTIAL METHODS FOR Fe(II) OXIDATION
IN ACID MINE DRAINAGE TREATMENT
Category
Process
References
Chemical (air)
oxidation
Mechanical aerators
Diffused aerators
Cascade aerators
Possibly activated carbons (27)
(35, 47, 72, 73, 74)
Commercial oxidants
Hydrogen Peroxide
Ozone
Chlorine
Potassium Permanganate
(11, 17, 31, 60)
Electrochemical
oxidation
Carbon electrodes
(28, 37)
Biological oxidation
Activated sludge
Trickling filters
Rotating biological
contactors
Possibly activated
carbon
(29, 46, 48, 54, 65, 70)
(27)
independent of pH and Fe(II) oxidation proceeds very slowly. For instance,
Singer and Stumm (62) indicate half times on the order of 1,000 days for
purely chemical Fe(II) oxidation at pH 3. Therefore, prior to the chemical
Fe(II) oxidation step, mine drainage pH must be elevated to above pH 7 where
Fe(II) oxidation can be accomplished by conventional diffused, mechanical,
or cascade aeration. The neutralizing chemical must be able to elevate the
pH to these levels and lime [CaO or Ca(OH)2] is generally employed for this
purpose. Limestone (CaC03), which has been reported to have certain
advantages over lime (47), cannot elevate the pH to levels high enough to
stimulate rapid chemical Fe(II) oxidation (72). Wilmoth (73) investigated a
combination of lime and limestone for neutralization and found a potential
for cost savings. Experiments on use of limestone alone, however, were not
feasible with mine drainages containing high Fe(II) concentrations (35, 47,
72).
12
-------
o
TJ
o»
o
3.0
+2.0
+ 1.0
0.0
-1.0
1 T
-2.0
-3.0
-4.0-
-5.0-
„ -d log CFedDl
k = dt
P0z = 0.20 atm.
Temp. 25°C
Figure 3. Chemical oxidation rate of Fe(II) as a function of pH (62)
13
-------
The alternative processes for Fe(II) oxidation listed in Table 2 may
take place under acid conditions, thus allowing the use of limestone as the
sole neutralizing agent. This reversal of plant unit processes (Figure 4)
may be advantageous in process control and also provide chemical savings.
In chemical Fe(II) oxidation, pH must be elevated to levels higher than
minimum discharge requirements (pH 6) in order to provide rapid oxidation
and combat post-oxidation acid release. This practice necessitates an
increase in chemical costs.
Commercial oxidants and electrochemical systems have been investigated
as alternative processes for Fe(II) oxidation. In general, either high
electrical operating costs (electrochemical oxidation) or high chemical costs
(commercial oxidants) have restricted use of these processes for most
applications.
Jasinski and Gaines (37) developed an electrochemical process to
oxidize Fe(II) in acid waters. The high acid and ionic strength of mine
drainage allowed use of the ionic conductivity to promote oxidation of iron
on carbon anodes. Type 316 stainless steel was used as the cathode and
application of 0.8 volts oxidized 95 percent of the initial Fe(II) in a batch
synthetic mine drainage solution. In addition to providing the desired
oxidation, electrolytic hydrogen was produced as a by-product. The process
was deemed feasible for practical application and cost estimates appeared
attractive. However, all costs were estimated based on laboratory experi-
ments with no supporting field data. Franco and Balouskus (28) further
evaluated electrochemical processes at laboratory and pilot scale using
natural mine drainage. An 18.9 liter/min pilot plant was operated with mine
drainage feeds containing 40 and 250 mg/£ Fe(II) at pH 2 and 5. At the lower
pH approximately 86 percent of influent Fe(II) was oxidized. At pH 5,
however, the electrodes coated with Fe(III) precipitates which resulted in
decreased Fe(II) oxidation efficiency to 55 to 78 percent. Although the
process appeared technically feasible for low pH mine drainage, capital and
operating costs for oxidation of 700 m^/d of mine drainage containing
250 mg/£ Fe(II) were estimated to be higher than those for conventional air
oxidation by factors of 5.0 and 1.7, respectively.
In 1942, Hann (31) considered the use of ozone as a means of oxidizing
Fe(II) in near-potable quality water. Beller, Waide, and Steinberg (11)
conducted a conceptual engineering design and economic study of the use of
ozone in mine drainage treatment. The study involved analysis of available
methods of ozone production and identified process configurations with lime-
stone as the final neutralizing chemical which appeared to provide an
attractive mine drainage treatment method. Conclusions of the.report,
however, were not verified by actual laboratory or pilot plant investiga-
tions. Simpson and Rozelle (60) reported preliminary results of laboratory
and field studies on use of ozone for Fe(II) oxidation. Ozone was capable
of providing appreciable Fe(II) oxidation (greater than 90 percent) although
somewhat less oxidation occurred in natural mine drainage as compared to pure
FeSO^ solutions and this observation was attributed to oxidation of
manganese (II). Recently, Cole et al. (17) investigated the use of hydrogen
peroxide with some success to improve the efficiency of a cascade aerator at
an existing mine drainage treatment plant. Such application of an auxiliary
14
-------
INFLUENT
ACID MINE DRAINAGE
Fe(H) OXIDATION
COMMERCIAL OXIDANTS
ELECTROCHEMICAL OXIDATION
BIOLOGICAL OXIDATION
NEUTRALIZATION
LIMESTONE ONLY
LIME ONLY
CLARIFICATION
SLUDGE HANDLING
T
TREATED
EFFLUENT
NEUTRALIZATION
LIME ONLY
LIMESTONE AND LIME
Fe(H) OXIDATION
CHEMICAL OXYGENATION
COMMERCIAL OXIDANTS
CLARIFICATION
SLUDGE HANDLING
TREATED
EFFLUENT
Figure 4. Location of alternative Fe(II) oxidation process
and neutralization options in two acid mine
drainage treatment schemes.
15
-------
oxidant may prove cost-effective by eliminating the need for treatment plant
expansion in a manner an .lagous to polymer use to improve clarification
processes.
BIOLOGICAL Fe(II) OXIDATION
In the preceeding discussion on chemical Fe(II) oxidation and its
dependency upon pH, the term "purely chemical" was repeatedly stated. Its
use was to clearly distinguish chemical activity from the interaction of
microorganisms which influence the overall rate of transformation of Fe(II)
to Fe(III). The oxidation of Fe(ll) by iron-oxidizing bacteria may be
illustrated by Equation (4.5).
Fe2+ = Fe3+ + e~ + 11.5 kcal (4.5)
The liberated electron can be used by the iron-oxidizing bacterium,
Thiobacillus ferrooxidans, as the sole energy source. The energy is
indirectly utilized for ultimate reduction of C02 into new cell material.
Organisms which fix C02 as a primary carbon source and utilize inorganic
chemicals or radiant energy are called autotrophs. This terminology allows
differentiation from heterotrophic organisms which require preformed
organic carbon for implementation of vital processes. Organisms which gain
major energy from mineral sources as do the iron-oxidizing bacteria, are
termed chemolithotrophs.
The iron-oxidizing bacteria related to T_. ferrooxidans are ecologically
compatible and indigenous to acid waters characterized by low pH and high
oxidation-reduction potential (Eh). It is evident that these bacteria
perform a substantial contribution to the production of acid mine drainage
constituents by completing the cycle in the chemistry of mine drainage.
Other microorganisms have been reported to exist in acid drainage and mine
water polluted streams, e.g., gram-positive and gram-negative heterotrophic
bacteria, algae, yeast, protozoa, and fungi (19, 23, 43).
BIOLOGICAL TREATMENT OF ACID MINE DRAINAGE
Initial attempts to utilize iron-oxidizing bacteria in mine drainage
treatment were patterned after the conventional activated sludge sewage
treatment process (29).
Glover (29) reported limited success with pilot scale biochemical
Fe(II) oxidation. Nutrient supplemented acid mine drainage was fed in
series to three, 140-X. aerated vessels. The continuous flow system required
about 80 days to establish an equilibrium oxidation rate of 83 mg/£ Fe(II)
per hour at a mine water temperature of 15 to 20°C. Attempts to improve the
oxidation rate involved recycling combined bacterial-Fe(III) solids to
maintain a high iron-oxidizing bacterial population in the aeration chambers.
The overall activity was increased ten-fold by sludge recycle although few
viable microorganisms were observed in the return solids. The oxidation
rate was found to decrease ten-fold at freezing mine water temperatures.
Glover considered these rates of Fe(II) oxidation to be sufficiently high to
16
-------
warrant construction of a pilot plant to include neutralization using
limestone. The plant was operated for 28 months and major problems were
encountered with biological sludge deposition in reactors and piping. The
main advantage cited for the process was the ten-fold reduction in sludge
volume produced in comparison to the conventional lime process. Extremes in
temperature, acidity and Mn(II) concentration were suggested as potential
limitations in use of the process for acid mine drainage treatment.
Whitesell, Huddleston, and Allred (70) studied microbiologically
mediated acid mine water treatment at laboratory and pilot scale and failed
to achieve a rate of Fe(II) oxidation in 3,800 m3 reactors similar to that
observed in the laboratory. The reduction in surface-to-volume ratio
through scale-up of the process was thought to be responsible for inability
to achieve laboratory results in the field.
Lovell (47) compared a mechanically aerated activated sludge type
system to a trickling filter for microbial Fe(II) oxidation. The fixed-film
biological oxidation approach utilized in the trickling filter proved far
superior in performance to the suspended growth system. Oxidation perfor-
mance for the 10.7-m diameter trickling filter ranged from 60 to 94 percent
Fe(II) oxidized during the six months in which continuous operation was
reported. By comparison, the deep tank system was observed to be more
sensitive to changes in Fe(II) concentration and flow rates although similar
efficiencies were experienced by use of longer liquid retention times.
During winter operation of the trickling filter, influent mine water froze
at the surface of the reactor, preventing flow through the media. A plastic
film placed over the reactor permitted continued operation in winter.
Following one year of operation, hydraulic flow rates were reduced somewhat
by Fe(III) deposits in the voids of the reactor media. The control and
removal of these deposits were considered essential and Lovell suggested the
need for further study. Additional experiments by Lovell (48) revealed that
a polypropylene filter medium eliminated media disintegration problems and
substantially reduced the problem of clogging.
Unz and Lieberman (65) studied the microbial populations and transfor-
mations of mine drainage percolated through columns of trickling filter
media and postulated that, under suitable conditions, a fixed-film biological
method was feasible for Fe(II) oxidation in acid mine drainage treatment.
Baskets containing columns of either argillite stone or polypropylene filter
media were inserted into cased cores in the prototype trickling filter
previously examined by Lovell (47). Mine water was allowed to percolate
through 3Q- and 60-cm columns of trickling filter media. The polypropylene
filter medium was found to be superior to argillite stone under all condi-
tions. At a hydraulic loading of 0.05 m3/d«m2, oxidation efficiencies ranged
from 87 to 96 percent for percolation through a 30-cm column of polypropylene
media. Agrillite stone was observed to fracture and disintegrate following
contact with mine water, apparently due to its inability to withstand the
acidic nature of the drainage. Under the conditions encountered, it was
concluded that Fe(II) oxidation in the columns was microbially mediated.
Ferruginous solids formed on the filter media following continuous applica-
tion of mine water contained gelatinous organic matter with embedded iron-
oxidizing and heterotrophic bacteria. It was concluded that further
17
-------
microbiological and engineering research was desirable to develop and
optimize the method, and it was suggested that other biological treatment
devices may be more effective in this regard.
Olem and Unz (54) conducted small pilot scale experiments with the RBC
wastewater treatment device to provide optimized fixed-film biological Fe(II)
oxidation in acid mine drainage treatment (see Section 5).
18
-------
SECTION 5
ROTATING BIOLOGICAL CONTACTOR TREATMENT SYSTEM
PRINCIPLES
There are many types of fixed-film biological treatment methods that
have been applied to the oxidation of biodegradable pollutants. These
processes include trickling filtration, contact aeration, activated bio-
filtration, biological towers, and other modifications of basic systems
which rely on the attachment of microorganisms to an inert surface which in
some way contacts the wastewater to be treated. Grieves (30) presented a
detailed review of the history and application of fixed-film biological
reactors for wastewater treatment. The RBC process will be discussed
specifically in the following passage owing to its application in this study
for oxidation of Fe(II) in mine drainage.
The RBC is a wastewater treatment system that provides for oxidation
of biodegradable pollutants principally through aerobic microbiological
action, although in treatment of organic wastes anaerobic transformations may
well occur at the base of films. The device consists of a series of large
diameter plastic discs which are mounted on a horizontal shaft and suspended
in a trough through which the wastewater flows (Figures 5 and 6). The discs
are approximately 40 percent immersed in the wastewater and slowly rotated.
Following start-up, microorganisms naturally present in the incoming waste-
water colonize the surfaces of the discs. Rotation provides intermittent
contact with the wastewater for adsorption and assimilation of constituents
and exposure to an atmospheric oxygen supply necessary for microbial
respiration. The thin film of wastewater which adheres to the discs trickles
down the surfaces and absorbs oxygen from the air. Rotation also provides
continuous mixing of the wastewater and allows shearing of excess biomass
from the discs which remain in suspension and pass through the system with
the flow of wastewater. The biological film which develops on the inert
plastic surfaces has been reported to be normally 1 to 4 mm in thickness for
most domestic and industrial wastewater treatment applications (4).
HISTORY
According to Antonie (4), the RBC was conceptualized by Weigand in
1900 in Germany as a sewage treatment process for removal of biodegradable
organic matter. A German patent was issued to Weigand for a device which
consisted of a cylinder of wooden slats. A prototype was not developed
until the 1930's and experiments at this time were discontinued when the
19
-------
FEED BUCKET
DRIVE SYSTEM
FEED
CHAMBER
INFLUENT
STAGES OF DISCS
EFFLUENT
Figure 5. Schematic of rotating biological contactor package unit
showing disc assembly suspended in contoured tank.
Figure 6. Large-scale rotating biological contactor installation
showing disc assemblies suspended in below-ground
concrete basins (courtesy of Autotrol Corporation).
20
-------
unit experienced clogging problems. Doman (20) reported on the development
of a rotating metal plate filter in the United States in 1929. In the same
year, Allen (1) described the "biologic wheel" which consisted of rotating
paddle wheels. Results were not encouraging on further process developments
and no further work was reported in the popular technical journals of the
time. In 1956, Morgan et al. (53) began pilot plant studies of the "Gresham
biological filter", which consisted of rotating screen-wound reels or plywood
discs. Results were encouraging in terms of organic matter removals, but
construction materials available at that time were not adequate and no
further developmental work was reported.
At approximately the same time work was being conducted in the U.S.,
Hartman and PBpel at The Technical University of Stuttgart, West Germany,
carried out experiments on the use of plastic discs for biological oxidation.
Further development of the process using expanded-polystyrene discs resulted
in commercial application in Europe. In 1957, the J. Conrad Stengelin
Company of Tuttlingen, West Germany manufactured the first commercial 2-m and
3-m diameter polystyrene disc units for wastewater treatment. The process
received relatively wide use but was mostly restricted to small installations
(less than 1,000 population equivalent).
In the U.S., efforts toward commercial development of the RBC system
for wastewater treatment began in the mid-1960fs at Allis-Chalmers of
Milwaukee, Wisconsin. In 1969, the first U.S. installations began operation
at a small cheese factory and at a municipal sewage treatment plant at
Pewaukee, Wisconsin.
Following purchase of the "bio-disc" operation from Allis-Chalmers,
Autotrol Corporation, also of Milwaukee, Wisconsin, replaced the expanded-
polystyrene disc system with corrugated sheets of high-density polyethylene.
This modification increased surface area density from 53 m2/m^ to 120 m2/m3.
Further increases in disc diameter and shaft length to 3.6 and 7.6 m,
respectively, provided up to 9,290 m2 of available surface area per unit.
Autotrol branded their equipment "bio-surf" and recently other manufacturers
have begun production of similar units under trade names such as "RBS"
(rotating biological surface) and "bio-drum". The term rotating biological
contactor or RBC has gained general acceptance as a generic term for
rotating fixed-film biological reactors and will be used in subsequent
sections.
PROCESS OPERATION
Wastewater, pretreated if necessary by either primary clarification,
nutrient addition, or pH adjustment, is fed on a once-through basis to the
RBC system for biological oxidation of waste matters. Recycling of RBC
effluent was examined in early stages of development; however, results
indicated that this option provided insufficient increase in treatment
efficiency to justify the increased piping and pumping costs (4).
Within the RBC trough, wastewater passes through several stages of
discs which are separated by baffles. Experience has indicated that staged
operation improves residence time distribution and more closely approximates
21
-------
"plug flow" conditions (4). Staging is particularly important when it is
desired to achieve both organic matter removal and ammonia nitrogen oxidation
(nitrification). Fixed microbial cultures which degrade organic matter
develop on initial stages of discs while nitrifying bacteria actively occur
in the latter stages (4).
Rotational speed of the discs is an important consideration for a
variety of reasons. An increase in rotational disc velocity improves the
rate of aeration, increases the degree of contact between the microbial film
and the wastewater, and more completely mixes the contents of each stage.
Antonie and Koehler (7) suggested that peripheral disc velocity should be the
design parameter of choice when it was discovered that different diameter RBC
units provided similar results when operated at equal peripheral velocities
even though the revolutions per minute (rpm) were different. Recently,
however, Chesner and Molof (15) reported that the peripheral disc velocities
used in pilot studies may produce unexpected results when applied in scale-up
due to the high wastewater dissolved oxygen (DO) which may be generated in
troughs of smaller diameter pilot units because of the higher rpm. Full size
units normally rotate at 1 to 2 rpm. Higher rotational speeds may increase
treatment capacity but the accompanying logarithmic increase in power costs
have been shown to offset the improved performance capability in many
instances (4).
Because solids are continually sheared from disc surfaces at
equilibrium conditions, clarification of RBC effluent is generally required.
Treatment of certain industrial wastes, however, results in low production
of solids and these effluents do not require final clarification. In other
instances, intermediate clarification may be required in order to remove
excess solids having an oxygen demand which cannot be met by the agitation
and aeration provided by the rotating discs.
The major parameter used for RBC design has been hydraulic loading,
measured as cubic meters per day of wastewater flow per square meter of disc
surface area (m-Vd-m^). This parameter does not aid in sizing the volume of
the trough. Antonie (4) found that decreases in the ratio of disc surface
area-to-trough volume improved performance by increasing liquid retention
time. It was recommended that a surface-to-volume ratio of about 200 wr/m
be employed in construction of RBC units. For units which have an equal
surface-to-volume ratio, changes in hydraulic loading will directly
affect liquid retention time. Increases in hydraulic loading will decrease
retention time and vice-versa. These factors affect the probability of t
contact of microorganisms on disc surfaces with oxidizable matters in the
wastewater flowing through the trough.
Because the RBC process relies on active microbial cultures, wastewater
temperature has been shown to directly affect treatment performance (24).
The degree to which efficiency is affected is a function of the nature of
the resident microbial populations and the reaction of interest. Sharma and
Ahlert (57), in a review of biological nitrification, revealed evidence of
operational problems in wastewater nitrification systems at temperatures
below 8.3°C.
22
-------
Potentially toxic materials in the incoming wastewater may also
decrease treatment capacity by affecting the fixed biological population.
Several researchers have noted increased ability of fixed-film biological
systems to withstand the effects of potentially toxic or shock loads in
comparison with suspended growth processes.
In order to protect the biological growth on the rotating surfaces
from freezing temperatures and excess wastewater heat loss, enclosures usually
have been employed for the RBC. The enclosure may be a molded plastic cover
(Figure 6) or a separate building or frame structure. Insulation alone has
been shown to be adequate to maintain air temperatures within the enclosure
equal to that of the wastewater.
APPLICATIONS
At present, there are about 1,000 RBC installations treating organic
wastewaters in Europe, located primarily in West Germany, France, and
Switzerland, and almost as many plants in the U.S. (A). Size of plants vary
from those treating the sewage from single-family dwellings to municipal
and industrial facilities with wastewater flows up to 200,000 m3/d (4). The
short-term large-scale application of the RBC in the U.S. has provided little
information on life of hardware and process stability. As more facilities
are constructed for different applications, operating experience will provide
additional information on the longevity characteristics of the RBC.
Pilot studies have been performed with the RBC to examine treatment
potential for organic wastes such as those from meat and other food process-
ing plants (13), pulp and paper manufacture (12), beverage and distillery
industries (41), dairy farms (5), refineries (18), and textile industries (4).
The process has been applied to pretreatment of industrial wastewaters prior
to entry into municipal sewers as well as treatment of combined domestic and
industrial wastewaters (4). Its use may be combined with other biological
treatment devices such as activated sludge processes or trickling filters.
In the latter case, RBC treatment is added in order to satisfy requirements
for upgrading existing wastewater treatment facilities through increased
organic removals or enhanced nitrification.
Application of the RBC to the myriad of industrial wastewaters requires
varied design considerations unique to the industry. Consideration for
design of industrial and domestic wastewater treatment facilities may include
requirements for primary treatment, secondary clarification, metal or
phosphorus removal, chlorination, sludge handling and disposal, and appli-
cable advanced wastewater treatment. Several mathematical models have been
developed recently (15, 16, 25, 26, 30) to aid in prediction of RBC
performance under varied conditions.
INORGANIC REACTIONS
Nitrification is included in domestic wastewater treatment in order
to reduce ammonia nitrogen concentrations in effluent discharges. Ammonia
causes an oxygen demand in the receiving stream, stimulates eutrophication
and has been shown to be toxic to many species of aquatic life. Nitrifica-
23
-------
tion may also be applied as a first step in a total nitrogen removal scheme
(nitrification-denitrification) (3). The RBC has been used for domestic
wastewater nitrification following other biological processes or as the sole
biological treatment method whereby nitrifying organisms develop on latter
stages of RBC disc surfaces (4).
Certain industrial wastes contain ammonia in excessive concentrations.
Nitrification has been required for wastewaters from such industries as latex
polymer production and electronics manufacture (4). Treatment of high
ammonia sludge supernatant has also been successfully accomplished with the
RBC (50).
The various biological reactions which have been discussed previously
are considered to proceed optimally at neutral or alkaline pH conditions.
Severe upsets in performance have been observed when pH values shift into
the acid range, in part, due to the growth of undesirable microbial species,
such as yeasts and fungi (4). Processes which require the oxidation of
sulfide, thiosulfate, and Fe(II), on the other hand, have been applied
successfully for treatment of low pH wastes from depilatory manufacture (4),
base metal mining (4), and coal mine drainage (54), respectively. The micro-
organisms which are operative in the oxidation of these inorganic ions
require a low pH environment. The oxidized species are either suitable for
effluent discharge (sulfate in the case of sulfide and thiosulfate oxidation)
or more easily handled in a separation step due to its lower solubility
[Fe(III) as described in Section 4].
Fe(II) OXIDATION - STUDIES AT HOLLYWOOD, PA.
Previous studies on biological Fe(II) oxidation were conducted with a
0.5-m diameter RBC pilot unit on a single source of mine water at the
Experimental Mine Drainage Treatment Facility at Hollywood, Pa. (54). The
mine drainage characteristics included a naturally stable pH of approximately
3, relatively constant temperature (9-12°C) and only slight (seasonal)
variations in flow and certain physical and chemical characteristics. These
experiments were mainly performed to examine the variables in RBC design
which may permit appreciable transformation of Fe(II) to Fe(III). Variables
examined included hydraulic loading and peripheral disc velocity.
A disc velocity of 19 m/min was observed to be more effective (above
90 percent Fe(II) oxidized) than 10 m/min. Higher disc velocities were not
examined in view of the very high DO levels maintained in the reactor, high
efficiency of Fe(II) oxidation observed, and a previously reported logarith-
mic increase in power consumption required for subsequent increases in disc
velocity.
Five different hydraulic loading rates were applied to the RBC pilot
unit (0.11, 0.16, 0.22, 0.31, and 0.44 m3/d-m2) and each successive increase
resulted in corresponding decreases in Fe(II) oxidation efficiency. Based on
these observations, it was concluded that by application of proper design
hydraulic loading, any degree of treatment down to 1 mg/£ effluent Fe(II)
could be achieved for Hollywood mine drainage.
24
-------
Kinetics of Fe(II) oxidation reaction were found to follow a concentra-
tion dependent, first order relationship at all hydraulic loadings and disc
rotation rates examined. Start-up characteristics and time to reach equili-
brium oxidation and solids buildup conditions were examined along with an
investigation of the chemical and physical characteristics of disc solids and
viable iron-oxidizing bacterial populations in disc solids. No other mine
drainages were investigated. It is important to note that the RBC and other
fixed-film reactors previously observed to be capable of providing appreciable
Fe(II) oxidation in field studies were examined only at the Hollywood
facility. These limitations in previous studies inspired the present work
reported in the remaining sections.
25
-------
SECTION 6
FIELD STUDIES: PENNSYLVANIA AND WEST VIRGINIA
LOCATION OF FIELD SITES
A continuous flow of mine water was maintained through RBC units in
experiments conducted during (a) May 22 to July 13, 1977 at the U.S.
Environmental Protection Agency Crown Field Site near Morgantown, W. Va.,
(b) September 8 to November 19, 1977 at the Commonwealth of Pennsylvania
Experimental Mine Drainage Treatment Facility, Hollywood, Pa., and (c)
January 6 to April 12, 1978 at the Commonwealth of Pennsylvania Hawk Run
Mine Water Treatment Plant near Philipsburg, Pa.
Crown, West Virginia
At the Crown Facility two acid waters were available for study. The
first water was raw acid mine drainage from an active coal mine pumped from
an 85-m deep borehole located along Stewart's Run. A portion of this mine
discharge was passed through an experimental reverse osmosis (RO) unit and
provided the second water. The RO treatment system was capable of producing
approximately 50-percent recovery of water of near potable quality. The
remaining portion was continuously rejected from the unit as a concentrate.
This "brine" contained approximately twice the original concentration of
ionic constituents, including Fe(II), and had a low pH value due to the
necessity of pretreatment with sulfuric acid (^SO^) in order to prevent iron
fouling of the RO membrane. These characteristics, along with temperature
changes through the RO unit, provided a highly polluted, unique quality mine
water for the RBC study.
Hollywood, Pennsylvania
The Proctor No. 2 coal mine at Hollywood, Pa. provided acid drainage
which was pumped a distance of 180 m from the discharge point to the RBC
building. This inactive mine produced a relatively constant flow of
approximately 3,000 m-Vd.
Hawk Run, Pennsylvania
The Hawk Run facility is located near two mine holes which produce
acid discharges. A third water resulted from the mixing of these two
discharges. Mine water from the two mine holes and the combined flow were
each examined separately. The drainage from Mine Hole 3 originated 3 km
from the plant site and flowed in a shallow, wide stream (Hawk Run) which
26
-------
was readily subject to change in water temperature with ambient conditions.
This mine hole provide^ nearly all the flow in Hawk Run. Mine Hole 1 is
located at the research facility and the discharge blended with Hawk Run at
the site. A schematic of the building which housed the RBC units and the
mine drainage streams is presented in Figure 7 and includes the locations at
which mine water was pumped to the building from each of the three sources.
MINE WATER CHARACTERISTICS
In order to more precisely interpret the findings of experiments on
Fe(ll) oxidation with the RBC, it was important to characterize the various
mine waters using parameters which are related to the process. Mine waters
were analyzed regularly (two to five times per week) during the respective
periods of experimentation. Analytical methods are outlined in Appendix B.
A summary of mine water characteristics for each of the locations is shown
in Tables 3, 4, and 5 and includes statistical analyses for mean and
standard deviation.
Crown, West Virginia
The Crown drainage had a higher pH (4.40 to 5.50) and temperature
(15.5 to 18.8°C) than other acid mine drainage waters analyzed. The higher
temperature may have been due to the effect of pressure on the groundwater
and the insulation of the soil at the depth of the borehole (85 m). An
extremely high iron-oxidizing bacterial population was observed for the
Crown drainage. Wilmoth (73) conducted a more complete chemical analysis of
the Crown mine drainage and noted a relatively high sodium ion content
(280-670 mg/£). The unusual characteristics of higher pH and sodium content
than experienced at other mine drainage locations suggests that the active
mine drains through or mixes with a high ionic strength aquifer which is
possibly highly buffered and nonacid. A saline (NaCl type) aquifer is not
likely because of the very low chloride content (17 mg/£) of the drainage.
The unusual characteristics of the mine drainage at Crown, W. Va. provided a
high Fe(II)-containing water for experiments in RBC treatment.
The RO brine had a very low pH (2.18 to 2.66) and had about double the
ionic strength of the original mine water. It was evident from the low
Fe(III) content of the brine that Fe(II) oxidation could not occur during
passage of Crown drainage through the RO unit. Hence, treatment of the brine
required Fe(II) oxidation. At a 50-percent clean water recovery rate for
the RO process, one-half the original mine drainage flow would be rejected
as brine which must be then treated prior to stream-release. Clearly, if
the RO process becomes a feasible system for potable water production from
mine drainage, particularly, with improved clean water recovery rates, very
high Fe(II) concentrations present in the brine would require an economical
Fe(II) oxidation step in the treatment. It was of interest, therefore, to
determine the capability of the RBC for biological Fe(II) oxidation and
correspondingly to determine the microorganism survival through the high
pressures which developed in the RO modules (approximately 400,000 kg/m^).
Apparently many, but not all, iron-oxidizing bacteria survived the pressures
and were rejected along with the other constituents (Table 3). An analysis
27
-------
DIRECTION
OF FLOW
RBC
BUILDING
INFLUENT
PIPING
Figure 7. Schematic of mine drainage streams available at Hawk
Fa. showing locations of mine water feed lines.
28
-------
TABLE 3. CHARACTERISTICS OF ACID MINE DRAINAGES STUDIED
AT THE U.S. EPA CROWN FIELD SITE NEAR MORGANTOWN, W. VA.
Parameter
Temp., °C
pH
Fe(II), mg/Jl
Fe(III), mg/Jl
Al, mg/Jl
DO, mg/Jl
Total N, mg/£
Org-N
NH3-N
(N02 + N03)-N
Total P, mg/Jl
Iron-oxidizing
bacteria,
MPN/100 mJl
Crown
Mean
16.7 (32) c
4.83 (32)
195 (32)
7 (30)
9.9 (31)
2.4 (10)
2.44 (3)
0.76 (3)
1.83 (3)
0.03 (3)
0.41 (3)
45,000
drainage a
Std. dev.
0.7
—
28
16
4.4
1.1
1.12
0.81
0.38
0.02
0.34
17
RO
Mean
23.9 (10)
2.34 (10)
356 (10)
10 (9)
19.7 (9)
2.2 (6)
3.18 (6)
0.54 (6)
2.62 (6)
0.01 (6)
0.11 (6)
5,200
brineb
Std. dev.
2.1
—
47
16
12.5
0.8
1.11
0.53
0.66
0.02
0.12
2
Samples were collected from May 23 to July 2, 1977.
Samples were collected from July 2 to July 13, 1977.
cNumber of analyses applied to statistical determinations.
29
-------
TABLE 4. CHARACTERISTICS OF ACID MINE DRAINAGE STUDIED AT
THE EXPERIMENTAL MINE DRAINAGE TREATMENT FACILITY AT HOLLYWOOD, PA.
Parameter
Temp., °C
PH
Fe(II), mg/£
Fe(III), mg/£
Al, mg/£
DO, mg/£
Total N, mg/£
Org-N
NH3-N
(N02 + N03)-N
Total P, mg/£
Q
Proctor No. 2 mine drainage
Mean Std. dev.
10.4
2.74
146
121
57.5
2.2
1.21
0.14
1.03
0.04
0.05
(23)b
(23)
(23)
(13)
(13)
(20)
(15)
(15)
(15)
(15)
(15)
0.6
—
69
31
1.6
0.8
0.28
0.21
0.18
0.03
0.06
Iron-oxidizing
bacteria,
MPN/100 m£
Heterotrophic
bacteria0,
cells/100 rnH
3,400
610
Samples were collected from September 8 to November 19, 1977.
Number of analyses applied to statistical determinations.
"The term "heterotrophic bacteria" refers to organisms which were shown not
to be able to oxidize Fe(II); however, these organisms could not be
subcultured on organic media.
30
-------
TABLE 5. CHARACTERISTICS OF ACID MINE DRAINAGES STUDIED AT THE HAWK RUN
MINE WATER TREATMENT PLANT NEAR PHILPSBURG, PA.
Parameter
Temp., °C
PH
Fe(II), mg/4
Fe(III), mg/A
Al, mg/4
DO, mg/£
Total N, mg/£,
Org-N
NH3-N
(N02 + N03)-N
Total P, mg/£
Iron-oxidizing
bacteria,
MPN/100 m£
Mine Hole la
Mean Std. dev
10.0
3.77
62.8
6.9
6.2
2.2
1.04
0.37
0.58
0.09
0.06
(16)d
(16)
(16)
(10)
(12)
(14)
(6)
(6)
(6)
(6)
(6)
2,000
0.1
—
8.8
2.4
1.4
0.3
0.29
0.36
0.03
0.05
0.05
4
Hawk Runb
Mean Std. dev
4.6
3.15
32.8
11.4
10.1
10.8
0.96
0.37
0.31
0.27
0.06
440
(7)
(7)
(7)
(6)
(6)
(6)
(6)
(6)
(6)
(6)
(6)
3
4
5
2
1
0
0
0
0
.0
.6
—
.3
.6
.3
.7
.37
.34
.08
.05
.05
1
Hawk Run + MH 1°
Mean Std. dev.
8.2
3.42
53.2
9.6
8.6
7.4
0.55
0.01
0.42
0.13
0.10
630
(18)
(18)
(18)
(18)
(18)
(14)
(9)
(9)
(9)
(9)
(7)
I
2
2
0
0
0
0
0
0
0
.0
—
.0
.8
.4
.5
.67
.01
.04
.03
.17
4
Samples were collected from March 8 to April 12, 1978.
Samples were collected from February 23 to March 8, 1978.
c
Samples were collected from January 6 to February 23, 1978.
mimber of analyses applied to statistical determinations.
31
-------
of the product water indicated the presence, although low, of iron-oxidizing
bacteria (300 MPN/100 mi).
Hollywood, Pennsylvania
Results of analyses of Proctor No. 2 mine drainage (Table 4) in the
present study were very similar to those previously reported by Olem and Unz
(54) in earlier work with Proctor No. 2 mine water. Unlike the drainage at
the Crown Site, a large portion of the iron exists in the Hollywood drainage
as Fe(III).
Dissolved oxygen (DO) was measured by probe in the feed well of the RBC
units as described in Appendix B. Measurements taken from the point of mine
water discharge prior to pumping to the treatment units showed DO values less
than 0.5 mg/fc. This observation suggests that underground mine waters, such
as the Crown and Hollywood drainages, contain little DO prior to surface
discharge.
Hawk Run, Pennsylvania
The mine waters at the Hawk Run facility had a generally lower iron
content than the drainages at other locations. Mine Hole 1 water contained
iron mainly as Fe(II) and temperatures (9.6 to 10.3°C) were similar to
those reported for most underground mine waters just following surface
discharge (15). DO values less than 0.5 mg/£ were measured prior to pumping
as was the case of the other underground mine waters analyzed. Other than
the lower iron and aluminum content, Mine Hole 1 discharge appeared similar
to the Proctor No. 2 drainage including nutrient and iron-oxidizing bacteria
content.
Analysis of samples collected from Mine Hole 3 at the point of surface
discharge revealed chemical and microbiological characteristics similar to
those of Mine Hole 1. Apparently, dilution of drainage by surface runoff to
the stream and the precipitation of insoluble ferric iron (evident from
ferruginous deposits on stream bed) resulted in a lower iron content in the
water when collected and analyzed at the downstream location depicted in
Figure 7.
Total nitrogen content of the Hawk Run water was similar to the levels
present in Mine Hole 1 discharge; however, a greater percentage of nitrogen
was in the form of N02 or NO 3 in the Hawk Run water. Biological nitrifica-
tion was not likely due to the low pH of the environment (57). Therefore,
the small surface drainage most likely contributed NOo to Hawk Run. A few
localized septic tank tile field systems may have contributed a nitrified
effluent to the stream.
Exposure to ambient climatic conditions over the distance traveled by
Hawk Run from its origin at Mine Hole 3 to the point of pumping and subse-
quent analysis (3 km) resulted in changes in the water temperature, which
was initially consistently 10°C. During the period February 23 to March 8,
1978, continuous temperature monitoring revealed water feed temperatures to
the RBC units ranged from 0.4 to 9.8°C. At the lower temperatures, DO levels
32
-------
were 8.4 to 12.8 mg/£ due to natural aeration of the shallow, fast moving
stream. Hawk Run provided a Fe(II)-containing water which differed from the
other drainages examined in several of the characteristics listed in Table 5.
A mixture of Mine Hole 1 water and Hawk Run discharge was pumped to the
RBC building from a point sufficiently below the confluence to assure
homogenization of the sources. Experiments were conducted using this mixture
in order to study a "synthesized" natural mine water having characteristics
different from the individual waters. Influent characteristics were
determined as listed in Table 5. The flows of Mine Hole 1 and Hawk Run were
nearly equal (25,000 m^/d) and it was anticipated that each conservative
constituent of the mixture in any section of the stream under consideration
would be the mean of the concentrations of each constituent from the waters
prior to mixing. This mass balance seemed to hold true for most of the
parameters analyzed (Table 5), even though most are non-conservative.
Mass balance relationships were used to determine the extent of lateral
mixing which occurred in the 50-m section below the confluence of the two
streams. This information was needed to determine whether there was
sufficient mixing in the reach so that the influent mixture pumped to the
RBC units would be significantly different in quality from the individual
waters. Experimental procedures and results of the study are presented in
Appendix C. The experiment also was designed to determine whether backmixing
of Mine Hole 1 discharge affected the quality at the point where mine drainage
was being pumped from Hawk Run. Results of this work indicated that back-
mixing was not occurring at this point and that the location of the feed line
provided water representative of Hawk Run. Although the intake point in the
combined stream (Hawk Run + Mine Hole 1) was not located far enough down-
stream from the confluence to provide complete mixing, sufficient mixing did
occur such that mine drainage influent differed in quality from either of
the contributing flows.
EXPERIMENTAL APPARATUS AND PROCEDURES
Rotating Biological Contactors
Field studies were conducted with commercial RBC pilot units (Autotrol
Corporation, Milwaukee, Wisconsin) equipped with 0.5-m and 2.0-m diameter
discs (Figure 8). The 2.0-m RBC unit was not used in experiments at Crown,
W. Va. because of shipment delays encountered. Each unit consisted of four
stages of closely spaced and corrugated high-density polyethylene discs
suspended on a shaft in a corrosion-proof trough. The troughs of the 0.5-
and 2.0-m units were constructed of epoxy-coated aluminum and bitumastic-
coated steel, respectively. The individual stages were separated by baffles.
Flow through the stages of the 0.5-m unit was facilitated by a 2.5-cm
diameter hole in each baffle. A serpentine flow pattern was formed in the
2.0-m unit by an arrangement of piping on the outside of stage compartments.
A rotating bucket mechanism fed mine water from the influent chamber of each
unit to stage compartments and thus controlled flow. Detailed information on
the characteristics and design parameters of the experimental RBC units may
be found in Appendix D.
33
-------
Figure 8. RBC units with disc diameters of 2.0 m (top) and
0.5 m (bottom) employed in field experiments.
34
-------
Other Equipment
At each site, the RBC units were enclosed in a building for protection
from the general environment. At the Crown Field site, mine water was
continuously pumped to the pilot plant building as described by Wilmoth (73).
Mine drainage flowed continuously during the respective study periods except
for two occasions at Crown, W. Va. in which borehole pumping was interrupted.
In these instances, mine water was stored prior to shutdown in a lO-m^
capacity tank to maintain flow over 48-hr periods. At Hollywood and Hawk
Run, mine waters were delivered from the source by two 0.25-kw centrifugal
pumps in connection with a 7.6-cm diameter foot valve and appropriate lengths
of flexible polyethylene piping.
Flow rates were checked by collecting effluent samples for one minute
in 3-H and 100-& containers, respectively, fo'r the 0.5- and 2.0-m units.
Water samples were collected manually or with portable automatic, discrete
twenty-four bottle samplers (Manning model S-4040, Manning Environmental
Corporation, Santa Cruz, Calif). Instruments were located on site for
measurement of those constituents which required immediate determination
[Fe(II), pH, temperature, and dissolved oxygen]. Temperature of the
influent mine water was recorded continuously for experiments at Hawk Run,
Pa. by use of a Linear Instruments model 261 chart recorder (Linear Instru-
ments Corporation, Irvine, Calif.). Other analytical equipment was located
either at the Crown Field Site or The Pennsylvania State University
laboratories (see Appendix B).
Experimental Procedures
Mine water was mecha'nically pumped from each source to the feed
chambers of the pilot units at a rate slightly greater than the desired
119 mj/d. Excess flow from the feed chamber was discharged through an
overflow pipe in order to maintain a constant liquid level of fresh mine
water in the feed bucket chamber. The static level of mine water in the
chamber and the number of feed buckets employed were determined in accordance
with the desired flow rate.
All experiments were performed at an equivalent peripheral disc velocity
of 19 m/min and a hydraulic loading of 0.16 m^/d-m2 in order to compare
performance of the RBC units at each location. Attainment of a peripheral
disc velocity of 19 m/min required 13 and 2.9 rpm for the 0.5- and 2.0-m
units, respectively. The peripheral velocity of 19 m/min was selected on
the basis of results in previous experiments on rotational effects for the
RBC in Fe(II) oxidation (54). Where applicable, alteration of the standard
hydraulic loading rate was made only after equilibrium data had been
collected at a particular site and in the interest of improving performance
(Hawk Run).
Treatment units were cleaned and disinfected prior to start-up at each
location with 10-percent (v/v) HC1 and 5-percent (w/w) sodium hydrosulfite-
sodium bisulfite blend (Nalclean 2564, Nalco Chemical Co., Paulsboro, N.J.)
for the 0.5- and 2.0-m units, respectively. Solutions were prepared with
tap water, added to the troughs of each unit, and the discs were rotated for
35
-------
one week. Start-up of the units began with the initiation of disc rotation
and intake of mine water to the influent chamber. At no time during the
field studies were mine waters seeded with bacteria.
Sampling
Water samples were collected from the influent line and from each
stage compartment (see Appendix D for location of sampling lines). Sampling
lines were flushed for one to two minutes prior to sample collection depend-
ing on the length of the hose. Water samples representative of influent mine
water were initially collected from the feed chamber of each unit. Analyses
of these samples were made to insure that no significant Fe(II) oxidation
occurred in the feed chamber (less than 1 percent occurred) and that samples
from each unit were similar in water quality for the parameters analyzed.
The grab samples from influent and each stage of RBC units were either
manually collected (two to five times per week) or automatically taken
(twenty-four samples per day) depending on the particular experiment.
Premeasured plastic sampling squares affixed to the outer-most disc on
each stage prior to start-up were carefully removed (one to three times per
week) for appropriate chemical and microbiological analyses (Appendix B).
This sampling schedule was employed both during start-up and during
equilibrium Fe(II) oxidation periods which lasted a total of two to four
months. Equilibrium conditions were considered to be the period during which
Fe(ll) oxidation efficiency was maximum and relatively constant. Following
sufficient data collection at equilibrium, mine water feed was replaced, if
available, with another source at the same location.
RESULTS
Start-Up
Performance of the 0.5-m RBC in the oxidation of Fe(II) was related to
the source of mine drainage treated (Figure 9). At Hollywood, Pa., Fe(II)
oxidation increased steadily after an initial 3-day lag period and exceeded
90 percent after 10 days of continuous operation. Fe(II) oxidation continued
to increase steadily and an equilibrium point was reached in approximately
30 days. A similar pattern of Fe(II) oxidation was observed for the 2.0-m
RBC except that equilibrium oxidation required approximately 35 days. A
short lag in oxidation (3 days) with the 0.5-m unit occurred also in the
initial treatment of combined Hawk Run and Mine Hole 1 drainage. By
comparison to Proctor No. 2 oxidation, Fe(II) oxidation proceeded at a
slightly slower rate following the lag period. Fe(II) oxidation approached
75 percent after 16 days and improved steadily thereafter until about 85
percent oxidation was reached in 40 days. The larger 2.0-m RBC required 50
days to reach an equilibrium level of oxidation. In treatment of Crown
drainage, the 0.5-m RBC reached almost 8 percent oxidation within the first
3 days of continuous operation. During this period, pH increased from 4.5
(influent) to 6.0 (effluent). Concurrently, aluminum was continually being
precipitated from solution and apparently onto disc surfaces. The effluent
contained 43.1 percent less Al than the influent, presumably due to the
lower solubility of Al at higher pH values (9). Following two days at
36
-------
100
80
0)
o
t_
Q)
0.
60
o
x 40
O
20
CROWN, W. VA.
• HOLLYWOOD, PA.
• HAWK RUN, PA.
0
10
20 30
DAYS
40
50
Figure 9. Start-up and time required to obtain equilibrium Fe(II)
oxidation with the 0.5-m RBC in treatment of Crown mine
drainage at the Hollywood Proctor No. 2 mine water and
combination Hawk Run + MH 1 discharge.
37
-------
elevated pH and low effluent aluminum concentration, pH values of the
effluent were comparable to influent values, and no change was observed in
Al content in passage of mine water through the RBC. Fe(II) oxidation
efficiency during this period was less than 2 percent. Thereafter, Fe(II)
oxidation efficiency increased rapidly to an equilibrium level of 90 to 95
percent in 19 days.
Colonization of RBC discs by iron-oxidizing bacteria appeared to occur
at approximately the same time during the treatment of influent mine water at
Hollywood and Hawk Run (Figure 10). In contrast, bacterial densities on disc
surfaces after 17 days treatment of Crown mine drainage were at least 1,000-
fold greater than population densities developed on Hollywood and Hawk Run
drainages for the same period of time. Nevertheless, Fe(II) oxidation
performance for the RBC treating the various mine waters was not proportional
to the populations of iron-oxidizing bacteria present on discs (Figure 9).
The inorganic salts-agar medium of Manning (51) was used to cultivate
noniron-oxidizing microorganisms. Organisms recovered and examined by
phase microscopic examination were found to be rod-shaped bacteria which
would not be subcultured on the primary isolation medium or in a glucose-yeast
extract medium and would not grow autotrophically in Fe(II) medium. On the
primary isolation medium, colonies were nonpigmented, punctiform, and did not
show any evidence of Fe(II) oxidation through deposition of brown Fe precipi-
tates. It was suspected that the organisms were not nitrifying bacteria due
to the excessively low pH of the medium (pH 3). Although the true nature of
the noniron-oxidizing bacteria enumerated on the salts-agar medium could not
be determined, they were presumed to be fastidious heterotrophs and will be
referred to in this manner in all further discussions. A comparison of the
development of heterotrophic and iron-oxidizing bacterial populations on disc
surfaces with time at Hollywood revealed a much higher initial heterotrophic
bacterial population per unit surface area (Figure 11). After 15 days, viabl^
heterotrophic and chemolithotrophic bacterial numbers became parallel and the
absolute density of heterotrophs exceeded that of the iron-oxidizing
bacteria.
Equilibrium Conditions
The Fe(II) oxidation performance of the 0.5- and 2.0-m RBC units was
compared for each drainage treated and related to corresponding mean influent
mine water pH, temperature, and Fe(II) concentration (Table 6). The 0.5-m
RBC performance was slightly superior to the prototype 2.0-m unit for the
four mine waters treated. Hawk Run drainage contained the lowest mean
influent Fe(II) and temperature and exhibited a greater variation in Fe(II)
oxidation than did the other mine waters. Day-to-day variability in
oxidation performance at a particular temperature was less than 5 percent
for both units.
Changes in the dissolved oxygen concentration were observed as the
water passed through each stage of the two RBC units (Table 7). With one
exception, aeration created by turbulence during disc rotation and exposure
to atmospheric oxygen resulted in a substantial increase in DO through the
stages to near saturation levels in the final effluent. During treatment of
38
-------
CM
LU
CD
CD
Z
M
9 *
X
o
i
o
cr
CJ>
o
CROWN, W, VA.
HOLLYWOOD, PA.
HAWK RUN, PA.
I I
10
20 30
DAYS
40
50
Figure 10. Iron-oxidizing bacteria at surface of 0.5-m RBC discs
for each mine drainage location. Data points represent
mean value for samples taken from influent-side disc of
all stages.
39
-------
HETEROTROPHIC
BACTERIA, cells/cm2
IRON-OXIDIZING
BACTERIA, MPN/cm
20 30
DAYS
Figure 11. Iron-oxidizing and heterotrophic bacteria at surface
of 0.5-m RBC discs at Hollywood, Pa. Data points
represent mean value for samples taken from influent-
side disc of all stages.
40
-------
TABLE 6. COMPARATIVE PERFORMANCE OF 0.5- and 2,0-M RBC UNITS AT EQUILIBRIUM Fe(II)
OXIDATION FOR EACH MINE DRAINAGE TREATED
Mean influent
characteristics
Mine Drainage
Crown (17) a
RO brine (10)
Hollywood (8)
Mine Hole 1 (5)
Hawk Run (5)
Hawk Run (4)
+ MH 1
PH
4.93
2.38
2.75
3.92
3.23
3.38
Temp., °C
16.6
23.9
10.0
9.9
3.9
7.6
Fe(II), mg/£
190
356
116
70
32
53
0
Mean
93.3
97.7
98.3
84.7
75.3
85.4
Fe(II) oxidation, percent
.5-m RBC 2.0-m RBC
Std. dev. Mean Std. dev.
2.0 — b
1.1
0.9 92.8 1.9
1.4 76.9 0.3
7.6 72.3 4.1
1.9 75.1 1.6
of influent and effluent measurements applied to statistical determinations.
RBC unit not examined for mine drainage listed.
-------
NJ
TABLE 7. DISSOLVED OXYGEN IN MINE DRAINAGE INFLUENT AND INDIVIDUAL STAGE COMPARTMENTS
OF 0.5- AND 2.0-M RBC UNITS AT EQUILIBRIUM Fe(II) OXIDATION EFFICIENCY
Mean dissolved oxygen, mg/£
0.5-m RBC 2.0-m RBC
Mine Drainage
Crown (10)a
RO brine (6)
Hollywood (7)
Mine Hole (3)
Hawk Run (4)
Hawk Run (3)
+ MH 1
Inf.
2.4
2.2
1.9
2.4
11.7
7.1
SI
5.9
3.2
6.4
6.8
10.8
8.1
S2
7.0
5.1
8.4
8.0
10.6
8.6
S3
7.7
6.7
9.4
8.5
10.5
8.7
Eff. Inf. SI
8.2 — b
7.6
9.8 2.9 5.5
9.2 1.2 5.8
10.9 11.9 11.0
9.3 6.4 8.0
S2 S3 Eff.
—
—
7.1 8.3 9.1
7.1 8.1 9.4
11.1 11.1 11.4
8.2 8.6 9.4
munber of measurements applied to calculation of mean values.
RBC unit not examined for mine drainage listed.
-------
Hawk Run drainage, DO decreased slightly through the stages. The lower
temperature mine water streams (Hawk Run and Hawk Run + MH 1) were initially
higher in DO. The DO profiles were similar in both the 0.5- and 2.0-m RBC
units.
Changes in the viable iron-oxidizing bacteria content of influent and
effluent mine water were observed (Table 8). An increase was noted in iron-
oxidizing bacterial numbers in the effluents of RBC units as compared to
influent mine waters for the Crown RO brine, Mine Hole 1, and Hawk Run mine
waters. The reverse situation occurred for the Hollywood mine drainage with
both RBC field units and for the Hawk Run and combination drainages with the
2.0-m RBC unit. Heterotrophic bacteria were enumerated in influent and
effluent mine drainage of the 0.5-m RBC at Hollywood. Mean influent and
effluent populations at equilibrium Fe(II) oxidation were 340 and 540
cells/100 mS,, respectively.
Effect of Temperature
The Hawk Run mine drainage exhibited the greatest fluctuation in
temperature in comparison to the other five influent mine waters (0.4 to
9.8°C). During winter, influent and effluent mine water samples were auto-
matically collected every hour over a 24-hr period for Mine Hole 1, Hawk Run,
and combination Hawk Run and Mine Hole 1 drainages. Samples were analyzed
for Fe(II) and the results were examined in relation to the continuous
influent mine water temperature measurements. Since the RBC building was
maintained at an ambient air temperature near the influent mine water
temperature, variation in temperature as the mine water passed through the
units was less than 1°C. Mine Hole 1 water temperature was constant at 10°C
and Fe(II) oxidation was stable at 76.8 to 78.4 percent efficiency (Figure 12,
top). Hawk Run + MH 1 mine water exhibited fluctuations in temperature which
closely paralleled changes in Fe(II) oxidation (Figure 12, bottom). In the
Hawk Run stream (Figure 13), Fe(II) oxidation lagged behind both a tempera-
ture decrease and increase. Both 0.5- and 2.0-m RBC units behaved similarly
in Fe(II) oxidation with changes in temperature over a 24-hr period.
At Hawk Run, Pa. there existed good statistical correlation (1-percent
confidence level) between Fe(II) oxidation efficiency and mine water tempera-
ture for the 0.5- and 2.0-m RBC units (Figure 14 and 15). A higher correla-
tion was observed for the 0.5-m RBC (R = 0.853) than for the 2.0-m RBC
(R = 0.668) due to an unexplained greater effect of temperature on Fe(II)
oxidation in the smaller unit. The reverse situation was expected because
the temperature of mine water effluent from the 0.5-m unit had increased
slightly more than the temperature of the effluent from the 2.0-m RBC (0.9°C
difference vs. 0.4°C difference).
Although the temperatures of Crown, W. Va. mine waters were higher than
the temperature of Hollywood mine drainage, Fe(II) oxidation efficiency was
not observed to be greater at the Crown site (Table 6). The temperatures of
Hawk Run drainages were, at times, equivalent to the temperature of the
Hollywood discharge; however, Fe(II) oxidation efficiency for field units in
operation under equivalent conditions was 10 to 15 percent lower for the Hawk
43
-------
TABLE 8. IRON-OXIDIZING BACTERIA IN MINE DRAINAGE INFLUENT AND EFFLUENT
OF 0.5- AND 2.0-M RBC UNITS AT EQUILIBRIUM Fe(II)
OXIDATION EFFICIENCY
Mean iron-oxidizing bacteria, MPN/100 mJl
Mine
Drainage
Crown (5)a
RO brine (4)
Hollywood (4)
Mine Hole 1 (3)
Hawk Run (4)
Hawk Run (3)
+ MH 1
Influent
290,000
5,200
3,300
4,700
440
200
Effluent
0.5-m RED
160,000
6,600
990
14,000
1,100
310
2.0-m RBC
__b
—
1,900
5,900
420
110
TJumber of measurements applied to calculation of mean values.
RBC unit not examined for mine drainage listed.
44
-------
IUU
•«—
c
0)
o
k.
0)
1 1 1 1 1 1 1 1
"" —
•"
•»
O
S 80
X *
o
S 70
0) x
"- 0
"* — _
— A ^L^ A _
1^ ^•"" A
—
""" ^,^
<•
L-J rJ 1 1 1 1 1 L
6PM
EMPERATURE, °C
I2M
6AM
I2N
«. 100
o
*.
0)
<
o
X
o
£
H 70-
T . 1 1
AFe(H) OXIDATION
•TEMPERATURE
6PM
12 M
6AM
I2N
14
12
o
o
Figure 12. Effect of mine drainage temperature fluctuations over
one 24-hr period on Fe(II) oxidation with 2.0-m RBC
in treatment of MH 1 discharge (top) and combination
Hawk Run and MH 1 drainage (bottom).
-------
_ 100
6PM
I2M
6AM
I2N
100
OXIDATION
•TEMPERATURE
6PM
I2M
6AM
I2N
UJ
Figure 13. Effect of temperature fluctuations over one 24-hr period
on Fe(II) oxidation with 2.0-m RBC in treatment of Hawk
Run mine water.
46
-------
100
90
tso
x 70
60
I
Y = I.79X +68.4
R=0.853
I
I
468
TEMPERATURE,°C
10
12
Figure 14. Effect of mine drainage temperature on Fe(II) oxidation for
the 0.5-m RBC In treatment of mine waters at Hawk Run, Pa.
Actual data collected within the temperature range of 0.4
to 10.3°C (inclusive).
47
-------
100
c
0>
o
k.
-------
Run drainages. Therefore, mine water temperature does not appear to be the
principal reason for the lower Fe(II) oxidation efficiency at Hawk Run.
Effect of pH
Over the range of influent mine water pH values observed, including
Crown, which fluctuated from pH 4.40 to 5.50, no effect of pH on Fe(II)
oxidation performance was evident at each location with each RBC unit.
Correlation of pH with Fe(II) oxidation revealed a coefficient (R) of less
than 0.5 for each mine drainage examined.
Changes in pH occurred in the mine water at each stage of the RBC
(Figure 16). When influent mine water was above pH 3.0, a mine water
buffering system resulted in decreasing individual stage and effluent pH
values to near pH 3. On the other hand, when influent mine water pH was
below 3.0 (RO brine and Hollywood mine drainage), individual stage and
effluent pH values increased to values near pH 3. Observations of pH
characteristics were seen to be independent of the RBC studied. These
occurrences are consistent with the nature of iron chemistry which is
discussed in Section 8.
Effect of Influent Fe(II)
Influent Fe(II) content for all mine drainages ranged from a low of
27 mg/£ for the Hawk Run stream to a high of 433 mg/£ for the concentrated
RO brine. The Hollywood mine drainage exhibited the greatest fluctuation
in influent Fe(II) concentration and a decrease in influent Fe(II) resulted
in a subsequent increase in oxidation efficiency (Figure 17). For the
limited data collected, a correlation (R) of -0.838 and -0.832 was observed
for the 0.5- and 2.0-m units, respectively (significant at the 1-percent
confidence level). The 0.5-m RBC was less affected by increased influent
Fe(II) than the 2.0-m unit. Influent Fe(II) did not fluctuate as greatly at
the other treatment sites and minor fluctuations showed poor correlation
with efficiency (not significant at the 5-percent level).
Kinetics of Fe(II) Oxidation
Although several mine water characteristics varied in magnitude for
each drainage, Fe(II) oxidation followed a concentration dependent, first-
order relationship with time in all instances. The straight line, semi-
logarithmic reactions depicted for the 0.5- and 2.0-m RBC units are
illustrated in Figures 18 and 19, respectively.
The slopes of these lines (k values) indicated the rate of Fe(II)
removal in accordance with first order kinetics:
F » FQe~kt (6.1)
49
-------
pH
1 1
A CROWN DRAINAGE
O RO BRINE
Q HOLLYWOOD
A MINE HOLE I
• HAWK RUN + MH I
• HAWK RUN
••V0 20 40
THEORETICAL RETENTION TIME, min
Figure 16. Changes in mine water pH with retention time in the
0.5-m RBC. Each data point represents the mean of
four to seventeen observations.
50
-------
100
c
0>
o
»_
0)
Q.
R = -0.838
95
X
o
fi
£
90
1
50
R =-0.832
0.5m RBC
2.0m RBC
I
\
100 150
INFLUENT FedC), mg/j?
200
Figure 17. Effect of influent Fe(II) concentration on Fe(II)
oxidation efficiency in treatment of Hollywood
mine drainage.
51
-------
500c
CROWN
DRAINAGE
i ' i
RO BRINE
i I i
HAWK RUN
I T I r
HAWK RUN
MH I
k = 0,035
I L_
20 40 60 0 20 40
THEORETICAL RETENTION TIME, min
60
Figure 18. Kinetics of Fe(II) oxidation with the 0.5-m RBC showing
values of k (Base e). Each stage data point represents
the mean of four to seventeen observations.
52
-------
UJ
cr
S
500c
100'^
I ' I ' -
HOLLYWOOD -
20 40 60 0 20 40
THEORETICAL RETENTION TIME, min
60
Figure 19. Kinetics of Fe(II) oxidation with the 2.0-m RBC showing
values of k (Base e). Each stage data point represents
the mean of four to eight observations.
53
-------
where
F = Fe(II) concentration at time, t, mg/£
F = initial Fe(II) concentration, mg/£,
t = theoretical liquid retention time, min, and
k = decay coefficient (base e) for Fe(II), min
Equation (6.1) can be made more useful by taking the natural logarithm of
both sides and letting Y = An F and a = £n F :
Y = -kt + a (6.2)
Values for the decay rate shown in Figures 18 and 19 were obtained from
Equation (6.2) by solving for the unknown, k. A similar decay rate (k =
0.054) was observed in previous studies (54) in treatment of Hollywood mine
drainage with the 0.5-m RBC at a hydraulic loading equivalent to that used in
this study.
Effect of Hydraulic Loading
The lower Fe(II) oxidation observed for the mine waters treated at
Hawk Run, Pa. prompted attempts to improve oxidation efficiencies to the
levels observed at other locations. After reaching equilibrium Fe(II)
oxidation in treatment of Mine Hole 1 discharge at a hydraulic loading of
0.16 m^/d-m^ with RBC field units, the hydraulic loading was decreased to
0.11 m^/d-m^, allowed to reach equilibrium, and decreased again to
0.08 m^/d-m^. Decreases in hydraulic loading resulted in corresponding
increases in Fe(II) oxidation efficiency (Table 9). A hydraulic loading
of 0.08 m-Vd-m^ allowed a sufficient increase in residence time in the trough
of each RBC unit to obtain Fe(Il) oxidation efficiencies similar to that nor-
mally observed at double the hydraulic loading applied at the other mine
water treatment locations (Table 6).
Effect of Unanticipated Shutdown
After approximately 50 days of continuous operation of the 2.0-m RBC
at Hollywood, Pa., malfunction of the drive motor which activates disc rota-
tion occurred. Mine water was permitted to flow continuously through the
unit; however, only 35 percent of the disc surface area remained wetted.
Consequently, it was possible to observe the effect of a temporary shutdown
on the operation of the biological process. During this period (4 days),
Fe(II) oxidation efficiency decreased from 95.0 to 1.5 percent and DO levels
in the trough of the unit decreased from 7.6 mg/fc just prior to shutdown to
0.5 mg/£ at the quiescent condition. Following this period, repair of the
motor was completed and passage of mine water equivalent to one tank volume
(4.6 m^) resulted in a return of Fe(II) oxidation efficiency to 85.0 percent.
Some desiccation, presumably injurious to the iron-oxidizing bacteria, was
visibly apparent at the disc surfaces exposed to the air. Interior surfaces
between discs, however, which provide the majority of available surface area,
54
-------
TABLE 9. EFFECT OF HYDRAULIC LOADING ON Fe(II) OXIDATION EFFICIENCY WITH
THE 0.5- AND 2.0-M RBC UNITS IN TREATMENT OF MINE HOLE 1 DISCHARGE
Fe(II) oxidation, percent
Hydraulic
loading, m3/d-m2
0.08 (4)a
0.11 (4)
0.16 (5)
Mean
96.8
91.7
84.7
0.5-m RBC
Std. dev.
2.2
1.3
2.7
Mean
94.4
83.9
76.9
2.0-m RBC
Std. dev.
1.1
0.6
2.3
Number of analyses applied to statistical determinations.
55
-------
may protect the bacteria from evaporation of water from solids.
Following six days of renewed operation of the 2.0-m RBC at a Fe(II)
oxidation of 85.0 to 86.0 percent, drive motor malfunction recurred. At
this time, mine water flow was interrupted and the troughs of both units
drained, thereby exposing the entire disc ensemble to the atmosphere. The
units remained in this condition without disc rotation for 8 days. When a
new drive motor was installed on the larger unit (the original motor was
defective), flow was returned to both units and disc rotation was restored.
After processing 0.14 and 4.6 m^ (one tank volume) of mine water in the
0.5- and 2.0-m RBC units, respectively, Fe(II) oxidation efficiency was
estimated at 71.7 percent for the smaller RBC and 70.0 percent for the
larger unit. Five days later and prior to the final shutdown at the
Hollywood facility, Fe(II) oxidation efficiencies of 93.5 and 84.3 percent
were reached for the 0.5- and 2.0-m units respectively. Under the limited
conditions observed in the field, it appeared that the unit supporting an
established biomass active in Fe(II) transformation may be expected to
recover oxidative capacity readily.
Characteristics of Biological Film
Solids formed on disc surfaces of the 0.5- and 2.0-m RBC units were
examined at equilibrium Fe(II) oxidation for all mine waters treated. At
Hollywood, Pa., (Figure 20, top) and Crown, W. Va., disc solids consisted
of a bright orange surface layer and a darker inner layer. Solids were
translucent and gelatinous, particularly on discs at stages close to the
influent end of the RBC. The gelatinous outer layer was not seen on the
disc surfaces during treatment of mine waters at Hawk Run, Pa. (Figure 20,
bottom).
Solids thickness on stage 1 discs of the 0.5-m RBC treating first
Crown mine drainage and later RO brine was approximately 2 and 3 mm,
respectively. At Hollywood, stage 1 disc solids attained a maximum thickness
of 2 mm with the 0.5-m RBC and 1 mm with the 2.0-m RBC. During equilibrium
Fe(II) oxidation at Hawk Run, Pa., stage 1 solids had a maximum thickness of
1 mm for the 0.5-m RBC and 0.5 mm in thickness for the 2.0-m unit. The
thickness of solids on discs visibly decreased in the direction of flow of
mine water through both the 0.5- and 2.0-m units with all mine drainages
treated. Differences in solids thickness at various stages of treatment were
less evident with the high Fe(II)-containing RO brine and the outer gelatinous
layer prevailed on the latter stages of the RBC to a greater extent than when
Crown or Hollywood mine drainages were treated.
Microscopic observation of solids, previously freed of iron (see
Appendix B), under phase-contrast optics indicated the presence of a
gelatinous matrix with embedded bacteria (zoogloeae) (Figure 21), The matrix
was more readily perceived in samples taken from RBC units at Hollywood and
Crown, although zoogloeal aggregates were observed in Hawk Run solids,
especially in samples retrieved from stage 1.
Glass cover slips attached to disc surfaces from stage 1 of the 2.0-m
RBC prior to start-up at Hawk Run were removed at various intervals and
56
-------
Figure 20.
Solids formation on stage 1 discs of 0.5-m RBC in treatment
of mine drainage at Hollywood, Pa. (top) and Hawk Run, Pa.
(bottom). Cut-away sections reveal thickness of solids layer,
-
-------
Figure 21. Phase-contrast photomicrograph showing bacteria embedded
in gelatinous matrix (zoogloea) from stage 1 of 0.5-m RBC
in treatment of mine drainage at Hollywood, Pa.
58
-------
prepared for observation under the scanning electron microscope (SEM) (see
Appendix B). Straight and curved bacilli were the only microbiological forms
detected in observation of the iron-free samples. Observations by the SEM of
a specimen prepared from a solids sample taken after 54 days of RBC mine
drainage treatment clearly demonstrated bacterial cells in a slime layer
(Figure 22, top). It was apparent, however, that the extensive preparation
required prior to SEM analysis (fixation in OsO^, alcohol drying, and
critical point drying) resulted in the loss of many cells from specimens.
Microorganisms recovered on 0.4-ym Nucleopore polycarbonate filters (General
Electric Corp., Pleasanton, Calif.) from the influent mine drainage (Figure
22, bottom) were similar in morphology to those samples from disc solids
observed under the SEM.
Certain chemical and microbiological characteristics of disc solids
taken from the 0.5- and 2.0-m RBC units at equilibrium Fe(II) oxidation are
given for each mine drainage in Tables 10 and 11, respectively. The dry
weight of total solids and total Fe and Al per unit surface area (mg/cnr) was
highest on discs of the 0.5-m RBC which treated RO brine (Table 10). Iron-
oxidizing bacterial densities were similar for the Crown mine drainage and
RO brine even though influent Fe(II) content and corresponding Fe(II)
oxidation efficiency were higher in treatment of the RO brine. Similarly,
iron-oxidizing bacterial densities on disc surfaces were lower in treatment
of Hollywood mine drainage as compared to the drainages of Hawk Run, even
though efficiency of Fe(II) oxidation was higher at Hollywood at equivalent
flow and disc rotation. A similar relationship existed for the 2.0-m RBC
(Table 11). Approximately double the weight of total solids and total Fe per
unit surface area formed on discs of the 0.5-m RBC in treatment of RO brine
in comparison with other drainages. Similar relationships developed with the
2.0-m RBC although no data on this unit are available for Crown mine water
treatment. The larger solids mass obtained in treatment of RO brine most
probably relates to the high ferrous iron present in the influent brine.
Solids on disc surfaces of the 0.5- and 2.0-m RBC units contained a
very low percentage of Al regardless of the source of mine drainage. Total
Fe and Al comprised approximately 25 to 50 and 0.05 to 0.1 percent,
respectively, of total disc solids dry weight for both RBC units. The
Hollywood Proctor No. 2 mine drainage contained a much higher Al content
than the drainages at Hawk Run (see Section 6) which probably accounts for
the higher Al content in solids formed on Proctor No. 2 mine drainage. The
relationship appeared to hold for both the 0.5- and 2.0-m RBC, although
total solids content was higher for the 0.5-m RBC.
Disc solids characteristics presented in Tables 10 and 11 are tabulated
for the influent-side disc of each individual stage (Tables 12 and 15). A
general decrease per stage was observed in weight per unit disc surface area
for each chemical characteristic (total solids, total Fe, and Al) (Tables
13, 14, and 15). Iron-oxidizing bacterial densities exhibited a greater
variation in disc solids from stage to stage than chemical characteristics,
although similar trends were apparent for both the 0.5- and 2.0-m RBC units
(Table 12).
59
-------
Figure 22. Scanning electron micrographs. Top: Glass cover slips
removed from stage 1 disc of 2.0-m RBC after 54 days of
continuous treatment at Hawk Run. Bottom: Microorganisms
filtered from Hawk Run mine water onto a 0.40-ym diameter
filter. Iron solubilized by addition of 1-percent oxalic
acid to prepare samples for observation of microorganisms.
60
-------
TABLE 10. CHARACTERISTICS OF SOLIDS AT SURFACE OF
DISCS IN 0.5-m RBCa
Iron-oxidizing
Mine bacteria,
Drainage MPN/cm^
Crown (5)b
RO brine
Hollywood (4)
Mine Hole 1 (3)
Hawk Run (4)
Hawk Run (3)
+ MH 1
15,000
3,800
64
220
290
70
Total solids,
mg/cm^
20.9
45.3
19.2
22.2
21.2
16.7
Total iron,
rag/cm^
7.7
14.9
10.2
4.0
5.0
7.9
Aluminum,
mg/cm^
0.026
0.029
0.024
0.009
0.011
0.013
Samples collected from influent-side disc of each stage of RBC during
equilibrium Fe(II) oxidation.
"Wtmber of analyses used to calculate mean values.
TABLE 11. CHARACTERISTICS OF SOLIDS AT SURFACE OF
DISCS IN 2.0-m RBCa
Iron-oxidizing
Mine
Drainage
Hollywood (4)b
Mine Hole 1 (3)
Hawk Run (4)
Hawk Run (3)
+ 1
bacteria,
MPN/cm2
51
1,200
82
410
Total solids,
rag/cm^
10.6
12.6
13.9
11.2
Total iron,
mg/cm
4.7
3.1
3.2
3.3
Aluminum,
mg/cm^
0.021
0.011
0.010
0.012
Samples collected from influent-side disc of each stage of RBC during
equilibrium Fe(II) oxidation.
DNumber of analyses used to calculate mean values.
61
-------
TABLE 12. IRON-OXIDIZING BACTERIAL DENSITIES OF DISC SOLIDS AT EACH STAGE OF THE 0.5- AND 2.0-M RBC3
2
Iron-oxidizing bacteria, MPN/cm
Mine
Drainage
Crown (5)b
RO brine (4)
Hollywood (4)
£ Mine Hole 1 (3)
Hawk Run (4)
Hawk Run (3)
+ MH 1
SI
25,000
510
150
450
300
72
0.5-m
S2
12,000
3,100
60
310
430
77
RBC
S3
11,000
10,000
58
91
220
62
S4
17,000
12,000
33
190
240
69
2.0-m RBC
SI S2 S3
c
—
91 77 38
850 1,600 640
94 56 74
650 220 420
S4
—
—
26
950
120
450
a
Samples collected from influent-side disc of each stage of RBC during equilibrium Fe(II) oxidation.
Number of analyses for each stage used to calculate mean values.
RBC unit not examined for mine drainage listed.
-------
TABLE 13. TOTAL SOLIDS CONTENT OF DISC SOLIDS AT EACH STAGE OF THE 0.5- AND 2.0-M RBCa
Mine
Drainage
Crown (5)b
RO brine (4)
Hollywood (4)
2 Mine Hole 1 (3)
Hawk Run (4)
Hawk Run (3)
+ MH 1
SI
20.6
81.3
30.8
31.7
24.8
18.9
0.5-m
S2
22.3
40.5
22.6
19.8
23.7
19.7
Total
RBC
S3
17.5
34.0
13.6
17.9
21.4
15.9
o
solids dry weight, rag/cm
2.0-m RBC
S4 SI S2 S3
9.9 — C
26.8
10.0 13.4 14.3 8.4
19.4 10.4 13.4 14.7
14.8 13.7 14.8 13.3
12.1 14.6 12.6 8.3
S4
—
—
6.2
12.0
13.6
9.1
aSamples collected from influent-side disc of each stage of RBC during equilibrium Fe(II) oxidation.
Number of analyses for each stage used to calculate mean values.
RBC unit not examined for mine drainage listed.
-------
TABLE 14. TOTAL IRON CONTENT OF DISC SOLIDS AT EACH STAGE OF THE 0.5- AND 2.0-M RBCa
Total iron, mg/cm^
Mine
Drainage
Crown (5)b
RO brine (4)
Hollywood (4)
Mine Hole 1 (3)
Hawk Run (4)
Hawk Run (4)
+ MH 1
SI
11.8
14.1
10.8
4.0
4.5
18.1
0.5-m
S2
8.1
19.9
17.9
4.2
7.2
7.0
RBC
S3
7.1
13.5
5.7
4.3
4.8
4.9
S4
3.6
11.9
6.2
3.7
3.5
1.5
2.0-m
SI S2
c
—
5.1 6.6
1.8 3.9
3.9 4.8
3.0 3.6
RBC
S3
—
—
5.1
3.5
3.1
3.2
S4
—
—
2.1
2.7
1.0
3.2
aSamples collected from influent-side disc of each stage of RBC during equilibrium Fe(II)
oxidation.
Number of analyses for each stage used to calculate mean values.
°RBC unit not examined for mine drainage listed.
-------
TABLE 15. ALUMINUM CONTENT OF DISC SOLIDS AT EACH STAGE OF THE 0.5 and 2.0-MRBC£
Mine
Drainage
Crown (5)b
RO brine (4)
Hollywood (4)
Mine Hole 1 (3)
Hawk Run (4)
Hawk Run (3)
+ MH 1
Aluminum
, mg/cm2
0.5-m RBC
SI
0.047
0.045
0.025
0.010
0.011
0.027
S2
0.024
0.028
0.027
0.010
0.011
0.008
S3
0.017
0.022
0.022
0.010
0.011
0.007
S4
0.017
0.029
0.023
0.008
0.010
0.010
2.0-m RBC
SI S2 S3
c
— —
0.024 0.022 0.022
0.007 0.016 0.011
0.010 0.010 0.010
0.007 0.010 0.025
S4
—
—
0.017
0.012
0.009
0.006
Samples collected from influent-side disc of each stage of RBC during equilibrium Fe(II) oxidation.
Number of analyses for each stage used to calculate mean values.
•*
"RBC unit not examined for mine drainage listed.
-------
Hydraulic Characteristics of RBC
In order to characterize the hydraulic efficiency of the RBC, a concen-
trated tracer solution (1.0 g/fc LiCl) was added instantaneously to one feed
bucket of each RBC unit just prior to discharge into stage 1. The volume of
tracer solution added to the 2.0-m RBC unit was proportionately larger on the
basis of trough volume. Samples collected from each stage and effluent at
specified time intervals were analyzed for lithium and the results were used
to construct plots of Li recovered versus time (Figure 23). Samples
collected from stage 1 within 30 seconds of tracer addition revealed peak Li
content. Higher Li concentrations were recovered from stage 1 of the 0.5-m
RBC (Figure 23, top) than from the same stage of the 2.0-m RBC (Figure 23,
bottom) which suggested more short-circuiting of flow to subsequent stages
for the larger sized RBC. Time for 50 percent of tracer to pass through the
tank (mean retention time) was nearly identical to the theoretical retention
times calculated on the basis of flow and trough volume for the respective
units (Table 16). Although RBC units were sized for equivalent hydraulic
loading (flow per unit disc surface area) for all experiments described in
this study, different ratios of surface area-to-trough volume for the units
(Table D-l) resulted in different theoretical retention times. Villemonte
and Rohlich (68) described certain dimensionless ratios which may be employed
to evaluate the hydraulic efficiency of continuous flow reactors. Application
of these indices to the results of tracer addition revealed near ideal dis-
persion of reactor contents in individual stages of each unit, some short-
circuiting between individual stage compartments, and overall flow through
the tank which approached "plug flow" conditions (Table 16).
Analysis of total Fe and Al in influent and effluent of RBC units at
Hollywood and Hawk Run indicated that approximately 10 percent of influent
total Fe and Al was continually being deposited on disc and trough surfaces.
Mass balance calculations for total Fe and Al in influent and effluent mine
water and on disc surfaces suggested that approximately 5 percent of the
total influent Fe was continually being deposited in the trough with the
remainder accumulating on disc surfaces. Following shutdown at the Hollywood
Facility, total solids deposited in the trough of the 0.5-m and 2.0-m RBC
units were found to have reduced the liquid volume of the trough by 14
percent (19 liters) and 1.5 percent (57 liters), respectively. The 2.0-m
RBC was equipped with four perforated tubes attached to the periphery of
disc surfaces of each stage to aid in turbulence and the prevention of
solids deposition in the reactor compartments (see description of tubes in
Appendix D). The 0.5-m unit, however, did not provide means of controlling
solids deposition except for the turbulence created by disc rotation.
COST ANALYSIS
Estimated Costs
Costs were estimated for Fe(II) oxidation with the RBC for mine
drainage flows of 500 and 5,000 m-Vd (Table 17). Costs for pumping, neutral-
ization, and solids handling were not listed. Only unit process costs were
included in the estimates. Cost estimates were quoted by Autotrol Corpora-
tion, one of several manufacturers of RBC equipment. Equipment needs were
66
-------
T
* STAGE I
o STAGE 2
a STAGE 3
EFFLUENT
0 20 40 60 80 100 120 140 160
TIME, min
20 40 60 80 100 120 140 160
TIME, min
Figure 23. Recovery of tracer in each state of RBC field units.
Top: 0.29 g Li added to influent of 0.5-m RBC at
flow rate of 3.4 m3/d; Bottom: 8.2 g Li added to
influent of 2.0-m RBC flow rate of 115 m3/d.
67
-------
TABLE 16. HYDRAULIC CHARACTERISTICS OF 0.5- AND 2.0-M RBC UNITS
Lithium,
Para-
meter
P
ho
^0
'90
T
JS.
T
ho
T
mg/5.
0.5-m RBC
SI
3
1
10
33
13
0.
17
0.
.9
.3
.9
.6
22
.8
76
S2
S3
15
8.
23.
54.
27.
1
5
3
2
0.55
6.7
0.86
36
17.3
39
77
40
0.
4
0.
.2
.0
.8
88
.5
96
S4
47
28.
56.
98.
54.
9
8
7
4
0.86
3.4
1.04
SI
3
1.8
10.4
36.1
12.0
0.25
20.1
0.87
2.0-m RBC
S2 S3
6 27
5.1 15.0
20.5 38.1
53.4 76.3
24.0 36.0
0.25 0.75
10.5 5.1
0.85 1.06
S4
37
20.1
49.4
96.5
48.0
0.77
4.8
1.03
P
ho =
ho =
ho =
T
t
T !
ho
time
time
time
time
for
for
for
for
peak
tracer concentration to
10 percent
50 percent
90 percent
of
of
of
theoretical retention
Measures
volume .
tracer
tracer
tracer
to
to
to
pass
pass
pass
pass through tank.
through
through
through
tank.
tank.
tank.
time (trough volume/ flow rate).
average short-circuiting, dead spaces
It is 1.0 for ideal settling and zero
T*. -t „ 1
n
•C~_ 4 -
, and effective tank
for ideal mixing.
i-i_~ ™_j 01 n *~~ j
t J~«1
• • • *^*»*h* v**- WM -u A w jv«**> M^iira* • » -fc *«r 4. AM-W^*-^* «J ^» te ^ At^&A^ fc*&&^» *• •
10 mixing.
C50
-=~ : It is 1.0 for ideal settling and 0.693 for ideal mixing.
68
-------
TABLE 17. ESTIMATED COSTS FOR Fe(II) OXIDATION WITH
ROTATING BIOLOGICAL CONTACTORS3
Item
Initial cost
RBC shafts with drive package
Enclosures (fiberglass)
Concrete basins (installed)
Freight and installation
Total
Annual amortization costc
Annual operating cost
Electrical (5
-------
sized on the basis of surface area required for obtaining the hydraulic
loading normally used in this study (0.16 m3/d«m2). Normal large-scale
application of the RBC would employ 3.6-m diameter discs. Autotrol Corpora-
tion recommended use of 2.0-m discs because of the greater mass density of
iron solids in comparison to solids adhered to disc surfaces in organic
wastewater treatment.
Costs per cubic meter treated for the 500 and 5,000 m^/d installations
were estimated at 3.8 and 2.9C, respectively, for amortization of capital
expenditures and 1.8 and 1.9C, respectively, for operation and maintenance.
The larger facility may realize an economy of scale in comparison of amortized
cost estimates per cubic meter treated for the two design flows. The compar-
able operating costs were due, primarily, to the constant electrical cost per
kwh applied to the estimates for both installations.
Cost Comparison
Costs for Fe(II) oxidation by conventional chemical oxidation were
estimated for design flows equivalent to those used in cost estimates for the
RBC (Table 18). Estimates were obtained with the aid of current equipment
manufacturer literature and information contained in publications by Mihok
(52) and Doyle et al. (22) which described mine drainage pollution control
costs. Required aeration basin volume was determined for each flow rate by
use of a design retention time of 90 min. Wilmoth et al. (74) found retention
times in aeration basins used in six chemical mine drainage treatment plants
to range from 10 to 370 min.
For both design flows, amortized capital costs were estimated to be less
than one-half those estimated for the RBC. Operating costs were almost double
those of the RBC estimates. Total costs per cubic meter treated were lower,
however, for the conventional process in use today for mine drainage Fe(II)
oxidation. In simply comparing total unit process costs, the lower unit costs
estimated for electricity, maintenance, and personnel with the RBC do not com-
pensate for higher capital cost of equipment and installation. The overall
economics would favor the RBC method only when electrical power costs double
the value used in the estimates.
Direct comparison of the RBC and conventional chemical oxidation
methods may not provide a true comparison of total estimated treatment costs.
The process chosen for the Fe(II) oxidation may change equipment and chemical
needs for other required processes and these modifications could change the
overall economics. Several advantages have been cited for oxidation systems
which precede neutralization. Use of limestone has proved to be feasible and
lower in cost than lime when mine water iron exists as Fe(III) (72). Lime-
stone has been reported to be impractical for neutralization when mine water
Fe(II) is greater than 100 mg/Jl because the pH cannot be elevated to levels
high enough to stimulate rapid chemical Fe(II) oxidation (47). Fe(II)
.oxidation, if performed prior to neutralization, would allow use of limestone.
Over-treatment would not be likely with limestone treatment because pH values
above 8 are not attainable. Process control would be simpler when neutrali-
zation follows Fe(II) oxidation because addition of excess neutralizing
chemical would not be necessary to account for acidity production from post
70
-------
TABLE 18. ESTIMATED COSTS FOR Fe(II) OXIDATION BY
CONVENTIONAL CHEMICAL OXYGENATION WITH MECHANICAL AERATORS3
Item
Initial cost
Mechanical aerators
Concrete basins (installed)^*
Freight and installation
Total
Annual amortization costc
Annual operating cost
Electrical (5
-------
Fe(II) oxidation and hydrolysis. Lovell (47) observed better sludge settle-
ability for limestone treatment in comparison to hydrated lime. These
considerations may be more important in choosing a method for Fe(II)
oxidation than a simple comparison of costs for the unit process alone.
Within the precision of treatment costs estimates, it appeared that
total costs for the RBC system would be similar to costs for a well designed
chemical oxidation process. The RBC system was estimated to be a more
capital intensive process, while the chemical oxidation method was found to
be more energy intensive and higher in total operating costs.
Costs estimated for the two Fe(II) oxidation systems were based on
liberal use of design safety factors. Sufficient cost savings may be accrued
with the conventional method by use of clay-lined ponds in place of concrete
basins. In addition, smaller size basins would be required for the oxidation
system by elevating pH prior to oxidation to values above the lower limit
for current effluent discharge regulations. In this way, it would be
possible to take advantage of the more rapid oxidation rates which would
result (see Figure 3). Holland et al. (35) indicated that a pH of 10.5 was
desired to assure complete and rapid Fe(II) oxidation by chemical oxygena-
tion. Studies by Wilmoth (73) and Lovell (47) showed elevation of mine water
pH to a value of 8 was adequate for efficient oxidation. The additional
dosage of neutralizing chemical required to attain these pH values may offset
its potential savings.
For the RBC system, capital and operating costs may be lowered by use
of larger diameter disc assemblies. The RBC equipment recommended by
Autotrol Corp. and used in estimating costs included 2.0-m discs in place of
their conventional full-scale equipment (3.6-m diameter discs). Other
suppliers of RBC devices manufacture mechanically supported disc assemblies.
The 3.6-m assemblies available from these manufacturers may be suitable for
use in a mine drainage treatment application and should result in lower
total costs. For example, a smaller number of RBC shafts, enclosures, and
concrete basins would be required in order to provide the desired hydraulic
loading for application to a 5,000 m^/d design flow and their use may save
an estimated $200,000 in total initial costs.
72
-------
SECTION 7
BENCH-SCALE STUDIES
INTRODUCTION
Bench-scale RBC experiments permitted examination of the effect on
Fe(II) oxidation efficiency of supplementation or alteration of mine drainage
feed which could not be conveniently achieved in field studies. At the Crown
field site, two identical 15-cm diameter RBC units (Figure 24, top),
constructed and described in previous studies (54), were fed mine drainage
having a natural pH range of 4.05 to 5.74. One unit was supplemented with
0.25N H2SO^ to lower the pH to 2.8 to 3.4. Schnaltman et al. (56) observed
stimulation of Fe(II) oxidation by T_. ferrooxidans at this pH range. The
remaining unit received natural, unadjusted pH mine drainage and served as a
control.
Miniaturized versions of RBC field units with discs measuring 10 cm in
diameter were constructed and operated under controlled environment conditions
(Figure 24, bottom). Synthetic mine drainage was used as the main feed source
for the two identical units in order to provide more control over mine
drainage variables of temperature, ionic concentration, pH, and microorganism
content. One RBC received a supplement of natural mine drainage (5 percent
of total flow) to examine the effect on Fe(II) oxidation of constituents in
drainage not present in synthetic mine water.
For both bench-scale studies, Fe(II) oxidation, viable iron-oxidizing
bacteria density both in mine water and on disc surfaces, and other para-
meters of interest were analyzed throughout the length of the study (see
Appendix B).
MINE WATER CHARACTERISTICS
The Crown drainage was used as mine water feed to the 15-cm RBC units.
Mine water was analyzed during the period of study for selected physical,
chemical, and microbiological parameters (Table 19). Mine water temperature
was somewhat higher than that measured in field studies. The difference was
apparently due to the slow velocity of travel of mine water through the small
diameter tubing. Other characteristics of the mine water used in the bench-
scale work were similar to those observed for the mine water during field
studies (Table 3).
Synthetic mine water applied in controlled laboratory experiments was
produced by dissolving reagents in deionized water (Table 20). Fe(II)
content and pH of the synthetic mixture were measured regularly (five to
73
-------
Figure 24. Bench-scale RBC units. Top: Experimental 15-cm diameter
units used in pH studies at Crown, W. Va. Bottom: 10-cm
units used in laboratory studies on the supplementation
of natural mine drainage to synthetic mine water feed.
74
-------
TABLE 19. CHARACTERISTICS OF ACID MINE DRAINAGE EMPLOYED IN BENCH-SCALE
EXPERIMENTS AT THE U.S. EPA CROWN FIELD SITE NEAR MORGANTOWN. W. VA.a
Parameter
Temp., °C (21)b
PH (21)
Fe(II), mg/Jl (21)
Fe(III), mg/£ (11)
Iron-oxidizing (11)
bacteria,
MPN/100 mi
Unsupplemented
mine drainage
Mean Std. dev.
19.2 2.7
4.34
182 35
36 50
3,000 13
Acidified
mine drainage
Mean
19.3
2.95
181
35
6,500
Std. dev.
2.5
—
35
44
21
aSamples were collected from May 24 to June 22, 1977.
Number of analyses applied to statistical determinations.
75
-------
TABLE 20. COMPOSITION OF SYNTHETIC MINE DRAINAGE FEED SOLUTION
EMPLOYED IN LABORATORY STUDIES
Constituent
Intended final
ionic concentration
CaCl
MgS04-7H20
A12(S04)3-18H20
250 mg/£ Fe(II)
50 mg/fl, Ca
50 mg/2, Mg
25 mg/£ Al
5 mg/fc Mn
1 mg/«, NH3-N
0.5 mg/£ P04
to pH 3
76
-------
a
seven times per week). A final pH of 2.8 to 3.2 was desired to reflect
typical acid mine drainage pH and was maintained approximately 85 percent of
the time. Fe(II) concentrations were maintained at a mean value of 233 mg/£
and a standard deviation of 17 mg/£. Hollywood mine drainage served as the
natural mine drainage supplement (Table 4). Iron-oxidizing bacterial
populations were monitored for each container of drainage (15 liters)
collected weekly by sampling and analyzing just prior to use and at the point
of exhaustion of the feed. These analyses were performed to check the
accuracy of dilution into synthetic mine water. The mean iron-oxidizing
bacteria in drainages were found to be 2,100 MPN/100 mA. The calculated mean
density of iron-oxidizing bacteria following dilution was 100 MPN/100 mH.
EXPERIMENTAL APPARATUS AND PROCEDURES
Crown. West Virginia
The two 15-cm diameter RBC units were constructed from acrylic plastic
and designed to conform to a surface area-to-trough volume ratio similar to
commercial field units (see Appendix D). Peripheral disc velocity was
equivalent to that used in field studies (19 m/min) and rotation (38 rpm) was
maintained by a small gear-reduced electric motor. Hydraulic loading to the
units was the same as normally used in field studies (0.16 rnVd-m2) and mine
water was applied to each unit with a Gorman-Rupp model 7128 dual head
metering pump at the rate of 100 mA/min per channel. A small amount of 0.25N
H2SO/i (approximately 1 mJl/min) was injected into the feed line of one unit
with the aid of a Beckman model 746 metering pump.
Laboratory
The two identical 10-cm diameter RBC units were constructed from
acrylic plastic and scaled to the 15-cm RBC units, but were equipped with
only two stages (see Appendix D). Peripheral disc velocity was maintained
at 19 m/min (57 rpm) with an electric motor.
Synthetic mine drainage flow was provided by a system (Figure 25) of
laboratory pumps (two Sage model 375A four channel tubing pumps, and one
Pharmacia model P-3 three channel peristaltic pump) which metered separate
concentrated solutions of inorganic salts, FeSO^^O, and 0.1N HoSO* along
with deionized water prepared by continuous feed of tap water through a
commercial laboratory RO unit (Millipore Corporation, Bedford, Mass.). Final
mine water flows (10 m£/min) were equivalent to a hydraulic loading to each
unit of 0.16 m3/d-m2. The controlled temperature room which housed the RBC
units and the feed solutions was maintained at approximately 10°C during the
Period of study (44 days).
Extra pump channels allowed separate metering of Hollywood drainage to
both RBC units for initial start-up and to one RBC for supplementation of
synthetic feed. At start-up, the units received approximately 1 m£/min of
Hollywood drainage to produce a population of microorganisms on RBC disc
surfaces. After 14 days, significant Fe(II) oxidation occurred (greater than
80 percent) and mine drainage feed was replaced with the synthetic mixture
and the flow rate increased to 10 mA/min. Following a period of initial
77
-------
DE-
IONIZED1
H20
JSONC.
SALT
LUTION
[RJMPJZr
RBC I
TO RBC 2
JDONC.
• S04
-UTION
Figure 25. Apparatus and concentrated solutions employed in continuous
metering of synthetic mine drainage to 10-cm bench-scale
RBC units.
78
-------
acclimation to synthetic feed (24 days), one unit was fed Hollywood drainage
at 5 percent of total feed while the other unit received no supplement and
served as a control.
RESULTS
Adjustment of pH of Natural Mine Water
The 15-cm RBC (RBC A) which received natural Crown mine drainage
behaved similarly to the 0.5-m RBC pilot unit in establishing Fe(II) oxida-
tion (Figure 26). Three days after start-up, Fe(II) oxidation efficiency was
approximately 6 percent in the bench-scale unit and mine water effluent pH
increased to 6.55. Thereafter, effluent pH was below 4 and oxidation
efficiency increased steadily to above 90 percent after 13 days. RBC B, which
received mine drainage acidified with H2S04 to a mean pH of 2.95, did not
exhibit the rapid onset of Fe(II) oxidation demonstrated by RBC A (Figure 26).
After 26 days of continuous operation, both units achieved an equilibrium
Fe(II) oxidation efficiency of 95.3 to 99.2 percent.
Iron-oxidizing bacterial densities and total solids content on disc
surfaces were compared for the two RBC units (Figure 27). Initially, iron-
oxidizing bacterial numbers on disc surfaces were lower for RBC B. After 20
days, however, iron-oxidizing bacterial densities were nearly equal and the
efficiency of Fe(II) oxidation was greater than 80 percent in both units.
RBC B had a lower total disc solids dry weight during the 30-day period.
.Supplementation of Synthetic Mine Water with Natural Mine Drainage
Initial acclimation of two bench-scale 10-cm RBC units to pH 3
synthetic mine drainage (24 days) revealed a maximum Fe(II) oxidation
efficiency of 49.4 and 40.6 percent after 17 days of operation in a controlled
(10°C) room (Figure 28). Following this period, oxidation efficiency de-
creased steadily. Just prior to supplementation with natural mine drainage,
Fe(ll) oxidation in the two units was 19.8 and 28.6 percent. The difference
observed in oxidation with the two identical units at the time of Fe(II)
analysis apparently was due to chance because Fe(II) concentrations fluctua-
ted between the units.
On the 26th day of continuous synthetic mine drainage feed, one unit
(RBC 1) was supplemented with natural mine drainage (5 percent). Fe(II)
oxidation efficiency increased over 20 percent after 1 day and continued to
increase thereafter. The control unit (RBC 2) was consistently less
efficient than the unit which received mine drainage supplement. Twelve
days after initiation of mine drainage supplement, samples were periodically
collected for enumeration of iron-oxidizing bacteria in influent and effluent
mine water and on disc surfaces (Table 21). Iron-oxidizing bacteria were not
detected in the synthetic mine drainage, but a mean MPN of 1,100 cells/100 mfc
were recovered in the effluent. A higher effluent iron-oxidizing bacterial
density was observed for the RBC which received mine drainage supplement.
Iron-oxidizing bacterial populations on stage 1 and 2 discs were higher,
however, for the unit which received synthetic feed (RBC 2), possibly
Deflecting the higher proportion of solids-associated bacteria on RBC 2 which
79
-------
100
DAYS
Figure 26. Fe(II) oxidation in 15-cm RBC units operated on Crown mine
drainage with a natural pH range of 4.05 to 5.74 (RBC A)
and acidified to pH 3 (RBC B).
80
-------
IRON-OXIDIZING BACTERIA
RBC A
RBCB
TOTAL DISC SOLIDS
o RBC A
RBC B
DAYS
Figure 27. Iron-oxidizing bacterial densities at surface of discs and
total disc solids dry weight. Bench-scale (15-cm) RBC
units in treatment of Crown drainage with a natural pH
range of 4.05 to 5.74 (RBC A) and an adjusted pH of 3
(RBC B).
81
-------
100
80
• RBC I
A RBC 2
c
o»
o
t_
4>
z
o
x
o
60
40
20
10
20
30
40
50
DAYS
Figure 28. Fe(II) oxidation in bench-scale (10-cm) RBC units in
treatment of synthetic mine water at 10°C and initial
pH of 3. Arrow denotes point of addition of 5 percent
Hollywood mine drainage to RBC 1.
82
-------
TABLE 21. IRON-OXIDIZING BACTERIA IN INFLUENT AND EFFLUENT MINE WATER
AND DISC SOLIDS OF 10-CM RBC UNITS IN TREATMENT OF SYNTHETIC
MINE WATER SUPPLEMENTED WITH HOLLYWOOD MINE DRAINAGE
(RBC 1) AND UNSUPPLEMENTED (RBC 2).
Iron-oxidizing bacteria3
RBC 1 RBC 2
Sample (supplemented) (no supplement)
Mine water, MPN/100 m£
Influent 38 NDb
Effluent 3,200 1,100
Disc solids, MPN/cm2
Stage 1 420 470
Stage 2 230 710
aEach analysis represents the mean of three determinations.
None detected.
must have become established by adsorption of the bacterial inoculum to the
feed rather than continuous adsorption of cells from the influent. In order
to preserve sufficient active solids on disc surfaces of the small 10-cm
units only very small areas (4 cm^) on the discs could be sampled for solids.
The limited solids which could be collected from the surfaces of discs
contributed to inaccuracies in bacterial enumeration of solids.
83
-------
SECTION 8
DISCUSSION
ROLE OF RBC IN MINE DRAINAGE TREATMENT
The RBC is not intended to provide complete treatment of acid mine
drainage. The function of the RBC is to achieve effective oxidation of
Fe(II) through microbial action. Neutralization of RBC effluent and
separation of precipitated iron solids would be necessary in order to
produce a water suitable for stream-release. Several investigators have
reported advantageous use of limestone in place of lime for neutralization
of acid mine drainage (33, 47, 72, 73). It was concluded, however, that
chemical oxidation of Fe(II) would not be feasible for mine drainage
treatment when limestone was applied as the neutralizing chemical. The time
required for sufficient Fe(II) oxidation was reported to be extremely long
because the pH values attainable with limestone were not sufficiently high to
promote rapid oxidation. Application of the RBC to oxidation of Fe(II) prior
to mine drainage neutralization would allow the use of limestone for effluent
pH adjustment.
A flow diagram of a proposed acid mine drainage treatment process is
presented in Figure 29. Lovell (47) suggested that preneutralization-Fe(II)
oxidation systems should produce an effluent Fe(II) concentration below
10 mg/£. He observed that agitation during neutralization provided further
Fe(II) oxidation which produced undetectable Fe(II) content in the final
effluent. Detailed design of the proposed treatment system may require
bench-scale studies to evaluate Fe(II) oxidation in neutralization of low
Fe(II) (less than 10 mg/Jl) acid water to simulate treatment of RBC effluent.
Fe(II) oxidation efficiencies in RBC treatment of mine waters at Hawk
Run, Pa. were observed to be lower than those for mine drainage treated at
the other two locations under similar temperature, pH, and RBC operating
conditions. The mine waters at Hawk Run, however, were usually lower in
Fe(II) content. Therefore, Fe(II) concentration of Hawk Run RBC effluents
was similar to Fe(II) in RBC effluents for the other two locations [about
10 mg/& Fe(II)]. If observed lower efficiencies are not due to lower initial
Fe(II) content, it may then be desirable to conduct pilot or bench-scale
studies at the actual mine drainage site in preparation for full-scale
design. Obviously, these studies could not be performed to obtain design
information prior to mining because the actual mine drainage may not be
available. It may be possible to obtain mine water from a nearby mine, if
available, although care must be exercised in application of results. The
results of any small scale RBC studies should be used with caution.
84
-------
INFLUENT
ACID MINE DRAINAGE
I
Fe(E) OXIDATION
ROTATING BIOLOGICAL CONTACTOR
1
NEUTRALIZATION
LIMESTONE
CLARIFICATION
SEDIMENTATION
POND OR BASIN
SLUDGE
HANDLING
AND DISPOSAL
TREATED
EFFLUENT
Figure 29. Proposed treatment process for acid mine drainage,
85
-------
RELATIONSHIP OF TEMPERATURE TO Fe(II) OXIDATION
A microbial population which mediated continuous Fe(ll) oxidation at
temperatures outside the range reported to be optimum for pure cultures of
the iron-oxidizing bacterium, Thiobacillus ferrooxidans, developed on disc
surfaces of RBC units in field experiments with natural mine drainage.
Silverman and Lundgren (59) reported maximum growth and Fe(II) oxidation of
Ferrobacillus ferrooxidans Or. ferrooxidans) at 28 and 37°C, respectively;
values representative of the warmer region of the mesophilic temperature
range. Field RBC units developed populations of iron-oxidizing bacteria
which, in all probability, included strains of TD. ferrooxidans. Mine water
temperatures as low as 0.4°C did not inhibit Fe(II) oxidation in field RBC
units, although Fe(II) oxidation efficiency was reduced by 10 to 20 percent
from that obtained in treatment of the same mine water at 10°C. These mine
waters are well within the psychrophilic temperature range for micro-
organisms and the results are at least suggestive of the activity of
psychrotrophic iron-oxidizing bacteria in the mine water environment.
Discrepancies in chemolithotrophic activity by pure and mixed cultures
in respect to environmental temperatures have been noted elsewhere. Sharma
and Ahlert (57) distinguished between nitrification by pure cultures and in
soils, streams, and treatment plants. It was recognized that nitrification
in natural environments and treatment plants can proceed at temperatures
below that recorded as minimal for nitrogen transformations by pure cultures
in the laboratory. Duddles et al. (22) observed 90 percent efficiency for
nitrification with a full-scale fixed-film biological tower at wastewater
temperatures as low as 4°C. Ammonia and nitrite oxidation serve as energy
yielding reactions for autotrophic bacteria of the genera Nitrosomonas and
Nitrobacter, respectively. A variety of microorganisms, both heterotrophic
and autotrophic, however, may contribute to nitrification. Similarly, other
microorganisms in addition to the autotrophic bacterium, ^T. ferrooxidans,
may promote Fe(II) oxidation in acid mine waters. The discrepancy between
characteristics of laboratory cultures of T_. ferrooxidans and biological
Fe(II) oxidation in mixed culture suggests the possibility of strain
differences.
RELATIONSHIP OF pH TO Fe(II) OXIDATION
Schnaitman et al. (56) observed the optimum pH for microbial Fe(It) ...
oxidation to be 3.0, with activity substantially reduced below pH 2.5 and
above 3.8. Influent mine water pH values as high as 5.50 and as low as 2.16
did not retard Fe(II) oxidation in field and laboratory RBC units. It
appeared that Fe(II) oxidation at equilibrium and subsequent hydrolysis of
Fe(III) caused the pH in the RBC reactor to buffer nearer to the pH values
reported as optimum for microbial Fe(II) oxidation. The observation is
explainable by a simplified view of Fe(II) oxidation [Equation (8.1)] and
Fe(III) hydrolysis [Equation (8.2)].
Fe2+ + ^ 02 + H+ - Fe3+ + I H20 (8.1)
Fe3+ + 3H20 = Fe(OH)3(s) + 3H+ (8.2)
86
-------
It may be observed from the preceding equations that Fe(II) oxidation
consumes one mole of H+ for every mole of Fe(II) oxidized and should result
in an increase in pH. In contrast, subsequent hydrolysis of Fe(III) results
in production of three moles of H+ for every mole of Fe(III) hydrolyzed which
would lead to a lower pH. Thermodynamic equilibria indicate that a large
percentage of Fe(III) would be hydrolyzed only at initial pH values above
about pH 3 (see Section 4). Therefore, when influent mine water is above
pH 3, oxidation and hydrolysis would produce a net decrease in pH. Theoret-
ically, one mole of IT1" would be consumed and three moles of H+ would be
produced for each mole of Fe(II) oxidized and mole of Fe(III) hydrolyzed.
This possible explanation would account for the lowering of pH observed for
the Crown mine drainage in passage through the RBC unit (see Figure 16). On
the other hand, when the influent is below approximately pH 3, relatively
little hydrolysis would occur and Fe(II) oxidation would produce a net slight
increase in pH (though slight considering pH is a log function and 3 is near
the end of the scale). Because Fe(II) oxidation is more important in
comparison to Fe(III) hydrolysis at these pH values, one mole of H* would be
consumed for every mole of Fe(II) oxidized and the pH would rise, as observed
in Hollywood mine drainage and Crown RO brine. As suggested in Section 3,
several complexes of Fe(II) and Fe(III) are possible in acid mine waters and
these may affect hydrogen ion activity. Also, acid mine drainage may be a
potent solvent for many minerals, depending largely on the geology and
hydrology of the drainage area, and dissolved constituents may react with H+.
SYNTHETIC VERSUS NATURAL MINE DRAINAGE
The difference between RBC performance in Fe(II) oxidation when
synthetic mine water is treated in the presence and absence of natural mine
drainage appears to demonstrate the importance of some natural stimulatory
constituent in mine drainage (qualitative or quantitative microbial
characteristics, growth factor, detoxicant, etc.) which was not present in
synthetic feed. Although it may appear that the absence of indigenous
microorganisms would preclude treatment of a specific wastewater by a
biological treatment process, foreign organisms acclimated to the treatment
of certain industrial wastewaters, particularly those from chemical
manufacture, have been inoculated and, if necessary, nutrient supplemented,
to permit a biological purification to take place. Multiplication of iron-
oxidizing bacteria, initially established in bench-scale RBC units treating
mine water, was demonstrated, although rates of Fe(II) oxidation were slower
without mine drainage supplement which provided a continuous inoculum of
bacteria. Possibly, a combination of adsorption of mine water bacteria onto
discs and synthesis of new cells accelerate Fe(II) oxidation with the RBC.
DEVELOPMENT OF SOLIDS ON RBC DISC SURFACES
Inspection of RBC solids from disc surfaces at all field locations
revealed two distinct layers (Figure 30) ; an inner dark brown layer and an
orange-brown outer layer which may or may not be gelatinous. Atmospheric
oxygen must dissolve in the mine water film on discs during rotation and,
subsequently, penetrate the solids layers along with soluble Fe(II) and
other essential nutrients for growth of microorganisms (C02, N, P, and trace
elements). In addition, diffusible waste products leave the film and collect
87
-------
ABC
—*\
A- DISC SURFACE
B - INNER SOLIDS LAYER
C - OUTER SOLIDS LAYER
D -LIQUID FILM
E - AIR
DIFFUSION
FLOW
0.5-3 mm
Figure 30. Functions of RBC disc surface, disc solids, liquid
film, and atmosphere in mine water treatment.
88
-------
in the water. Antonie (4) reported solids buildup on RBC discs reached a
maximum in as little as one week in treatment of domestic wastewater. The
time to reach equilibrium disc solids weight was reported by Olem and Unz
(54) in Fe(II) oxidation experiments with the RBC to be approximately five
months. Periodic cleaning of the discs did not appear necessary. Hoehn and
Ray (34) examined the structure and function of the solids layer which
developed in fixed-film treatment of domestic wastewater. These organic
films were unlike the primarily inorganic disc solids reported herein.
Langmuir and Wittemore (45) described the relationship between age of
Fe(III) deposits in mine drainage streams and certain characteristics of
precipitates. Fresh amorphous Fe(III) precipitates are typically light brown,
orange, or yellow in color, depending on exact composition, and have maximum
solubility in comparison to aged Fe(III) deposits which may be partially or
totally crystalline in composition, darker brown, and have a much lower
solubility. It appeared that the solids which initially developed on disc
surfaces of field RBC units did not slough off and become replaced as quickly
as occurs in treatment of many organic wastes with the RBC (4). In time,
Fe(III) deposits may age and appear at the surface of discs as dense, dark
brown, impure oxyhydroxides. Bacterial colonization and activity in this
layer is probably minimal. The added weight of the coating may place
additional stress on structural capacity of discs.
A gelatinous surface covering was not observed for RBC treatment of any
of the mine waters examined at the Hawk Run facility. At mine water tempera-
tures similar to the Hollywood mine drainage (10°C), Fe(II) oxidation
efficiencies were still 10 to 15 percent lower at Hawk Run than at other
sites. Fe(II) concentrations were relatively low for the Hawk Run mine
waters (under 100 mg/£); however, the Hollywood drainage was at times as low
in Fe(II), and still Fe(II) oxidation greater than that observed at Hawk Run
was noted. This observation may suggest that the iron-oxidizing bacterial
population was sufficiently high at Hollywood prior to onset of lower Fe(II)
levels and permitted the high efficiencies to continue at the lower Fe(II)
concentrations. Estimates of iron-oxidizing bacterial densities revealed
no difference in cells per unit surface area for the Hawk Run drainage as
compared to that of Hollywood. Absence of a thick gelatinous surface
coating was readily apparent in RBC treatment of mine drainage at Hawk Run
as compared to other locations. A fibrous (gelatinous) surface layer may
provide more surface area for improved 02 transfer due to its shaggy nature
and thus may increase effective disc surface area.
COMPARISON OF DIFFERENT SIZE RBC UNITS
The 0.5-m RBC performance was slightly superior in Fe(II) oxidation
efficiency compared to the prototype unit under all conditions. Lower mine
water temperatures, however, did not seem to affect the performance of the
larger unit as greatly as the smaller RBC. There does exist a sizeable
difference in available surface area for the two RBC units (Table 22).
Design RBC hydraulic loadings calculated in this report were based solely on
disc surface area, although, the smaller RBC unit had a greater proportion
of the total surface area due to the troughs than the prototype and full-
scale units. Olem and Unz (54) observed similar viable iron-oxidizing
89
-------
TABLE 22. COMPARISON OF AVAILABLE SURFACE AREA FOR DIFFERENT SIZE
RBC UNITS
RBC unit
Bench-scale
Pilot-scale
Prototype
Full-scale
Disc
diameter
15 cm
0.5 m
2.0 m
3.6 m
Surface
Disc
0.438
21.8
738.1
9,290
2
area, nr
Trough3
0.10
2.14
18.8
50
Portion of total
surface area due
to trough, percent
19.1
8.9
2.5
0.5
Q
Available trough area was calculated from dimensions of bench-scale and
field RBC units and estimated for full-scale RBC by use of one 7.6-m
shaft placed in a contoured basin.
bacterial densities in comparison of trough and disc surfaces of the same
stage. Therefore, calculation of the effective surface area of small diameter
RBC units should include the trough surfaces. It is likely that if the units
had been sized on the basis of trough surface area, closer agreement would
have been obtained in Fe(II) oxidation performance with the large and small
RBC units.
Trough surface area of the pilot unit should be included in design of
the RBC. Only on this basis may the procedures described by Olem and Unz
(54) be used with confidence. The method uses the Fe(II) loading in terms
of grams of Fe(II) applied per day per square meter of available surface
area. "The loading is calculated using the influent Fe(II) concentration
(mg/fc) and a design mine water flow (m3/d). This information will allow
determination of required disc area when a desired effluent Fe(II) content
is chosen.
90
-------
REFERENCES
1. Allen, K. The Biologic Wheel. Sew. Works Jour., 1:560, 1929.
2. Anonymous. Acid Mine Drainage: An Expensive Problem to Solve.
Appalachia, 2:20-25, 1969.
3. Antonie, R. L. Nitrification of Activated Sludge Effluent:
BIO-SURF Process. Water and Sew. Works, 121:44-47, 1974.
4. Antonie, R. L. Fixed Biological Surfaces - Wastewater Treatment:
The Rotating Biological Contactor. CRC Press, Inc., Cleveland,
Ohio, 1976. 200 pp.
5. Antonie, R. L. and R. J. Hynek. Operating Experience with BIO-SURF
Process Treatment of Food Processing Wastes. In: Proc. 28th Ind.
Waste Conf., Purdue University, W. Lafayette, Ind., 1973. pp. 849-860.
6. Antonie, R. L., D. L. Kluge, and J. H. Mielke. Evaluation of a
Rotating Disc Wastewater Treatment Plant. Jour. Water Poll. Cont. Fed.,
46:498-511, 1974.
7. Antonie, R. L. and E. J. Koehler. Application of Rotating Disc Process
to Municipal Wastewater Treatment. 17050 DAM 11/71, U.S. Environ.
Protection Agency, Washington, D.C., 1971. 75 pp.
8. B. C. Research. Development of a Biological System for the Treatment
of Mill Waste Waters Containing Thiosulfate. Project 1525, Report to
Canada Dept. of the Environ., Ottawa, 1973.
9. Baes, C. F., Jr. and R. E. Mesmer. The Hydrolysis of Cations. John
Wiley and Sons, New York, 1976. 489 pp.
10. Baker, R. A. and A. G. Wilshire. Mlcrobial Factor in Acid Mine
Drainage Formation. II. Further Observation from a Pilot Plant
Study. SCi. Total Environ., 1:411-426, 1970.
11. Seller, M., C. Waide, and M. Steinberg. Treatment of Acid Mine
Drainage by Ozone Oxidation. 14010 FMH 12/70, U.S. Environ. Protection
Agency, Washington, D.C., 1970. 87 pp.
12. Bennett, D. J., T. Needham, and R. E. Summer. Pilot Application of the
Rotating Biological Surface Concept for Secondary Treatment of Insulat-
ing Board Mill Effluents. TAPPI, 56:182-187, 1973.
91
-------
13. Blrks, C. W. and R. J. Hynek. Treatment of Cheese Processing Wastes by
the Bio-Disc Process. In: Proc. 26th Ind. Waste Conf., Purdue
University, W. Lafayette, Ind., 1971. pp. 89-105.
14. Carruccio, F. T. and R. R. Parizek. An Evaluation of Factors Affecting
Acid Mine Drainage Production and the Ground Water Interactions in
Selected Areas of Western Pennsylvania. 2nd Symp. Coal Mine Drainage
Res. Preprints, Pittsburgh, Pa. 1968. pp. 107-152.
15. Chesner, W. H. and A. H. Molof. Biological Rotating Disk Scale-Up
Design: Dissolved Oxygen Effects. Prog. Water Tech., 9:811-819, 1977.
16. Clark, J. H., E. M. Moseng, and T. Asano. Performance of a Rotating
Biological Contactor Under Varying Wastewater Flow. Jour. Water Poll.
Cont. Fed., 50:896-911, 1978.
17. Cole, C. A., A. E. Molinski, N. Rieg, and F. Backus. Peroxide Oxidation
of Iron in Coal Mine Drainage. Jour. Water Poll. Cont. Fed., 49:1616-
1620, 1977.
18. Congram, G. E. Biodisk Improves Effluent-Water-Treating Operation.
Oil and Gas Jour., 74:126, 1976.
19. Cooke, W. B. The Occurrence of Fungi in Acid Mine Drainage. In:
Proc. 21st Ind. Waste Conf., Purdue University, W. Lafayette, Ind.,
1966. pp. 258-274.
20. Doman, J. Results of Operation of Experimental Contact Filter with
Partially Submerged Rotating Plates. Sew. Works Jour., 1:555, 1929.
21. Doyle, F. J., H. G. Bhatt, and J. R. Rapp. Analysis of Pollution
Control Costs. EPA-670/2-74-009, U.S. Environ. Protection Agency,
Washington, D.C., 1974. 347 pp.
22. "Buddies, G. A., S. E. Richardson, and E. F. Barth. Plastic-Medium
Trickling Filters for Biological Nitrogen Control. Jour. Water Poll.
Cont. Fed., 46:937-945, 1974.
23. Dugan, P. R., C. B. MacMillan, and R. M. Pfister. Aerobic Hetero-
trophic Bacteria Indigenous to pH 2.8 Acid Mine Water: Predominant
Slime Producing Bacteria in Acid Streamers. Jour. Bacteriol.,
101:982-988, 1970.
24. Ellis, K. V. and S. E. I. Banaga. A Study of Rotating-Disc Treatment
Units Operating at Different Temperatures. Water Poll. Cont., 75:73-
89, 1976.
25. Famularo, J., J. A. Mueller, and T. Mulligan. Applications of Mass
Transfer to Rotating Biological Contactors. Jour. Water Poll. Cont.
Fed., 50:653-671, 1978.
92
-------
26. Filion, M. P., K. L. Murphy, and J. P. Stephenson. Performance of a
Rotating Biological Contactor Under Transient Loading Conditions.
Presented at 50th Annual Conf. of Water Poll. Cont. Fed., Philadelphia,
Pa., 1977.
27. Ford, C. T. and J. F. Boyer. Treatment of Ferrous Acid Mine Drainage
with Activated Carbon. EPA-R2-73-150, U.S. Environ. Protection Agency,
Washington, D.C., 1973. 123 pp.
28. Franco, N. B. and R. A. Balouskus. Electrochemical Removal of Heavy
Metals from Acid Mine Drainage. EPA-670/2-74-023, U.S. Environ.
Protection Agency, Washington, D.C., 1974.
29. Glover, H. G. The Control of Acid Mine Drainage Pollution by Bio-
chemical Oxidation and Limestone Neutralization Treatment. In: Proc.
22nd Ind. Waste Conf., Purdue University, W. Lafayette, Ind., 1967.
pp. 823-847.
30. Grieves, C. R. Dynamic and Steady State Models for the Rotating
Biological Disc Reactor. Ph.D. Thesis, Clemson University, Clemson,
S. Carolina, 1972.
31. Hann, V. Ozone Treatment of Water. Jour. Am. Water Works Assn.,
35:585-591, 1943,
32. Hanna, G. P., J. R. Lucas, C. I. Randies, E. E. Smith and R. A. Brant.
Acid Mine Drainage Research Potentialities. Jour. Water Poll. Cont.
Fed., 35:275-294, 1963.
33. Hill, R. D. Mine Drainage Research - State of the Art and Research
Needs. 14010 12/68, Fed. Water Poll. Cont. Admin., Washington, D.C.,
1968.
34. Hoehn, R. C. and A. D. Ray. Effects of Thickness on Bacterial Film.
Jour. Water Poll. Cont. Fed., 45:2302-2320, 1973.
35. Holland, C. T., J. L. Corsaro, and D. J. Ladish. Factors in the Design
of an Acid Mine Drainage Treatment Plant. 2nd Symp. Coal Mine Drainage
Res. Preprints, Pittsburgh, Pa., 1968. pp. 274-290.
36. Holmes, J. and K. Schmidt. Ion Exchange Treatment of Acid Mine
Drainage. 4th Symp. Coal Mine Drainage Res. Preprints, Pittsburgh,
Pa., 1972. pp. 179-200.
37. Jasinski, R. and L. Gaines. Electrochemical Treatment of Acid Mine
Waters. 14010 FNQ 02/72, Environ. Protection Agency, Washington, D.C.,
1972.
38. Kellenberger, E., A. Ryter, and J. Se"chaud. Electron Microscope Study
of DNA-Containing Plasms. Jour. Biophysic. and Biochem. Cytol., 4:671-
678, 1958.
93
-------
39. Klingensmith, R. S., A. F. Miorin, and J. R. Saliunas. At-Source
Control Through the Application of Several Abatement Techniques.
6th Symp. Coal Mine Drainage Res. Preprints, Louisville, Ky., 1976.
pp. 270-284.
40. Krause, R. R. Mining and Reclamation Techniques to Control Mine
Drainage. 4th Symp. Coal Mine Drainage Res. Preprints, Pittsburgh,
Pa., 1972. pp. 425-430.
41. LaBella, S., I. Thacker, and J. Teehan. Treatment of Winery Waste by
Aerated Lagoon, Activated Sludge Process, and Rotating Biological
Contactors, or RBC's. In: Proc. 27th Ind. Waste Conf., Purdue
University, W. Lafayette, Ind., 1972. pp. 803-813.
42. Lachman, R. J. and H. L. Lovell. An Investigation of the Natural
Beneficiation of Coal Mine Drainage. Special Res. Report SR-76, Coal
Res. Board of the Comm. of Pa., Harrisburg, Pa., 1970.
43. Lackey, J. B. The Flora and Fauna of Surface Waters Polluted by Acid
Mine Drainage. Public Health Reports, U.S., 53:1499-1507, 1938.
44. Langmuir, D. The Gibbs Free Energy of Substances in the System
Fe-02-H20-C02 at 25°C. U.S. Geol. Survey Prof. Paper 650-B,
1969, pp. 180-184.
45. Langmuir, D. and D. 0. Whittemore. Variations in the Stability of
Precipitated Ferric Oxyhydroxides. In: Nonequilibrium Systems in
Natural Water Chemistry. F. R. Gould, ed., Advan. Chem. Series No.
106, 1971. pp. 209-234.
46. Lovell, H. L. Experience with Biochemical Iron Oxidation Limestone
Neutralization Process. 4th Symp. Coal Mine Drainage Res. Preprints,
Pittsburgh, Pa., 1972. pp. 292-1 to 292-2.
47. vLovell, H. L. An Appraisal of Neutralization Processes to Treat Coal
Mine Drainage. EPA-670/2-73-093, U.S. Environ. Protection Agency,
Washington, D.C., 1973. 247 pp.
48. Lovell, H. L. Studies in the Treatment of Coal Mine Drainage by
Biochemical Iron Oxidation and Limestone Neutralization. Special
Res. Report SR-98. Coal Research Board of the Comm. of Pa.,
Harrisburg, Pa., 1974.
49. Loy, L. D., Jr., J. W. Gunnett, J. D. Robins, and J. B. Warg.
Description of New, Innovative and Theoretical Mine Drainage Abatement
Techniques. 5th Symp. Coal Mine Drainage Res. Preprints, Louisville,
Ky., 1974. pp. 146-159.
50. Lue-Hing, C., A. W. Obayshi, D. R. Zenz, B. Washington, and B. M.
Sawyer. Biological Nitrification of Sludge Supernatant by Rotating
Disks. Jour. Water Poll. Cont. Fed., 48:25-46, 1976.
94
-------
51. Manning, H. L. New Medium for Isolating Iron-Oxidizing and Hetero-
trophic Acidophilic Bacteria from Acid Mine Drainage. Appl. Microbiol.,
30:1010-1016, 1975.
52. Mihok, E. A. Mine Water Research - Plant Design and Cost Estimates for
Limestone Treatment. Report of Inv. 7368, U.S. Bur. Mines, Washington,
D.C., 1970.
53. Morgan, P. F., P. J. Neuspiel, and E. A. Jeffrey. The Gresham
Biological Filter. Unpublished Progress Report, September 1956.
4 pp.
54. Olem, H. and R. F. Unz. Acid Mine Drainage Treatment with the
Rotating Biological Contactor. Research Report 93, Inst. for Res.
on Land and Water Resouces, Pa. State University, University Park,
Pa., 1976. 70 pp.
55. Public Law 92-500. Water Pollution Control Act Amendments of 1972,
1972.
56. Schnaitman, C. A., M. S. Korczynski, and D. G. Lundgren. Kinetic
Studies of Iron Oxidation by Whole Cells of Ferrobacillus ferrooxidans.
Jour. Bacteriol., 99:552-557, 1969.
57. Sharma, B. and R. C. Ahlert. Nitrification and Nitrogen Removal.
Water Res., 11:897-925, 1977.
58. Silverman, M. P. Mechanism of Bacterial Pyrite Oxidation. Jour.
Bacteriol., 94:1046-1051, 1967.
59. Silverman, M. P. and D. G. Lundgren. Studies on the Chemolithotrophic
Iron Bacterium Ferrobacillus ferrooxidans. I. An Improved Medium and
a Harvesting Procedure for Securing High Cell Yields. Jour.
Bacteriol., 77:642-647, 1959.
60. Simpson, D. G. and R. B. Rozelle. Studies on the Removal of Iron from
Acid Mine Drainage. 1st Symp. Coal Mine Drainage Res. Preprints,
Pittsburgh, Pa., 1965. pp. 64-68.
61. Singer, P. C. and W. Stumm. Acidic Mine Drainage: The Rate Limiting
Step. Science, 167:1121-1123, 1970.
62. Singer, P. C. and W. Stumm. Oxygenation of Ferrous Iron. 14010
06/69, Fed. Water Poll. Cont. Admin., Washington, D.C., 1970.
63. Standard Methods for the Examination of Water and Wastewater. 14th
Edition. Amer. Public Health Assn., New York, 1974.
64. Todd, R. L. and T. J. Kerr. Scanning Electron Microscopy of Microbial
Cells on Membrane Filters. Appl. Microbiol., 23:1160-1162, 1972.
95
-------
65. Unz, R. F. and M. T. Lieberman. The Microbiology of Acid Mine Water
Treatment in Packed Bed Columns. Research Report 77, Inst. for Res.
on Land and Water Resources, Pa. State University, University Park,
Pa., 1973. 73 pp.
66. U.S. Bureau of Mines. Implications of the Water Pollution Control Act
of 1972 for the Mineral Resources Industry: A Survey. PB 245 936,
NTIS, U.S. Department of Commerce, Springfield, Va., 1972. 23 pp.
67. U.S. Environmental Protection Agency. Effluent Guidelines and
Standards: Coal Mining Point Source Category. Fed. Register,
42:21380-21390, April 26, 1977.
68. Villemonte, J. R. and G. A. Rohlich. Hydraulic Characteristics of
Circular Sedimentation Basins. In: Proc. 17th Ind. Waste Conf.,
Purdue University, W. Lafayette, Ind., 1962. pp. 682-702.
69. Wagman, D. D., W. H. Evans, V. B. Parker, and R. H. Schumm. Selected
Values of Chemical Thermodynamic Properties. U.S. National Bureau
Standards Tech. Notes 270-3 and 270-4, 1968. 144 pp.
70. Whitesell, L. B., Jr. R. L. Huddleston, and R. C. Allred. Microbio-
logical Treatment of Acid Mine Waters. 14010 ENW 09/71, U.S. Environ.
Protection Agency, Washington, D.C., 1971. 78 pp.
71. Wilmoth, R. C. Applications of Reverse Osmosis to Acid Mine Drainage
Treatment. EPA-670/2-73-100, U.S. Environ. Protection Agency,
Washington, D.C., 1973. 159 pp.
72. Wilmoth, R. C. Limestone and Limestone-Lime Neutralization of Acid
Mine Drainage. EPA-670/2-74-051, U.S. Environ. Protection Agency,
Washington, D.C., 1974. 92 pp.
73. Wilmoth, R. C. Combination Limestone-Lime Neutralization of Ferrous
vIron Acid Mine Drainage. EPA-600/2-78-002, U.S. Environ. Protection
Agency, Washington, D.C., 1978. 52 pp.
74. Wilmoth, R. C., J. L. Kennedy, and R. D. Hill. Observations of Iron-
Oxidation Rates in Acid Mine Drainage Neutralization Plants.
5th Symp. Coal Mine Drainage Res. Preprints, Louisville, Ky., 1974.
pp. 246-261.
75. Wilmoth, R. C., D. G. Mason, and M. Gupta. Treatment of Ferrous Iron
Acid Mine Drainage by Reverse Osmosis. 4th Symp. Coal Mine Drainage
Res. Preprints, 1972. pp. 115-156.
76. Yastsimirskii, K. B. and V. P. Vasil'ev. Instability Constants of
Complex Compounds. Van Nostrand Co., Inc, New York, 1966. 214 pp.
96
-------
APPENDIX A
CHEMICAL EQUILIBRIA OF IRON
Calculations for constructing the solubility diagrams for Fe(OH) (s)
and Fe(OH)2(s) are summarized-below according to equilibrium constants at
25 °C (Table A-l).
Fe(III) - Fe(OH)3(s) Solubility Diagram
a. Fe(OH)3(s) = Fe3+ + 30H~ K = lO'37'1
[Fe3+] = 10+4'9°[H+]3
b. Fe3+ + H20 = FeOH2+ + H+ K - 10"2'17
[FeOH2+] = 10~2-17[Fe3+]/[H+]
c. Fe3++2H20 = Fe(OH)2++2H+ K = 10~7'17
[Fe(OH)2+] = 10"7<17[Fe3+]/[H+]2
d. Fe3+ + 3H20 = Fe(OH)3°(aq) + 3H+ K = 10~13'70
[Fe(OH)3°] = 10-13'7°[Fe3+]/[H+]3
e. Fe3++4H20 = Fe(OH>4" + 4H+ K = 10'21'88
[Fe(OH)4~] = 10'21'88[Fe3+]/[H+]4
f. Fe3+ + S042" = FeS04+ K = 10+4'13
[FeS04+] = 10+4'13[Fe3+][S042-]
g. EFe(III) = [Fe3+] = [FeOH2+] + [Fe(OH)2+]
+ [Fe(OH)3°] + [Fe(OH)4~] + [FeS04+]
97
-------
TABLE A-l. CHEMICAL EQUILIBRIUM DESCRIBING SOLUBILITY AND
COMPLEXATION OF Fe(II) AND Fe(III)
Fe(OH)
Fa2+4
**«
Fe2+ +
»» +
Fe(OH)
»» +
Fe3+ +
Fe* +
Fe3+ +
Fe3+ +
Equilibrium
Reaction constant at 25°C
2(s) = Fe2+ + 20H~ 10"15'1
• H20 = FeOH+ + H+ lO'10'11
• 2H20 = Fe(OH)2°(aq) + 2H+ 10"19'66
— 4- —^1 Q7
• 3H20 = Fe(OH)3 + 3H „ 10 J1*y/
S042- = FeS04°(aq) 10+2'2°
3(s) = Fe3+ + 30H~ 10~37>1
• H20 = FeOH2+ + H+ 10~2*17
2H20 - Fe(OH)2+ + 2H+ 10~7*17
3H20 = Fe(OH)3°(aq) + 3H+ 1Q-13.70
4H20 = Fe(OH)4~ + 4H+ 1Q-21.88
SO 2 - FeSO.+ 10+4'13
4 4
Reference
(9)
(76)
(76)
(76)
(69)
(45)
(69)
(69)
(44)
(44)
(69)
98
-------
h. ZS04 = [FeS04+] +
= 10"1'98(1000 mg/Jt)
3+
i. For each pH, [Fe ] and each Fe(III)-OH complex are calculated. The
two remaining equations, which contain two unknowns (FeSO^"*" and S0,^~),
are then solved simultaneously.
Fe(II) - Fe(OH)2(s) Solubility Diagram
a. Fe(OH)2(s) = Fe2+ + 20H~ K - lO""15*1
[Fe2+] = 10+12-9[H+3
b. Fe2+ + H20 = FeOH+ + H+ K - ID"10'11
[FeOH+] = 10-10-n[Fe2+]/[H+]
c. Fe2"*" + 2H20 = Fe(OH)2°(aq) + 2H+ K = 10~19'66
[Fe(OH)2°] = 10-19-66[Fe2+]/[H+]2
d. Fe2+ + 3H20 = Fe(OH)3~ + 3H+ K = 10~31'97
[Fe(OH)3~] = lCT31-97[Fe2+]/[H+]3
e. Fe2++S042 = FeS04°(aq) K = 10+2'20
[FeS04°] = 10+2'2°[Fe2+][S042-]
f. EFe(II) = [Fe2+] + [FeOH+] = [Fe(OH)2°]
+ [Fe(OH)3~] + [FeS04°]
g. ES04 = [FeS04°] + [S042"]
2+
h. For each pH, [Fe ] and each Fe(II)-OH complex are calculated. The two
remaining equations which contain two unknowns (FeSCL and SO^") are
then solved simultaneously.
99
-------
APPENDIX B
ANALYTICAL PROCEDURES
WATER SAMPLES
Temperatures
An electric thermometer with a stainless steel probe (YSI model 43TD,
Yellow Springs Instrument Co., Inc., Yellow Springs, Ohio) was used for
in situ temperature measurements.
2H
Measurements of pH were determined with an Orion model 801 digital
pH/mv meter (Orion Research Inc., Cambridge, Mass.).
Fe(II)
A dichromate titration procedure for Fe(II) was followed according to
procedures described by Lachman and Lovell (42) and is presented below.
Reagents—
a. Phosphoric-sulfuric acid mixture. Concentrated H^SO^ and
concentrated HgPO^ (150 m& each) are added to 700 m2, distilled
water.
b. Diphenylamine indicator. Barium diphenylamine sulfonate (0.32 g)
is dissolved in 100 mil distilled water. Sodium sulfate (0.500 g)
is then added and the BaSO^ precipitate is filtered. The filtrate
is the indicator.
c. 0.01 N potassium dichromate (K^CrjOy).
Procedure—
a. The sample (25 mi) is pipetted into a flask.
b. Acid mixture (10 m£) is added to sample.
c. Two drops of diphenylamine indicator is added.
d. Titration with K^Cr^O- is performed to a purple endpoint.
100
-------
Calculations—
Fe(II), mg/Jl = C x P x 2,234
where
C = m£ K2Cr207 used, and
N - normality of K2Cr207
Fe(III)
Fe(III) was considered the difference between total Fe and Fe(II).
Total Fe was determined using a Perkin-Elmer model 403 and 305A atomic
absorption spectrophotometer (Perkin-Elmer Corp., Norwalk, Conn.) at Crown,
W. Va. and The Pennsylvania State University, respectively. Samples were
acidified prior to analysis with 50 percent HCl.
Aluminum
Acidified samples were analyzed for Al using a Perkin-Elmer model 403
and atomic absorption spectrophotometer at Crown, W. Va. and The Pennsylvania
State University, respectively.
Dissolved Oxygen
DO was determined in situ with a YSI model 54B oxygen meter and probe.
Measurements of the DO of feed well contents were considered representative
of influent mine water DO.
Ammonia Nitrogen
NH3-N was determined colorimetrically with the aid of a Technicon
Autoanalyzer II (Technicon Industrial Systems, Tarrytown, N.Y.) and was based
on the procedure in Standard Methods for the Examination of Water and Waste-
water (63) .
Organic Nitrogen
Org-N was considered the difference between NH.J-N and total Kjeldahl
nitrogen (TKN). TKN was determined colorimetrically using a Technicon
Autoanalyzer II according to an automated version of the procedures in
Standard Methods.
Nitrate plus Nitrate Nitrogen
(N09 + NO-)-N was measured colorimetrically by the cadmium reduction
method (63) using a Technicon Autoanalyzer II. Samples received preliminary
cation exchange treatment and neutralization to pH 7.
101
-------
Total Phosphorus
The stannous chloride method with persulfate digestion (63) was employed
for determination of total P.
Iron-oxidizing Bacteria
The most probable number (MPN) of iron-oxidizing bacteria in mine
waters was estimated with a combination multiple tube and microtitre plate
procedure. An inorganic salts culture medium was employed which contained
1,500 mZ Fe(II) (Table B-l). Five, 1.0-m portions of mine water were
added to tubes which contained 2.0 m£ of culture media. Five, 0.1-mJl
aliquots of a serial dilution of mine water were transferred to a presteril-
ized plastic microtitre plate (Cooke Engineering Co., Alexandria, Va.) which
contained 0.2 mi of culture media per well. Plates were covered with plastic
film and presterilized plastic lids and incubated with tubes at 28°C for
three weeks. The number of wells per dilution giving positive results, as
evidenced by the appearance of an orange color, was recorded and used to
calculate the MPN per 100 mfc (63).
Heterotrophic Bacteria
Enumeration of heterotrophic bacteria was based on an adaptation of the
membrane filter technique for coliform bacteria (63) using an iron salts and
purified agar culture medium (51) (Table B-2). Mine water samples (0.01, 0.1,
1, 10, and 50 m£ ) were each filtered through a 0.45 urn membrane filter
(Millipore Corp., Bedford, Mass.) and placed in a 47-mm petri plate (Millipore
Corp.) which was one-half full of culture media. Plates were incubated at
28°C for two weeks. The number of plates which contained 30 to 300 colonies
was used to calculate the number of cells of heterotrophic bacteria per
100 mJl of original sample.
Lithium
Following addition of lithium chloride (LiCl) as a hydraulic tracer
(see Section 6), Li was measured with a Perkin-Elmer model 305A atomic
absorption spectrophotometer.
DISC SOLIDS SAMPLES
Premeasured plastic sampling tabs attached to disc surfaces prior to
start-up were removed from discs and placed in sterile containers. Sterile
distilled water, previously adjusted to pH 3 with H2SO^, was added to the
container. The contents were then mechanically homogenized for two minutes
and apportioned for appropriate microbiological and chemical analyses.
Total Solids Dry Weight
Total disc solids were quantitated using 10 to 25 mi portions of the
original homogenized sample. Samples were dried overnight at 103°C, cooled
in a desiccator, and weighed.
102
-------
TABLE B-l. CULTURE MEDIUM USED TO RECOVER IRON-OXIDIZING BACTERIA
FROM MINE WATER AND RBC DISC SOLIDS
Constituent Concentration, g/fc
S04 0.15
KC1 0.15
0.15
7IU3 3.36
CaCl2 0.97
2.25
MnS04-H20 0.12
FeS04-7H2Oa 7.46
alron source added separately, then medium adjusted to pH 3.0 with IN H2S04
and sterilized by passage through 0.20-)jm membrane filter (Mlllipore Corp.)
103
-------
TABLE B-2. CULTURE MEDIUM USED TO RECOVER HETEROTROPHIC BACTERIA
_ FROM MINE WATER AND RBC DISC SOLIDS3 (51)
Solution .Constituent Concentration,
Ab FeS04-7H20 33.4
BC (NH)S0 6.0
KC1 0.2
1.0
Ca(N03)2 0.02
Cd Purified Agar L28e 7.0
aSolution B was added to C at room temperature and gently mixed. Then
solution A was added to the combination and mixed well.
Solution A (300 m£) was adjusted to pH 2.5 with 6M H2S04 and sterilized by
passage through a 0.20-nm membrane filter (Millipore Corp.).
°Solution B (550 M) was adjusted to pH 3.0 with 6 M H2SO, and autoclaved at
121°C for 15 min.
A suspension of agar (150 mi) was soaked for 15 min and autoclaved at 121°C
for 15 min.
g
Flow Laboratories, Rockville, Md.
104
-------
Total Iron
Total Fe in disc solids was determined with the aid of a Perkin-Elmer
model 305A atomic absorption spectrophotometer using dilute samples which had
been dissolved in 50-percent HC1.
Aluminum
A Perkin-Elmer model 305A atomic absorption spectrophotometer was used
to determine Al concentration using acidified solutions of disc solids
samples.
Iron-oxidizing Bacteria
Iron-oxidizing bacterial densities in disc solids were estimated using
the procedure outlined previously for their enumeration in mine water.
Heterotrophic Bacteria
The enumeration of heterotrophic bacteria in disc solids was performed
using the procedure previously described for their determination in mine
water.
MICROSCOPY
Phase-contrast Microscopy
Living cells were observed and photographed by preparation of glass
slide wet mounts under a Zeiss Standard Universal microscope with 35-mm
camera (Carl Zeiss, Oberkochen/Wuert., West Germany).
Scanning Electron Microscopy
Glass cover slips were attached to clean discs on the first stage of
the 2.0-m RBC unit prior to start-up at Hawk Run, Pa. The samples were
removed at various intervals and prepared for observation under the SEM.
For initial preparation, samples were dipped in a one-percent solution of
osmium tetroxide (OsO^) in Kellenberger buffer (38) and then exposed to OsO/
vapors for 1.5 hours. Fixed samples were dehydrated by successive soakings
in 25, 50, 75, 95, and 100 percent ethyl alcohol, transferred to fresh 100
percent alcohol, and then dried with a Pelco model H critical point dryer
(Ted Pella Co., Tustin, Calif.). Dried samples received a sputter coating
of gold and were placed in a desiccator for future observation under an ISI
Super IIIA scanning electron microscope with 35-mm camera (International
Scientific Instruments, Inc., Santa Clara, Calif.).
Membrane filters recovered microorganisms from mine water and received
the same preparation prior to viewing under the SEM as did glass cover slips.
Nucleopore brand polycarbonate filters which contained 0.4-ym pores were used
as suggested by Todd and Kerr (64).
105
-------
APPENDIX C
MIXING CHARACTERISTICS OF MINE DRAINAGE STREAMS
AT HAWK RUN, PA
The extent of lateral mixing which occurred in the 50-m section below
the confluence of Hawk Run and Mine Hole 1 was evaluated to determine
whether there was sufficient mixing in the reach so that the influent mixture
pumped to the RBC units would be significantly different in quality from the
individual waters to represent a third type of drainage. Also determined was
whether backmixing of Mine Hole 1 discharge affected the water quality at the
point where mine drainage was being pumped from Hawk Run. Stations were
established at eleven points downstream from the confluence and four points
upstream (Figure C-l). Water samples were collected from each station at
three equally spaced point across the stream. Measurements of Fe(II)
concentration were taken at each of these points (Table C-l). Fe(II) is not
a conservative substance; however, Fe(II) concentrations were equal for
stations 1 and 2, which covered approximately the same distance as was
investigated in the analysis of mixing. Therefore, Fe(II) was considered
within the reach under study.
Fe(II) concentrations across the stream at station 3 were similar;
therefore, backmixing did not affect the mine water quality at the intake
line for Hawk Run. Fe(II) measurements for station 12 (Hawk Run + 1 intake)
indicated that complete mixing had not yet occurred. Complete mixing was
evident at some location downstream from the intake for the combined stream
at either station 14, 15, or beyond.
106
-------
DIRECTION
OF FLOW
15
RBC
BUILDING
INFLUENT
PIPING
Figure C-l. Schematic of mine drainage streams at Hawk Run, Pa.,
showing locations of sampling stations and mine
water feed lines.
107
-------
TABLE C-l. MINE WATER Fe(II) CONCENTRATIONS AT STATIONS
ON STREAMS AT HAWK RUN. PA.a
Fe(II) concentration, mg/£
Station13 Left
1
2
3 32
4 32
5 31
6 34
7 31
8 32
9 32
10 31
11 32
12 32
13 34
14 38
15 38
Middle
31
31
32
32
34
57
55
47
44
40
36
36
35
38
37
Right0
—
—
31
53
57
53
45
45
45
44
44
40
40
40
40
Fe(II) concentration of mine hole 1 = 57 mg/Jl.
See Figure C-l for location of stations.
'Corresponds to side of Hawk Run Stream where RBC units were located.
108
-------
APPENDIX D
CHARACTERISTICS OF EXPERIMENTAL RBC UNITS
The RBC field units used in this study were commercial devices equipped
with 0.5-m and 2.0-m diameter discs (Autotrol Corporation, Milwaukee, Wise.)
designed specifically for experimental pilot plant use. Bench-scale RBC
units were designed to scale from pilot units, except the 10-cm unit, which
was equipped with only two stages instead of the normal four (Table D-l).
DRIVE SYSTEM
Disc rotation for the 10-cm, 15-cm, and 0.5-m units was supplied by
93, 47, and 186 watt, respectively, 115 v, single phase, gear-reduced motors.
Discs of the 2.0-m RBC were rotated by a 0.746 kw, 230/460 v, three phase,
variable speed motor. All RBC units were equipped with sprocket and chain
drive systems.
FLOW PATTERN
An influent chamber and bucket feed system supplied a constant flow of
mine water to the stages of the RBC field units. Flow through each stage
was formed in a serpentine pattern for the 2.0-m RBC and by a 2.5-cm diameter
hole in each baffle for the 0.5-m unit (Figure D-l). Flow was metered to the
bench-scale units and influent feed chambers were not used. Flow through the
stages of bench-scale units was facilitated by small holes between baffles
in a fashion similar to the 0.5-m field RBC.
RBC CONSTRUCTION
Trough and disc material used in construction of bench-scale units were
acrylic plastic. Corrosion-resistant aluminum and steel were used for the
0.5- and 2.0-m RBC troughs, respectively. Disc construction material for
commercial RBC units was composed of vacuum-formed, 1-mm thick, high-density
polyethylene which was heat-welded, shaped in a honeycomb configuration, and
stacked on square corrosion-resistant steel shafts. CPVC tubing served as
shafts for both bench-scale units.
The 2.0-m RBC was equipped with four perforated tubes attached to the
periphery of discs on each stage and spaced at 90° intervals. Each 0.5-m
length of PVC pipe was 7.6 cm in diameter and contained approximately ten
2.5-cm diameter holes which release air below the liquid surface and apply
mine water to the discs during rotation out of the liquid. The tubes were
designed in this manner to improve turbulence and minimize solids deposition
in the reactor compartment.
109
-------
TABLE D-l. DESIGN AND NORMAL OPERATIONAL PARAMETERS OF RBC EXPERIMENTAL
UNITS
Parameter
Number of stages
Disc diameter, m
o
Disc surface area, m
Trough volume, liters
Surface-to-volume ratio,
Disc rotation
Revolutions per min
Peripheral velocity,
m/min
Hydraulic
Flow rate, m-Vd
Loading, m^/d-m^
10- cm
Bench-scale
RBC unit
0.10
0.090
0.60
157
57
19
0.014
0.16
15- cm
Bench-scale
RBC unit
0.15
0.438
3.0
146
38
19
0.072
0.16
0.5-m
Pilot
RBC
unit
0.5
21.8
140
156
13
19
3.41
0.16
2.0-m
Prototype
RBC unit
2.0
738.1
3,800
194
2.9
19
115
0.16
110
-------
INFLUENT
ff
INFLUENT
•F?
FEEDWELL
~ T
; ;
S2 1 (D
i !
— ~
S3 1 (D
; ;
S4
la
EFFLUENT
k 0 /
L-
1
FEED
— S
•— S
o
WELL
©
"*"
2 -*-
5 -*•
4_^
•^-
— N
J
_-X
k
w
__x
EFFLUENT
A
Figure D-l. Plan and side view showing flow pattern and stage sampling
locations for trough of 0.5-m RBC (left) and 2.0-m RBC
(right). Drawing is not to scale.
Ill
-------
APPENDIX E
FORMULAS FOR STATISTICAL ANALYSIS
MEAN X AND Y
N
X
X =
1=1
and
Y =
N
£ Y
1=1
N N
where X., Y. = values of datum point, 1, and
N = total number of data points.
MEAN pH
Mean pH
N
Z pH
1=1
N
where pH. = value of datum point, 1.
MEAN BACTERIAL POPULATIONS
Mean Bact. = antilog
N
E log bact..
1=1 I
N
where bact.. = value of datum point, i.
112
-------
STANDARD DEVIATION
-i 2
Std. Dev. =
N 2
Z X. -
1=1
N
I X
1=1
L N J
N-l
LINEAR REGRESSION EQUATION
Y = mX + b
SLOPE OF THE REGRESSION LINE, (m)
m =
N
£ X Y
1=1
N
- X Y
I X
where „ .
variance of X values,
N
ORDINATE INTERCEPT OF THE REGRESSION LINE, (b)
b - Y - mX
CORRELATION COEFFICIENT, (R)
R =
N
I X Y
1=1 x
N
- X Y
a a
x y
N
E Y
where a •
y N
- variance of Y values,
113
-------
STUDENT'S t-TEST FOR SIGNIFICANCE OF CORRELATION COEFFICIENT
R
0
_RJ
N - 2
114
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/7-80-006
3. RECIPIENT'S ACCESSI Ol> NO.
4. TITLE ANDSUBTITLE
ROTATING DISC BIOLOGICAL TREATMENT OF ACID MINE
DRAINAGE
5. REPORT DATE
January 1980 issuing date
6. PERFORMING ORGANIZATION CODE
'. AUTHOR(S)
Harvey Olem and Richard F. Unz
8. PERFORMING, ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
The Pennsylvania State University
University Park, Pennsylvania 16802
10. PROGRAM ELEMENT NO.
EHE623
11. CONTRACT/GRANT NO.
R 805132
12. SPONSORING AGENCY NAME AND ADDRESS
Industrial Environmental Research Lab,
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
- Cinn, OH
13. TYPE OF REPORT AND PERIOD COVERED
Final 5/77-10/78
14. SPONSORING AGENCY CODE
EPA/600/12
15. SUPPLEMENTARY NOTES
16. ABSTRACT Pilot scale (0.5-m diameter) and prototype (2.0-m diameter) rotating
biological contactors (RBC) were investigated for oxidation of ferrous Fe(II) iron con-
tained in six heterogeneous mine waters located at three coal mining sites in Pennsyl-
vania and West Virginia. Continuous biological oxidation of Fe(II) and Fe(III) was
accomplished at natural mine water temperatures as low as 0.4°C at Hawk Run, Pa. and
as high as 29°C at Crown, W. Va. Reduction of Fe(II) oxidation efficiency at 0.4°C
amounted to 10 to 20 percent of that achieved at 10°C. Oxidation efficiency was above
80 percent at mine water temperatures of 10 to 29°C. Microbiological oxidation with
the 0.5-m RBC was unaffected at influent mine water pH values in the range of 2.18 to
5.50 (Crown, W. Va.). Fe(II) oxidation was an average 10 percent less efficient for
a mine water treated under similar operating conditions with the 2.0-m than with the
).5-m RBC. The observed decrease may be due to nonmicrobiological factors such as
increased short-circuiting, lower residence time, and a smaller effective surface area
which may be increased through proper design. Costs for Fe(II) oxidation with the RBC
were estimated to be about twice the amortized capital costs and one-half the operating
costs compared to a conventional chemical oxidation process. Neutralization of RBC
effluent and separation of precipitated iron solids is required to produce water of
suitable quality for stream-release. Both iron-oxidizing and heterotrophic bacteria
existed in a gelatinous matrix present on disc surfaces of RBC units operating at,
Pa.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Bacteria, Biological Treatment, Cost
Comparison, Iron, Oxidation, Water
Pollution Control
Acid Mine Drainage
Autotrophic Bacteria
Coal Mine Drainage
Ferric Iron
Ferrous Iron
Iron-oxidizing Bacteria
Rotating Biological Con-
tactor
13B
8. DISTRIBUTION STATEMEN1
Release to public
19. SECURITY CLASS (ThisReport/
unclassified
21. NO. OF PAGES
131
20. SECURITY CLASS (This page)
unclassified
22. PRICE
EPA Form 2220-1 (9-73)
115
* U.S. GOVERNMENT BUNTING WFICE: 1WO-657-146/5552
------- |