1. The Passive Treatment of Coal Mine Drainage
      2. Abstract
      3. Acknowledgements
      4. Contents
      5. Figures
      6. Tables
      7. Introduction
      8. Background
      9. Passive Treatment Processes
      10. Materials and Methods
      11. Removal of Contaminants by Passive Unit Operations
      12. Designing Passive Treatment Systems
      13. Conclusions
      14. Abbreviations and Acronyms
      15. References

Exhibit I
:
Biological Stream Ratings of Significance
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Cook Co.
Pike Co.
Lee Co.
Will Co.
McLean Co.
LaSalle Co.
Iroquois Co.
Ogle Co.
Fulton Co.
Henry Co.
Bureau Co.
Adams Co.
Knox Co.
Shelby Co.
Livingston Co.
Vermilion Co.
Fayette Co.
Hancock Co.
Champaign Co.
Peoria Co.
Wayne Co.
Macoupin Co.
Edgar Co.
Logan Co.
Madison Co.
Ford Co.
DeKalb Co.
Sangamon Co.
Kane Co.
Coles Co.
St Clair Co.
Macon Co.
Mercer Co.
Christian Co.
Clay Co.
Mason Co.
Lake Co.
Tazewell Co.
Clark Co.
Jackson Co.
Jasper Co.
Greene Co.
Marion Co.
White Co.
Whiteside Co.
Piatt Co.
Kankakee Co.
Morgan Co.
JoDaviess Co.
Perry Co.
Warren Co.
Jefferson Co.
Pope Co.
McHenry Co.
Cass Co.
Randolph Co.
Clinton Co.
Carroll Co.
Montgomery Co.
Bond Co.
Union Co.
Woodford Co.
Jersey Co.
McDonough Co.
DeWitt Co.
Stephenson Co.
Saline Co.
Monroe Co.
Grundy Co.
Douglas Co.
Washington Co.
Crawford Co.
Winnebago Co.
Schuyler Co.
Franklin Co.
Hamilton Co.
Effingham Co.
Marshall Co.
Johnson Co.
Brown Co.
DuPage Co.
Stark Co.
Boone Co.
Lawrence Co.
Rock Island Co.
Williamson Co.
Moultrie Co.
Richland Co.
Scott Co.
Menard Co.
Kendall Co.
Gallatin Co.
Cumberland Co.
Massac Co.
Wabash Co.
Pulaski Co.
Hardin Co.
Putnam Co.
Henderson Co.
Calhoun Co.
Alexander Co.
Edwards Co.
Biologically Significant Stream
Major River
Third Order or Larger Stream
R
o
c
k
R
i
v
e
r
M
i
s
s
i
s
s
i
p
p
i
R
i
v
e
r
K
a
s
k
a
s
k
i
a
R
i
v
e
r
K
a
s
k
a
s
k
i
a
R
i
v
er
O
h
i
o
R
iv
e
r
C
a
c
h
e
R
i
v
e
r
L
a
M
o
i
n
e
R
i
v
e
r
M
a
c
o
u
p
i
n
C
r
e
e
k
B
e
a
u
c
o
u
p
C
r
e
e
k
B
i
g
M
u
d
d
y
R
i
v
e
r
S
a
l
i
n
e
R
i
ve
r
S
k
i
l
l
e
t
F
o
r
k
L
i
t
t
l
e
W
a
b
a
s
h
R
i
v
e
r
E
m
b
a
r
r
a
s
R
i
v
e
r
W
a
b
a
s
h
R
i
v
e
r
W
a
b
a
s
h
R
i
v
e
r
O
h
i
o
R
i
v
e
r
M
i
s
s
i
s
s
i
p
p
i
R
i
v
e
r
I
l
l
i
n
o
i
s
R
i
v
e
r
S
a
n
g
a
m
o
n
R
i
v
er
S
p
o
o
n
R
i
v
e
r
M
a
c
k
i
n
a
w
Ri
v
e
r
G
r
e
e
n
Rive
r
I
l
l
i
n
o
i
s
R
i
v
e
r
V
e
r
m
i
l
l
i
o
n
R
i
v
e
r
M
i
s
s
i
s
s
i
p
p
i
R
i
v
e
r
P
e
c
a
t
o
n
i
c
a
R
i
v
e
r
F
o
x
R
i
v
e
r
I
l
l
i
n
o
i
s
R
i
v
e
r
D
e
s
P
l
a
i
n
e
s
R
i
v
e
r
V
e
r
m
i
l
i
o
n
R
i
v
e
r
K
a
n
k
a
k
e
e
R
i
v
e
r
Biological Stream Ratings
for Significance
September, 2008
Office of
Resource Conservation
Biologically
Significant Streams
(BSS): streams that
have a high rating or
score based on data
from at least two
taxonomic groups. This
can
be
achieved
by
obtaining an "A" rating either
for diversity or for integrity
that is based on data from two or
more taxonomic groups. A second way
to achieve this status is for a stream
segment to have class scores in the highest
class for at least two different taxonomic
groups when considering the combined data
from the diversity and integrity ratings.
Stream segments identified as biologically significant
are unique resources in the state and the biological
communities present must be protected at the stream
reach, as well as upstream of the reach. Therefore
BSS reaches were extrapolated from site-specific
information to upstream stream segments to arrive
at the segments identified as biologically
significant.
Biologically
Significant
Streams
(BSS)
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Exhibit J
:
Hunt Letter, August 2
nd
, 2007

Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Exhibit K
:
Memorandum re: Viable Technologies for Treating Coal Mine
Stormwater Runoff

CARPENTER ENVIRONMENTAL ASSOCIATES, INC.
307 Museum Village Road
P.O.Box 656
Monroe, New York 10950
Phone: 845-781-4844
Fax: 845-782-5591
Senders E-mail: r.pape@cea-enviro.com
MEMORANDUM
Date:
October 15, 2008
To:
Sierra Club, Illinois Chapter
From:
Robert J. Pape, P.E.
Re:
Viable Technologies for Treating Coal Mine Stormwater Runoff
CEA No. 08047
Carpenter Environmental Associates has been retained by Sierra Club, Illinois Chapter, to research viable
technologies for treating stormwater runoff from coal mines. Mine stormwater runoff can contain a
variety of metals as well as sulfate, chlorides, and total suspended solids (TSS). The following paper
identifies technologies that have shown to be effective in removing these pollutants.
1) Bioremediation
Constructed wetlands and bioreactors are two types of bioremediation processes that are
employed to treat mine runoff containing metals and sulfate.
1
In general, constructed wetlands (CW) can be single or multi-basins designed to cultivate
biology in an environment void of oxygen (anaerobic), or cultivate biology in an oxygen rich
environment (aerobic). A multi-basin system can consist of both aerobic and anaerobic basins.
Aerobic ponds and wetlands are effective for removing iron from net-alkaline mine water.
2
Anaerobic CWs commonly aim to establish a sulfate reducing bacteria (SRB) population.
Production of SRB results in the precipitation of dissolved metals, including those found in
mine run-off, as metal-sulfide complexes as well as reducing sulfate and increasing pH.
3
Successive Alkalinity Producing Systems (SAPS) is a specific type of CW that consists of an
organic mulch layer, limestone layer and a drainage system that has flushing capabilities to
insure that iron and aluminum precipitates, which may be
contained in the mine runoff,
will
not clog the cell.
4
SAPS can remove iron, aluminum, manganese, and sulfate as well as trace
metals including barium, cadmium, chromium, copper, nickel, zinc and lead
as can be found
in mine runoff
.
5
Runoff from a former coal mining operation in Gowen, Oklahoma, was
treated using SAPS. Although actual data was not obtainable for the Gowen SAPS, the report
regarding the Gowen SAPS found on the USEPA’s website indicated that “concentrations of
iron, aluminum, and manganese have decreased significantly” and the trace metals were
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

reduced to near or below detection levels.
6,7
The success of this project spurred Oklahoma to
use the Gowen SAPS design in at the Tar Creek superfund site in Ottawa, Oklahoma, and is
being investigated for use in several watersheds nationwide and should be evaluated for coal
mines in Illinois.
8
A multi-basin wetland system was used in Albright, West Virginia, and in Springdale, West
Virginia. These wetlands consisted of limestone rock drains and manganese-oxidizing bacteria
to treat metal-contaminated leachate from a closed coal combustion byproduct landfill. The
leachate entering these systems contain pollutants
similar to mine runoff
, specifically, iron,
manganese and total suspended solids (TSS). The Albright CW treatment system reduced an
influent stream containing: 2.5 mg/L total iron, 8.5 mg/L manganese, and 8 mg/L TSS to 0.3
mg/L, 1.1 mg/L, and 1 mg/L, respectively.
9
For comparison to the situation in Illinois,
Albright’s effluent are at or below the Deer Run Mine Draft NPDES Permit (Permit) levels for
iron, manganese, and TSS. The Springdale CW treatment system reduced an influent stream
containing: 12.5 mg/L total iron, 6.1 mg/L dissolved iron, 2.7 mg/L manganese, and 25 mg/L
TSS to 0.3 mg/L, 0.1 mg/L, 0.2 mg/L, and 8mg/L, respectively.
10
Springdale’s effluent
concentrations are below the Permit levels for iron, manganese, and TSS.
Bioreactors are lined trenches, pits, or aboveground tanks that can contain a variety of
materials as packing or have no packing and are entirely suspended growth reactors.
11
Bioreactors are either anaerobic or aerobic and can be a standard technology or proprietary
technology. Like CWs, aerobic bioreactors can remove iron and aluminum.
12
Bioreactors that
cultivate SRB biology can be used to eliminate metals commonly found in mine runoff,
including iron and manganese, as seen with the SRB bioreactor system in Champagne Creek,
Idaho which removed up to 91% of the iron in the influent waste stream.
13,14
A pilot study on a standard technology type bioreactor investigated the effects of an aerated
dolomite packed bed bioreactor and an aerated quartzite packed bed bioreactor on the
reduction of manganese in mine water that is net-alkaline.
15
Each reactor was fed net-alkaline
mine water containing 15 – 30 mg/L manganese.
16
The study concluded that these bioreactors
reduced the manganese by 90 – 97%.
17
One proprietary bioreactor system is ABMet®. ABMet® systems are configured for site
specific waste streams and have been successfully used to treat mining water.
18,19
ABMet®
has shown to remove selenium, arsenic, mercury, chromium, cadmium, copper, zinc, cobalt,
nickel, antimony and nitrate.
20,21
Depending upon the ABMet configuration, these metals can
be reduced 99+% and some metals can be reduced to 10 part per billion (ppb) or less.
22, 23
Bioremediation has also been extensively used in the neutralization of acidic water (pH
control), nutrient reduction and cyanide reduction.
2)
Reverse Osmosis
Vibratory Shear Enhanced Process (VSEP) is a proprietary RO type process.
24
In addition to
employing the technology of the pressure gradient across a membrane, VSEP adds
“…torsional vibration of the membrane surface, which creates high shear energy at the surface
of the membrane. The result is that colloidal fouling and polarization of the membrane due to
concentration of rejected materials are greatly reduced.”
25
A case study of VSEP treating acid
mine water showed VSEP reduced the following constituents: total dissolved solids, calcium,

magnesium, sodium, iron, manganese, copper, zinc and sulfates. Specifically, a mine water
stream containing iron at 1,100 mg/L and sulfate at 8,000 mg/L were reduced to <0.1 mg/L,
and 100 mg/L, respectively by VSEP.
26
A case study to treat an RO reject stream (the RO unit
treated brackish well water) used VSEP as stage one of a two stage process and used a
conventional spiral RO system as stage two. The RO reject stream contained similar
pollutants such as chloride, sulfate and total dissolved solids
as does mine water runoff.
The
study showed that VSEP reduced chloride, sulfate, nitrate, total dissolved solids, boron, and
sodium. Specifically, an influent waste stream containing chloride at 3,285 mg/L, sulfate at
304 mg/L, and total dissolved solids at 7,314 mg/L were reduced to 628 mg/L, 25 mg/L, and
1,617 mg/L, respectively by VSEP.
27
The study also showed that the conventional spiral RO
system produced a final treated water stream containing chloride at 11 mg/L, sulfate at 0 mg/L,
and total dissolved solids at 51 mg/L. Both stages produce an effluent significantly below the
Permit levels for sulfate and the two stage process reduces chlorides significantly below
Permit levels.
3) Chemical treatment includes the addition of hydroxide or lime to the waste stream prior to the
streams entry to a settling tank. Hydroxide and lime treatment removes metals using pH.
Hydroxide and lime are added to a waste stream to achieve pHs that specifically correspond to
the minimum solubilities of the metals being removed. Since different metals have their own
minimum solubility at different pHs, several treatment stages may be necessary if multiple
metals are to be removed from the waste stream.
Chemical treatment also includes the addition of proprietary chemicals to reduce heavy metals
or other constituents. Organosulfide TMT 15™ is a priority treatment chemical that can be
added to the waste steam prior to a settling tank.
28
Organosulfide TMT 15™ promotes
precipitation of cadmium, copper, lead, mercury, nickel, and silver from wastewater streams.
29
Organosulfide TMT 15™ is typically used as a second step reagent after the bulk of the metals
are removed with hydroxide precipitation because Organosulfide TMT 15™ dosage is a
function of stoichiometry. In an application in the coal burning Pleasant Prairie Power Plant,
Organosulfide TMT 15™ is anticipated to reduce mercury in a flue gas desulfurization
wastewater stream from <2,000 ug/L to 0.5 ug/L.
30
4) Ion exchange removes unwanted ions from water by transferring them to a solid material
(resins), in an ion exchanger, which accepts them while giving back an equivalent number of
desirable ions contained in the ion exchanger. In the simplest terms, water softening is a form
of ion exchange in which sodium, from salt, is exchanged with the calcium responsible for
water "hardness." Selective Metal Ion Exchange has resins that have been tailored to removal
of heavy metals including: copper, uranium, vanadium, mercury, lead, nickel, zinc, cobalt and
cadmium.
31
Ion exchange has been used to treat mine wastewater for metals and nitrate
removal. An example of this method being utilized for sulfate removal is at the Sierrita copper
mine in Arizona.
32
Attachments
1
Acid Mine Drainage: Innovative Treatment Technologies, Christine Costello, October 2003.
2
The Passive Treatment of Coal Mine Drainage, U.S. Department of Energy National Energy Technology Laboratory,
Watzlaf, et. al., 2004

3
Acid Mine Drainage: Innovative Treatment Technologies, Christine Costello, October 2003.
4
Acid Mine Drainage: Innovative Treatment Technologies, Christine Costello, October 2003.
5
Acid Mine Drainage: Innovative Treatment Technologies, Christine Costello, October 2003.
6
Acid Mine Drainage: Innovative Treatment Technologies, Christine Costello, October 2003.
7
EPA Website Detailing Gowen, OK Alkalinity-Producing System Results, August 5, 2004.
http://www.epa.gov/owow/nps/Section319III/OK.htm.
8
Acid Mine Drainage: Innovative Treatment Technologies, Christine Costello, October 2003.
9
Applications of Passive Treatment to Trace Metals Recovery, Hoover, et.al.,
10
Applications of Passive Treatment to Trace Metals Recovery, Hoover, et.al.,
11
Acid Mine Drainage: Innovative Treatment Technologies, Christine Costello, October 2003.
12
Acid Mine Drainage: Innovative Treatment Technologies, Christine Costello, October 2003
13
Acid Mine Drainage: Innovative Treatment Technologies, Christine Costello, October 2003
14
Acid Mine Drainage: Innovative Treatment Technologies, Christine Costello, October 2003.
15
Rapid Manganese Removal from Mine Waters Using an Aerated Packed-Bed Bioreactor, Johnson K. L., Younger, P.
L., May 11, 2005
16
Rapid Manganese Removal from Mine Waters Using an Aerated Packed-Bed Bioreactor, Johnson K. L., Younger, P.
L., May 11, 2005
17
Rapid Manganese Removal from Mine Waters Using an Aerated Packed-Bed Bioreactor, Johnson K. L., Younger, P.
L., May 11, 2005
18
Treatment Technology Summary for Critical Pollutants of Concern in Power Plant Wastewaters, Chu, P., January
2007.
19
Water Online.com description of the Zenon ABMet process at http://www.gewater.com/products/equipment/
other_equipment/ABMet.jsp
20
Treatment Technology Summary for Critical Pollutants of Concern in Power Plant Wastewaters, Chu, P., January
2007.
21
Water Online.com description of the Zenon ABMet process at http://www.gewater.com/products/equipment/
other_equipment/ABMet.jsp
22
Treatment Technology Summary for Critical Pollutants of Concern in Power Plant Wastewaters, Chu, P., January
2007.
23
Water Online.com description of the Zenon ABMet process at http://www.gewater.com/products/equipment/
other_equipment/ABMet.jsp
24
2006 El Paso Desalination Conference, El Paso, Texas. A Comparison of Conventional Treatment Methods and
VSEP, a Vibrating Membrane Filtration System, Johnson, Greg, Larry Stowell, and Michele Monroe, March 2006.
25
2006 El Paso Desalination Conference, El Paso, Texas. A Comparison of Conventional Treatment Methods and
VSEP, a Vibrating Membrane Filtration System, Johnson, Greg, Larry Stowell, and Michele Monroe, March 2006.
26
VSEP Filtration of Acid Mine Drainage: A cost-effective and efficient processing solution, Case Study, New Logic
Research.
27
2006 El Paso Desalination Conference, El Paso, Texas. A Comparison of Conventional Treatment Methods and
VSEP, a Vibrating Membrane Filtration System, Johnson, Greg, Larry Stowell, and Michele Monroe, March 2006.
28
Treatment Technology Summary for Critical Pollutants of Concern in Power Plant Wastewaters, Chu, P., January
2007.
29
Treatment Technology Summary for Critical Pollutants of Concern in Power Plant Wastewaters, Chu, P., January
2007.
30
Treatment Technology Summary for Critical Pollutants of Concern in Power Plant Wastewaters, Chu, P., January
2007.
31
Treatment Technology Summary for Critical Pollutants of Concern in Power Plant Wastewaters, Chu, P., January
2007.
32
Sulphate removal demonstration plant using BioteQ's proprietary Sulf-IX ion-exchange technology (www.bioteq.ca)

Exhibit L
:
Acid Mine Drainage: Innovative Treatment Technologies,
October 2003
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage:
Innovative Treatment Technologies
October 2003
Prepared by
Christine Costello
National Network of Environmental Management Studies Fellow
for
U.S. Environmental Protection Agency
Office
of Solid Waste and Emergency Response
Technology Innovation Office
Washington, DC
www.clu-in.org
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage: Innovative Treatment Technologies
NOTICE
This document was prepared by a National Network of Environmental Management Studies
grantee under a fellowship from the U.S. Environmental Protection Agency. This report was not
subject to EPA peer review or technical review. EPA makes no warranties, expressed or implied,
including without limitation, warranties for completeness, accuracy, usefulness of the information,
merchantability, or fitness for a particular purpose. Moreover, the listing of any technology,
corporation, company, person, or facility in this report does not constitute endorsement, approval,
or recommendation by EPA.
i

Acid Mine Drainage: Innovative Treatment Technologies
FOREWORD
About the National Network for Environmental Management Studies (NNEMS) Program
The U.S. Environmental Protection Agency (EPA) established the NNEMS program in 1986 to
foster
a growing interest among higher education students in environmental careers. The
NNEMS program is a comprehensive fellowship program that provides undergraduate and
graduate students an opportunity to participate in a fellowship project that is directly related to
their field of study. The NNEMS program is sponsored by EPA’s Office of Environmental
Education.
Students who are awarded NNEMS fellowships are offered a unique opportunity to gain research
and
training experience directly linked to their undergraduate or graduate studies. NNEMS
fellows conduct research projects to augment their academic studies, which EPA supports with
financial assistance.
Each year, the NNEMS program offers approximately 50 to 60 research projects, developed and
sponsored
by EPA Headquarters in Washington, D.C. and EPA’s ten regional offices throughout
the U.S. The projects allow students to conduct research while working full-time at EPA during
the summer or part-time during the school year.
ii

Acid Mine Drainage: Innovative Treatment Technologies
CONTENTS
Page
Purpose ................................................................... 1
1.0
Introduction .......................................................... 1
1.1 Background ....................................................... 2
1.2 Chemistry ......................................................... 3
1.2.1 Acid Generation and Metal Leaching ............................. 3
1.2.2 Neutralization and Metals Removal .............................. 5
1.3 Environmental Concerns ............................................. 6
2.0
Treatment ........................................................... 8
2.1 Traditional ........................................................ 8
Case Study: California Gulch, Leadville, CO ............................ 9
2.2 Innovative ....................................................... 11
2.2.1 Limestone Drains ........................................... 12
Case Study: North Pennine, Orefield, UK ....................... 13
2.2.2 Constructed Wetlands ....................................... 13
Case Study: Burleigh Tunnel, CO
.......................... 14
2.2.3 Bioreactors ............................................... 15
A. Case Study: Silver Bow County, MT ........................ 16
B. Case Study: Champagne Creek, Butte, ID .................... 17
2.2.4 Successive Alkalinity Producing Systems ......................... 19
A. Case Study: Oven Run, PA ................................ 20
B. Case Study: #40 Gowen, Gaines Watershed,OK ................ 20
2.2.5 Permeable Reactive Barriers .................................. 21
2.2.6 Biosolids ................................................. 23
A. Case Study: Frostburg, MD ............................... 24
B. Case Study: Leadville, CO ................................ 26
2.2.7 Phytoremediation ........................................... 27
3.0
Conclusion .......................................................... 29
4.0
References .......................................................... 29
Appendix A - Description of AML Programs in Selected Western States ................. 37
Appendix B - Brief Case Description of Case Studies Found In This Report .............. 47
FIGURES
Figure 1. Air Compressor at an abandoned mine site in Leadville, CO .................... 7
Figure 2. Capped waste pile in Leadville, CO ..................................... 10
iii
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage: Innovative Treatment Technologies
Figure 3. Burleigh Tunnel, 7/2003 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Figure 4. Schematic of Champagne Creek Bioreactor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Figure 5. General Schematic of a SAPS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Figure 6. Oven Run, SAPS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Figure 7. SAPS cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Figure 8. Schematic of Non-pumping well PRB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Figure 9. Plant biomass yields two years after biosolids application . . . . . . . . . . . . . . . . . . . . . 25
Figure 10. Relationship between total plant metals and rates of application
of biosolids and other amendments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Figure 11. Metal salt accumulation on soils in Leadville, CO on the banks of the
Arkansas River . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
iv

Acid Mine Drainage: Innovative Treatment Technologies
Purpose
Currently there is no comprehensive survey about the types of remediation technologies being
used to treat abandoned mines. The purpose of this paper is to provide information about this
topic with a particular focus on hard rock mining sites. Hard rock mines can be loosely defined as
non-coal, metal mines, in the United States these mines are located in the Mid-West and Western
states. This paper provides an overview of treatment technologies being used to remedy
environmental problems at abandoned mine sites, with a focus on innovative treatment techniques.
1.0 Introduction
This report aims to identify abandoned mine sites that utilize innovative technologies to treat mine
drainage or contaminated soils and to put that information into a database. Therefore, this paper
is not a highly detailed description of a single technology, but rather an introduction to a variety
of technologies currently used to treat mine sites in the country. A database was created to
compliment this paper. It contains all of the case studies highlighted in this paper and quite a few
other. It is available through the Technology Innovation Program website www.cluin.org. The
goal of this database is to allow parties interested in implementing innovative treatments at AMD
sites to learn from past successes and failures to advance these technologies.
A variety of sources were consulted to identify sites. Government agencies were the main targets
as they
are the most likely group to be addressing abandoned mines that presumably do not have a
linked responsible party. Also, unlike consulting firms and private industry, i.e. the companies
themselves, the government will generally disclose most technical information. The internet was
used to look up information about the following agencies:
!
Environmental Protection Agency - specifically the Comprehensive Environmental
Response, Compensation, and Liability Act (CERCLA, a.k.a. Superfund) and the
Clean Water Act (CWA) programs;
!
Department of the Interior - specifically the Office of Surface Mining, Bureau of
Land Management (BLM), and United States Geological Survey (USGS);
!
State environmental departments;
!
Local environmental committees and community groups.
Another research avenue was grant distributions from CWA programs to local organizations. For
example,
Pennsylvania’s Growing Greener grant program funds many Abandoned Mine Land
(AML) reclamation efforts in the name of improving local watersheds. This did not prove
beneficial for every state. Where possible individuals were contacted through email and telephone
for more specific information about programs and sites. As with any project an extensive
literature review was done. Science Direct was searched, the Library of Congress catalog and
on-line databases were also utilized and numerous conference proceedings were perused.
Due to the universe of abandoned mines, constantly evolving programs and projects this report
does
not imply a complete picture of all projects and programs in the nation. The report is,
however, a start toward understanding what technologies are being used and some of the barriers
1
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage: Innovative Treatment Technologies
to remediation, especially for non-traditional treatments. Appendix A includes brief summaries on
the efforts being taken at the state level to address AML sites for selected states.
1.1 Background
Mining practices, present and past, cause environmental problems that can damage ecosystems
and
human health. Mining disturbs geologic formations that took millions of years to form;
likewise, related natural systems and processes are disturbed, e.g. hydrology. Once disruption
has taken place a variety of problems may arise, from physical hazards to pollution of water and
soil. The most severe and widespread environmental problems almost always have to do with
water, indeed all of the treatment technologies that will be discussed in this paper are designed for
water or the prevention of water contact with solids.
There has been a lot of effort to quantify the universe of abandoned mines, the results vary. Some
of
the problem lies in definition. While some agencies define a site as a particular opening; others
define a site as all of the openings at a particular location as one mine site. The Bureau of Land
Management claims that estimates from Federal land management agencies, state and privately
owned lands have ranged from about 80,000 to hundreds of thousands of small to medium-sized
sites (U.S. Dept. of the Interior, 2003b). The Office of Surface Mining describes the problem in
terms of money, “of the $8.2 billion of high priority [physical hazards] coal related AML problems
in the AML inventory, $6.6 billion, 80%, have yet to be reclaimed; furthermore, “almost ninety
percent of the $2.0 billion of coal related environmental problems in the AML inventory are not
reclaimed. And this represents only a small part of the total problem as no systematic effort has
been made to inventory these problems” (U.S. Dept. of the Interior, 2002a). To give one last
perspective, the Mineral Policy Center, a non-profit organization, claims that there are 557,000
abandoned mines - mostly in the western United States (2003). Although it is difficult to say
exactly how many sites exist, the number of abandoned mine sites in the US is enormous.
For roughly 25 years there have been efforts to address the dangers created by the past 250 - 300
years
of large-scale mining in this country. The Surface Mining Coal and Reclamation Act
(SMCRA), passed in 1977, requires a tax on coal production to be set aside in a fund for
remediation efforts at abandoned coal mines. However, many abandoned mines are hard rock
mines and are typically not eligible for SMCRA funding, though there are some exceptions. Other
sources of funding may come from CWA grants, CERCLA grants or State funding. While there
has been significant progress, there are still many sites without adequate funding. For example,
California has no abandoned coal mines, therefore ineligible for SMCRA funding. A multi-
stakeholder task force in California identified lack of funding as a key impediment to cleanup of
abandoned mines in the state (see Appendix A). Some states have started to lobby for funding,
for example, Colorado House Representative Mark Udall is seeking legislation that would create
a fund for hard rock sites similar to that created by SMCRA.
Many states and agencies have only recently finished inventorying the number of sites and begun
to
evaluate sites to determine priorities for cleanup. States and other agencies that are doing
remediation under SMCRA must address Priority 1 & 2 problems - those dealing with physical
dangers - before they are able to use funding to address Priority 3 problems - environmental
2

Acid Mine Drainage: Innovative Treatment Technologies
problems and/or high priority non-coal sites. The priority number system was defined by the U.S.
Department of the Interior.
Due to limited resources, especially in the case of hard rock mines, innovative technologies can
offer a plausible solution to the environmental threats created by abandoned mines. Traditional
water treatments are modeled after wastewater treatment plants, which are machine intensive,
chemical dependant, and require continuous operations and maintenance (O & M) staff.
Traditional solid mine waste remediation tactics involve covering of piles and water diversion
tactics which do not treat wastes but rather mitigate their impacts. The innovative technologies
that will be discussed in this paper are largely passive treatment systems. Passive treatment
systems are described as having little O & M costs, require little chemical application, and few if
any mechanical devices (Hedin et al., 1994). Passive treatment systems can be a good solution for
small drainage sites that might otherwise have few treatment options.
1.2 Chemistry
1.2.1 Acid Generation and Metals Leaching
Acid generation and metals dissolution are the primary problems associated with pollution
from
mining activities. The chemistry of these processes appears fairly straightforward, but becomes
complicated quickly as geochemistry and physical characteristics can vary greatly from site to site.
This paper will not describe these variables and their affects on chemistry, it will give an overview
of the most common scenario found at coal and hard-rock sites with environmental concerns.
Pyrite (FeS
2
) is responsible for starting acid generation and metals dissolution in coal and hard
rock sites alike. When pyrite is exposed to oxygen and water it will be oxidized, resulting in
hydrogen ion release - acidity, sulfate ions, and soluble metal cations, equation 1. This oxidation
process occurs in undisturbed rock but at a slow rate and the water is able to buffer the acid
generated. Mining increases the exposed surface area of these sulfur-bearing rocks allowing for
excess acid generation beyond the water’s natural buffering capabilities.
2FeS
2
(s) + 7O
2
(aq) + 2H
2
O –> 2Fe
+2
+ 4SO
4
-2
+ 4H
+
(1)
Further oxidation of Fe
+2
(ferrous iron) to Fe
+3
(ferric iron) occurs when sufficient oxygen is
dissolved in the water or when the water is exposed to sufficient atmospheric oxygen.
2Fe
+2
+ ½ O
2
+ 2H
+
–> 2Fe
+3
+ H
2
O
(2)
Some acidity is consumed in this process, however, the stage is set for further hydrogen ion
release
that will surpass these benefits. Ferric iron can either precipitate as ochre (Fe(OH)
3
the
red-orange precipitate seen in waters affected by acid mine drainage) or it can react directly with
pyrite to produce more ferrous iron and acidity.
2Fe
+3
+ 6H
2
O <–> 2Fe(OH)
3
(s) + 6H
+
(3)
3

Acid Mine Drainage: Innovative Treatment Technologies
14Fe
+3
+ FeS
2
(s) + 8H
2
O –> 2SO
4
-2
+ 15Fe
+2
+ 16H
+
(4)
When ferrous iron is produced as a result of equation 4 and sufficient dissolved oxygen is present
the
cycle of equations 2 & 3 is perpetuated (Younger, et al, 2002). Without dissolved oxygen
equation 4 will continue to completion and water will show elevated levels of ferrous iron
(Younger, et al, 2002).
Once the waters are sufficiently acidic, acidophilic bacteria - bacteria that thrive in low pH - are
able
to establish themselves. Microorganisms can play a significant role in accelerating the
chemical reactions taking place in mine drainage situations.
Thiobacillus Ferroxidans
, a bacteria,
is commonly referenced in this case. These bacteria catalyze the oxidation of ferrous iron, further
perpetuating equations 2 through 4. Another microbe belonging to the Archaea kingdom, named
Ferroplasma Acidarmanus
, has recently been discovered to also play a significant role in the
production of acidity in mine waters (Lauzon, 2000).
Though not a major source of acidity, the generation of hydrogen ions when certain metals form
precipitates
, must be taken into account when considering treatment options.
Al
+3
+ 3H
2
O <–> Al(OH)
3
+ 3H
+
(5)
Fe
+3
+ 3H
2
O <–> Fe(OH)
3
+ 3H
+
(see equation 3)
(6)
Fe
+2
+ 0.25 O
2
(aq) + 2.5 H
2
O <–> Fe(OH)
3
+ 2H
+
(7)
Mn
+2
+ 0.25 O
2
(aq) + 2.5 H
2
O <–> Mn(OH)
3
+ 2H
+
(8)
Other metals commonly found in mine drainage waters exist because they are present in the rocks,
similar
to pyrite. For example, there are a variety of other metal sulfides that may release metal
ions into solution, but may not generate acidity (Younger et al., 2002) the reasons for this are not
clear. Including:
Sphalerite
ZnS(s) + 2O
2
(aq) –> Zn
+2
+ SO
4
-2
(9)
Galena
PbS(s) + 2O
2
(aq) –> Pb
+2
+ SO
4
-2
(10)
Millerite
NiS(s) + 2O
2
(aq) –> Ni
+2
+ SO
4
-2
(11)
Greenockite CdS(s) + 2O
2
(aq) –> Cd
+2
+ SO
4
-2
(12)
Covellite
CuS(s) + 2O
2
(aq) –> Cu
+2
+ SO
4
-2
(13)
Chalcopyrite CuFeS
2
(s) + 4O
2
(aq) –> Cu
+2
+ Fe
+2
+ SO
4
-2
(14)
Metals are naturally dissolved from weathering slowly over time. The dissolution process is sped
up when
the pH of the water strays from near-neutral, that is at either high or low pH - in the case
of mine drainage low pH is the more plausible scenario (Younger et al., 2002; Blowes et al.,
2000). For more information see Chapter 2 of Mine Water: Hydrology, Pollution, Remediation
by Younger et al., 2002.
4
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage: Innovative Treatment Technologies
1.2.2 Neutralization and Metals Removal
The ways by which metals precipitate have seemingly endless possibilities and are not always well
understood. By far the most common application for reducing acidity and adding alkalinity is
lime. There are many ways to treat mine drainage through enhanced natural processes which
form the basis for passive treatments. There are many aerobic and anaerobic process that lead to
metals precipitation that are commonly practiced. Though not complete the following
information should provide some insight about the technologies that will be discussed shortly.
It is very important to gain control of the pH of the drainage because pH effects many things
including
the solubility of metals and the kinetics of the oxidation and hydrolysis processes (EPA,
Vol.4). In addition, the relationship between pH and metal removal processes varies among
metals and also between biotic and abiotic processes (EPA, Vol. 4)
Limestone (calcium carbonate), rich in calcite, increases the pH of water by consuming hydrogen
ions
and adding alkalinity through bicarbonate ions (Younger et al., 2002).
CaCO
3
+ 2H
+
= Ca
+2
+ H
2
O + CO
2
(15)
CaCO
3
+ H
2
CO
3
= Ca
+2
+ 2HCO
3
-
(16)
Once the pH of the acidic water has been raised metals can precipitate more easily to form
hydroxides
and oxyhydroxides, in some cases the pH alone will change the metal ion to an
insoluble form, this is true in the case of aluminum.
Other commonly used alkaline agents are hydrated lime (calcium hydroxide), soda
ash (sodium
carbonate), caustic soda (sodium hydroxide), and in some cases ammonia (U.S. Dept. of the
Interior, 2002b).
The processes involving metals more common to coal mining regions (iron, aluminum, and
manganese)
are fairly well understood. The removal of iron is better understood than other
metals common to drainage sites, which may be one of the reasons why passive treatments are
more common in the East. Iron can form oxyhyroxides (FeOOH) or hydroxides (Fe(OH)
3
) under
aerobic conditions or a sulfide solid under anaerobic conditions. Iron and manganese (Mn)
precipitation processes are related in that the precipitations are sequential in aerobic conditions
(EPA, Vol. 4). Iron oxidizes and precipitates more quickly than Mn because oxidized Mn solids
are unstable in the presence of Fe
+2
therefore the levels must be reduced significantly before Mn
can be converted to stable solid precipitates (EPA, Vol. 4). Manganese under aerobic conditions
can form an oxyhydroxide (MnOOH) and oxides (Mn
3
O
4
and MnO
2
) and in alkaline environments
a carbonate (MnCO
3
) (EPA, Vol. 4). Manganese sulfide is highly soluble and therefore highly
unlikely to remain precipitated if it should form under anaerobic conditions (EPA, Vol. 4).
Aluminum is removed from waters by maintaining the pH between 5 and 8, where Al(OH)
3
is
highly insoluble; the passage of mine water through highly oxidized or reduced environments has
no effect on Al concentrations (EPA, Vol. 4).
5
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage: Innovative Treatment Technologies
Technologies designed to remove metals common to hard rock mining almost always involve the
establishment of sulfate reducing bacteria (SRB), which can be difficult in cold climates. Sulfate
reducing bacteria remove metals from solution as precipitates as a result of their survival (Zaluski
et al., 2000). SRB reduce sulfate to soluble sulfide when provided with an organic carbon source,
i.e. compost; as a result of this process acetate and bicarbonate ion are also produced. The
soluble
sulfide reacts with the dissolved metals to form insoluble metal sulfides, equation 18; the
bicarbonate ions increase the pH and alkalinity of the water, equation 17 (Zaluski et al., 2000).
Bicarbonate also allows for the possible production of Zn, Cu, or Mn carbonates (Macalady,
1998). Metals likely to form insoluble sulfide precipitates include: Cu, Zn, Cd, Pb, Ag, and Fe(II)
(Macalady, 1998). These processes are summarized by the following reactions:
SO
4
-
2-
+ 2CH
2
O —> H
2
S + 2HCO
3
(17)
H
2
S + M
+2
—> MS + 2H
+
(18)
In addition to precipitation processes, metals can be removed from water through a variety of
methods
common to wetlands, and seen in technologies utilizing organic matter and/or vegetation:
!
filtering suspended and colloidal material from the water
!
uptake of contaminants into the roots and leaves of live plants
!
adsorption or exchange of contaminants onto inorganic soil constituents, organic solids,
dead plant material or algal material
!
neutralization and precipitation of contaminants through the generation of HCO
3
-
and NH
3
by bacterial decay of organic matter.
!
destruction or precipitation of chemicals in the anaerobic zone catalyzed by the activity of
bacteria
!
destruction or precipitation of contaminants in the aerobic zone catalyzed by the activity
of bacteria (EPA, 1993b).
It is not in the scope of this paper to describe all of the potential considerations related to each
metal
of concern; a few examples have been mentioned to illustrate the necessity of carefully
analyzing all of the metal contaminants and the surrounding hydrologic, geologic, chemical, and
biologic situation in order to properly design for removal.
1.3 Environmental Concerns
Environmental damage or pollution associated with mines nearly always has to do with a decrease
in
pH and/or elevated concentrations of heavy metals in nearby waters and soils. There are
instances were one problem occurs without the other, for example circum-neutral pH and elevated
metals concentrations or vice versa. Debris from waste piles may be blown and contaminate
surrounding areas with metals. Silt and sediments may run-off into nearby streams and obstruct
water flow. Other sources of pollution that may not initially come to mind are abandoned
buildings and industrial equipment that contribute to pollution, including waste drums, heavy
equipment, batteries, etcetera.
6

Acid Mine Drainage: Innovative Treatment Technologies
While all of these problems are serious, the main focus will be on polluted water resulting from
mine drainage. Indeed, with the exception of a few new means of revegetation, most of the
innovative technologies in the literature address water
treatment.
Younger et al., provide information about the impacts that
mining
has on the water environment, they have defined six
distinct impacts (2002). Not all of these impacts deal with
pollution, but it is useful to consider all of the potential
problems for a holistic view useful for designing an
effective remediation plan.
Figure 1. Air compressor at an
1)
“The mining process itself
”(pp. 55) Which is
abandoned mine site in Leadville, CO
.
associated with the disruption of groundwater hydrology.
It has been pointed out that, “the miner and the water
resource manager share a common interest in avoiding
the ingress of fresh water into a mine void; the water manager’s loss of resource is the miner’s
increase in nuisance” (pp. 57).
2)
“Mineral processing operations
” (pp.
57). For example, cyanide leaching operations, gold-
mercury amalgamation. Contaminated abandoned leach pads can contribute to polluted runoff
from the mine site. Active mines today, at least in the U.S., have regulatory obligations to
prevent this type of contamination.
3)
The dewatering which is undertaken to make mining possible
.
Some of the problems that can
arise from pumping water out of mining shafts include: water table depression resulting in
reduction in water availability for residents and the surrounding hyrdologic system, i.e. wetlands,
streams, lakes; land subsidence or collapse; and, surface or groundwater pollution if mine water is
of poor quality and runs to nearby waterways. However, mining industry today takes some
measures to reduce these impacts through: compensation flows, in which water is added to
sensitive surface waters, and may even be treated and pumped to specific locations; local re­
injection of groundwater; alternate water supplies might be provided for affected residents; and/or
waters that are unaffected by the mining operation itself, but are not of good quality might be
treated before discharge.
4)
Seepage of contaminated leachate from
waste rock piles and tailings dams
. For example,
waste rock piles may not have had enough metal present to be economical to recover; however,
the rock material might have sufficient pyrite present to generate acidity and mobilize metals.
5)
“The flooding of abandoned mine workings after
mining has ended” or, “water table
rebound”
(pp.59). While the water table is recessed and pyrite is able to oxidize causing a build­
up of “acid-generating salts,” when the water table rises these salts are dissolved causing an
increase in pH and dissolved metals (pp. 60). There are other possible hazards like erosion of
support columns in the mine tunnels leading to subsidence and also, the converse rising ground
levels due to rehydration of soils, especially clays.
7
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage: Innovative Treatment Technologies
6)
Discharge of untreated mine water after flooding of workings
can lead to: surface water
pollution, pollution of overlying aquifers, localized flooding and overloading and clogging of
sewers.
For more information about environmental concerns associated with abandoned mines see EPA,
2001, chapters
2 & 3 and Younger et al., 2002.
2.0 Treatment
Treatment of mine sites generally requires pH adjustment, oxidizing or reducing (redox)
conditions, and/or stabilization of wastes. Treatment technologies will be broken up into this
categories: traditional and innovative. It is difficult to assign absolute definitions, the following
distinctions will help to clarify the meanings.
2.1 Traditional
Traditional treatments rely on conventional, well-recognized technology to raise pH or create
redox
conditions. The types of technologies considered traditional in this paper include: water
treatment plants, relocation of wastes, covering of waste piles, water diversion tactics, and in
some cases revegetation.
Traditional or conventional treatments for mine waters are those that follow the pattern of an
ordinary
wastewater treatment plant often referred to as active treatment. Younger et al. define
“active” treatment as “...the improvement of water quality by methods which require ongoing
inputs of artificial energy and/or (bio)chemical reagents” (2002, pp. 271). There are a variety of
methods that are considered “active,” the most predominate one is “ODAS” - oxidation, dosing
with alkali, and sedimentation (Younger, et al., 2002). The process is similar to that of traditional
wastewater treatment plants. Others traditional or “active” treatments common to wastewater
treatment plants include: sulfidization, biosedimentation, sorption and ion exchange, and
membrane processes like filtration and reverse osmosis (Younger, et al, 2002). The waters are
removed from their course, treated and then discharged.
Depending on the situation it may be advantageous to install a traditional water treatment system
as described
above, in some cases it might even be the only option. One of the advantages is
precision. For the most part an engineered system can be altered to obtain desired discharge
regardless of the changes in the incoming water characteristics. This can be useful for active
mining sites with frequently changing water characteristics (Younger, et al., 2002). For instance,
Russ Forba who works on the Berkeley Pit Superfund site in Montana, stated that after evaluation
of the options they did not feel comfortable with the reliability and contaminant removal
efficiencies associated with innovative treatments due to the seven million gallon-per-day flow and
complex characteristics of the wastewater (R. Forba, personal communication, 6/20/2003).
Another benefit is that the land required to establish a plant for large flows is much smaller than
the space required for comparable passive treatment systems (Younger et al., 2002). Finally,
traditional wastewater treatment plants are accompanied by a large body of experience and
information, making the expertise easier to find and with a higher confidence level in performance.
8
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage: Innovative Treatment Technologies
Traditional treatment approaches to handling solid mine wastes include a variety of excavation,
landscape adjustment, and stabilization techniques. Again, in most cases the main concern is to
avoid water infiltration of the solid wastes. Solid wastes can be contained on-site in a variety of
ways: lined pits, un-lined pits, clay or plastic caps, etc. Similarly, wastes can be shipped off-site to
landfills, treatment plants, or recovery operations; however, this option may be costly as prices are
by the ton. Covering or “capping” wastes is a fairly common choice, the cover can be multiple
layers of plastic, cement, soil, compost, rock, vegetation, etc. The idea behind these methods is
that the solid materials high in metals and/or acid-producing materials will not be exposed to the
elements and will not cause typical problems associated with mine wastes. While these solutions
may be a reasonable option for reducing potential harm to humans and the environments they
generally do not reduce the toxicity or volume of the metals present in the soils (Pioneer
Technical Services, 2002).
Re-grading is a common term used in describing remediation efforts. Re-grading is simply
reducing
the slope of a waste rock or tailings pile to prevent erosion by reducing water runoff and
to provide a more stable surface to enhance revegetation efforts. Another tactic to control water
flow near a waste pile is to attempt to divert water from the pile by installing trenches and
culverts.
Whether revegetation is traditional or innovative is somewhat obscure. It is not a new idea and
has
been done for many years. However, some new methods have made it possible to revegetate
areas that were previously thought to be a lost cause. For example, biosolids and lime
applications have been proven to be a viable method for establishing self-sufficient vegetative
cover. The distinction between innovative and traditional lies between the goal and outcomes of
establishing vegetation. If the goal or outcome is to reduce toxicity or to recover metals then it
would probably be considered an innovative treatment; if the goal is to prevent metals
contaminated soils from being blown into nearby yards, but the metals are still present in the same
quantities it would probably be considered traditional.
For more information about available treatment technologies please see: EPA’s Abandoned Mine
Site
Characterization and Cleanup Handbook, Chapter 10.
Case Study: California Gulch, Leadville, Colorado
The California Gulch Superfund site located in Leadville, Colorado utilizes nearly every
traditional
treatment option described above and even some innovative applications. Some of the
treatments include: two water treatment plants, consolidation and stabilization of piles, water
diversion, capping, revegetation, and biosolids application. Mining for gold, silver, copper, zinc,
manganese, and lead began in 1859. The site is approximately 16.5 square miles, divided into
twelve Operable Units (OUs) (EPA, 2003). Each OUs is managed by a different party including
EPA, the U.S. Bureau of Reclamation, the State of Colorado, and ASARCO, the Resurrection
Mining Company, a subsidiary of Newmont Gold Company, and the ASARCO-Resurrection joint
venture (EPA, 2003).
9

Acid Mine Drainage: Innovative Treatment Technologies
The two water treatment plants are located at the outfall of
abandoned mine tunnels: the Yak Tunnel and the Leadville
Mine Drainage Tunnel. Tunnels were built to transport
ores out of mines and sometimes drain groundwater to
allow access to the underground. Rock in these tunnels is
highly disturbed and exposed to water and oxygen
therefore, pyrite oxidation and metals leaching is likely and
effluent from the tunnels is highly problematic. The
treatment plant at Yak Tunnel is managed by the
ASARCO-Resurrection joint venture. The flow to the
plant is highly dependent on season, during summer months
there can be little more than a trickle of water but in the
spring during snowmelt the flow increases dramatically.
Before the plant was in operation nearly 210 tons of metals
entered the Arkansas River annually (EPA, 2003). The
second treatment plant is at the end of the ten mile long
Leadville Mine Drainage Tunnel. This plant is managed by
the Bureau of Reclamation. In addition to treating water
that has made its way into the tunnel, it receives runoff
Figure 2. Capped waste pile in
from tailings piles located near the origin of the tunnel
Leadville, CO. Photo taken by author.
from tailings & waste rock piles. Furthermore, it is
believed that nearly two-thirds of the water reaching the
tunnel is runoff and groundwater that is uncontaminated
before
entering the tunnel (M. Holmes, personal communication, 7/22/2003). If this groundwater
could be diverted from the tunnel the treatment plant would be more efficient in treating the
drainage. However, this is easier said than done given the depth of the tunnel and the complexity
of the hydrology at the site. The current thought is to install a plug to block flow of clean water
into the tunnel; but, in order to do this a shaft would have to intersect the tunnel at a depth of 500
feet below ground level (M. Holmes, personal communication, 7/22/2003). Aside from being
expensive it is difficult to drill accurately enough to intersect the tunnel at an appropriate location
(M. Holmes, personal communication, 7/22/2003). Further complications would arise with the
construction
and performance of the plug.
Given the highly variable flow patterns and difficult climate at the elevation of 10,162 feet
water
treatment plants are a good option for treatment at the California Gulch site. In addition, when
passive treatment was considered using wetlands it was determined that the space needed for
construction would consume the entire town of Leadville (M. Holmes, personal communication,
7/22/2003).
Consolidation of waste piles is another large effort taken at the site to reduce water quality
impacts
. More than 350,000 yards of contaminated soils, sediments, and mine processing wastes
have been consolidated on site (EPA, 2003). Once consolidated, a variety of measures have been
taken including diversion trenches and culverts, evaporation ponds, and capping to minimize
contaminated runoff leaving the site. Diversion trenches attempt to catch runoff before it comes
into contact with the waste pile thereby avoiding contamination of the water so that it might reach
the river or other water body in a “clean” state. Evaporation ponds collect runoff from piles and
10
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage: Innovative Treatment Technologies
allow water to evaporate leaving metal precipitates, mostly iron, to stay in the pond. A couple of
piles have been capped to prevent water infiltration and runoff. Revegetation efforts are also
underway at many locations on the site, some of the locations have used the application of
biosolids which will be describe in the Innovative Technologies section of this paper. All of these
efforts have helped to reduce water quality issues in the Arkansas River.
2.2 Innovative
What is considered innovative? The Encarta English Word Dictionary provides some insight, the
technology
should be attempting to change the properties or form of a chemical, here the
hydrogen ion and metal ions, in a way that has not been attempted in recent years:
innovative
is defined as: “new and original: new and original or taking a new
and
original approach”
treatment
is defined as: “technology: treating something with agent: an act of subjecting
something
to a physical, chemical or biological process or agent” (2003).
Though “innovative treatment” could surely describe a wide range of technologies, for example
chemical
encapsulation of wastes, the discussion here is limited to full-scale implementation of
new technologies that have been installed at multiple abandoned mine sites. A variety of “passive
treatments” have become the most predominate innovative treatments applied aside from
traditional choices. Passive treatments are considered to be those that treat waters or solids using
enhanced natural processes, in-situ and require minimal upkeep (Hedin et al., 1994; Younger et
al., 2002). Research into these techniques began as early as thirty years ago and has been
growing ever since.
The beginning of this movement developed out of the observation that wetlands naturally remove
metals
from contaminated water (Gusek, 1998). Through trial and error it was discovered that in
many instances plants were not necessary to treat the waters, rather other biochemical and
geochemical reactions were responsible for water quality improvements (Gusek, 1998). For
metals common to hard rock mining (Zn, Pb, Cd, As, Mo, Au, Ag, to name a few) sulfate
reduction by bacteria is usually the premise behind the design of passive treatment with the goal of
inducing metal precipitation as sulfides. For metals common to coal mining (Fe, Al, and Mn)
aerobic processes, with or without an alkaline agent are the most commonly applied applications.
Another major player in passive treatment are alkaline agents, most commonly lime, although the
application of lime to reduce acidity it not particularly innovative, some of the ways to expose the
acidic waters to the alkaline agent are innovative.
Many of the innovative technologies in operation are based on the same principles. Permeable
Reactive
Barriers (PRBs), bio-reactors, and constructed wetland technologies can all utilize
alkaline agents and sulfate reducing bacteria to treat mine drainage. The majority are in-situ
applications that manipulate natural processes to treat acidic and/or metals contaminated water,
the exception is the use of iron in PRBs to treat uranium, see pages 25-26. Their differences lie in
construction and water source. PRBs have a subsurface reactive section that groundwater flows
through following its natural course to be treated, in some cases there are impermeable walls to
direct the flow of the water to the reactive section. The reactive media is usually compost
11
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage: Innovative Treatment Technologies
material that hosts sulfate reducing bacteria, though there are a few others. Bioreactors are
somewhere between a PRB and a wetland, water - ground or surface - flows through and natural
reactions work to remove metals. Whether subsurface or exposed to the atmosphere, bioreactors
are generally lined, filled with composted materials and/or alkaline agents, and in some cases
include vegetation. Constructed wetlands are very similar to both PRBs and bioreactors, they are
often lined ponds filled with organic matter and/or alkaline agents and sometimes vegetation.
Organic matter and vegetation allow an opportunity for metals to absorb and/or adsorb to organic
surfaces, this is true for bioreactors and PRBs as well. Anaerobic wetlands aim to promote the
growth of sulfate-reducing bacteria and raise pH. Aerobic wetlands are most often used for net
alkaline waters, oxygen infiltration is encouraged and metals precipitate as oxyhydroxides,
hydroxides, and carbonates. Both bioreactors and wetlands almost always include collection and
piping systems, while PRBs are simply placed in the flow path.
Lime-based applications considered innovative in this paper, are anoxic limestone drains and
Successive
Alkalinity Producing Systems (SAPS). The latter is very similar in construction and
theory to wetlands/bioreactors and is also an improvement to the anoxic limestone drain
technology.
2.2.1 Limestone Drains
Anoxic Limestone Drains (ALDs) treat acidic and potentially metals-laden waters by sending
them
through an underground pathway that is packed with crushed limestone. ALDs typically
outlet into a settling pond or wetland to allow metals an opportunity to precipitate and settle
(Cravotta, 2002).
The problem with ALDs is that they often experience armoring - described as strong adhesion and
complete pacification by encrustation - causing the limestone to become inactivated and
potentially cause clogging of the drain (Cravotta, 2002; Sasowsky, 2000). To effectively install
an ALD many suggest that dissolved oxygen, Fe
+3
, and Al
+3
concentrations be less than 1 mg/L;
some authors have suggested that Fe
+3
and Al
+3
concentrations can be higher, between 1 and 5
mg/L (Cravotta, 2002). In either case this is a very low threshold when dealing with mine
drainage.
A study by Sasowsky et al., suggests that the armoring of limestone
can be substantially offset by
incorporating sandstone into the drain (2000). Sasowsky et al., observed that when acidic and
metals contaminated drainage at Big Laurel Creek at the East Fork Obey River in Tennessee
discharged onto both exposed limestone and sandstone the majority of metallic oxides precipitated
onto sandstone rocks (2000). This suggested a preferred precipitation media. In order to validate
that the observed precipitation was not merely coincidence or mechanical, laboratory and field test
at another mine drainage location in Silver Creek, Ohio were conducted. Similar results were
recorded, and sandstone had an order of magnitude higher iron precipitation than limestone
(Sasowsky et al., 2000). If this preference is fairly consistent, the addition of crushed sandstone
to limestone drains could reduce armoring of limestone. It might also be noted that these studies
were not conducted at oxygen deficit locations, and so behavior in anoxic conditions should be
investigated.
12

Acid Mine Drainage: Innovative Treatment Technologies
Case Study: North Pennine Orefield, UK
An emerging potential use for ALDs is for zinc removal.
Nuttall & Younger (2000) conducted a field-scale test to
use an ALD to remove zinc from net alkaline waters. The
pilot scale ALD was place in the Nent Valley within the
North Pennine Orefield, United Kingdom, the area had been
mined for over two centuries for lead and zinc (Nuttall &
Younger, 2000).
Metals leach from spoil heaps and tailing
dams;
contaminated land drainage and five abandoned mine adits
also discharge metals into the River Nent. The waters have
high hardness values, high alkalinity, and pH in the range of
7.4 to 8.0. The dissolution of sphalerite, ZnS (see equation
5),
results in zinc concentrations in the range of 3 to 8
mg/L; there are also concentrations of lead, cadmium (both
well below 1 mg/L), and arsenic (Nuttall & Younger,
2000). Geochemical modeling and laboratory tests revealed
photFiguro
e
ta3. keBn
urby
leigh
the
Tunneauthor l, 7/2003,
that raising the pH from approximately 7.5 to 8.2 would
result in the optimal reduction of zinc in solution (Nuttall &
Younger, 2000).
Aerobic processes that aim to result in hydroxide or sulfite solids have not been successful in this
case because in hard, net-alkaline waters zinc is predominantly present as carbonate complex
(ZnCO
3
0
) and will not readily form non-carbonate solids (Nuttall & Younger, 2000). Therefore,
an anoxic limestone drain was chosen as a possible way to raise the pH to roughly 8.2 for optimal
removal (Nuttal & Younger, 2000). The results of the pilot test show 22-percent reduction in
zinc concentrations after passing through the anoxic limestone conditions, with a retention time of
14 hours (Nuttall & Younger, 2000).
This is not the typical example, most ALD installations have been at coal drainage sites. It is
particularly
interesting because it does not rely on microorganisms which tend to be more
temperature dependent, so the application might be possible at colder temperature sites.
2.2.2 Constructed Wetlands
There are two types of wetlands used to treat mine drainage, aerobic and anaerobic/compost. As
mentioned
previously, observations by ecologists that wetlands are capable of treating water
and/or retaining toxics forms the basis of most passive treatment technologies.
It is possible for mine drainage to be net alkaline. If the metal of concern is iron an aerobic
wetland
is the best treatment option; aerobic treatment alone is rarely successful with other types
of metals. Net alkaline waters are able to buffer the additional hydrogen ions released during
metal hydrolysis reactions, for example: Fe
+3
+2H
2
O --> FeOOH + 3H
+
(EPA, Vol. 4). The
13
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage: Innovative Treatment Technologies
precipitation of metals is a purely chemical reaction and is not as temperature dependent as sulfate
precipitation common to anaerobic wetlands (EPA, Vol. 4). The main limiting factor for these
systems is metal precipitate build-up, these deposits may need to be removed to allow for
continued wetland operation. Robert Hedin has started a company that dredges this build-up and
sells it for use as pigment in dyes and paints.
When waters are net acidic, the pH must be raised and ideally the waters will be bought to net
alkaline
conditions. When iron and aluminum are the main contaminants then alkaline addition
followed by an aerobic settling pond is often used to precipitate metals and raise pH. The most
common wetland application for hard rock mines aims to establish sulfate-reducing bacteria under
anaerobic conditions and, as a result of the bacteria’s metabolic needs, metals are precipitated as
sulfides, see equations 17 &18. Anaerobic wetlands generally consist of organic substrate, often
compost, and can be mixed with lime to increase alkalinity (EPA, Vol. 4).
There are a variety of considerations when designing a constructed wetland, more information can
be
obtained in EPA's Volume 4: Coal Mine Drainage; Younger et al., 2002; and, Macalady, 1998.
Case Study: Burleigh Tunnel, part of the Clear Creek/Central City Superfund Site,
Colorado
This site is located in Idaho Springs, Colorado in a narrow valley with very harsh cold winters and
limited
sunlight year-round. This project was in operation for about 3 years before treatment
failed for a variety of reasons and was decommissioned.
The water exiting the Tunnel is roughly neutral with a pH of 6.5, with discharge averaging 60
gallons
per minute, elevated concentrations of bicarbonate buffer the mine water, and zinc is the
metal of most concern (J. Lewis, personal communication, 7/7/2003). The pilot system installed
is described as two “anaerobic compost wetlands in both upflow and downflow configurations,”
they were not designed to treat the entire flow, but only one-fourth, or 15 gpm - approximately
7.5 gpm in each cell (EPA, 2002b).
Each wetland was a 0.05-acre (2178 ft
2
) cell (a.k.a. “pit”) filled four feet deep with a mixture of
an organic-rich compost (96 percent) and alfalfa hay (4 percent). The cells were installed below
grade to reduce freezing and the earthen side walls were bermed.
The base of each cell was made up of a gravel subgrade, a 16 ounce geofabric, a sand layer, a clay
liner
, and a high-density polyethylene liner (EPA, 2002b). Geonets and geofabrics were applied in
order to: separate influent and effluent piping; hold compost in place in the upflow cell; separate
perforated PVC piping from the compost (EPA, 2002b). The geonet and perforated piping
ensured even distribution of the influent water into treatment cells and prevented short-circuiting
of water through the cells. For more details consult the EPA “SITE Technology” publication
listed in the bibliography as EPA 2002b.
The hydraulic system for the cell involved concrete v-notch weirs, one for influent and effluent for
each
cell. At some point the valves in the downflow cell became locked-up and could no longer
14

Acid Mine Drainage: Innovative Treatment Technologies
be operated; the time and reason are unknown (J. Lewis, personal communication, 7/16/2003).
Water entered the upflow cell under pressure at the bottom of the compost and exited from the
top; water entered the downflow cell at the top and flowed down by gravity, exiting at the bottom
(EPA, 2002b). A drainage collection structure was built within the Tunnel to build sufficient
hydraulic head to drive flow through the two cells (EPA, 2002b). A bypass system was also
constructed, though was not always effective (J. Lewis, personal communication, 7/16/2003).
During its three years in operation the upflow wetland removed an average of 93 percent of zinc
the
first year and 49 and 43 percent during the second and third years (EPA, 2002b). The
downflow wetland removed a mean of 77 percent of the zinc during the first year and 70 percent
the second year; flow was discontinued in the third year (EPA, 2002b). Based on aqueous
geochemical modeling, observations of cell compost, results of the sulfate-reducing bacteria
count, and acid volatile sulfide data, biological sulfate reduction was not the main removal
mechanism. Primary removal is thought to have occurred due to precipitation of zinc oxides,
hydroxides, and carbonates in aerobic portions of the cell. The upflow cell during the first six
months of operation had effluent levels of less than 1mg/L; concentrations began to increase near
the end of 1994 into 1995, by May 1997 concentrations had reached 60.1 mg/L (EPA, 2002b).
The cell suffered a significant blow in the spring of 1995. Heavy runoff increased the flow
through the cell to 20 gpm of aerobic water, and the increased flow also apparently mobilized
more zinc and substantially increased the zinc concentrations. After the increased flow, removal
efficiencies were around 43 to 49 percent, whereas before removal efficiency were more than 90
percent. In 1997 a visibly obvious preferential flow path developed and was eliminated. The
upflow cell was decommissioned in June of 1999. It is believed that the initial high removal rates
in the upflow cell are the result of adsorption and absorption along with biological sulfate
reduction; decline in removal rates is speculated be related to the decline in SRBs.
Currently there is no treatment being done at the Tunnel. The water seems to be entering the
subsurface,
it is unclear whether it is building up on the site, draining from the site, or traveling as
groundwater; however, sampling of water is indicating that zinc concentrations are within
regulatory standards of less than 200 micrograms/liter, the reason is undefined (J. Lewis, personal
communication, 7/16/2003).
This example is interesting because the design was to precipitate metal sulfides under anaerobic
conditions
, yet the predominate form of precipitate was that common to aerobic conditions. It
would be useful to gather information on the potential precipitates under aerobic conditions,
especially abiotic reactions.
2.2.3 Bioreactors
Passive bioreactors are lined trenches or pits that can contain a variety of materials, most
commonly
a mixture of cobbles, compost, other organic matter, and/or an alkaline agent.
Sometimes above ground tanks containing any variety of materials including those described
above and other trickling filter types of materials - common in bio-treatment of municipal waste­
15

Acid Mine Drainage: Innovative Treatment Technologies
water treatment used to establish appropriate microorganisms to precipitate metals and adjust pH
- are referred to as “bioreactors.” The tank type of bioreactor will not
be discussed in this paper,
though they are used to treat acid mine drainage. They are both legitimate in using the term,
“bioreactor” as they are using biological reactions to treat the waters. Arguably, the term
"bioreactor" would in fact include PRBs, SAPS, and wetlands. The distinction between them has
been made because the literature does so.
A. Case Study: Silver Bow County, Montana
Sulfate Reducing Bacteria (SRB) are the key to these bioreactors installed at the Calliope
abandoned mine site in Silver Bow County, Montana in the Fall of 1998 (Zaluski et al., 2000).
This project was funded by the EPA and jointly administered by the EPA and the Department of
Energy; the project was implemented by MSE Technology Applications in Butte, MT. Water that
flows through a collapsed adit discharges onto a large waste rock pile, upon exiting the pile the
water has an average pH of 2.6 and elevated metals concentrations; this water then flows into a
pond resulting in a pH of 3 to 5.5 depending on mixing ratios largely determined by the season.
In order to treat the mine drainage and conduct research to obtain knowledge about optimal
design characteristics three SRB reactors (II, III, and IV) with different attributes were designed.
Two of the three reactors were placed below grade (ground) to minimize temperature changes
and one above to study the effects of freezing. The reactors were filled with a combination of
organic carbon, cobbles, and/or crushed limestone . Each reactor had a fifty foot section of
cobble preceded by organic matter and/or limestone. Two of the three reactors had
“pretreatment” sections, which consisted of an additional five foot section of organic carbon and a
five foot section of crushed limestone; while the third one had only a five foot section of organic
matter.
The most notable obstacle to the success was when the flow through reactor II ceased due to
biofouling
and consequent plugging. The problem was quickly addressed within a month.
The reactors were monitored monthly for sulfate, alkalinity, SRB count, heterotrophic bacteria
count
, dissolved oxygen, Eh (a measure of redox potential), and metals including: aluminum, zinc,
cadmium, copper, iron, and manganese.
Overall, the results were positive, pH was increased and metals concentrations were reduced.
Comparison
of reactors shows that “initial increase of pH can largely be attributed to alkalinity
present within the organic substrate rather than to limestone” (Zalusk, et al., 2000). Once SRB
were established their metabolic reactions also contributed to pH increase.
Some of the more interesting findings when comparing the bioreactors included (Zalusk, et al.,
2000):
!
More organic matter leads to more organic matter fermentation reactions resulting
in an increase in temperature; this could be critical in cold climates.
16
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage: Innovative Treatment Technologies
!
Increased temperature leads to greater microbial activity.
!
Prior to SRB activity, adsorption of metals to organic substrate seems to be the
cause of concentration reductions.
B. Case Study: Champagne Creek - Butte, Idaho
The mine drainage from Moran Tunnel contributes low pH, metals-laden water discharging to
Champagne Creek. The project is being handled through the Bureau of Land Management’s
Abandoned Mine Lands Program in the State of Idaho. The watershed is 9.2 square miles,
moderately steep, at an elevation of 6060 feet, mostly covered by semi-arid rangeland. The
stream itself is only 4.5 miles long, it is consumed by alfalfa hay irrigation and does not reach a
receiving stream. The annual average precipitation is about 16 inches, the majority of runoff is
due to snowmelt; in times of extreme drought the stream will run dry (Moore & Kotansky, 2002).
Figure 4. Schematic of Champagne Creek Bioreactor (Moore & Kotansky, 2002)
Mining around Champagne Creek began around 1883 with the discovery of silver ores, this first
mine operation ended around 1887. In the late 1920s deeper base-metal sulfide ores were mined
for lead and zinc. Around this same time the Moran Tunnel was constructed with the hope of
intersecting the Last Chance vein at around 450 feet below the surface; the vein was never found.
The area was last mined in 1946.
The site underwent a Preliminary Assessment in 1985 and
a Site Investigation in 1988. The
Bureau of Land Management completed its own study in 1989. It was this report that required
17

Acid Mine Drainage: Innovative Treatment Technologies
additional water quality monitoring and a review of which passive treatment wetland system might
be able to be used (Moore & Kotansky, 2002).
In 1999 cleanup at the Moran Tunnel began. The first actions were removal of waste rock piles, a
17,500 cubic yard pile was placed in a repository above the flood plain; additional waste rock
totaling 2700 cubic yards from surrounding areas was also placed in this repository (Moore &
Kotansky, 2002). A four-cell passive bioreactor system was constructed based on SRB and lime
treatment. The cells consisted of organic material (manure and hay) to encourage SRB
establishment and limestone to neutralized acidic discharges (Moore & Kotansky, 2002). Berms
were also put into
place leading to the passive system; they were made up of lime and materials to encourage SRB
growth.
The system was effective in improving water quality for the first few months of operation (S.
Moore,
personal communication, 7/30/2003). The pH from pond 1 to pond 2 increased from 3.3
to 6.4. The first winter (1999-2000) after installation revealed lower metal discharges and a
decrease in SRB activity common during cold weather (Moore & Kotansky, 2002). The first
berm, made of “limestone and SRBs,” initially led to a decrease in aluminum and copper of nearly
100-percent, 92-percent of cadmium, 77-percent of zinc, and 65-percent of iron (Moore &
Kotansky, 2002). By May 0f 2000 the removal rates were nearly 100-percent aluminum and
copper, 91-percent iron, and 56-percent zinc (Moore & Kotansky, 2002).
In 2001 the passive treatment system was enhanced with the addition of an anaerobic treatment
tank
. It was added between the discharge from the Moran Tunnel and the first treatment pond.
The tank was put into place because water quality data indicated that the high concentration of
iron on the first pond was interfering with the effectiveness of the bioreactor berms in removing
zinc, copper, and buffering pH (Moore & Kotansky, 2002). Eventually, the tank also clogged and
performance of the system declined (S. Moore, personal communication, 7/30/2003).
The system has required recharge of the berms with “SRBs and limestone” and the addition of the
anaerobic
tank (Moore & Kotansky, 2002). This system has not yet proved to be a walk-away
solution, but BLM-Idaho are working on improving the system and carefully documenting efforts
so that lessons may be learned for future projects (S. Moore, personal communication,
7/30/2003).
18

Acid Mine Drainage: Innovative Treatment Technologies
2.2.4 Successive Alkalinity
Producing Systems
Successive Alkalinity Producing
Systems have the following basic
elements: organic mulch layer,
limestone layer, and a drainage
system - most include a flushing
system as well. This technology
was created in the early 1990's by
Kepler and McCleary (Younger et
al., 2002). The idea is that mine
Figure 5. General Schematic of a SAPS Available at:
drainage flows into the tops of the
http://sudan.cses.vt.edu/prp/Research_Results/SAPS.html
cell creating a top layer of water
which prevents the infiltration of
oxygen into the bottom layers
(water is also used in this way in tailings holding dams). The organic layer serves to remove
dissolved oxygen from the water, farther down anaerobic conditions support the establishment of
sulfate-reducing bacteria. The anaerobic environment is a reducing environment that changes
Fe
+3
to Fe,
+2
thereby reducing the likelihood of iron hydroxide precipitation, see equation 3. Since
these units encourage reducing conditions and establishment of SRB, a major contribution to the
treatment of the water, these units are sometimes referred to as RAPS - Reducing and Alkalinity
Producing Systems (Younger et al, 2002). Finally, the water enters the limestone region,
essentially devoid of oxygen preventing the armoring of limestone. Upon leaving the SAPS the
water is usually directed to an aerobic settling pond or wetland to allow metals to form
precipitates and further water polishing (Kepler & McCleary, 2003).
Many SAPS include flushing systems because as one would imagine oxidation and reduction of
Fe and Al leads to precipitates that can clog the cell (Rees et al., 2001). The flushing systems
generally operate by generating head differences that move water rapidly through the system
(Kepler & McCleary, 2003).
SAPS tend to be more efficient than anaerobic wetlands and require less space to provide the
same
level of treatment (Younger et al., 2002). SAPS require some maintenance, not only for
periodic flushing, but also to prevent or correct the development of preferential flow paths,
possible in any of these passive systems (Kepler & McCleary, 2003; Rees et al., 2001). If
preferential flow paths develop the water short circuits the system. They also require driving head
and freeboard resulting in topographic relief requirements of greater than five meters (Younger et
al., 2002).
19
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage: Innovative Treatment Technologies
A. Case Study: Oven Run, Pennsylvania
The watershed protection group Stoneycreek - Conemaugh River Improvement Project (SCRIP)
located
in western Pennsylvania, has initiated and completed multiple projects to improve their
watershed. During a phone conversation with Dave , who is directly involved in this project, it
was revealed that many of their remediation projects utilize SAPS (personal communication,
6/18/2003). Oven Run is one of the larger sites handled by SCRIP, it has six sources of highly
acidic, metals-laden drainage totaling
720,000 gallons a day (Oven Run). Projected costs were
$5 million, actual costs were $4.1 million (D. Steel, personal communication, 6/18/2003). Five of
the six sources are treated using SAPS, the sixth has been backfilled. The first SAPS was
installed in 1995, the last in 2003.
So far treatment has been successful in
removing metals and acidity, while
generating alkalinity. Not including the
most recently installed SAPS, pH at
downstream monitoring points have
increased from the 3 to 4 range to the 5 to
6 range and 200 tons of iron and 200 tons
of aluminum are removed each year. In
addition, the samples showed some
alkalinity, which is particularly impressive
because other acidic waters drain into the
creek after Oven Run, so the treated
waters are able to buffer some of the
additional pollution.
Figure 6. Oven Run, SAPS, Available at:
http://www.ctcnet.net/scrip/stoven.htm
B. Case Study: #40 Gowen, Gaines
Watershed. Oklahoma
This is a former coal mining site experiencing
the
typical aliments, acidity and elevated metals
concentrations, mostly iron. This project was
commissioned in 1998 with the help of an EPA
Region 8 Section 319 grant and the Oklahoma
Conservation Commission, by the University
of Oklahoma (EPA, 2002c). This site,
amongst other AMD sites, was designated as
having the greatest impact on Pitt Creek, a
Figure 7. SAPS cell (“Section 319 Success Stories”
Available at:
http://www.epa.gov/owow/nps/Section319III/OK
.htm
20
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage: Innovative Treatment Technologies
tributary to Gaines Creek which drains into Lake Eufaula (EPA
2002c). The Gaines Creek Watershed is located in Pittsburg
and Latimer counties. The treatment design is a four-cell
system with alternating vertical flow wetland (though figure
shows little more than ornamental plant-life) and surface flow
aerobic ponds. The project budget was $125,000 and was
installed in 1998. To avoid confusion, the vertical flow cell is
what would typically be defined as a SAPS though some define
the entire alternating system as a SAPS as well. The vertical
flow cell consists of a layer of water on top, followed by 1
meter of compost mixed with limestone and flyash and a
cobble-fill pipe drainage system (EPA, 2002c).
There was not enough space at this site to construct a system
that
would be able to treat the entire flow; the system treats
approximately 7600 gallons per day (EPA, 2002c). The
desand ign
7,000
was
grbamases
d
per
on “day
conotaf maciindianty t lanoadid
inrogs n”
of
(EabPAou,
t
2002c)18,000 .
pumFigurpie ng
8.
wSecll
hePmRBat
.
ic of Non-
Removal rates for acidity are estimated to be 30 - 40
gram/meter
2
- day; the total surface area is approximately
750 meter
2
.
The system has been in operation for two years and monitored every two weeks. Though actual
data
could not be obtained, the report about the project on EPA’s website indicates that
“concentrations of iron, aluminum, and manganese have decreased significantly,” pH of the final
effluent is at 6 and alkalinity is above 150 milligram/liter (EPA, 2002c). Trace metals - barium,
cadmium, chromium, copper, nickel, zinc, and lead - were reduced to near or below detection
levels. A recent biological survey counted 2000 healthy populations of fish and
macroinvertebrates in three of four cells.
This project is of particular interest because it is the first successful passive treatment AMD
treatment
project carried out in Oklahoma (EPA, 2002c). The success of this project has spurred
the state to use this wetland design at the Tar Creek Superfund site in Ottawa county, Oklahoma,
and is being investigated for application in several watershed nationwide (EPA, 2002c).
2.2.5 Permeable Reactive Barriers
Permeable Reactive Barriers (PRBs) are exactly what they sound like: barriers that react with
specific
chemicals of concern that are placed in the path of groundwater flow allowing the water
to flow through easily (Blowes et. al, 2000). In PRBs designed to treat acid mine drainage
(AMD) with metals contamination the barrier is generally composed of solid organic matter, like
municipal compost, leaf compost, and wood chips/sawdust (Blowes, et. al., 2000). Organic
matter encourages the proliferation of sulfate-reducing bacteria that will reduce sulfate to sulfide
21

Acid Mine Drainage: Innovative Treatment Technologies
and will result in the subsequent formation of insoluble metal sulfides which has been described
with regards to bioreactors, please see equations 17 & 18. Research has been done to evaluate
the efficiency of using PRBs to remove uranium contamination at abandoned mine sites; possible
reactive materials are zero-valent iron, bone char phosphate, and amorphous ferric oxyhyroxide
(Naftz, et al., 1999).
One important consideration in the design of a PRB to treat AMD is the stability of the metal
sulfides
(Blowes, et. al., 2000). Sulfides have low solubility in anaerobic conditions, if oxidation
were to occur, metals could be released from their metal sulfide form into the environment
(Blowes, et. al., 2000). An example of designing to prevent oxidation is illustrated by a project at
Nickel Rim Mine, Sudbury, Ontario. The designers considered the implications of an oxidizing
agent in the flow of groundwater and the PRB was covered by a 20cm saturated clay cap to
prevent oxygen infiltration (Blowes, et. al., 2000).
Although not discussed much in this paper, former uranium mines are also a serious concern
Naftz
et al. conducted a field demonstration using six different PRBs to study the removal
efficiencies of uranium at a site in southeastern Utah (1999). There were four different reactive
media and two design types. Three of the PRBs were “funnel and gate” types, the gate is where
the reactive media is located and the funnel is two impermeable walls directing groundwater to the
gate. Each gate was consisted of a different material: (1) bone char phosphate (PO
4
) pellets, the
phosphate source facilitates the formation of insoluble uranyl phosphate compounds; (2) zero
valent iron (ZVI) pellets which induce the reduction of uranium (VI) to the less soluble uranium
(IV); and (3) pelletized amorphous ferric oxyhydroxides (AFO) which remove uranium by
adsorption to the ferric oxide surface (Naftz et al., 1999). The other three PRBs were six-inch
diameter non-pumping wells consisting of different proportions of bone char phosphate and
foamed iron oxide pellets; the phosphate will adsorb to the iron pellets to allow access for
formation of uranyl phosphate compounds (Naftz, et al., 1999). The hypothesis is that wells will
allow for contact with deeper plumes and will be more suitable for remote locations (Naftz et al.,
1999).
Results of the field demonstration were positive. After one year of operation and seven sampling
events
the funnel the ZVI barrier removed >99.9 percent consistently, the PO
4
barrier removed
>90 percent on all but two of the seven sampling events, and the AFO barrier varied the most but
still removed an average of 88.1 percent (Naftz, et al., 1999, Table 1). Data from the wells spans
only three months and the results are not quite as impressive, but still reasonable; the average of
removal rates overall was 67 percent (Naftz et al., 1999)
PRBs are a relatively new technology and work is continually being done to optimize installations.
As
it is often helpful to learn from past error a brief discussion of common problems of PRB
performance is presented (Blowes, et. al., 2000).
22

Acid Mine Drainage: Innovative Treatment Technologies
1) Although barriers often have very long theoretical treatment lifetimes when only the material
and the contaminants of concern are considered, actual lifetimes are considerably shorter due to
the presence of other reactive substances in the environment;
2) Chemical reactions can be slowed due to depletion of reactive component of the barrier;
3) Precipitation of a secondary reactive precipitate can reduce the reactive surface area;
4) Physical clogging or preferential path flow.
2.2.6 Biosolids
Biosolids are treated municipal sewage sludge; the EPA defines biosolids as follows:
“...the nutrient-rich organic materials resulting
from the treatment of sewage sludge (the
name for the solid, semisolid or liquid untreated residue generated during the treatment of
domestic sewage in a treatment facility). When treated and processed, sewage sludge
becomes biosolids which can be safely recycled and applied as fertilizer to sustainably
improve and maintain productive soils and stimulate plant growth.”
Biosolids have a growing number of useful applications and the search for more continues as
population
and hence sludge production increases. Biosolids are being used to reclaim mine lands
(Murray et al., 1981; Sopper, 1993; Toffey, 2003) and have also been used for agricultural
purposes. There are federal standards, namely Section 103C of the Clean Water Act and state
standards that have to be met in order to apply biosolids to land. Over a twenty-five year period,
the field experience with biosolids continues to demonstrate clear environmental benefits and
negligible adverse effects (Sopper, 1993; EPA; Toffey, 2003).
When reclaiming mine sites biosolids are almost always applied with lime, either pre-mixed or in
stages
(R. Bastian, personal communication, 6/2/2003). Lime serves to increase the pH of the soil
rapidly. Lime application alone may not be sufficient for long term improvement in the soil
characteristics because the pH will eventually decline as sulfur-bearing minerals are oxidized
(Sopper, 1993). However, biosolids application without lime has in some cases raised the soil pH
and decreased availability of metals (Sopper, 1993).
Biosolids also show advantages over chemical fertilizers (Sopper, 1993) because they provide a
source
of carbon and capacity for moisture retention which are conducive to microbial and plant
growth. This is important for the establishment of a long-term self-sustaining system. Sopper
summarizes that biosolids application re-establishes a functioning microbial population comparable
to undisturbed levels within two or three years of application, much more quickly than with
traditional chemical treatment (1993).
23
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage: Innovative Treatment Technologies
The application of biosolids does not necessarily reduce the amount of metals present in the soil.
In a draft report for the EPA, Maxemchuk explains that tailings sites treated with biosolids do not
experience a reduction in total metals, rather metals availability is reduced (2001). Metals are
immobilized through precipitation as carbonates, phosphates, sulfides, silicates and sorption by
organic matter, and hydroxides (Sopper, 1993). In some cases vegetation may be responsible for
immobilizing the metals, or might even remove the metals from the soil, also known as phyto­
extraction.
There is ample evidence to support the use of biosolids in reclaiming
mine lands. It is a cost-
efficient method for reducing potential harm to the environment and its occupants. It is
particularly attractive when the other options are removal and/or capping. Removal is generally
expensive, especially when sites are very large and this approach just relocates the waste material,
posing a potential problem at a new location. Capping alone can prevent further exacerbation of
the problem, but will not help to re-establish a functioning ecosystem at the site unless natural soils
are used. The use of natural soils as caps on large area sites is impractical, expensive and leaves
“borrowed” areas highly disturbed and subject to intense erosion. Biosolids provide an apparently
indefinite solution to contaminated sites because metals of concern are complexed, reducing their
bioavailability, and the health of the A-horizon in the soil profile is improved. This allows
vegetation to replenish itself - stabilizing and improving the health of the ecosystem in the area.
A. Case Study: Frostburg, Maryland
This project is a testament to the longevity of the use of biosolids in the reclamation of mine lands.
The
field plot experiments were installed in 1974 on a former strip mine. The site had been
completely devoid of vegetation for four years (Griebel et al., 1979). The overburden and rock
wastes resulted in a dark-colored, acidic - pH of 2.9, spoil material (Griebel et al., 1979).
A total of nine test plots 3.6m x 4.5m were installed. There were three basic applications tested:
biosolid
compost alone, biosolid compost with rock phosphate, and biosolid compost with
dolomitic limestone (Griebel et al., 1979). For each of these three scenarios biosolids were applied
in three different amounts: 56 metric tons per hectare (mt/ha), 112 mt/ha, and 224 mt/ha (Griebel
et al., 1979). The biosolids were supplied by the Blue Plains Wastewater Treatment Plant in
Washington, DC. They were then composted at ARS-MES Composting Facility in Beltsville,
Maryland. When sewage sludge is composted the material becomes more humus-like and excess
heat and water are driven off and decreases in the availability of certain metals results (Griebel et
al., 1979). Both rock phosphate and dolomitic limestone were applied at the rate of 11 mt/ha. In
addition, each plot received 110 kg/ha of nitrogen in the form of NH
4
NO
3
. A grass legume seed
mixture was applied at 40 kg/ha and Empire birdsfoot trefoil,
Lotus Coniculatus L.
, was applied at
10 kg/ha (Griebel et al., 1979)
24

Acid Mine Drainage: Innovative Treatment Technologies
After two years the vegetation was harvested and analyzed for yield and metals uptake into the
plants. Soil conditions were also analyzed.
After two years the control plot had a pH of 3.1, the test plots had pH’s as seen in the table below.
The plots with the maximum biosolid application had the most improved pH; interestingly the
difference between compost alone and compost with alkaline amendments did not differ
significantly.
pH of the Soil Two Years After Biosolids Application
56 mt/ha of biosolid
112 mt/ha of biosolid
224 mt/ha of biosolid
compost alone
4.2
3.9
5.0
compost & rock
phosphate
4.5
4.8
5.1
compost & dolomitic
limestone
4.6
4.8
5.1
Adapted from Griebel et al., 1979
Plant yields, shown in figure 9, below are from a single harvest taken during the second growth
season (Griebel et al., 1979). Overall the results are positive, and certainly better than the control
plot with no amendments.
Figure 9. Plant biomass yields two years after
biosolids application (Griebel et al., 1979, figure
25-1)
25

Acid Mine Drainage: Innovative Treatment Technologies
Figure 10. Relationship between total plant metals and rates of application of
biosolids and other amendments (Griebel et al., 1979, figure 25.8)
One of the more interesting things to note is that the “lowest compost treatment (56 mt/ha), used
in combination with either rock phosphate or dolomite, provides yields equal to those obtained
with 112 mt/ha compost alone” (Griebel et al., 1979. p. 296). Furthermore, the highest rate of
compost alone (224 mt/ha) was exceeded only when the same amount was applied with dolomitic
limestone (Griebel et al., 1979).
A common concern when applying biosolids at metals contaminated sites with the intention of
establishing
vegetation is that the vegetation will accumulate high levels of metals that could
potentially be a hazard to wildlife. In this study the observed metal concentrations of Cu, Zn, Ni,
and Cd in vegetation were well within the range of concentrations found in vegetation produced on
regular agricultural soils (Griebel et al., 1979). Figure x shows the concentrations of the metals
with respect to each amendment combination. It is obvious from studying the graphs that the
addition of limestone or phosphate rock reduces the amount of metals taken up by the plants,
generally to around the levels found in the control plot’s vegetation.
B. Case Study: Leadville, Colorado
Biosolids were applied at the Leadville site to revegetate the alluvial tailings deposits that were
washed
in and around the Arkansas River. The tailings have been deposited at various locations
26
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage: Innovative Treatment Technologies
along an 11-mile stretch of the river. This has created a
variety of environmental problems including acidic soils in
the range of pH 1.5-4.5, Zn and Pb salt formation on the
soil surface, sedimentation in the river of up to two feet in
some spots, and death of vegetation leading to erosion of
river banks (“Upper Arkansas,
2000).
Biosolids provided by Denver Metro were applied to
portions
of the site at a rate of 100 dry tons/acre in August
of 1998 (“Upper Arkansas,” 2000). Approximately 100
tons/ac of lime were also applied; both were tilled into the
soil at a depth of twelve inches (“Upper Arkansas,” 2000).
A variety of soil amendment combinations were also
bapiopsloield ids
to
anted
st
alpkallots
ine
at
agenthe
st iwote,
ulto d
dpreteormmoitne e
twhe
himch
osmt
ix of
Fisogurils e
in
11.
LeaMdveilletal ,
salCO
t accumon
the
ulbanks
ation onof
the
vegetation. During a visit to the site in July of 2003 it
Arkansas River, (Upper Arkansas, 2000).
appeared that the applications were working quite well.
2.2.7 Phytoremediation
Phytoremediation suggests the use of plants to treat or remove contamination. Wong defines the
term
as, “the use of green plants and their associated microbiota, soil amendments, and agronomic
techniques to remove, contain, or render harmless environmental contaminants (2003). Though
there are a wide variety of subcategories in the field of phytoremediation only four will be
discussed in this paper, phytoextraction/phtyomining, phytostabilization, rhizofiltration, and
phytovolatilization. For more information about other technologies, consult US EPA’s
Introduction to Phytoremediation.
Phytoextraction, or phytomining if metals can be recovered, is defined as:
“the uptake of contaminants by plant roots and translocation within the plants. This
concentration technology leaves a much smaller mass to be disposed of than does
excavation of the soil or other media” (EPA, 2000, p. 143).
There are a limited number of plants known to be capable of this and climate determines what
species
can be used. Phyto-mining requires that the plants be “hyperaccumulators,” i.e., they will
uptake more than the average concentration of metals. According to Brooks et al., there are about
300 species that hyperaccumulate nickel, 26 cobalt, 24 copper, 19 selenium, 16 zinc, 11
manganese, one thallium and one cadmium (1998). Although these numbers are encouraging there
are few field applications. An important consideration in applying phytoextraction, especially with
the use of hyperaccumulators, is whether the resulting vegetation will be hazardous to local
animals; this possibility varies from site to site (Wong, 2003).
27

Acid Mine Drainage: Innovative Treatment Technologies
Phytostabilization is fairly common with regards to mining sites, it is a common practice to
revegetate spoiled mine lands to prevent soil erosion and deposition of contaminated soils in
streams and nearby lands. The EPA defines it as:
“(1) immobilization of a contaminant in soil through adsorption and accumulation
by roots,
adsorption onto roots, or precipitation within the root zone of plants, and (2) the use of
plants and plant roots to prevent contaminant migration via wind and water erosion,
leaching, and soil dispersion” (EPA, 2000, p. 21).
Ideal plants for this technique use metal-tolerant, drought-resistant, fast growing crops that can
also
grow in nutrient deficient soils (Wong, 2003). The advantages are that it is a relatively
inexpensive technique, soils do not need to be removed, ecosystem restoration is enhanced, and
disposal of hazardous materials or biomass is not required (EPA, 2000). Disadvantages are that
the contaminants remain in place - care must be taken to ensure that the vegetation continues to
stabilize the metals; extensive fertilization or soil modification may be necessary; plant uptake and
translocation of metals must be prevented; root zones, root exudates, contaminants, and soil
amendments must be monitored to prevent an increase in metal solubility and leaching; it may only
be considered a temporary measure; stabilization might be due primarily to the effects of soil
amendments, with plants only contributing to stabilization by decreasing the amount of water
moving through the soil and by physically stabilizing the soil against erosion (EPA, 2000). The
application of biosolids fits well with this phytoremediation technique as it provides necessary
fertilizing agents and aids in microorganism establishment.
Rhyzofiltration involves the removal of contaminants in solution through adsorption or
precipitation
onto plant roots or absorption into the roots, this can also be achieved by the
microorganisms associated with the rhizosphere (EPA, 2000; Wong, 2003). This technology is
applied in water, that is the plants are either aquatic plants or terrestrial plants on a floating
platform (EPA, 2000). Contaminants can be physically removed by removing the plants
themselves. Some of the disadvantages to this technology include a need for good control over
pH, and a clear understanding of the chemical speciation and interaction of all species in the
influent (EPA, 2000). In addition to this, control over influent concentration and flow rate may be
necessary, plants may need to be grown and then translocated to the site (especially terrestrial
plants), periodic harvesting and disposal are required, and laboratory results might not be
achievable in the field (EPA, 2000). Phytovolatilization has been identified as a potential
treatment for mercury and selenium contaminated soils (Chaney, et al., 1997; EPA, 2000).
Phytovolatilization is defined by the EPA as,
“...uptake and transpiration of a contaminant by plant, with release of the contaminant or a
modified form of the contaminant to the atmosphere from the plant through contaminant
uptake, plant metabolism, and plant transportation” (EPA, 2000).
This process is beneficial if the contaminants of concern will be transformed to less-toxic forms, for
example
elemental mercury and dimethyl selenite gas. The disadvantages are uncertainty about
28

Acid Mine Drainage: Innovative Treatment Technologies
metabolites, unhealthy plant accumulation, and uncertainty about other constituents at the site, i.e.,
where there is one form of contamination there could be many more and one must understand how
they will react with the plants as well. For more information including references and plant species
appropriate for the different technologies discussed here please see the following reference:
Introduction to Phytoremediation, 2000; Wong, 2003; Brooks et al., 1998; Brooks, 1998; and,
Madejon et al., 2003.
3.0 Conclusion
Given the seriousness and scale of mine drainage it is important to continue to work towards
affordable and effective treatment options. The passive treatments discussed in this paper are
exhibiting mixed success, results are encouraging but not the “walk-away,” cheap solution that
they are sometimes described to be. Still the innovative treatments discussed here are showing
progress and with further research and performance analysis these technologies may become more
widely used in the future. While there are drawbacks to traditional treatments, there are some
benefits that make them widespread and in some cases the preferred alternative.
As with almost any topic, there is need for more work, some of the more pressing areas include
communication
, funding, and research about fundamental processes that cause precipitation of
metals. Many of the people contacted for this report expressed their desire for a better line of
communication and access to information about similar projects. The database created in
conjunction with this report began to create a place to access project information, though it is not a
complete picture of the efforts being done. Regional communication between parties working with
similar geology and climate would probably be the most useful effort as the performance of passive
treatments seems to be greatly affected by these factors. Funding, or lack thereof, is a serious
issue especially for hard rock sites. As previously mentioned, non-coal states are not eligible for
SMCRA funds and states that are eligible must address coal-related issues before hard rock.
Considering the number of hard rock sites it does not seem reasonable to rely on existing federal
and state environmental funds, for example CERCLA and CWA, to sponsor remediation efforts.
As mentioned earlier in this paper, some states are making efforts to identify and remedy this
funding problem. Research about the precipitation of metals has been done, however, the
differences from site to site in geology, hydrology, climate, and chemistry make general application
of this information quite difficult. Each site must be addressed separately to account for the
variations that occur in nature. As experience builds and information is shared the application of
passive technologies will improve.
4.0 References
Bastian, Robert (EPA-Office of Solid Waste and Emergency Response). Personal communication.
2 June 2003.
29
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage: Innovative Treatment Technologies
Blowes, David W., Ptacek, Carol J, Benner, Shawn G., McRae, Che, Bennett, Timothy A., & Puls,
Robert W. (2000) Treatment of inorganic contaminants using permeable reactive barriers.
Journal of Contaminant Hydrology, 45 (1-2), 123-137.
Brooks, R. R. (ed.) (1998). Plants that hyperaccumulate metals their role in phytoremediation,
microbiology
, archaeology, mineral exploration, and phytominig. CAB International: 1998.
Brooks, Robert R. Chambers, Michael F., Nicks, Larry J., and Robinson, Brett H. (1998)
Phytomining
. Trends in Plant Science, 3 (9), 359-362.
Chaney, Rufus L., Malik, Minnie, Li, Yin M., Brown, Salley, Brewer, Eric P., Angle, J. Scott, and
Baker
, Alan JM. (1997). Phytoremediation of soil metals. Current Opinion in
Biotechnolgoy, 8 (3), 279-284.
Cravotta, Charles A. & Watzlat, George R. (2002) “Design and Performance of Limestone Drains
to
Increase pH and Remove Metals From Acid Mine Drainage.” In: Handbook of
Groundwater Remediation Using Permeable Reactive Barriers: applications to
radionuclides, trace metals, and nutrients. (See Naftz, et. al. 2002 below).
Dempsey, Brian A. & Jeon, Byong-Hun. (2001) Characteristics of sludge produced from passive
treatment
of mine drainage. Geochemistry: Exploration, Environment, Analysis, 1, 89-94.
Encarta®® World English Dictionary [North American Edition] ©© & (P)2003 Microsoft
Corporation. All rights reserved. Developed for Microsoft by Bloomsbury Publishing Plc.
www.encarta.msn.com
Forba, Russ (EPA-Region 8). Personal communication. 20 June, 2003.
Griebel, George, E., Arminger, Walter H., Parr, James, E., Steck, D.W., and Adam, J.A.. (1979).
“Use
of Composted Sewage Sludge in Revegetation of Surface-Mined Areas.” Utilization
of Municipal Sewage Sludge Effluent and Sludge on Forest and Disturbed Land. Sopper,
William E. and Kerr, Sonja, N. (eds.) Pennsylvania: The Penn State University Press.
Gusek, James, J. and Wildeman, Dr. Thomas R (1995). New Developments in Passive Treatment
of Acid Rock Drainage. This article is from the South African Mining Delegation Seminar.
Denver, CO, USA. Aug. 24 - Sept. 4, 1998. (First presentated at: Engineering Foundation
Conference on Technical Solutions For Pollution Prevention in the Mining & Mineral
Processing Industries, Palm
Coast, Florida. Jan 22-27 1995.)
30
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage: Innovative Treatment Technologies
Gusek, James J. (1996). Passive Treatment of Acid Rock Drainage: What is the Potential Bottom
Line? International Journal of Rock Mechanics and Mining Science & Geomechanics
Abstracts, 33, (2) 95A.
Gusek, James J. (1998). (Knight Piesold, LLC; 1050 Seventeenth St., Suite 500; Denver, CO
80265-0500)
. Three Case Histories of Passive Treatment of Metal Mine Drainage. This
article is from the South African Mining Delegation Seminar. Denver, CO, USA. Aug. 24 ­
Sept. 4, 1998.
Hedin, Robert S., Narin, Robert W., Kleinmann, Robert L. P. (1994). Passive Treatment of Coal
Mine Drainage. Bureau of Land Mines Information Circular 9389.
Holmes, Michael. (EPA-Region 8). Personal communication. 22 July, 2003.
Hopey, Don. (2001, August 10). Seeing Red and Profit from Mine Iron Oxide Touted for Paint,
Hair
Dye. Pittsburgh Post – Gazette. pp. B-4.
Kepler, D. & McCleary, E. (Damariscotta). (2003). Successive Alkalinity Producing Systems ­
for
the 2003 West Virginia Surface Mine Drainage Task Force Symposium. Available at:
http://www.wvu.edu/~agexten/landrec/2003TFS/Kepler03.pdf.
Lauzon, Shelly (Woods Hole Oceanographic Institution). (2002, March). New Iron Eating
Microbe
Major Component of Mining Pollution and Iron and Sulfur Cycling.
Available at: http://www.whoi.edu/media/IronEatingMicrobe.html
Lewis, Jim (Colorado Department of Public Health and Environment). Personal communication.
16 July, 2003.
Macalady, Donald L. (Ed.). (1998) Passive Bioremediation of Metals and Inorganic
Contaminants. Perspectives in Environmental Chemistry. New York: Oxford University
Press. p 473-495.
Madejon, P., Murillo, J. M., Maranon, T, Cabrera, F, and Soriano, M.A. (2003,
May) Trace
elements and nutrient accumulation in sunflower plants two years after the Aznalcollar
mine spill. The Science of the Total Environment, 307. (1-3) pp. 239-257.
Welch, E. B. (1980). Ecological Effects of Waste Water.
New York, NY: Cambridge University
Press.
31

Acid Mine Drainage: Innovative Treatment Technologies
Matlock, Matthew M., Howerton, Brock S., & Atwood, David A. (2002). Chemical Precipitation
of heavy metals from acid mine drainage. Water Research 36, 4757-4764.
Matlock, Matthew M., Howerton, Brock S., & Atwood, David A. (2003). Irreversible binding of
mercury
from contaminated soil. Advances in Environmental Research 7, 347-352.
Matlock, Matthew M., Howerton, Brock, S., Aelstyn, Michael Van, Henke, Kevin, R, and
Atwood
, David A. (2003). Soft metal preferences of 1,3-benzenediamidoethanethiol.
Water Research, 37, 579-584.
Maxemchuk, Amanda. (2002). DRAFT of Final Report: Assessment of Ecological Risk From
Untreated
Mine Tailings and the California Gulch Superfund Site in Leadville, Colorado,
Phase IV: August 2001 Sampling Event. [Prepared for Harry Compton & Mark Sprenger
of EPA].
Mineral Policy Center. (2003). Cleaning Up Abandoned Hardrock Mines.
Available
at: http://www.mineralpolicy.org/us/abandoned.php
Moore, Steve. (Bureau of Land Management). Personal communication. 30 July, 2003.
Moore, Steve and Kotansky, Dan (US Dept. of the Interior, Bureau of Land Managment,
Abandoned
Mine Lands, Idaho) (Feb. 2002). AML Project Notebook: Project Name:
Champagne Creek Restoration. Available at:
http://www.id.blm.gov/aml/projects/champagne/index.htm
Murry, Douglas T., Townsend, Steven A., and Sopper, William E. (1981)
Using Sludge to
Reclaim Mineland. BioCycle, 22 (3), 48-55.
Naftz, David L., Morrison. Stan J., Davis, James A., & Fulter, Christopher (Eds.). (2002).
Handbook of Groundwater Remediation Using Permeable Reactive Barriers: applications
to radionuclides, trace metals, and nutrients. San Diego, California: Academic Press.
[ISBN: 0-12-513563-7]
Naftz, David L., Davis, J.A., Fuller, C.C., Morrison, S.J., Freethey, G.W., Feltcom, R.G.,
Wilhelm
, M.J., Piana, J. Joye, and Rowland, R.C. (1999).
Field Demonstration of
Permeable Reactive Barriers to Control Radionuclide and Trace-Element Contamination
in Ground Water from Abandoned Mine Lands
. U.S. Geological Survey Toxic Substances
Hydrology Program - Proceedings of the Technical Meeting, Charleston, South Caroline,
March 8 -12. USGS Water Resources Investigations Report 99-4018A, Volume 1,
Contamination from Hard-Rock Mining. Available at: http://toxics.usgs.gov/pubs/wri99-
4018/
32

Acid Mine Drainage: Innovative Treatment Technologies
Nash, J. Thomas (2002). Hydrogeochemical Investigations of Historic Mining Districts, Central
Western Slope of Colorado, Including Influence on Surface-Water Quality. U.S. Dept. of
the Interior: U.S. Geological Survey. USGS Digital Data Series DDS-73 Version 1.0.
Available at: http://pubs.usgs.gov/dds/2002/dds-073/dds-073.pdf
Nuttall, Charlotte A., Younger, Paul L. Zinc removal from hard, circum-neutral mine waters using
a novel closed-bed limestone reactor. Water Research, 34 1262-1268.
Oklahoma Department of Environmental Quality. Governor Frank Keating’s Tar Creek Superfund
Task
Force, Water Quality Subcommittee Task 2 Report, Water Quality Improvement
Alternatives. (2000, Sept. 7). Available at:
http://www.deq.state.ok.us/LPDnew/Tarcreek/WQTask2FinalReport.PDF
Oven Run Project: Project Key to Making Stonycreek a Net Alkaline System.
Available
at: http://www.ctcnet.net/scrip/stoven.htm
Pioneer Technical Services, Inc. (2002). Draft Final Expanded Engineering Evaluation/Cost
Analysis(EEE/CA)
For the McLaren Tailings Site; Cooke City, Montana. Prepared for:
Mr. John Koerth, Montana Dept. of Environmental Quality - Mine Waste Cleanup Bureau.
http://www.deq.state.mt.us/Rem/mwc/download.asp
Rees, Ben., Bowell, Rob, Dey, Matt, & Williams, Keith. (2001). Performance of a Successive
Alkalinity Producing System (SAPS). Available at:
http://www.srk.co.uk/news/SAPS_Manuscript.pdf
Sasowsky, Ira D., Foos, Annabelle, & Miller, Christopher M. (2000).
Lithic controls on the
removal of iron and remediation of acidic mine drainage. Water Research, 34. (10) pp.
2742-2746.
Sopper, William E. (1993). Municipal Sludge Use in Land Reclamation. Lewis Publishers: Boca
Raton.
Steele, Dave (District Manager of Somerset County Conservation District). Personal
communication
. 18 June. 2003.
Stephens, James F., Rogers, Robert S., Fields, Phillip K. (2001, March)
Conservation Practices
for the Reclamation of Surface Mines in Arkansas. Arkansas Department of Environmental
Quality, Surface Mining and Reclamation Division. Available at:
http://www.adeq.state.ar.us/mining/pdfs/handbook.pdf
33
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage: Innovative Treatment Technologies
Successive Alkalinity Producing Systems for Renovation of Acid Mine Drainage. [associated with
Virginia Tech, Powell River Project]. Available at:
http://sudan.cses.vt.edu/prp/Research_Results/SAPS.html
Toffey, William. (2003) Twenty-five Year of Mine Reclamation with Biosolids in Pennsylvania.
From Conference Proceedings : RevTech: Cleaning Up Contaminated Properties for Reuse
and Revitalization: Effective Technical Approaches and Tools, USEPA Technology
Innovation Office. July 22-24, 2003. Marriott City Center, Pittsburgh, PA.
Upper Arkansas River Alluvium Remediation: Biosolids Demonstration, Leadville, CO (2000)
.
Available at : http://faculty.washington.edu/clh/leadville.html
U.S. Department of the Interior (2002a) : Office of Surface Mining. Abandoned Mine Land
Program.
Unreclaimed Problems: Public Health & Safety Coal Related High Priority
Problems (Priority 1 & 2). Available at: http://www.osmre.gov/aml/remain/zintroun.htm.
U.S. Department of the Interior (2002b) : Office of Surface Mining. Acid Mine Drainage
Treatment Techniques and Costs. Available at : http://www.osmre.gov/amdtcst.htm
U.S. Department of the Interior (2003a) : Office of Surface Mining Reclamation and Enforcement.
White paper on: “People Potentially at Risk From Priority 1 & 2 AML Hazards”
Available: http://www.osmre.gov/pdf/wp041703.pdf
U.S. Department of the Interior (2003b) : Bureau of Land Management: Abandoned Mine Lands
Cleanup Program. Frequently Asked Questions. Available at:
http://www.blm.gov/aml/faqs.htm
U.S. General Accounting Office. (1996). Federal Land Management: Information on Efforts to
Inventory Abandoned Hard Rock Mines. Washington, DC. Available at:
http://frwebgate.access.gpo.gov/cgi-bin/useftp.cgi?IPaddress=162.140.64.21&filename=rc
96030.pdf&directory=/diskb/wais/data/gao
U.S. Environmental Protection Agency (1993a). [Wildeman, Thomas, Dietz, John, Gusek, James,
Morea,
Susan.] Colorado School of Mines - Constructed Wetland-Based Treatment.
EPA/540/R-93/523. Full Report available at:
http://www.epa.gov/ORD/SITE/reports.html.
U.S. Environmental Protection Agency. (1993b). Emerging Technology Summary: Handbook for
Constructed
Wetlands Receiving Acid Mine Drainage. EPA/540/SR-93/523.
34

Acid Mine Drainage: Innovative Treatment Technologies
U.S. Environmental Protection Agency (1999). Review of Passive Systems for Treatment of Acid
Mine Drainage: MEND (Mine Environment Neutral Drainage Program) Report 3.14.1.
May 1996: revised 1999. Available at:
http://www.epa.gov/NE/superfund/sites/elizmine/43547mend.pdf
U.S. Environmental Protection Agency (2000), National Risk Management Research Laboratory,
Office of Research and Development. Introduction to Phytoremediation. EPA/600/R-
99/107.
U.S. Environmental Protection Agency: Solid Waste and Emergency Response. (2001)
.
Abandoned Mine Site Characterization and Cleanup Handbook. EPA530-R-01-002.
U.S. Environmental Protection Agency (2002a). Biosolids.
Available
at: http://www.epa.gov/owm/mtb/biosolids/index.htm
U.S. Environmental Protection Agency (2002b). SITE Technology Capsule: Anaerobic Compost
Constructed
Wetlands Technology. EPA/540/R-02/506a.
U.S. Environmental Protection Agency (2002c). Section 319 Success Stories, Vol. III: Acid
Mine
Drainage Treatment Wetlands: A Sustainable Solution for Abandoned Mine Problems.
Available at: http://www.epa.gov/owow/nps/Section319III/OK.htm
U.S. Environmental Protection Agency (2003). Active NPL Sites: California Gulch. Available at:
http://www.epa.gov/region8/superfund/sites/co/calgulch.html
U.S. Environmental Protection Agency. A Handbook of Constructed Wetlands: Volume 4 Coal
Mine
Drainage. 843F00003. [Search for document title or number at
http://www.epa.gov/clariton/clhtml/pubtitleOW.html]
Wong, M.H. (2003) Ecological Restoration of Mine Dredged Soils, With Emphasis On Metal
Contaminated
Soils. Chemosphere Issue 6, 775-780.
Younger, Paul, Banwart, Steven A, Hedin, Robert, S. (2002) Mine Water: Hydrology,
Pollution, Remediation. The Netherlands: Kluwer Academic Press.
Zaluski, Marek, Trudnowski, John, Canty, Marietta, Baker, Mary Ann Harrington (MSE
Technology
Applications, Inc. Butte, Montana). Performance of Field-Bioreactors with
Sulfate-Reducing Bacteria to Control Acid Mine Drainage
.
Fifth International Conference
35

Acid Mine Drainage: Innovative Treatment Technologies
on Acid Rock Drainage, 20-26 May, 2000, Denver, CO Society for Mining, Metallurgy,
and Exploration, Inc. (SME), Littleton, CO. ISBN: 0-87335-182-7. Vol 2, p 1169-1175.
36

Acid Mine Drainage: Innovative Treatment Technologies
Appendix A: State Mine Reclamation/Remediation Status
This section presents state programs and activities that address abandoned mines. The focus was
limited to Western states dealing with hard rock mines. Note that this list is not comprehensive of
all of the programs and activities occurring in each state.
Alaska
Department of Natural Resources
Division of Mining, Land & Water
http://www.dnr.state.ak.us/mlw/mining/aml.htm
Funding for this program comes from SMCRA funds. The state is only able to generate $200,000,
but
has $2 million in reclamation needs; therefore, the state qualifies for Minimum Program Status
from the SMCRA fund, entitling them to $1.5 million plus emergency funds annually until the
work remaining on the inventory drops below $2 million.
Coal and non-coal mining abandoned mines were inventoried. The coal inventory is complete, and
340 sites
were identified. The non-coal inventory is incomplete with a count of 432. Each site
was evaluated to determine funding eligibility. Priority 1 and 2 coal projects must be completed
first, so only priority non-coal projects can be reclaimed. Priority 3 projects can be worked on in
conjunction with Priority 1 and 2 projects or after all Priority 1 and 2 projects have been
completed.
The state developed a variety of priorities to select sites for remediation; they came up with 224
coal
projects and 32 - 123 non-coal projects. Initial inventories estimated costs at $52 million and
non-coal costs at $2.7 million. To date, 36 AML projects have been completed at a cost of
$8,880,980. Most of the projects involved preventing physical hazards.
U.S. Department of the Interior - Bureau of Land Management - Alaska
http://www.ak.blm.gov
http://www.ak.blm.gov/amines/amlindex
.html
About 15 to 20 projects are either active or have been completed. Projects are selected using
water
shed approach (i.e., projects that will have the greatest impact on water quality in the
watershed are chosen first). As with many of the programs, funding is an issue. The Web page
states, “Because there is never enough money, the BLM must first consider watersheds damaged
by
abandoned mines.”
37

Acid Mine Drainage: Innovative Treatment Technologies
Arizona
Arizona State Mine Inspector
http://www.asmi.state.az.us
http://www.asmi.state.az.us/abandoned.html
Part of this state office’s mission is to review and monitor all mine reclamation activities. This
office established the Abandoned and Inactive Mine (AIM) Survey to inventory abandoned and
inactive mines throughout Arizona. The majority of the funding for this program comes from the
Bureau of Land Management. The program began inventorying sites in 1992 and estimates that
there are at least 125,000 abandoned or inactive openings in the State of Arizona.
As of January 1999, 7,844 mines have been surveyed, with 288 mines with some type of
Environmental Hazards and 1149 mines with Significant Public Hazards.
U.S. Department of the Interior - Bureau of Land Management - Arizona
http://www.az.blm.gov/
California
U.S. Department of the Interior - Bureau of Land Management - California
http://www.ca.blm.gov
http://www.ca.blm.gov/pa/aml/ - for specific AML activity
The California BLM manages 15 Resource Areas (RA’s, Field Offices) comprising over 16 million
acres
in California and Northwest Nevada. Over 12,000 mine properties in California and
Northwest Nevada are listed in the Bureau of Land Mines Mineral Industries Location System
(MILS) database as on BLM land. An estimated additional 5000 sites likely to be on BLM land
are not recorded in this database are. Of these 17,000 sites, an estimated 3000 significant
properties contain hazardous substances or physical features and/or have environmental problems.
No comprehensive AML inventory has been conducted on any RA in the state and six RA’s have
no recorded inventory of mine sites.
“The California State Office, (with limited staff) from mid-2000, has been conducting watershed-
based projects that have and will continue to identify mine sites with environmental and/or safety
issues” (http://www.ca.blm.gov/pa/aml/). To date about 40 sites have been identified as “high
priority,” more than 170 sites have been added to the Abandoned Mine Land Identification System
- a database of AMLs on BLM lands. According to the website 7 projects have been completed as
of
April 21, 2003.
California Environmental Protection Agency
38

Acid Mine Drainage: Innovative Treatment Technologies
State Water Resources Control Board
http://www.swrcb.ca.gov/
One of the things found via this website was a document entitled, “The Abandoned Mine Technical
Advisory Committee’s Report on Abandoned Mines.”
[http://www.swrcb.ca.gov/nps/docs/tac_abandmin.doc] The document was created by the
Abandoned Mines Technical Advisory Committee (TAC). TAC spent six months discussing the
issues surrounding abandoned mines, past cleanup efforts, and desired future courses of action.
They prioritized courses of action and identified barriers to progress. TAC identified lack of
funding as a key impediment to cleanup of abandoned mines.
California is not a coal-mining state and therefore is ineligible to receive SMCRA funds.
Colorado
Colorado Department of Public Health and Environment
Hazardous Materials and Waste Division
http://www.cdphe.state.co.us/hm/hmhom.asp
This division regulates solid waste management, treatment, disposal facilities, and hazardous
waste
generation, storage, transportation, treatment, and disposal. The division also ensures compliance
with state hazardous waste regulations and permits and oversees remediation of contamination at
Federal Facilities located in the state. The division assists in the cleanup of hazardous waste sites
under the Superfund Program, and encourages brownfields redevelopment through implementation
of the Voluntary Cleanup and Redevelopment Act.
This state agency has dealt with the remediation of a few mine sites including Bonanza, Clear
Creek, Eagle, Idarado, Leadville (California Gulch), and Summitville Mine.
U.S. Department of the Interior - Bureau of Land Management
Colorado Abandoned Mine Land Program
http://www.co.blm.gov/mines/mine.htm
There are about 2,600 abandoned mines on Colorado’s public lands. The projects during 2002 are
listed
below. The projects are being addressed with a watershed approach.
Arkansas Watershed:
LakeFork Project - includes Nelson and Dinero Tunnel Projects
Mill Sap Gulch Project
Mount Robinson Project/Historic Rosita Mining District
Upper Animas Watershed
39
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage: Innovative Treatment Technologies
Elk Tunnel
Forest Queen
Joe and John
Lackawanna and Lark
Upper Gunnison Watershed
Palmetto Mine Remediation
Roy Pray #1 Remediation
Ute Ulay Mine/Mill Remediation and Mine Waste Repository
Wyoming Mine Remediation
Many of the project involve water diversion, materials removal, and revegetation.
United States Geologic Survey (USGS)
Toxic Substances Hydrology Program
http://toxics.usgs.gov/sites/upper_ark_page.html
The USGS has a few projects in Colorado and elsewhere that attempt to characterize metal
transport
in streams affected by mining. Work in the Upper Arkansas Toxic-Substance Hydrology
Project began in 1986. The approach is to study chemical processes within a hydrologic context,
using a two-step approach. First, we have employed in-stream experimentation to provide data
about the processes affecting metals. Second, they have used the resulting data sets to develop
and apply solute transport models that help quantify rates and processes. See the Web page for
more information about this and other projects.
Idaho
U.S. Department of the Interior - Bureau of Land Management
Idaho Abandoned
Mine Lands
http://www.id.blm.gov/aml/index.htm
The program stems from a 1982 report that four dozen livestock had been poisoned by ingestion
of lead tailings.
Significant effort has been put into Pine Creek, a tributary of the Coeur d’Alene River, in the Silver
Belt region of northern Idaho. Between 1996 and 1998 more than 30,000 cubic yards of tailings
were removed from the flood plain to prevent the deposition of the material in the river. Much of
the cleanup effort was accomplished through funding by the hazardous materials program, Central
Haz Mat Fund, and emergency flood funding.
40

Acid Mine Drainage: Innovative Treatment Technologies
Systematic AML site inventories began in mid-1990s.
Starting in the fiscal year 1999, Clean Water Action Plan funding enabled a more uniform national
effort to move from inventory to cleanup of AML sites. Project summaries of completed or active
projects can be found in the AML Project Notebook link at
http://www.id.blm.gov/aml/notebood.htm
In FY 2002, 2 projects using passive treatment were installed: Champagne Creek and Bridge
Creek.
At the beginning of fiscal year 2002, the focus was to better integrate AML with other statewide
Idaho priorities. “Lack of a national source of funding dedicated to addressing physical hazards
continues to be an issue. This year we are seeking a reallocation of some of Idaho’s BLM
program funding to better address priority sites, particularly in the proximity of recreation sites and
other public lands heavily visited by the public” (http://www.id.blm.gov/aml/program.htm
).
Basin Environmental Improvement Project Commission
http://www.basincommission
.com/
This organization was created by the Idaho legislature under the Basin Environmental
Improvement Act of 2001; the group became operation in March of 2002. It consists of
representatives of the state of Idaho, the three Idaho counties in the Basin, the Coeur d’Alene
Tribe, the state of Washington, and the United States of America (represented by the U.S. EPA).
It is the policy of the state to provide a system for environmental remediation, natural resource
restoration and related measures to address heavy metal contamination in the Coeur d’Alene Basin.
Montana / Dakotas
Department of Environmental Quality
Mine Waste Cleanup Bureau
http://www.deq.state.mt.us/rem/mwc/index.asp
The Mine Waste Cleanup Bureau (MWCB) focuses on two primary site types:
1) inactive mine sites addressed under the Surface Mining Coal and Reclamation Act
(SMCRA
1977).
2) mining related sites addressed under the Federal Comprehensive Environmental
Response and Liability Act (CERCLA) .
The MWCB divided its site-reclamation duties in this way because of distinctions between
applicable environmental laws and associated funding mechanisms.
41

Acid Mine Drainage: Innovative Treatment Technologies
The DEQ-MWCB must give priority to abandoned coal mines, and Montana has completed
reclamation of its abandoned coal mines and has now moved on to non-coal sites. The non-coal
sites are ranked in priority order based on a scoring system developed by the state. To date,
Montana’s abandoned mine reclamation program has overseen the completion of more than 283
projects totaling nearly 1174 acres.
U.S. Department of the Interior - Bureau of Land Management
Montana/Dakotas Abandoned Mine Land Program
http://www.mt.blm.gov/aml/index.html
Montana BLM has been working to clean up abandoned mines located on public lands utilizing a
watershed approach since 1995. An inventory of 1078 abandoned mines located on public lands
resulted in 65 sites that needed further investigation and potentially reclamation.
At least 15 projects are underway or completed. More information can be obtained from the Web
page cited above.
Navajo Nation
Division of Natural Resources
Navajo AML Reclamation/UMTRA [Uranium Mill Tailings Radiation Control Act] Department
http://www.navajoaml.osmre.gov
This program was certified to have reclaimed all Priority 1 and 2 abandoned coal mines by the
Secretary
of Interior as of May 4, 1994 . The program is now permitted to focus attention on non-
coal mines. The Navajo AML Program anticipates having all known and eligible abandoned mines
reclaimed by the end of 2004.
In 2000, the Navajo AML Program amended its AML Plan to incorporate the provisions of
SMCRA, Sections 411(e) and (f), which provide the authority for using AML funds to construct
public facilities as a means of mitigating current and past mining-related impacts to such
communities. Thus, the Navajo AML Program can now also use its AML funds for the
construction of Public Facility Projects (PFP’s). Navajo AML funded its first PFP in EY-2002.
In 2002, four reclamation projects were completed, all of the work done “minimized the need for
maintenance, promotes landscape stability, enhances re-establishment of natural vegetation,
enhances wildlife, and most importantly, adequately safeguards the physical and radioactive
hazards.”
- Office of Surface Mining Reclamation and Enforcement : Annual Evaluation Report ­
Evaluation Year 2002 (Oct. 1, 2001 through Sept. 30, 2002) - on the Navajo Abandoned
Mine Lands Reclamation Program.
http://www.navajoaml.osmre.gov/News_Rprts/AML/OSM_AER_Nav2002.pdf
42
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage: Innovative Treatment Technologies
Nevada
State of Nevada Commission on Mineral Resources
Division of Minerals
400 W. King Street, Suite 106
Carson City, Nevada 89703
(775) 684-7040
fax: (775) 684- 7052
http://minerals.state.nv.us/
The state’s first priority is to reduce hazards such as high walls, embankments, etcetera. It is
estimated that the state has 200,000 abandoned mine features. Approximately 50,000 present
physical safety hazards, including 9,244 hazardous mine openings throughout the state. Seven
thousand have been secured.
As for environmental problems, the State of Nevada has an Interagency Abandoned Mine Land
Environmental Task Force. In their Sept. 1999 report, an estimated 1 to 3 percent of 200,000 to
500,000 abandoned mine land features have the potential to impact ground or surface waters.
Even at 1 percent the numbers are very high--20,000 to 60,000 potential pollution sources. As of
1999 there were 33 sites identified for clean-up; 6 of these sites were considered high priority and
site characterization had begun. The report can be reviewed at:
http://minerals.state.nv.us/forms/aml/nvamlreport.pdf
U.S. Department of the Interior - Bureau of Land Management
Nevada
http://www.nv.blm.gov/AML/
In March of 1999, the Bureau of Land Management-Nevada State Office (BLM) initiated the
formation of an Interagency Abandoned Mine Land Environmental Task Force (IAMLET) to
begin remediation of abandoned mine land (AML) environmental problems associated with
watersheds in Nevada. The task force is comprised of federal and state agencies with a role in
abandoned mine lands in the state. Initial funding for the program is from the BLM through the
Soil, Water, and Air Management Budget, in accordance with the Clean Water Action Plan.
- From the report, 1999 Interagency Mined Land Environmental Task Force Report, found
on the webpage above.
Their accomplishments as of March 1999 included:
- Initiation of cleanup of two AML sites (Steward and Atronics millsites) in priority watershed;
- Establishment of site selection criteria for potential AML reclamation projects;
43

Acid Mine Drainage: Innovative Treatment Technologies
- Compilation of an initial list of 33 AML sites based on proximity and potential impacts to
watersheds and assignment of a priority rank to each site;
- Initiation of data compilation, including location and land status maps, existing site
characterizations
, and photographs for the 33 sites.
The groups involved with this Interagency are:
Bureau
of Land Management, BLM
United States Forest Service, USFS
United States Fish and Wildlife Service, USFWS
United States Geological Survey, USGS
Environmental Protection Agency, EPA
Nevada Division of Minerals, NDOM
Nevada Division of Environmental Protection, NDEP
Nevada Division of Wildlife, NDOW
Nevada Bureau of Mines and Geology, NBMG
Desert Research Institute, DRI
New Mexico
New Mexico Energy, Minerals and Natural Resources Department
Mining and Minerals Division
Abandoned Mine Land Program
http://www.emnrd.state.nm.us/Mining/aml/default.htm
The program was formed when SMCRA was passed in 1977. The description of this program
states
: “the fund is used to reclaim coal mines abandoned prior to the enactment of SMCRA.
Under certain conditions, abandoned noncoal mines may also be reclaimed.” The most common
mine hazards in NM are open adits and shafts. There are other concerns, including burning gob
piles and acid mine drainage.
New Mexico Energy, Minerals and Natural Resources Department
Mining and Minerals Division
Mining Act Reclamation Program
44
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage: Innovative Treatment Technologies
http://www.emnrd.state.nm.us/Mining/marp/default.htm
This program was created under the New Mexico Mining Act of 1993 to regulate hard rock
mining reclamation activities for all minerals except potash, sand, gravel, quarry rock used as
aggregate in construction, flagstone, calcite, clay, adobe, borrow dirt, activities regulated by the
Nuclear Regulatory Commission, and waste regulated under Subtitle C of the Federal Resource
Conservation and Recovery Act.
New Mexico Energy, Minerals and Natural Resources Department
Ground Water Quality Bureau
Mining Environmental Compliance Section
Active mines are handled through this office when water quality is an issue. Upon speaking with
Mark Phillip of this office, it became clear that most of their work involves water diversion and
water treatment plants.
Wyoming
Department of Environmental Quality (DEQ)
Abandoned Mine Land
http://deq.state.wy.us/aml
AML’s mission is to eliminate safety hazards and repair environmental damage from past mining
activities and to assist communities impacted by mining. AML pursues this mission in two ways:
1. The Traditional Reclamation Program which has reclaimed thousands of acres of
abandoned
coal, bentonite, and uranium open pit mines, and new projects are initiated each
year. AML has also closed several hundred hazardous gold and copper mine openings, and
has an ongoing program to mitigate subsidence risks. AML also makes subsidence
insurance available to property owners in affected communities.
2. The Public Facility Program, operating in conjunction with the State Loan and
Investment
Board, provides financial assistance for projects in communities with current or
past impact from mining. Applicants must first establish eligibility, then projects are ranked
and funded based on human health and safety issues.
U.S. Department of the Interior - Bureau of Land Management
45
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage: Innovative Treatment Technologies
Abandoned Mine Land Reclamation
http://www.wy.blm.gov/whatwedo/aml/aml_home.html
Wyoming BLM works closely with DEQ to share resources and pool funding. The projects listed
on the Web page did not use any innovative treatments.
Inventory of the sites and the work needed at each was expected to be done by the end of 2001.
46
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Acid Mine Drainage: Innovative Treatment Technologies
Appendix B: Brief Case Description of Case Studies Found in This Report
Site
Location
Mine Information
Pollution (media)
Treatment
Page
North Pennine Orefield
Nent Valley, UK
Metal/Hardrock
Zn (water)
Anoxic Limestone Drain
12
Burleigh Tunnel
Silver Plume, CO
Metal/Hardrock
Zn (water)
Constructed Wetland
14
Calliope Mine
Silver Bow, MT
Metal/Hardrock
Al, As, Cd, Cu, Fe, Mn, Zn, low pH
(
water)
Bioreactor
15
Champagne Creek
Butte, ID
Metal/Hardrock
Al, Cd, Cu, Fe, Zn, low pH (water)
Bioreactor
16
Oven Run
Pennsylvania
Coal
Fe, Al, low pH (water)
Successive Alkalinity
P
roducing System
18
Gowen Run
Gowen, OK
Coal
Fe, other metals, low pH (water)
Successive Alkalinity
P
roducing System
18
Frostburg, MD
Coal - strip mine
metals, low pH (soil)
Biosolids
21
Leadville, CO
Meta
l/
Hardrock
Zn, Pb, low
pH (
soil)
Biosolids
23
47

Exhibit M:
The Passive Treatment of Coal Mine Drainage
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

 
DOE/NETL-2004/1202
George R. Watzlaf
1
, Karl T. Schroeder
1
, Robert L. P. Kleinmann
1
,
Candace L. Kairies
1
, and Robert W. Nairn
2
1
U.S. Department of Energy
National Energy Technology Laboratory
P.O. Box 10940
Pittsburgh, PA 15236
2
University of Oklahoma
School of Civil Engineering and Environmental Science
202 West Boyd Street
Norman, OK 73019-0631
The Passive Treatment of Coal Mine Drainage
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

 
2
Abstract
Passive treatment of mine water uses chemical and biological processes to decrease metal
concentrations and neutralize acidity. Compared to conventional chemical treatment, passive
methods generally require more land area, but use less costly reagents, and require less
operational attention and maintenance. Currently, the three most common types of passive
technologies are aerobic ponds and wetlands, anoxic limestone drains (ALDs), and reducing and
alkalinity-producing systems (RAPS). Aerobic wetlands promote mixed oxidation and
hydrolysis reactions, and are effective when the raw mine water is net alkaline. Anoxic limestone
drains generate bicarbonate alkalinity and can be used to convert water that is net acidic into net-
alkaline water for treatment in aerobic ponds and wetlands. RAPS promote reducing conditions
and limestone dissolution. They extend the concept of ALDs by pre-treating the water before it
contacts the limestone, to eliminate dissolved oxygen and reduce dissolved ferric iron to ferrous
iron. These systems can generally be used to treat more acidic water than ALDs, and can better
treat water with significant aluminum concentrations.
In passive treatment systems, rates of metal and acidity removal and alkalinity generation
have been developed empirically. Aerobic wetlands remove iron from alkaline water at rates of
10 to 20 g m
&2
d
&1
. Anoxic limestone drains add 150 to 300 mg/L of alkalinity in about 15 hours
of contact, imparting 5 to 20 mg/L of alkalinity per hour of contact. Reducing and alkalinity-
producing systems add 15 to 60 g m
&2
d
&1
of alkalinity, depending on influent water quality and
contact time. Selection and sizing criteria for the design of passive treatment systems are
presented in this report.
Acknowledgements
The authors would like to thank the technical reviewers whose comments were greatly
appreciated: Arthur W. Rose, Paul L. Younger, Carl S. Kirby and Ben B. Faulkner. Laboratory
analyses were conducted by Mark Wesolowski, Joyce Swank, Dennis Viscusi, Harry Williams,
Hubert McDonald and Lillian Schlosser Balchus. John Kleinhenz, Randy Woods and John
Odoski assisted with field sampling and monitoring. We are indebted to the following
individuals who shared information on the construction of many of the passive treatment systems
discussed in this report: Doug Kepler, Eric McCleary, Dan Seibert, Margaret Dunn, Tim
Danehy, Connie Lyons and Joe Mills.

 
3
Contents
Introduction..................................................................................................................... 6
Treatment of Mine Water............................................................................................ 6
History of Passive Treatment...................................................................................... 7
Background ..................................................................................................................... 9
Formation of Polluted Mine Waters ........................................................................... 9
Chemical Characteristics of Mine Drainage............................................................. 11
Passive Treatment Processes ........................................................................................ 18
Limestone Dissolution .............................................................................................. 18
Sulfate Reduction...................................................................................................... 19
Metal Removal Processes ......................................................................................... 19
Materials and Methods.................................................................................................. 31
Collection of Water Samples .................................................................................... 31
Analysis of Water Samples....................................................................................... 31
Flow Rate Measurements.......................................................................................... 31
Analysis of Iron Sludge ............................................................................................ 32
Removal of Contaminants by Passive Unit Operations................................................ 32
Aerobic Wetlands and Ponds .................................................................................... 32
Anoxic Limestone Drains ......................................................................................... 33
Compost Wetlands .................................................................................................... 43
Reducing and Alkalinity-Producing Systems (RAPS) ............................................. 44
Other Types of Water Treatment Systems................................................................ 52
Designing Passive Treatment Systems ......................................................................... 54
Characterizing Mine Drainage Discharges ............................................................... 54
Selecting Unit Operations......................................................................................... 55
Sizing Passive Systems............................................................................................. 57
Constructing Passive Systems .................................................................................. 61
Operation and Maintenance ...................................................................................... 62
Conclusions................................................................................................................... 63
Abbreviations and Acronyms ........................................................................................... 65
References......................................................................................................................... 66

 
4
Figures
Figure 1. Calculated Versus Measured Acidity for Over 150 Coal Mine Discharges ........... 12
Figure 2a. Concentration of Iron and Field pH at the Emlenton Constructed Wetlands, which
Receives Net Acidic Water................................................................................... 22
Figure 2b. Concentration of Iron and Field pH at Cedar Grove Constructed Wetlands which
Receives Net Alkaline Water................................................................................ 23
Figure 3. Dissolved Ferric Iron Concentration Versus pH in Coal Mine Discharge.............. 24
Figure 4. Removal of Ferrous Iron from Acidic and Alkaline Mine Waters in a Laboratory
Experiment............................................................................................................ 25
Figure 5. Mean Concentration of Iron, Manganese and Magnesium at the Morrison Wetland
as the Mine Water Flows Linearly through the System........................................ 28
Figure 6. Changes in the Concentrations of Ferrous Iron and Manganese in (A) the Absence
of the MnOOH and (B) the Presence of MnOOH ................................................ 29
Figure 7. Dissolved Aluminum Concentration Versus pH in Coal Mine Discharges ............ 30
Figure 8. Bromide Concentration Versus Time in the Effluent of the Howe Bridge ALD 1
Resulting from a Pulse Input of a Bromide Tracer ............................................... 36
Figure 9. Alkalinity Concentration as Mine Water Flows through Selected ALD................. 38
Figure 10. Effluent Alkalinity Concentrations of Selected ALDs Over Time ....................... 39
Figure 11. Alkalinity Generation in the Howe Bridge RAPS................................................. 47
Figure 12. Selection of Passive Treatment Unit Operations................................................... 56
Tables
Table 1. Federal Effluent Limits for Coal Mine Drainage ....................................................... 6
Table 2. Contributions of Metal Concentration and pH to Acidity for Selected Mine
Discharges............................................................................................................. 13
Table 3. Proton Acidity Contributions at Various pH Values ................................................ 13
Table 4. Chemical Composition of Untreated Mine Waters Containing Alkalinity .............. 14
Table 5. Water Quality from 156 Coal Mine Discharges ....................................................... 17
Table 6. Equilibrium Concentrations of Alkalinity at Various P
CO2
Levels........................... 18
Table 7. Solubility Products of Selected Metal Sulfides ........................................................ 20
Table 8. Dimensions, Stone Size and Quality, and Source of Influent Water Quality
Data for ALDs....................................................................................................... 35
Table 9. Tracer Test Data for Two ALDs............................................................................... 37
Table 10. Initial and Current Conditions of ALDs ................................................................. 40
Table 11. Average Water Quality Before and After Contact with the Anoxic Limestone
Drain ..................................................................................................................... 40
Table 12. Additional Water Quality Parameters Before and After Contact with the
Anoxic Limestone Drain....................................................................................... 41
Table 13. Construction Specifications and Quantification of Alkalinity Generation
within RAPS ......................................................................................................... 46
Table 14. Water Quality Before and After Contact with Reducing and Alkalinity
Producing System ................................................................................................. 48
Table 15. Total Amounts of Retained Iron and Aluminum Prior to Flushes at the
DeSale II Site ........................................................................................................ 51
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

5
Table 16. Amount of Water, Iron, and Aluminum Flushed for the Two RAPS at the
DeSale II Site ........................................................................................................ 51
Table 17. Techniques Used for Treating Coal Mine Drainage.............................................. 54
Table 18. Classification of Mine Discharges.......................................................................... 55
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

 
6
Introduction
Treatment of Mine Water
In the United States, mining companies commonly treat contaminated drainage using
conventional chemical methods. In most conventional treatment systems, metal contaminants are
removed through the constantly measured addition of alkaline chemicals (e.g., NaOH, Ca(OH)
2
,
CaO, Na
2
CO
3,
or NH
3
) to meet Federal effluent limits. (See Table 1.) These are maximum
concentrations allowed for active coal mining operations. Some operators have much more
stringent effluent limits, based on the quality of the receiving stream. The chemicals used in
these treatment systems can be expensive, especially when required in large quantities. In
addition, there are operation and maintenance costs associated with aeration and mixing devices,
and additional costs associated with the disposal of the metal-laden sludge that accumulates in
settling ponds. It is not unusual for water treatment costs to exceed $10,000 per year at sites that
are otherwise successfully reclaimed. The high cost of water treatment places a serious financial
burden on active mining companies, and has contributed to the bankruptcy of many others.
The high cost of conventional chemical systems limit water treatment efforts at
abandoned sites. Thousands of miles of streams and rivers in Appalachia are currently polluted
by drainage from sites that were mined and abandoned before enactment of effluent regulations.
State and Federal reclamation agencies, local conservation organizations, and watershed
associations all consider the treatment of contaminated mine discharge to be a high priority.
However, insufficient funds are available for chemical water treatment, except in a few
watersheds of special value.
Table 1. Federal Effluent Limits for Coal Mine Drainage
Parameter
Maximum for Any
One Day
Average of Daily Values for
30 Consecutive Days
Iron, total (mg/L)
6.0
3.0
Manganese, total (mg/L)
4.0
2.0
Total suspended solids (mg/L)
70
35
pH (standard units)
between 6.0 and 9.0
During the past 20 years, the possibility that mine water might be treated passively has
developed from an experimental concept to full-scale field implementation at hundreds of sites
throughout the world (Younger et al. 2002, Brown et al. 2002). Passive technologies take
advantage of the natural chemical and biological processes that ameliorate contaminated water
conditions. Ideally, passive treatment systems require no constant input of chemicals, and little
maintenance. Passive treatment systems use contaminant removal processes that are slower than
conventional treatment and thus require longer retention times and larger areas to achieve similar
results. The goal of the passive mine drainage treatment system is to enhance natural
ameliorative processes, so that they occur within the treatment system, not in the receiving water
body. Two factors that determine whether this goal can be accomplished are the kinetics of the
contaminant removal processes, and the retention time of the mine water in the treatment system.
The retention time for a particular mine site is often limited by available land area. However, the

7
kinetics of contaminant removal processes can often be affected by manipulating the
environmental conditions that exist within the passive treatment system. Efficient manipulation
of contaminant removal processes requires an understanding of of each removal process and
their respective limitations..
History of Passive Treatment
Passive treatment of mine water can be traced to two independent research projects
which showed that natural wetlands were ameliorating mine drainage without incurring any
obvious ecological damage. Researchers at Wright State University studied a site in the
Powelson Wildlife Area in Ohio where
Sphagnum recurvem
had volunteered and was growing in
pH 2.5 water. As the water flowed through the boggy area, iron, magnesium, calcium, sulfate,
and manganese all decreased, while pH increased to 4.6. A natural outcrop of limestone located
at the downstream end provided sufficient neutralization to raise the effluent pH to between 6
and 7 (Huntsman et al. 1978)
.
Meanwhile, a similar study was being conducted by a group at
West Virginia University, working at a natural
Sphagnum
-dominated wetland, Tub Run Bog, in
northern West Virginia. They were looking at the ecological damage to the wetland as a result of
drainage water from an adjoining abandoned mine. They found no adverse ecological effects,
and that in fact, within 20 to 50 m of the influent, the pH of the water rose from between 3.05
and 3.55 to 5.45 and 6.05. Sulfate concentrations decreased to 15 mg/L or less, and iron
decreased to less than 2 mg/L (Wieder and Lang 1982). These field observations prompted the
idea that wetlands might be constructed for the intentional treatment of coal mine drainage. It
was thought that the small seeps present at many abandoned mine sites could be passively
treated in this manner. Research efforts were initiated by the United States Bureau of Mines, in
cooperation with Wright State University (Kleinmann et al. 1983, Kleinmann 1985).
Independently, West Virginia University, and subsequently, Pennsylvania State University
conducted research as well (e.g., Gerber et al. 1985, Stone and Pesavento 1985).
Initially, most of these experimental wetlands were constructed to mimic the
Sphagnum
wetlands. However,
Sphagnum
moss was not readily available, proved difficult to transplant, and
tended to accumulate metals to levels that were toxic to the
Sphagnum
after several months of
exposure to mine drainage (Huntsman et al. 1985, Spratt and Wieder 1989). Instead of
abandoning the concept, researchers experimented with different kinds of constructed wetlands.
Eventually a wetland design evolved that proved tolerant to years of exposure to contaminated
mine drainage and was effective at lowering concentrations of dissolved metals. Most of these
treatment systems consisted of a series of small wetlands (< 1 ha) that were vegetated with
cattails (
Typha latifolia
) (Girts et al. 1987, Stark et al. 1990). Although neither were as acid
tolerant or as effective in removing metals as the
Sphagnum
systems, the cattail systems proved
to be very hardy. We gradually learned that these systems were very cost effective in treating
circumneutral and net alkaline mine water, where the primary objective was to precipitate the
iron in the wetland, instead of downstream.
Some of these wetlands were constructed with a compost and limestone substrate to
provide a favorable environment for the cattails to root. Others were constructed without an
exogenous organic substrate; emergent plants were rooted in whatever soil or spoil substrate was
available on the site when the treatment system was constructed. Researchers soon realized that
the cattails were generally collecting only a small component of accumulated metals internally
(Sencindiver and Bhumbla 1988), and that its principal functions were dispersing the flow of the
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

8
water and filtering out the suspended floc of the precipitated metals (some recent research
indicates that plant uptake of iron at low concentrations may be critical to achieving very low
residual iron concentrations (Batty and Younger 2002)). Subsequently, some systems were
constructed that did not rely at all on the early wetland model . Ponds, ditches, and rock-filled
basins were constructed without emergent plants and, in some cases, without soil or organic
substrate.
In the late 1980s, two new approaches were developed that extended the treatment
capabilities of wetlands to more acidic mine water. In the first case, U. S. Bureau of Mines
researchers, assessing the performance of a wetland that had been constructed in an attempt to
treat very acidic water, found that in isolated locations, the mine water was being neutralized and
iron was being precipitated as a sulfide. Apparently, water was flowing down through the
compost/limestone substrate and then back up again, gaining alkalinity in the process (Hedin et
al. 1988). An approach was developed to optimize this effect and was evaluated in the field
(McIntyre and Edenborn 1990, Nawrot 1990), these anaerobic or compost wetlands added
alkalinity, but were not very efficient for iron removal, and required sequential placement of
aerobic and anaerobic systems. Currently, these systems are seldom constructed to treat coal
mine drainage, however, they can be useful for treatment of metal mine drainage, since they
provide a mechanism to remove metals such as cadmium, copper, lead, etc. (Wildeman et al.
1990, Wildeman et al. 1994).
The other new approach involved acidic water in contact with limestone in an anoxic
environment before flowing into a settling pond or wetland system. Although limestone had
previously been used many times to treat mine water, it typically became coated or “armored” by
iron hydroxide. Turner and McCoy (1990) reasoned that if the mine water could be intercepted
before it contacted the atmosphere, and was directed into a limestone-filled French drain, the
dissolved iron would not oxidize to ferric hydroxide to armor the limestone, and the water would
be neutralized. The water could then be discharged into an aeration pond and a wetland. A great
number of anoxic limestone drains (ALDs) were subsequently constructed, and soon, sizing
guidelines were developed (Hedin et al. 1994b). However, as discussed in more detail later in
this manual, these systems also had their limitations. They worked well for mildly acidic water
(pH > 4.5) that was anoxic, but more acidic water tended to contain dissolved aluminum, which
precipitated in the ALD and reduced permeability, often to the point of failure. In addition, if the
pH of the water was below about 3.5, the dissolved iron was often already oxidized (ferric), so
that armoring could occur even if no oxygen was present.
To compensate for dissolved oxygen and dissolved ferric iron, the concept of the ALD
and compost wetland were combined (Kepler and McCleary 1994, Kepler 1995). Compost was
placed up-gradient of the limestone. The bacterial activity in the compost consumed the
dissolved oxygen and reduced the ferric iron to ferrous iron, allowing the ALD component to
work as intended, even for very acidic water. They referred to these systems as sequential
alkalinity-producing systems (SAPS); others have preferred to use the term reducing and
alkalinity producing systems (RAPS) to more accurately describe the process, and to include
systems that did not put more than one unit in sequence. These systems are also called vertical
flow ponds, vertical flow wetlands or vertical flow systems. Aluminum is still retained in these
systems, so Kepler and McCleary (1997) suggested a simple gravity-powered flushing
mechanism to extend their effective life span.
It is difficult to argue with the long-term success of some of these passive treatment
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

 
9
systems. However, failures can be very damaging to the perceived effectiveness of the
technology. In general, we have found that the systems that were not effective or failed were
undersized, improperly designed, or both. The key is to understand the limitations of each unit’s
operation, to have reasonable expectations, and to use conservative sizing criteria to attain
specific water quality goals. Even undersized passive systems can be useful, discharging water
with significantly lower concentrations of metal contaminants than present in the inflow
drainage. These improvements in water quality decrease the costs of subsequent water treatment
at active sites, and decrease deleterious impacts that discharges from abandoned sites have on
receiving streams and lakes.
Researchers have recently developed additional passive treatment technologies, such as
steel slag leach beds (Simmons et al. 2002), which may prove to be useful additions to the
passive treatment arsenal. Research is being conducted on semi-passive approaches that have
the potential to significantly reduce the land requirements of passive treatment systems. Semi-
passive systems fall between conventional chemical treatment, which requires virtually around-
the-clock attention, and passive systems that ideally require very little maintenance and attention
(Younger et al. 2002). Semi-passive systems have been constructed using gravity-, wind-, and
water-powered aeration or neutralization processes, as well as some low-power demanding
devices.
Background
Formation of Polluted Mine Waters
The cause of most mine water degradation is the oxidation of iron sulfide minerals, such
as pyrite (FeS
2
). Equal amounts of acidity are produced by the oxidation of the sulfide to sulfate
(reaction A), and by the oxidation and hydrolysis of iron (reaction B) (Barnes and Romberger
1968):
FeS
2
+ 3.5O
2
+ H
2
O
÷
Fe
2+
+ 2SO
4
2-
+ 2H
+
(A)
Fe
2+
+ 2.5H
2
O + 0.25O
2
÷
Fe(OH)3(s) + 2H
+
(B)
Iron-oxidizing bacteria accelerate pyrite oxidation by two mechanisms: direct oxidation,
and oxidizing Fe
2+
to Fe
3+
, which in turn oxidizes the sulfide minerals (Beck and Brown, 1968,
Duncan et al. 1967, Groudev 1979, Silverman 1967). Direct oxidation is probably most
important during initial acidification, when complete hydrolysis of Fe
3+
and the resultant
precipitation of Fe(OH)
3
are too rapid to allow ferric iron to act as an important oxidant.
As the pH decreases, abiotic oxidation of Fe
2+
slows down dramatically, according to the rate
law:

10
-d (Fe
2+
)
(O
2
(aq)) (Fe
2+
)
= k
(C)
dt
(H
+
)
2
where (Fe
2+
), (O
2
(aq)), and (H
+
) are activities, k is the rate constant, and t is time (Stumm and
Morgan 1981). Below approximately pH 4 (Kirby et al. 1999), the iron-oxidizing bacteria
assume the primary role of oxidizing Fe
2+
, thereby allowing reaction B to continue producing
acidity and ferric hydroxide. Although the reaction stoichiometry remains the same, this is a
transition point from the primarily abiotic stage to the partially biological stage (Kleinmann et al.
1981). The pH decline typically continues to a stage where the reaction chemistry changes to a
biologically-mediated cycle of reactions D and E (Kleinmann et al. 1981, Temple and
Delchamps 1953):
Fe
2+
+ 0.25O
2
+ H
+
÷
0.5Fe
3+
+ 0.5H
2
O
(D)
FeS
2
+ 14Fe
3+
+ 8H
2
O
÷
15Fe
2+
+ 2SO
4
2-
+ 16H
+
(E)
As acidification proceeds and the pH in the immediate vicinity of the pyrite falls to less
than 3, the increased solubility of iron, and the decreased rate of Fe(OH)
3
precipitation result in
increased Fe
3+
activity (Silverman,1967). This is significant because as Fe
3+
aggressively attacks
pyrite, it is reduced to Fe
2+
(reaction E) for subsequent reoxidation by iron oxidizing bacteria,
such as
Acidithiobacillus ferrooxidans
(formerly called
Thiobacillus ferrooxidans
). Oxidation of
pyrite by Fe
3+
is about an order of magnitude faster than oxidation by equivalent concentrations
of dissolved oxygen, apparently because of different reaction mechanisms at the molecular level
(Luther 1987). When the pH in the immediate microenvironment of the pyrite falls to
approximately 2.5 (often corresponding to a drainage pH of 3.5 to 4.0), bacterial oxidation of
Fe
2+
and reduction of Fe
3+
by the pyrite (reactions D and E) combine to cause a dramatic
increase in acidity and iron concentrations (Kleinmann 1979).
As this solution moves through mine workings or spoils, it undergoes secondary
reactions that raise pH, decrease concentrations of iron, and increase the concentrations of other
cations. Contact with clays and other aluminosilicates releases aluminum, sodium, potassium,
and magnesium, while contact with carbonate minerals releases calcium, magnesium,
manganese, and additional iron (siderite). The various effects these reactions have on the
chemistry of the mine drainage depends on the volume of water, the amount of pyrite oxidized,
and the extent and variety of secondary chemical reactions. The secondary reactions can
produce a drainage with relatively high sulfate concentrations, butcircumneutral pH (Kleinmann
et al. 1983, Stone and Pesavento 1985), is low in metals, and fairly innocuous. Alternatively, the
mine water may have circumneutral pH, but contains elevated concentrations of dissolved iron
and manganese, and can become acidic (pH ~3) upon oxidation and precipitation of iron. In
other cases, the mine drainage is acidic; acid mine drainage often contains high concentrations of
dissolved iron, aluminum, and manganese. Both alkaline and acidic mine drainage may contain
other metals, namely zinc, nickel, and cobalt.
As contaminated mine drainage flows through receiving systems (streams, rivers, and
lakes), its toxic characteristics decrease naturally as a result of chemical and biological reactions,
and dilution with uncontaminated waters. Under the aerobic conditions found in most surface
waters, iron, aluminum, and manganese precipitate as oxides and hydroxides. Ferrous iron
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

11
oxidizes to ferric iron, which hydrolyzes and precipitates mainly as iron oxyhydroxides (e.g.,
FeOOH) or oxyhydroxylsulfates of various composition and crystallinity. These compounds
stain the bottom of many streams orange, often accumulating at sufficient depths to suffocate
benthic organisms. The rate of iron precipitation at low pH depends on the activity of the same
iron-oxidizing bacteria that catalyze pyrite oxidation (e.g.,
A. ferrooxidans
); the abiotic rate
increases a hundredfold for every unit increase in pH, and is also dependent on the amount of
oxygen dissolved in the water. (See reaction C.)
Aluminum generally hydrolyzes and precipitates as Al(OH)
3,
which is a white particulate.
Other aluminum compounds with silica and sulfate can also form, depending on the
environmental conditions. Oxidation is not required, and apparently bacterial activity is not a
factor. Precipitation of aluminum requires a pH above 4, and is generally observed at a pH of
4.5 or above. Aluminum solids will become soluble, as Al(OH)
4
-
, and at pH levels over 8.5.
This can occur in conventional chemical treatment systems that must increase pH to these higher
levels to remove manganese.
Manganese oxidizes and hydrolyzes to MnOOH or MnO
2
, and precipitates as a black
particulate. Ubiquitous manganese-oxidizing bacteria can influence the rate of removal, since
like iron, oxidation generally precedes precipitation. More important however, is that significant
oxidation and precipitation of manganese requires a pH greater than 6, and generally only occurs
in passive systems after virtually all of the iron has already precipitated. As a result, manganese
removal, if necessary, significantly increases the land area required for passive treatment.
Manganese precipitation is auto-catalytic; once precipitates form, their presence increases the
rate of manganese removal. In conventional chemical treatment systems, the pH is often raised
above 9 or 10 to remove manganese to desired levels.
Chemical Characteristics of Mine Drainage
Acidity
Acidity is a measurement of the base neutralization capacity of a volume of water. Four
types of acidity exist: organic acidity associated with dissolved organic compounds, carbon
dioxide acidity associated with dissolved carbon dioxide and carbonic acid; proton acidity
associated with pH (a measure of free H
%
ions); and mineral acidity associated with dissolved
metals (Hem 1985). Mine waters generally have very little dissolved organic carbon, so organic
acidity is very low. The amount of dissolved carbon dioxide in mine drainage varies with
geologic and environmental conditions and usually only contributes significantly to acidity at pH
levels > 5. In addition, carbon dioxide acidity can be thought of as temporary, because CO
2
-rich
waters will degas upon exposure to the atmosphere. The majority of acidity in coal mine
drainage arises from free protons (manifested in low pH) and the mineral acidity arising from
dissolved iron, aluminum, and manganese. These metals are considered acidic because they can
undergo hydrolysis reactions that produce H
%
.
Fe
2+
+ 0.25O
2
+ 1.5H
2
O
÷
FeOOH + 2H
+
(F)
Fe
3+
+ 2H
2
O
÷
FeOOH + 3H
+
(G)
Al
3+
+ 3H
2
O
÷
Al(OH)
3
+ 3H
+
(H)

12
Mn
2+
+ 0.25O
2
+ 1.5H
2
O
÷
MnOOH + 2H
+
(I)
These reactions can be used to calculate an estimate of the total acidity of a mine water
sample, and to partition the acidity into its various components. The expected acidity of a mine
water sample is calculated from its pH and the sum of the milliequivalents of the dissolved acidic
metals. For most coal mine drainages, the acidity is calculated as follows,
Acid
calc
= 50(2Fe
2+
/56 + 3Fe
3+
/56 + 3Al/27 + 2Mn/55 + 1000(10
-pH
))
(1)
where all metal concentrations are in mg/L, and 50 is the equivalent weight of CaCO
3
, and thus
transforming mg/L of acidity into mg/L as CaCO
3
equivalent. Simplifying the equation shows
the conversion factors to be applied to each dissolved metal and hydrogen ion concentration
(pH):
Acid
calc
= 1.79Fe
2+
+ 2.68Fe
3+
+ 5.56Al + 1.82Mn + 50,000(10
-pH
)
(2)
Equation 2 accurately characterizes mineral and proton acidity for most samples of actual
acid mine drainage. It must be emphasized that only dissolved metals add to acidity, not those
already precipitated. Figure 1 shows a very good correlation (R
2
= 0.9943 and slope = 1.026)
between measured and calculated acidity for mine drainage samples collected at over 150
different sites.
0
2000
4000
6000
8000
10000
0
2000
4000
6000
8000
10000
Measured Acidity, mg/L (CaCO
3
)
Calculated Acidity, mg/L (CaCO
3
)
Figure 1. Calculated Versus Measured Acidity for Over 150 Coal Mine Discharges
Equation 2 can be used to partition total acidity into its individual constituents. When the
total acidity of contaminated coal mine drainages is partitioned in this manner, the importance of
mineral acidity becomes apparent. Table 2 shows a breakdown of the acidic components of

13
three mine drainages. At each site, the acidity arising from protons (pH) was never the largest
contributor to total acidity. Only when pH is less than 3.5 does it contribute significantly to
acidity. (See Table 3.)
Table 2. Contributions of Metal Concentration and pH to Acidity for Selected Mine Discharges
Howe Bridge
Jennings
Oven Run E
Elklick
Value
A.C.
%
Value
A.C.
%
Value
A.C.
%
Value
A.C.
%
pH
5.38
<1
<1
3.35
22
6
2.74
91
34
5.79
<1
<1
Fe
2+
225
402
88
60
107
31
<1
<1
<1
55
98
93
Fe
3+
<1
<1
<1
4
11
3
20
54
20
<1
<1
<1
Al
3+
<1
<1
<1
30
167
48
18
100
37
<1
<1
<1
Mn
2+
29
53
12
22
40
12
13
24
9
4
7
7
pH in standard units. Fe
2+
, Fe
3+
, Al
3+
and Mn
2+
concentrations in mg/L. A.C. is acidity contribution in mg/L as
CaCO
3
.
Table 3. Proton Acidity Contributions at Various pH Values
pH
Acidity Equivalent
(mg/L as CaCO
3
)
6.0
0.05
5.0
0.5
4.5
1.6
4.0
5
3.5
16
3.0
50
2.5
158
2.0
500
Alkalinity
When mine water pH is greater than 4.5, it has acid neutralizing capacity and is said to
contain alkalinity. Alkalinity can result from hydroxyl ion (OH
&
), carbonate, silicate, borate,
organic ligands, phosphate, and ammonia (Hem 1985). The principal source of alkalinity in
mine water is dissolved carbonate, which can exist in bicarbonate (HCO
3
&
) or carbonate (CO
3
2&
)
form. Both can neutralize proton acidity (reactions J and K). In the pH range of most alkaline
mine waters, bicarbonate is the principal source of alkalinity (Wieder and Lang 1982, Stone and
Pesavento 1985).
2H
+
+ CO
3
2-
÷
H
2
O + CO
2
(J)
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

14
H
+
+ HCO
3
-
÷
H
2
O + CO
2
(K)
The presence of bicarbonate alkalinity in mine waters with elevated levels of metals is
not unusual, particularly in anoxic waters. Table 4 shows the chemical composition of six mine
waters in northern Appalachia that contain alkalinity, and are also contaminated with ferrous iron
and manganese. None are contaminated with significant levels of dissolved ferric iron or
aluminum because the solubilities of these metal hydroxides are low in mine waters with pH
greater than 5.0 (Hem 1985, Stumm and Morgan 1981).
Table 4. Chemical Composition of Untreated Mine Waters Containing Alkalinity
pH
Alkalinity
Fe
2+
Fe
3+
Al
3+
Mn
2+
Net Acidity
Calculated
Acidity
Penn
Allegh
6.64
470
76
<1
<1
2
-358
-330
Brinkerton
6.04
168
50
<1
<1
1
-101
-77
Scrubgrass
6.00
165
64
<1
<1
<1
-61
-50
Elklick
5.79
42
54
<1
<1
4
62
62
Howe
Bridge
5.38
35
225
<1
<1
37
395
435
Morrison
5.15
23
229
<1
<1
47
373
472
Alkalinity and pH were determined in the field. Metals and net acidity were analyzed in the lab. Calculated
acidity was calculated using Equation 2 subtracting field alkalinity. pH in standard units, alkalinity and
acidity in mg/L as CaCO
3
, metal concentrations in mg/L. Negative values of acidity indicate net alkalinity.
Alkalinity and acidity are not mutually exclusive terms. All of the mine waters shown in
Table 4 contain both acidity and alkalinity. When water contains both mineral acidity and
alkalinity, a comparison of the two measurements results in a determination as to whether the
water is net alkaline (alkalinity > acidity) or net acidic (acidity > alkalinity). Net alkaline water
contains enough alkalinity to neutralize the mineral acidity represented by dissolved ferrous iron
and manganese. As these metals oxidize and hydrolyze, the produced proton acidity is rapidly
neutralized by bicarbonate. For waters contaminated with Fe
2%
, the net reaction for the
oxidation, hydrolysis and neutralization reactions is:
Fe
2+
+ 0.25O
2
+ 2HCO
3
-
÷
FeOOH + 0.5H
2
O + 2CO
2
(L)
Reaction L indicates that net alkaline waters contain at least 1.8 mg/L alkalinity for each
1.0 mg/L of dissolved Fe
2+
. Waters that contain a lesser ratio are net acidic; the oxidation and
hydrolysis of the total dissolved iron content results in a net release of protons and a decrease in
the pH. For waters containing dissolved Fe
2+
, accurate determination of alkalinity must be
performed in the field, immediately upon the collection of water samples. Laboratory
determinations may lead to incorrect conclusions, due to reaction L occurring in the sample
bottle, thus decreasing measurable alkalinity concentrations.

15
Interpretation of Laboratory Analyses
There has been, and continues to be some confusion interpreting the results of net
alkaline or net acidic laboratory analyses (Kirby 2002). Selection of the most effective passive
treatment system design depends on whether the water is net acidic or net alkaline.
Interpretation confusion arises from the way laboratories report the acidity and alkalinity values.
They report acidity and alkalinity in mg/L as CaCO
3.
The analytical procedure in
Standard
Methods
(APHA 1998), however, actually measures net acidity for the acidity method, and gross
alkalinity for the alkalinity method, and offers no guidance for reporting acidity and alkalinity as
a net or gross value, respectively. The 20
th
edition of
Standard Methods
instructs the lab to
report “the acidity to pH ___ = ___ mg as CaCO
3
/L” and “the alkalinity to pH ___ = ___ mg as
CaCO
3
/L.” The wording was even less clear in previous editions (APHA 16th edition 1985),
which stated “if a negative value is obtained, determine the alkalinity according to [the chapter
on alkalinity].” It does not instruct the lab what to do with this negative number. Many labs
reported that acidity was zero or left a blank space for the acidity value. Some labs list the
acidity value as negative. The labs performed the alkalinity titration and recorded the value
obtained as alkalinity. The problem lies with the individual interpretation of these results. For
example, the table below shows two different water qualities:
Water
pH
Potential Acidity as Fe
2+
and/or Mn
2+
(mg/L as
CaCO
3
)
Alkalinity
(mg/L as CaCO
3
)
True Net Acidity
*
(mg/L as CaCO
3
)
A
6.2
100
105
-5
B
6.1
150
100
50
* Negative numbers denote net alkalinity.
If a laboratory analyzed this water and if they received a negative number for acidity (for water
A), they reported it as zero, and their lab sheet would look like this:
Water
pH
Acidity
(mg/L as CaCO
3
)
Alkalinity
(mg/L as CaCO
3
)
A
6.2
0
105
B
6.1
50
100
For water A, most labs would interpret the results as having a net alkalinity of 105 mg/L (as
CaCO
3
), when, in fact, the water is only barely net alkaline (5 mg/L). If the lab reported acidity
simply as negative, there was still confusion about whether the water was net alkaline or net
acidic. Most would interpret water B, which is truly net acidic (50 mg/L as CaCO
3
), as being net
alkaline (50 mg/L as CaCO
3
).
For water B, the important concept to remember is that the value obtained in the acidity
titration is a
net value
. Since the lab reported acidity as 50 mg/L, this indicates that the water is
net acidic (of 50 mg/L). For water A, if the laboratory reported the negative number that they
actually received for acidity, it would eliminate the confusion. In 1998, authors of
Standard
Methods
recognized this fact, and clarified their instruction in the 20
th
edition, instructing the
reader: “if a negative value is obtained, report the value as negative. The absolute value of this

16
negative value should be equivalent to the net alkalinity.” It is still not clearly stated to report
either positive or negative numbers as
net acidity
.
To complicate matters even further, some laboratories realized that if the water contained
alkalinity, that this was consumed in the acidity titration and functioned to lower the value
obtained in the titration. They then added the alkalinity value to the value obtained in the acidity
titration and entered that for the acidity value. In essence they were listing the true gross values
for acidity and alkalinity. For water A, they would list acidity as 100 mg/L asCaCO
3
and
alkalinity as 105 mg/L as CaCO
3
. For water B, they would list acidity as 150 mg/L and
alkalinity as 100 mg/L as CaCO
3
. The simple subtraction of these two values would result in the
correct interpretation, if the laboratory realized that gross values were being reported. If the
laboratory knew that the value for acidity in
Standards Methods
was typically a net value, they
would believe that both water samples were net acidic.
To ensure the correct interpretation of the values that laboratories provide, the laboratory
must be contacted to determine what values they reported. Laboratories should report the value
of any negative numbers they obtain in the acidity titration. If they follow the procedures
outlined in
Standard Methods
, the value for acidity is the
net value
with negative numbers
indicating
net
alkalinity, and the alkalinity value is the
gross
alkalinity.
As stated earlier, to ensure accurate measurement of alkalinity, the analysis should be
performed in the field. If ferrous iron is in the sample, oxidation and subsequent hydrolysis can
significantly lower alkalinity concentrations.
Concentrations of other constituents in coal mine drainage vary, depending on geologic
and environmental conditions. Table 5 lists the mean, median, and ranges of several chemical
parameters associated with 156 different coal mine drainage discharges.

17
Table 5. Water Quality from 156 Coal Mine Discharges
Parameter
Times
Reported
Mean
Median
Minimum
Maximum
Flow
54
601
71.5
5.50
15600
pH
156
4.03
3.37
2.18
7.80
Conductivity
64
2500
2000
320
8140
Alk, field
95
32.0
0
0
470
Acidity
151
909
315
-358
9220
Sulfate
156
1750
1220
67.5
1100
Aluminum
156
68.4
15.6
0
930
Antimony
120
0.006
0
0
0.200
Arsenic
142
0.040
0
0
2.95
Barium
135
0.010
0
0
0.200
Beryllium
140
0.017
0
0
0.270
Cadmium
146
0.006
0
0
0.200
Calcium
156
168
159
6.90
483
Chloride
50
64.1
8.15
0
849
Chromium
155
0.063
0
0
7.18
Cobalt
137
0.646
0.240
0
6.00
Copper
155
0.103
0
0
2.49
Iron, Ferric
140
96.7
4.35
0
2420
Iron, Ferrous
137
150
65.0
0
1610
Iron, Total
156
221
71.9
0
2440
Lead
144
0.009
0
0
0.433
Magnesium
156
104
80.0
2.75
638
Manganese
156
20.6
6.76
0
164
Nickel
150
0.962
0.400
0
10.0
Potassium
143
4.05
3.12
0.04
32.0
Selenium
136
0.013
0
0
0.369
Silver
22
0.0005
0
0
0.010
Sodium
156
45.3
8.70
0.33
712
Vanadium
20
0.115
0.050
0
0.660
Zinc
153
2.64
0.700
0
48.0
All concentrations in mg/L, flow in L/min, pH in standard units, acidity and alkalinity in mg/L as CaCO
3
,
negative acidity indicates net alkalinity.
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

 
18
Passive Treatment Processes
Limestone Dissolution
A major source of bicarbonate in many anoxic environments is the dissolution of
carbonate minerals, such as calcite.
CaCO
3
+ H
+
÷
Ca
2+
+ HCO
3
-
(M)
Carbonate dissolution can result in higher concentrations of bicarbonate in anoxic mine
water environments than oxic environments, for two reasons. First, the absence of ferric
hydroxide in most anoxic environments limits the formation of FeOOH coatings that may armor
carbonate surfaces and inhibit further carbonate dissolution in oxic environments (U.S. EPA
1983). Second, the solubility of carbonate compounds are directly affected by the partial
pressure of dissolved CO
2
(Stumm and Morgan 1996, Hem 1985, Butler 1991). Anoxic mine
water environments commonly contain high CO
2
partial pressures due to the decomposition of
organic matter and the neutralization of proton acidity. Table 6 shows how the partial pressure
of carbon dioxide affects the maximum level of potential alkalinity. At atmospheric levels
(~0.0003), only about 60 mg/L of alkalinity (as CaCO
3
) is capable of being dissolved. However,
CO
2
levels can be much higher within soil and mine spoil than in the atmosphere, from 0.01 to
0.10. At these CO
2
levels, alkalinity concentrations of 220 to 475 mg/L are possible.
Table 6. Equilibrium Concentrations of Alkalinity at Various P
CO2
Levels
P
CO2
(atm)
Alkalinity (mg/L as
CaCO
2
)
0.0003 (~atmospheric)
60
0.01
220
0.05
360
0.10
475
0.20
610
0.50
850
1.00
1085
The observation that limestone dissolution by mine water is enhanced under closed
conditions has resulted in the construction of anoxic limestone treatment systems. The first
demonstration of this technology was by Turner and McCoy (1990), who showed that when
anoxic acidic mine water was directed through a plastic-covered buried bed of limestone, it was
discharged in an alkaline condition.
Since Turner and McCoy described their findings in 1990, dozens of additional limestone
treatment systems have been constructed (e.g., Brodie et al. 1991, Skousen and Faulkner 1992).
These passive mine water pretreatment systems have become known as anoxic limestone drains
(ALDs). In an ALD, mine water is made to flow through a bed of limestone gravel that has been
buried to limit contact with atmospheric oxygen. The burial containment also traps CO
2
within
the treatment system, allowing the development of high CO
2
partial pressures, which in turn
allows additional limestone dissolution (Nairn et al. 1992).
Under oxic conditions, limestone dissolution may be enhanced by the active generation
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

19
of acidity (hydrogen ion) by ferric iron and aluminum hydrolysis. However, this process may
not lend itself to sustainable treatment systems, due to problems of armoring and clogging.
Sulfate Reduction
When mine water flows through an anaerobic environment that contains an organic
substrate, the water chemistry can be affected by bacterial sulfate reduction. In this process,
bacteria oxidize organic compounds using sulfate as the terminal electron acceptor and release
hydrogen sulfide and bicarbonate,
2CH
2
O + SO
4
2-
÷
H
2
S + 2HCO
3
-
(N)
where CH
2
O is used to represent organic matter. Bacterial sulfate reduction is limited to certain
environmental conditions (Postgate 1984). Bacteria requirea the presence of sulfate, suitable
concentrations of low-molecular weight carbon compounds, and the absence of oxidizing agents,
such as oxygen, Fe
3%
and Mn
4%
. These conditions are commonly satisfied in treatment systems
that receive coal mine drainage and are constructed with an organic substrate, such as a compost
material. High concentrations of sulfate (> 500 mg/L) are characteristic of contaminated coal
mine drainage. The oxygen demand of organic substrates causes the development of anoxic
conditions and an absence of oxidized forms of iron or manganese. The low molecular-weight
compounds that sulfate-reducing bacteria utilize (lactate, acetate) are common end-products of
microbial fermentation processes in anoxic environments. These sulfate reducing and
fermentative bacteria are more active above pH ~5, however, they can be very active in
drainages with lower pH levels, due to the presence of near-neutral pH microenvironments.
These microenvironments allow the sulfate reducing bacteria (SRB) to become established, and
because they generate alkalinity, these microenvironments become larger.
Metal Removal Processes
Coal mining can promote pyrite oxidation and result in drainage with high concentrations
of iron, manganese, and aluminum, as well as SO
4
, Ca, Mg, K and Na. The concentrations of
iron, manganese, and aluminum are generally very low in natural waters (< 1 mg/L) because of
chemical and biological processes that cause their precipitation in surface water environments.
The same chemical and biological processes remove iron, manganese, and aluminum from
contaminated coal mine drainage, but the metal loadings from abandoned mine sites are often so
high that the deleterious effects of these elements persist long enough to result in the pollution of
receiving waters.
Passive treatment systems function by retaining contaminated mine water long enough to
decrease contaminant concentrations to acceptable levels. The chemical and biological
processes that remove contaminants vary among metals and are affected by the mine water pH
and oxidation-reduction potential (Eh). Efficient passive treatment systems create conditions
that promote the processes that most rapidly remove target contaminants. Thus, the design of
passive treatment systems must be based on a solid understanding of mine drainage chemistry
and how different passive technologies affect this chemistry.
Reduction
Chemical and microbial processes in anaerobic environments differ from those observed
in aerobic environments. Because oxygen is absent, Fe
2%
and Mn
2%
do not oxidize, and
oxyhydroxide precipitates do not form. Hydroxides of the reduced iron and manganese ions,
Fe(OH)
2
and Mn(OH)
2
, do not form because of their high solubility under acidic or

20
circumneutral conditions. In passive treatment systems where mine water flows through
anaerobic environments, its chemistry is affected by chemical and biological processes that
generate bicarbonate and hydrogen sulfide.
Bacterial sulfate reduction not only improves water quality by the addition of bicarbonate
alkalinity, it can also lower the concentrations of dissolved metals, M
2+
, (e.g., Fe
2+
, Mn
2+
, Zn
2+
,
Ni
2+
, Cu
2+
, Cd
2+
, Pb
2+
) by precipitating them as metal sulfide solids.
M
2+
+ H
2
S + HCO
3
-
÷
MS + 2H
2
O + 2CO
2
(O)
For iron, the formation of iron monosulfide and even pyrite is possible:
Fe
2+
+ H
2
S + S
0
÷
FeS
2
+ 2H
+
(P)
The removal of dissolved metals as sulfide compounds depends on pH, the solubility
product of the specific metal sulfide, and the concentrations of the reactants. The solubilities of
various metal sulfides are shown in Table 7 (Ehrlich 1981). Laboratory studies have verified
that metal removal from mine water subjected to inflows of hydrogen sulfide occurs on an order
consistent with the solubility products shown in this table (Hammack et al. 1993). The first
metal sulfide that forms is CuS, followed by PbS, ZnS, and CdS. FeS is one of the last metal
sulfides to form. MnS is the most soluble metal sulfide shown, and is not expected to form.
Because of the low solubility of some of these metal sulfides relative to their solubilities as
oxides or hydroxides, sulfate reduction can be an important process to lower some metal
concentrations to acceptable levels, particularly for treating metal mine drainage.
Table 7. Solubility Products of Selected Metal Sulfides
Metal Sulfide
Solubility Product
CuS
4.0 x 10
-38
PbS
1.0 x 10
-29
ZnS
4.5 x 10
-24
CdS
1.4 x 10
-23
NiS
3.0 x 10
-21
FeS
1.0 x 10
-19
MnS
5.6 x 10
-16
For coal mine drainage, where metal contamination is generally limited to iron ,
manganese , and aluminum, the hydrogen sulfide produced by bacterial sulfate reduction
primarily affects dissolved iron concentrations. Aluminum does not form any sulfide
compounds in wetland environments, and the relatively high solubility of MnS makes its
formation unlikely.
The precipitation of metal sulfides in an organic substrate improves water quality by
decreasing mineral acidity without causing a parallel increase in proton acidity. Proton-releasing
aspects of the H
2
S dissociation process (H
2
S
÷
2H
+
%
S
2&
) are neutralized by an equal release of
bicarbonate during sulfate reduction. An organic substrate in which 100 percent of the H
2
S
(produced by sulfate reduction precipitated as FeS) would have no effect on the mine water pH
or alkalinity (although acidity would decrease). In fact, however, the chemistry of pore water in
wetlands constructed with an organic substrate characteristically has pH 6 to 8 and is highly
alkaline (Hedin et al. 1988, McIntire and Edenborn 1990). These alkaline conditions result, in
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

21
part, from reactions involving hydrogen sulfide that result in the net generation of bicarbonate.
Hydrogen sulfide is a very reactive compound that can undergo a variety of reactions in a
constructed wetland. In most wetlands (constructed and natural), surface waters are aerobic
while the underlying pore waters in contact with organic substrate are anaerobic. When sulfidic
pore waters diffuse from the organic substrate into zones that contain dissolved ferric iron,
dissolved oxygen, or precipitated ironand manganese oxides, the hydrogen sulfide can be
oxidized. These reactions affect the mineral acidity and the alkalinity in various manners.
Metal Oxidation and Hydrolysis
Oxidation and hydrolysis reactions commonly cause concentrations of Fe
2%
, Fe
3%
,
manganese , and Al to decrease when mine water flows through an aerobic environment.
Whether these reactions occur quickly enough to lower metal concentrations to an acceptable
level depends on the availability of oxygen for oxidation reactions, the pH of the water, the
activity of microbial and/or other catalysts and inhibitors, and the retention time of water in the
treatment system. The pH is an especially important parameter because it influences both the
solubility of metal hydroxide precipitates and the kinetics of the oxidation and hydrolysis
processes. The relationship between pH and metal-removal processes in passive treatment
systems is complex because it differs among metals and also between abiotic and biotic
processes.
The stoichiometries of the major metal removing reactions in passive treatment systems
are:
Fe
3+
+ 3H
2
O
÷
Fe(OH)
3
+ 3H
+
(Q)
Al
3+
+ 3H
2
O
÷
Al(OH)
3
+ 3H
+
(R)
Fe
2+
+ 0.25O
2
+ 2.5H
2
O
÷
Fe(OH)
3
+ 2H
+
(S)
Mn
2+
+ 0.25O
2
+ 1.5H
2
O
÷
MnOOH + 2H
+
(T)
The first two (Q and R) are simple hydrolysis reactions, which require only the presence
of water (and enough alkalinity to neutralize the H
+
produced). The last two reactions (S and T)
require the presence of oxygen to oxidize the metal prior to hydrolysis. All of the reactions
produce acidity, which was discussed previously. The goal of passive treatment systems is to
drive these reactions to completion and collect the resulting solids before the water enters a
receiving stream.
Iron
The most common contaminant of coal mine drainage is ferrous iron. In oxidizing
environments common to most surface waters, ferrous iron is oxidized to ferric iron (reaction S).
Ferrous iron oxidation occurs both abiotically and as a result of bacterial activity. The
stoichiometry of the reaction is the same for both oxidation processes.
From the stoichiometry, it can be seen that one mole of oxygen can oxidize 4 moles of
Fe
2+
. This corresponds to 7.0 mg of Fe
2+
oxidized per mg of O
2
. The solubility of oxygen in
water depends on both pressure and temperature. It can be as high as 13 mg/L (1 atm., < 5
o
C)
but under field conditions, a maximum practical DO level of 8 mg/L is a better estimate. At this
oxygen concentration, only about 55 mg/L of Fe
2+
can be oxidized without providing for
additional oxygenation of the water.
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

22
Because the net result of the oxidation and hydrolysis process is the production of
protons, the process can decrease pH. Thus, natural or constructed wetlands receiving
circumneutral net acidic water commonly decrease both iron concentrations and pH. An
example of this phenomenon is shown in Figure 2a. As water flowed through the constructed
wetland, dissolved iron concentrations decreased from 95 mg/L to 15 mg/L, and pH decreased
from 5.5 to 3.2. Figure 2b shows iron concentrations and pH in a wetland that received mine
water with a net alkalinity. Despite the removal of 60 mg/L Fe
2%
and the production of enough
protons to theoretically lower the pH to 2.7, the pH did not decrease because bicarbonate
alkalinity neutralized the proton acidity.
0
20
40
60
80
100
120
Influent
Cell 1
Cell 2
Cell 4
Cell 6
Cell 8
Effluent
Sampling Station
Fe, mg/L
3
4
5
6
7
pH
Iron (Fe)
pH
A
Figure 2a. Concentration of Iron and Field pH at the Emlenton Constructed Wetlands, which
Receives Net Acidic Water
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

23
0
30
60
90
120
150
Influent
Cell 1
Cell 3
Effluent
Sampling Station
Fe, mg/L
3
4
5
6
7
pH
Iron (Fe)
pH
B
Figure 2b. Concentration of Iron and Field pH at Cedar Grove Constructed Wetlands which
Receives Net Alkaline Water
As ferrous iron is converted to ferric iron, it is subject to hydrolysis reactions that can
precipitate it as a hydroxide (reaction Q). The hydrolysis reaction occurs abiotically; catalysis of
the reaction by microorganisms has not been demonstrated. Under equilibrium conditions, the
solubility of the ferric hydroxide solid is very low and little dissolved ferric iron (< 1 mg/L) is
predicted to exist, unless the pH of the water is less than 2.5. However, the rate of the hydrolysis
reaction is also pH dependent, and significant Fe
3%
can be found in mine water with a pH above
2.5. Figure 3 shows ferric iron concentrations for over 150 coal mine discharges. Significant
dissolved ferric iron is not generally present, unless the pH is less than 4. The highest
concentrations of ferric iron occurred when the pH is less than 3.
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

24
0
200
400
600
800
1000
012345678
pH, s.u.
Fe
3+
, mg/L
Figure 3. Dissolved Ferric Iron Concentration Versus pH in Coal Mine Discharge
Reaction rates are less well understood than stoichiometry and solubility. For the two
rates of importance to iron oxidation and hydrolysis, the oxidation rate is usually the slowest.
Metal hydrolysis reactions are much faster than oxidation reactions, and can be ignored at all but
the lowest pH values. At low pH values, Singer and Stumm (1970a) suggested a fourth-order
relationship with pH, which indicated that ferric iron hydrolysis processes shift from a very rapid
rate at pH greater than 3, to a very slow rate at pH less than 2.5.
Metal oxidation can be influenced by a number of factors that can accelerate (catalyze) or
retard (inhibit) the rate. In the simplest case where possible biological, catalytic, and inhibitory
mechanisms are ignored, the rate of ferrous iron oxidation can be described by equation 3.
d
[Fe(II)] /
d
t = -
k
[Fe(II)] [O
2
] [H
+
]
-2
(3)
As can be seen from the inverse second order dependence on [H
+
], the pH of the mine
water dramatically affects the kinetics of the oxidation process (Singer and Stumm 1970a, Singer
and Stumm 1970b). When oxygen is not limiting, the rate of homogenous abiotic iron oxidation
slows a hundredfold for every unit decrease in pH. At pH values greater than 8, the process is
fast (rates are measured in seconds), while at pH values less than 5, the process is slow (rates are
measured in days). The presence of bicarbonate alkalinity buffers mine water at a pH of 6 to 7, a
range at which homogeneous abiotic iron oxidation processes should dominate.
The effect that pH can have on the mechanism of iron oxidation is shown by the data in
Figure 4. Samples were collected from two mine drainages that were both contaminated with
ferrous iron but had different pH and alkalinity values. The samples were returned to the
laboratory and exposed to aerobic conditions. For the circumneutral waters, oxidation of ferrous
iron occurred at a rate of 18 mg L
&1
hr
&1
, while the rate for the raw acidic samples was only 1.4
mg L
&1
hr
&1
. In order to evaluate the significance of bacterial processes in iron oxidation, splits
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

25
of both samples were filter-sterilized (0.22 μm membrane filter) before the experiment was
begun. Removal of bacteria had no effect on the oxidation of ferrous iron for the circumneutral
water, but completely inhibited ferrous iron oxidation for the acidic water.
0
10
20
30
40
0
10
20
30
40
50
Time, h
Fe
2+
, mg/L
Filtered
Unfiltered
A
0
10
20
30
40
50
60
70
80
90
100
02
4
68
10
12
Time, h
Fe
2+
, mg/L
Filtered
Unfiltered
B
Figure 4. Removal of Ferrous Iron from Acidic and Alkaline Mine Waters in a Laboratory
Experiment
Untreated mine drainage was collected from the (A) acidic Latrobe site and (B) alkaline Cedar Grove site.
Splits of each sample were filter sterilized (0.22 micrometer filter).
In contrast to the uncatalyzed chemical rate, bacterial oxidation of ferrous iron peaks at
pH values between 2 and 3, while very little activity occurs at pH values greater than 5 (Nealson
1983a). Waters containing no alkalinity have a pH less than 4.5, and the removal of iron under
oxidizing conditions occurs primarily by bacterial oxidation, accompanied by hydrolysis and

26
precipitation (Kirby et al. 1999). Equation 4 gives the rate of loss of ferrous iron via the
microbial mechanism. Note that in the microbial case, the rate is directly proportional to the
hydrogen ion concentration and contains a term to account for the number of bacteria present,
[
Bact
].
d
[Fe(II)] /
d
t = -
k
bio
[
Bact
] [H
+
] [Fe(II)] [O
2
]
(4)
Solid surfaces, in particular the hydrous ferric oxide surface itself, can catalyze the
oxidation of ferrous iron that adsorbs to its surface. There have been recent attempts to exploit
this mechanism in a recirculated iron oxide reactor in an effort to increase iron removal rates
over those obtained in conventional ponds and wetlands (Dietz and Dempsey 2001). The rate of
heterogeneous catalysis is given in equation 5. In this case, the rate expression contains a term
to account for the amount of oxide present, [Fe(III)], and is proportional to the inverse of the
hydrogen ion concentration.
d
[Fe(II)] /
d
t = -
k
hetero
[Fe(III)][Fe(II)] [O
2
] [H
+
]
-1
(5)
Kirby and Elder Brady (1998) list several other factors that have been reported to affect
Fe(II) oxidation rate in natural waters: Cu(II), Co(II), anions that form complexes with Fe(III),
organic acids, Na
+
, presence of ferric hydroxide solids, ionic strength, sulfate, light intensity,
colloidal silica and aluminum oxide, and bentonite clay are all listed, together with literature
citations to the original work. It is likely that except for the presence of ferric hydroxide solids,
these other factors are not significant in passive treatment systems. Light intensity, which can
influence iron photoreduction (McKnight et al. 2001) may be significant, but conflicting results
(Wieder 1994) in the literature demonstrate that further study is needed.
Temperature is known to affect the rate in a number of ways. Because the dissociation
constant for water, K
w
, depends on temperature, this change must be taken into account during
the conversion from pH, the measured parameter, to hydroxide ion concentration, the rate
dependent variable. Alternatively, the rate constant can be determined using Equation 3 (often in
the integrated form) with the realization that it contains K
w
. However, because K
w
changes with
temperature, the former conversion of pH to [OH
-
] is preferred when rate constants determined at
different temperatures are to be used to determine an activation energy. The temperature affects
the Henry’s law constant, used to calculate the molar concentration of oxygen from its partial
pressure. However, in many studies, the dissolved oxygen concentration is measured directly.
Rate constants increase with increasing temperature. To quantitatively model the iron
loss in a system where the temperature is not constant, the temperature dependence is usually
expressed as the exponential given in equation 6 (Kirby et al. 1999, Watzlaf et al. 2001).
-E
act
k = A e
RT
(6)
Manganese
Manganese undergoes oxidation and hydrolysis reactions that result in the precipitation
of manganese oxyhydroxides. The specific mechanism(s) of Mn
2%
precipitation from aerobic
mine water in the absence of chemical additions are uncertain. Mn
2%
may be oxidized to either a
%3
or a
%4
valence, either one of which rapidly precipitates. (See reaction T.) If MnOOH
precipitates over time it likely oxidizes to the more stable MnO
2
. In alkaline environments,
Mn
2%
can precipitate as a carbonate (reaction U), which may be oxidized by oxygen to MnO
2
via

27
reaction V (Diehl and Stumm 1984).
Mn
2+
+ HCO
3
-
÷
MnCO
3
+ H
+
(U)
MnCO
3
+ 0.5O
2
÷
MnO
2
+ CO
2
(V)
Regardless of the mechanism by which Mn
2%
is oxidized to Mn
4%
, the removal of one
mole of Mn
2%
from solution results in the release of two moles of H
%
, or an equivalent decrease
in alkalinity (HCO
3
&
).
The kinetics of Mn
2%
oxidation reactions are strongly affected by pH. Abiotic oxidation
reactions are very slow at pH less than 8 (Stumm and Morgan 1981). Microorganisms can
catalyze Mn
2%
oxidation, but their activity is limited to aerobic waters with pH greater than 6
(Nealson 1983b).
Although the hydrolysis of manganese produces protons, the precipitation of MnOOH
does not result in large declines in pH, which can happen when FeOOH precipitates. This
difference between manganese and iron chemistry is due to the fact that no natural mechanism
exists to rapidly oxidize Mn
2%
under acidic conditions. If pH falls below 6, Mn
2%
oxidation
virtually ceases, the proton-producing hydrolysis reaction ceases, and the pH stabilizes.
The oxidation and precipitation of Mn
2%
from solution is accelerated by the presence of
MnO
2
and FeOOH (Stumm and Morgan 1981, Davies and Morgan 1989). Both solids
reportedly act as adsorption surfaces for Mn
2%
and catalyze the oxidation mechanism. While
additions of FeOOH to water containing manganese might accelerate manganese oxidation, the
direct precipitation of FeOOH from mine water that contains Fe
2%
does not generally stimulate
manganese removal processes in passive treatment systems. Figure 5 shows that concentrations
of manganese and iron in mine water markedly decreased as it flowed through a constructed
wetland. On average, iron decreased from 150 mg/L to less than 1 mg/L, while manganese
decreased from 42 mg/L to 11 mg/L. Removal of metals occurred sequentially, not
simultaneously. Two-thirds of the decrease in iron concentration occurred between the first and
second sampling stations. The wetland substrate in this area was covered with precipitated
FeOOH, and the water was turbid with suspended FeOOH. Despite the presence of large
quantities of FeOOH, little change occurred in the concentration of manganese between the first
and second sampling station. The slight decrease in manganese that occurred was
proportionally similar to the change in Mg, suggesting that dilution was the most likely cause of
the decrease in manganese concentrations. Between stations 3 and 5, there was little iron
present in the water and little visual evidence of FeOOH sludge on the wetland substrate. Most
of the observed removal of manganese occurred in this iron-free zone.
The absence of simultaneous precipitation of dissolved iron and manganese from aerobic
alkaline waters likely results from the reduction of oxidized forms of manganese by ferrous iron
as shown in reaction W, or reaction X.
MnO
2
+ 2Fe
2+
+ 2H
2
O
÷
2FeOOH + Mn
2+
+ 2H
+
(W)
MnOOH + Fe
2+
÷
FeOOH + Mn
2+
(X)
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

28
Figure 5. Mean Concentration of Iron, Manganese and Magnesium at the Morrison Wetland as the
Mine Water Flows Linearly through the System
Figure 6 shows the results of a laboratory study that demonstrates the instability of
manganese oxides in the presence of ferrous iron. Water samples and manganese oxides were
collected from a wetland that removed iron and manganese in a sequential manner. The wetland
influent was alkaline (pH 6.2, 162 mg/L alkalinity) and contaminated with 50 mg/L iron and 32
mg/L manganese . Two flasks of mine water received MnO
2
additions, while the controls did
not receive MnO
2
. Concentrations of dissolved iron and manganese were monitored in each
flask over a 73-hour period. In all flasks, concentrations of iron decreased to less than 1 mg/L.
In the control flasks, concentrations of iron decreased to less than 3 mg/L within 43 hours. In
flasks that received MnO
2
, concentrations of iron decreased to less than 3 mg/L in only 22 hours.
No change in concentrations of manganese occurred in the control flasks. Concentrations of
manganese in the MnO
2
flasks increased by 15 mg/L during the first 22 hours and did not change
during the remaining 50 hours of the experiment. The association of accelerated precipitation of
iron with solubilization of Mn
2%
suggests that the MnO
2
oxidized Fe
2%
in a manner analogous to
reaction K.
0
25
50
75
100
125
150
175
ALD
Ditch
Pond
Wetland
Effluent
Sampling Station
Fe & Mg, mg/L
0
15
30
45
60
75
Mn, mg/L
Fe
Mg
Mn
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

29
0
10
20
30
40
50
60
0
20
40
60
80
100
Time, h
Concentration, mg/L
Mn
Fe (II)
A
0
10
20
30
40
50
60
0
20
40
60
80
100
Time, h
Concentration, mg/L
Mn
Fe (II)
B
Figure 6. Changes in the Concentrations of Ferrous Iron and Manganese in (A) the Absence of the
MnOOH and (B) the Presence of MnOOH
The data presented in Figure 5 and Figure 6 demonstrate important aspects of iron and
manganese chemistry in passive treatment systems. Iron oxidizes and precipitates from alkaline

30
mine water much more rapidly than manganese. One reason for the differences in kinetics is that
the oxidized manganese solids are not stable in the presence of Fe
2%
. Concentrations of ferrous
iron must decrease to very low levels before Mn
2%
oxidation processes can result in a stable solid
precipitate. In the absence of Fe
2%
, manganese removal is still a very slow process under
laboratory conditions. Conditions in a wetland may either accelerate manganese removal
reactions or promote mechanisms that are not simulated in simple laboratory experiments.
However, both field and laboratory investigations indicate that, under aerobic conditions, the
removal of manganese occurs at a much slower rate than the removal of iron.
Aluminum
Aluminum has only one oxidation state in aquatic systems, +3. Oxidation and reduction
processes, which complicate iron and manganese chemistry, do not directly affect concentrations
of dissolved aluminum. Instead, concentrations of aluminum in mine waters are primarily
influenced by the solubility of Al(OH)
3
(Hem 1985, Nordstrom and Ball 1986). At pH levels
between 5 and 8, Al(OH)
3
is insoluble, and concentrations of dissolved aluminum are usually
less than 1 mg/L. At pH values less than 4, Al(OH)
3
is highly soluble and concentrations much
higher than 2 mg/L are possible. The amount of aluminum found in over 150 different mine
drainage samples are show in Figure 7. No significant amounts of dissolved aluminum were
found above a pH of 4.5, consistent with the expected behavior, based on solubility. The
kinetics of hydrolysis do not appear to play a role.
0
200
400
600
800
1000
012345678
pH, s.u.
Al, mg/L
Figure 7. Dissolved Aluminum Concentration Versus pH in Coal Mine Discharges
The passage of mine water through highly oxidized or highly reduced environments has
no effect on concentrations of aluminum unless the pH also changes. In those cases where the
pH of mine water decreases (due to iron oxidation and hydrolysis), concentrations of aluminum
can increase because of the dissolution of alumino-silicate clays by the acidic water. When

 
31
acidic mine water passes through anaerobic environments, the increased pH that can result from
carbonate dissolution or microbial activity can cause the precipitation of Al(OH)
3
. In addition to
Al(OH)
3
, aluminum hydroxysulfate minerals can form when the drainage contains aluminum and
SO
4
2-
at pH levels greater than 4.3 (Nordstrom and Ball 1986).
Robbins et al. (1996) found poorly crystalline aluminite [Al
2
(SO
4
)(OH)
4
.
7H
2
O] in an
ALD in West Virginia.
Materials and Methods
Collection of Water Samples
Water samples were collected at passive treatment systems from their influent and
effluent points, and, if applicable, between unit operations within the system. Raw and acidified
(1 to 2 mL of concentrated HCl) water samples were collected in 125 to 250 mL plastic bottles at
each sampling point. Acid was added to lower the pH to below 1.0. At sites where particulates
were visible in water samples, an additional sample was collected that was filtered through a
0.2 μm membrane filter prior to acidification. Samples were refrigerated in the analytical
laboratory at 4EC
until analysis. Measurements of pH and temperature were made in the field
with a calibrated portable pH/ISE meter. Alkalinity was measured in the field using a pH meter
and an Orion Total Alkalinity Test Kit or a Hach Digital Titrator.
Analysis of Water Samples
Concentrations of iron, manganese, aluminum, calcium, magnesium, sodium, cobalt,
nickel, and zinc were determined in the acidified samples using Inductively-Coupled Argon
Plasma-Atomic Emission Spectroscopy (ICAP-AES) (Instrumentation Laboratory Plasma 100
model or TJA Polyscan 61E). The acidified samples were at times filtered through a 0.45 μm
membrane filter to prevent clogging of the small diameter tubing in the system.
Ferrous iron concentrations were determined on acidified samples by the potassium
dichromate method (Fales and Kenny 1940). Sulfate concentrations were determined by one of
three methods: (1) reaction with barium chloride after first passing the raw sample through a
cation exchange resin with Thorin used as the end-point indicator (Kleinmann et al. 1988), (2)
ion chromatography, or (3) ICP-AES. The agreement among these methods was found to be
very good (within ~2 percent).
Acidity was determined by adding H
2
O
2
, heating and titrating the solution with NaOH
(American Public Health Assoc. 1985). An auto titrator was used to determine the inflection
point in the titration curve (i.e., first derivative mode). Acidity and alkalinity are reported as
mg/L CaCO
3
equivalent.
For each set of samples for a particular site, a duplicate, standard, and spike were
analyzed for quality control purposes. The relative standard deviation for duplicates were less
than 5 percent. Recovery for the standards were within 3 percent of the original standard. Spike
recoveries were within 5 percent of the expected values.
Flow Rate Measurements
Water flow rates were determined by one of three methods. Whenever possible, flow
was determined by the time necessary to collect a known volume of water using a bucket and
stopwatch. In all cases, three to five measurements were made at each sampling location, and
the mean flow rate of these measurements was reported. Flows were also measured with

 
32
permanently installed calibrated flumes, and portable calibrated pipe weirs.
Analysis of Iron Sludge
A sample of iron sludge was collected at the Morrison II site. Sludge at this site was
selected as a representative of iron sludge precipitated under alkaline conditions. Approximately
60 cm
3
of sludge was collected with a spatula and allowed to “drip dry” for about 1 minute
before placing it in a 125-mL plastic bottle. After transport back to the laboratory, it remained
undisturbed for about 6 weeks. The supernatant liquid was then withdrawn by pipette from the
top of the sludge (approximately 15 percent of the total volume). A graduated cylinder was
filled with 8.0 mL of distilled, deionized water. Sludge was added to the water until it rose to the
10.0 mL level, thereby collecting 2.0 cm
3
of sludge. The sludge/water mixture was transferred
into a volumetric flask, and nitric and hydrochloric acid were added to totally dissolve the
sludge. Distilled, deionized water was added, resulting in a final volume of 1.0 L. A portion of
this solution was analyzed for metal content as outlined above. It was found that this sludge
contained 0.17 grams of iron per cm
3
of sludge. This is consistent with previous measurements
of sludges precipitated from alkaline waters at other sites, and can be used to calculate how fast
systems will fill with iron precipitates.
Removal of Contaminants by Passive Unit Operations
Aerobic Wetlands and Ponds
To make reliable evaluations of wetland performance, a measure should be used that
allows comparison of contaminant removal between systems that vary in size and the chemical
composition and the flow rate of mine water they receive. In the past, concentration efficiency
(CE%) was a common measure of performance (Girts et al. 1987, Weider 1989). Using iron
concentration as an example, the calculation was:
CE% = [(Fe
in
- Fe
eff
)/Fe
in
] x 100
(7)
where the subscripts “in” and “eff” represented wetland influent and effluent sampling stations,
and iron concentrations were in mg/L.
Except in carefully controlled environments, concentration efficiency is a very poor
measure of wetland performance. The efficiency calculation results in the same measure of
performance for a system that lowers iron concentrations from 300 mg/L to 100 mg/L as one that
lowers concentrations from 3 mg/L to 1 mg/L. Neither the flow rate of the drainage nor the size
of the treatment system is incorporated into the calculation. As a result, the performance of
systems are compared without accounting for differences in flow rate (which vary from < 10
L/min to > 1,000 L/min), or for differences in system size (which vary from < 0.1 ha to > 10 ha)
(Weider 1989).
A more appropriate method for measuring the performance of treatment systems
calculates contaminant removal from a loading perspective. The daily load of contaminant
received by a wetland is calculated from the product of concentration and flow rate data. For
iron, the calculation of load in grams per day is:
Fe
in
(g/d) = 1.44 x flow (L/min) x Fe
in
(mg/L)
(8)
where 1.44 is the unit conversion factor to convert minutes to days, and milligrams to grams.
The daily mass of iron removed by the wetland between two sampling stations,
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

33
Fe(g/d)
rem
, is calculated by comparing contaminant loadings at the two points.
Fe
rem
(g/d) = Fe
in
(g/d) - Fe
eff
(g/d)
(9)
An area-adjusted daily iron removal rate is then calculated by dividing the load removed
by the surface area of the treatment system lying between the sampling points.
Fe
rem
(g d
-1
m
-2
) = Fe
rem
(g/d)/SA (m
2
)
(10)
Using this area-adjusted removal rate as the measure of treatment performance, Hedin et
al. (1994a) reported typical removal rates of 10 to 20 g d
-1
m
-2
for iron, and 0.5 to 1.0 g d
-1
m
-2
for manganese.
More recently, several groups have attempted to develop models that will more
effectively estimate the performance of treatment systems, especially for iron removal. Watzlaf
et al. (2001) were able to model a system consisting of an aerobic pond, an aeration cascade, and
a wetland using only equations 3 and 6. They found that the overall performance and the
performance of certain sections of the system fell within the 10 to 20 g d
-1
m
-2
range, but that
some sections did not. The model indicated that the pH was limiting the rate of removal.
Kirby et al. (1999) used a combination of the abiotic rate expression (equation 3), the
biological rate expression (equation 4), and the temperature dependence (equation 6) to model a
set of 17 ponds. They found that the relative importance of the biotic and abiotic mechanisms
was determined mainly by pH, with the abiotic path predominating at the higher pH values.
They suggest that pH and temperature are the most important variables for determining iron
oxidation rates, and therefore, iron removal rates. Little can be done to control temperature in a
passive treatment. The Kirby et al. (1999) work suggests that increasing pH from 6.1 to 6.4, for
example, greatly enhances oxidation, whereas doubling dissolved oxygen (as long as oxygen is
sufficiently high stoichiometrically to oxidize metals), pond volume, or retention time has
considerable less impact on oxidation rates.
Dempsey et al. (2001) modeled seasonal fluctuations of two systems using a combination
of the homogeneous rate (equation 3) and the heterogeneous rate (equation 5). They found that
oxygen transfer was rate limiting in one system, and that the amount of catalytic reaction
provided by ferric hydroxides was the determining factor at the second site. While
heterogeneous catalysis apparently plays a significant role in iron oxidation, it is difficult to
increase concentrations of iron solids in a completely passive system. Such catalysis could be
quite important in semi-passive or active treatment systems.
No one has yet tried a combination of all three rate expressions to apportion the relative
importance of the three mechanisms at a given site, but such studies are undoubtedly underway.
Because the three rate expressions contain different parameters (bacteria, ferric oxides) and have
different dependencies on the pH, it should be possible, in principle, to differentiate among the
three mechanisms.
It appears that the original estimate of Hedin et al. (1994a) of 10 to 20 g d
-1
m
-2
remains a
convenient pre-construction rule-of-thumb for estimating pond and wetlands sizes. Studies
undertaken since the publication of Hedin guidelines tend to support them in the majority of
cases (Younger et al. 2002).
Anoxic Limestone Drains
All sites are located in western Pennsylvania, with the exception of the Elklick site,
which is located in northwestern Maryland. Discharges are associated with Allegheny group

34
coals (mainly the Kittanning and Clarion seams) formed during the Pennsylvanian period.
Site Descriptions
Howe Bridge 1
- Mine pool discharge, which occurs through an abandoned gas well, is
captured and piped to the ALD. Influent water is sampled via a well prior to contact with
limestone. Four sampling wells are evenly spaced along the length of the drain.
Howe Bridge 2
- Mine pool discharge, which occurs through an abandoned gas well, is
treated in an S-shaped ALD. Influent water is sampled via a well as the water flows into the
beginning of the ALD. Two sampling wells are located along the length of the ALD.
Morrison
- Seepage is intercepted at the toe of the spoil of a reclaimed surface mine.
After the ALD was built, another seep, similar in quality to the pre-construction water, was
discovered, and is being used to represent influent water quality. Two sampling wells are
located along the length of the ALD.
Filson (R and L)
- Seepage is intercepted at the toe of the spoil. A seep, located between
the ALDs that is similar in quality to the pre-construction raw water, is used to represent influent
water quality.
Elklick
-Water from an abandoned borehole is collected in a bed (7.0 m x 1.8 m x 0.9 m)
of crushed, low-pyrite sandstone at the head of the ALD. Influent water is sampled at a well
located in this sandstone. Three sampling wells are equally spaced along the length of the ALD.
REM (R and L) and Schnepp
- ALDs were constructed down slope from collapsed
underground mine entrances. Influent water quality is based on historical data, which may
overestimate contaminant levels, since water quality elsewhere in the watershed has improved
significantly over the past decade.
Jennings
- An abandoned underground mine discharge was collected in a French drain
filled with inert river gravel and piped to the system. Influent water was sampled prior to contact
with limestone via a sampling well. The ALD consisted of a series of 6 buried limestone cells.
Water flowed into the bottom of each cell and exited through the top before being piped to the
next cell.
Detention times (t
d
) were calculated based on ALD volume (V) and average flow (Q),
using t
d
= 0.49V/Q. The porosity value was determined using containers of known volume filled
with the limestone used in the ALDs. The amount of water it took to fill these limestone-filled
containers to the top of the limestone was measured. An average value of 49 percent for porosity
was obtained. To confirm these calculated detention times, tracer tests were conducted at two
ALDs. The results of these tests are presented below. Additional details on the construction of
each ALD are presented in Table 8.
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

35
Table 8. Dimensions, Stone Size and Quality, and Source of Influent Water Quality Data for ALDs
Limestone
Influent Water
Quality Data
Source
ALD Site
ALD Dimensions:
Length x Width x Depth,
meters
size, cm
% CaCO
3
Howe Bridge 1
36.6
6.1
1.2
5.1 - 7.6
82
Well
Howe Bridge 2
13.7
4.6
0.9
5.1 - 7.6
82
Well
Elklick
36.6
3.1
0.9
5.1 - 20.3
85
Well
Jennings
1
228
1.0
1.0
15.2
90
Well
Morrison
45.7
0.9
0.9
5.1 - 7.6
92
Adjacent Seep
Filson - R
54.9
6.1
0.9
5.1 - 7.6
88
Adjacent Seep
Filson - L
54.9
6.1
0.9
5.1 - 7.6
88
Adjacent Seep
Schnepp
12.2
6.1
0.9
1.9 - 2.5
90
Historical
REM - R
13.7
7.6
0.9
7.6
82
Historical
REM - L
61.0
16.8
0.9
7.6
82
Historical
1
The Jennings ALD is composed of 6 sequential cells, each cell approximately 38 m x 1 m x 1 m.
Tracer Studies
To obtain information about the flow characteristics within the ALDs, tracer studies were
undertaken at two of the sites. Known amounts of concentrated sodium bromide solutions were
added to the influent flow. Samples were collected at the effluent after selected periodic
intervals (1to 8 hrs) using automatic samplers. Bromide concentrations were measured using a
specific ion electrode (in the field), and by ion chromatography (samples returned to lab) with
suppressed conductivity detection using a standard anion column. In analyzing the tracer data,
the effective (or mean) detention time (t
e
) was calculated by t
e
=
Σ
[(C
t
t)
t]
/
E(C
t
t), where C
t
is
the bromide concentration at time t, t is the time after tracer addition, and
t is the time between
samples.
The concentration profile obtained at the Howe Bridge site is shown in Figure 8. The
second experiment at the Morrison site, produced a similar profile. The profiles are
asymmetrical with rapidly rising concentrations at shorter times and gradually dropping
concentrations at longer times. Such profiles may be the result of a number of factors, such as
diffusion, channeling, back-mixing, adsorption, and mobile phase saturation, acting
simultaneously. In the case of ALDs, the first three factors presumably predominate.

36
5
10
15
20
25
30
B
r
o
m
i
d
e
C
o
n
c
e
n
t
r
a
t
i
o
n
0
25
50
75
100
125
150
175
200
Time, hours
Figure 8. Bromide Concentration Versus Time in the Effluent of the Howe Bridge ALD 1 Resulting
from a Pulse Input of a Bromide Tracer
In Table 9, a number of descriptive statistics that can be used to characterize the flow are
compared to the detention time calculated using the drain volume and the aforementioned 49
percent void volume. In both cases, the maximum concentration occurred fairly soon after the
first appearance of the bromide in the effluent. The time required for 50 percent of the material
to exit the drain—the median detention time—was considerably longer than the time to peak,
and most closely matched the calculated detention time. The time-weighted average—(i.e,
mean, or effective) —detention time was longer yet. Of these statistics, we consider the median
detention time to be the better measure of performance for ALDs. The ratio of the median (50
percent eluted) to the mean (effective) detention time is less than one. This indicates that a
disproportionate amount of material is eluting at times earlier than expected for an ideally
behaving plug-flow system, and is interpreted as an indication of channeling (i.e., short-
circuiting). In addition, concentrations of bromide above background levels continued to be
measured in ALD effluent for several days after tracer addition, indicating that more material is
eluting at times longer than would be ideally expected. This is taken as an indication that back-
mixing or dead areas exist within the drain. Thus, the ALDs appear to provide both shorter and
longer detention times than would be expected, based on simple plug-flow. Channeling is of
concern because it leads to inefficiencies in calcite dissolution. The longer residence times of
some of the mine water is not necessarily beneficial, because the concentration of alkalinity in
this water does not increase significantly after 15 hours of contact with the limestone.
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

37
Table 9. Tracer Test Data for Two ALDs
ALD
Howe Bridge 1
Morrison
Time to first appearance, hrs
7
4
Time to peak, hrs
16
10
Time for 50% of tracer to elute, hrs
30
61
Effective detention time1, hrs
40
87
Calculated detention time2, hrs
25
47
1
Effective detention time (t
e
) calculated by t
e
=
Σ
[(C
t
t)
t]
/
E(C
t
t), where C
t
= bromide concentration at time
t, t = time after tracer addition, and
t = time between samples.
2
Calculated detention time (t
d
), based on
limestone volume (V) and average flow rates (Q) by t
d
= 0.49V/Q, using 49% for porosity.
Limestone Dissolution and Alkalinity Production
Tables 8 through 12 show data describing the 10 ALDs discussed in this report. These
ALDs intercept flows ranging from about 10 to about 100 L/min. When possible, ALDs were
designed to provide a detention time of at least 15 hours. The importance of detention time is
seen in Figure 9, where the amount of alkalinity in the effluent ALD water is plotted as a
function of the time the water is in contact with the limestone (detention time). These data were
obtained at four sites where sampling wells had been installed at regular intervals along the
length of the ALD. The mine water increases in alkalinity as it travels through the ALD, until it
approaches a maximum after about 15 to 20 hours of contact. As the shape of the plots in
Figure 9 show, the ultimate level of alkalinity addition varies by ALD, but the rates at which the
alkalinity level increases appear to be nearly first order with a half-life of about 5 hours. A
minimum contact time of 15 hours ensures that at least 85 percent of the maximum achievable
alkalinity is realized in the ALD.
The variation in the level of alkalinity addition cannot be attributed to the size of the
limestone because it was the same for all four ALDs. There is a trend of increasing limestone
dissolution with decreasing pH for the data presented; however, the final concentration of
alkalinity produced in an ALD depends on factors other than just the pH of the water to be
treated. An empirical test has been developed to estimate the alkalinity concentration that will
be produced in an
ALD using the actual mine water and limestone in collapsible containers
(cubitainers) (Watzlaf and Hedin 1993). With this, we can determine limestone consumption
rates, the quantity of limestone needed for a desired design life, and whether the ALD will make
the mine water net alkaline.

38
0
50
100
150
200
250
300
0
10
20
30
40
50
Time (hrs)
Alkalinity Generation (mg/L as CaCO
3
)
Elklick
Howe Bridge 1
Howe Bridge 2
Morrison 1
Figure 9. Alkalinity Concentration as Mine Water Flows through Selected ALD
All of the ALDs successfully add alkalinity, increasing the effluent levels by 50 to 270
mg/L. (See Table 11.) The smallest increases, observed at REM-R and REM-L, are
undoubtedly the result of short detention times (7 to 8 hours) afforded by these ALDs. (See
Table 10.) At half of the sites, a single ALD was sufficient to convert net acidic to net alkaline
drainage. In the other five cases, the acidity produced from iron concentrations in excess of 200
mg/L was greater than the amount of alkalinity generated in the ALD. The increases in the
alkalinity measured between the inlet and outlet of each drain correlate with the increase in
calcium concentration. The average molar ratio of the increases in calcium and alkalinity, as
CaCO
3
, ((calcium out-calcium in)/(alkalinity out-alkalinity in)) was 1.02 for the seven cases for
which all the data were available. This compares well with the expected ratio of 1.00. In
general, ALDs receiving water low in aluminum and ferric iron concentrations, that have been
designed with detention times greater than 15 hours have generated alkalinity at a consistent rate
throughout their existence. (See Figure 10.) The effluent concentrations of alkalinity in the
ALDs indicate consistent performance over the past ten years. The significantly higher
alkalinity values for the last sample points for Filson R and L in Figure 10 were caused by very
high detention times reflecting very low flows caused by the drought of 1999, which affected the
entire state of Pennsylvania during that summer and into the fall. Detention times in these ALDs
were increased by more than a factor of four during this low flow period. Flows at the other
ALDs (Howe Bridge 1 and Elklick) shown in Figure 10 were not significantly affected by the
drought. It is also of interest to note that no seasonal variation was observed for these ALDs,
presumably because of the fairly narrow range of influent water temperatures, typical of
groundwater, and relatively consistent flow rates.
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

39
0
100
200
300
400
Sep-91 Jan-93 Jun-94 Oct-95 Mar-97 Jul-98 Dec-99 Apr-01 Sep-02
Date
Alkalinity, mg/L (as CaCO
3
)
HB-1
HB-2
Filson-R
Filson-L
Elklick
Morrison-1
Schnepp
REM-R
REM-L
Figure 10. Effluent Alkalinity Concentrations of Selected ALDs Over Time
The amount of calcium carbonate remaining in these ALDs was calculated using the
difference between the influent and effluent net acidity loadings over the period of time the
system has been in use. Based on the quantity of limestone remaining, and assuming that the
volume of the drain collapses around the shrinking core of limestone (i.e., void volume remains
at 49 percent), the current detention times were calculated. As would be expected, detention
times become shorter as the limestone is consumed. However, most ALDs are still operating at
near maximum efficiency because detention times remain in excess of 15 hours. As an estimate
of expected longevity, when the ALD detention time is expected to fall to the 15-hour minimum,
the year was calculated from a linear
extrapolation of the average rate of limestone consumption
to date (last column of Table 9). Over half of the ALDs are still expected to meet or exceed their
design life of 30 years. Resource constraints at the site resulted in the undersized construction of
three ALDs. The Jennings ALD is no longer in operation because of clogging failure, and is
described in more detail below.
Water Quality Changes
In addition to increases in calcium and alkalinity, changes in other effluent water quality
parameters (pH, sulfate, and metals) were observed. Influent and effluent water quality analyses
for the ALDs are presented in Table 11 and Table 12. At four of the ALDs (Howe Bridge 1,
Howe Bridge 2, Elklick, and Jennings), influent samples could be collected immediately before
the water flowed into the ALD. At three of the ALDs (Morrison, Filson-R, and Filson-L), seeps
in the immediate vicinity were sampled to represent influent water quality. The remaining three
ALDs (REM-R, REM-L, and Schnepp), based influent water quality on historical data.
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

40
Table 10. Initial and Current Conditions of ALDs
ALD Site
Year
Built
Initial Conditions
Avg.
Flow
Current
Conditions
Year when t
d
=
15 hours
Limestone
tonnes
t
d
1
hrs
L/min
Limestone
tonnes
t
d
1
hrs
Howe Bridge 1
1991
455
27
90.1
392
23
2024
Howe Bridge 2
1993
132
14
49.2
103
11
1993
Elklick
1994
165
25
35.8
147
22
2021
Jennings
1993
365
2
73.4
356
n/a
n/a
Morrison
1990
65
45
7.8
53
37
2035
Filson-R
1994
590
81
39.0
549
76
2100
Filson-L
1994
635
84
40.3
588
78
2086
Schnepp
1993
130
39
18.0
116
35
2047
REM-R
1992
125
6.0
112
92
4.4
n/a
REM-L
1992
125
7.1
94.5
75
4.3
n/a
The term t
d
represents detention time.
1
t
d
based on limestone volume (V) and average flow (Q) using t
d
= V/Q,
assuming 49% porosity. n/a - not applicable
Table 11. Average Water Quality Before and After Contact with the Anoxic Limestone Drain
ALD
Net Acidity,
1
mg/L as CaCO
3
Alkalinity,
mg/L as CaCO
3
Calcium,
mg/L
pH, s.u.
Sulfate
mg/L
In
Out
In
Out
In
Out
In
Out
In
Out
Howe Bridge 1
461
344
33.2
155
154
206
5.72
6.29
1294
1294
Howe Bridge 2
396
265
37.3
161
148
201
5.44
6.45
1171
1175
Elklick
54.1
-59.2
35.2
155
79.4
130
6.01
6.65
338
333
Jennings
280
-33.5
0
139
ND
201
3.23
6.16
633
620
Morrison
2
382
55.3
28.4
280
113
222
5.18
6.33
1246
1039
Filson-R
2
57.2
-154
48.1
300
69.6
186
5.61
6.41
411
427
Filson-L
2
57.2
-168
48.1
323
69.6
175
5.61
6.52
411
401
Schnepp
3
307
-22.7
0
165
69.2
199
3.28
6.16
980
768
REM-R
3
1148
819
0
56
258
228
4.28
5.41
2825
2338
REM-L
3
ND
256
ND
110
ND
202
ND
5.94
ND
1225
1
Negative net acidity values indicate net alkalinity.
2
A
In
@
concentrations based on water quality of a nearby
seep.
3
A
In
@
concentrations based on historical water quality data of the untreated mine drainage prior to
construction of the ALD. Numbers are not available for REM-L. ND = Not Determined.
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

41
Table 12. Additional Water Quality Parameters Before and After Contact with the Anoxic Limestone
Drain
ALD
Iron,
mg/L
Manganese,
mg/L
Aluminum,
mg/L
Cobalt,
mg/L
Nickel,
mg/L
Zinc,
mg/L
In
Out
In
Out
In
Out
In
Out
In
Out
In
Out
Howe Bridge 1
270
268
41.1
40.8
<0.2
<0.2
0.44
0.44
0.49
0.48
0.58
0.51
Howe Bridge 2
223
239
35.0
34.8
<0.2
<0.2
0.37
0.37
0.38
0.38
0.41
0.36
Elklick
56.0
54.2
4.72
4.82
<0.2
<0.2
0.07
0.07
0.10
0.09
0.14
0.08
Jennings
75.6
59.3
8.39
8.33
20.9
1.1
0.13
0.15
0.40
0.40
0.66
0.54
Morrison
1
208
157
48.3
40.4
0.6
<0.2
0.86
0.72
0.79
0.64
0.98
0.66
Filson-R
1
57.6
51.6
20.8
19.1
0.4
<0.2
0.24
0.23
0.19
0.18
0.23
0.22
Filson-L
1
57.6
68.1
20.8
16.9
0.4
<0.2
0.24
0.14
0.19
0.13
0.23
0.16
Schnepp
2
92.0
66.2
28.0
27.4
6.7
<0.2
ND
0.29
ND
0.35
ND
0.36
REM-R
2
589
437
136
123
4.5
3.2
ND
1.44
ND
1.46
ND
2.36
REM-L
2
ND
180
ND
50.6
ND
<0.2
ND
0.59
ND
0.64
ND
0.72
1
A
In
@
concentrations based on water quality of a nearby seep.
2
A
In
@
concentrations based on historical water
quality data of the untreated mine drainage prior to construction of the ALD. ND - Not Determined
In general, the pH increased as the alkalinity increased, until a pH of about 6.4 was
achieved above 160 mg/L. The seven pH measurements corresponding to alkalinities above 150
mg/L gave an average pH of 6.45 +
0.20 s.u. Thus, the effluent of an ALD resembles a
bicarbonate buffered solution, as would be expected for a mixture of mineral acid and carbonate
alkalinity.
Sulfate concentrations were not affected by the ALDs. The first 4 entries in Table 10,
which represent matched influent/effluent samples, never show more than a 15 mg/L loss of
sulfate. Subsequent entries do show some larger sulfate loses, but only historical or adjacent
seep data are available for the influents, making the apparent losses suspect, since sulfate losses
would not be expected in these ALDs. Chemical precipitation as gypsum is unlikely at these
concentrations. These ALDs do not contain added organic matter, which acts as ion exchange
material, or a source of carbon for sulfate-reducing anaerobes, such as occurs in RAPS.
Although sulfate reduction does not appear to be active in the systems studied here, it cannot be
ruled out for all ALDs. In some cases, mine drainage becomes associated with other pollution
sources, such
as feed lot runoff or contributions from leaking sewers or septic systems. In such
cases, a source of organic carbon would be present, which could provide an acceptable
environment for anaerobic, sulfate-reducing bacteria within the ALD.
In those cases where matched influent and effluent samples were obtainable (Howe
Bridge 1, Howe Bridge 2, Elklick, and Jennings), the iron balances (with the exception of
Jennings) indicated iron was not retained within the ALD. The influent water at Jennings
contained both ferrous and ferric iron. At the other three sites, all of the iron was in the ferrous
form (>99 percent). As already discussed, at the sites where inlet concentrations were estimated
from historical data, the listed “in” value probably overestimated the contamination actually
entering these ALDs. Manganese balances across the ALDs indicated little or no retention.
Chemical precipitation as an oxide or hydroxide would not be expected in water having a pH of
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

42
less than 7 under anoxic conditions. Although manganese carbonate precipitation is a
possibility, there was no indication of this at the ALDs in this study.
Only water at three of the sites water contained aluminum in excess of 1 mg/L. The
highest aluminum concentration was observed at the Jennings site (21 mg/L) and is most
probably the reason for the premature failure of this ALD. The REM-R and Schnepp ALDs
received 4.5 and 6.7 mg/L of aluminum (based on historical water quality). The REM-R ALD
recently failed after 10 years of treatment, with no water emanating from the effluent pipe. All
of the water is bypassing the ALD, presumably the result of a significant reduction in
permeability. The Schnepp ALD has continued to operate since 1993. Because the actual
influent water samples were unobtainable, the average aluminum concentrations at these two
ALDs is uncertain, but is presumed at somewhat less than the original 5 to 7 mg/L. The
untreated mine water quality at other sites in this watershed has shown a general and significant
improvement over the past ten years. Therefore, using historical data for the influent water
quality may bias the data toward the more contaminated water samples analyzed 10 or more
years ago.
The concentrations of cobalt, nickel, and zinc were low in these waters and seldom
exceeded 1.0 mg/L. Cobalt and nickel do not appear to be retained in the ALDs. The
appearance of Zinc diminishes in all of the effluents, but only by 0.1 mg/L or less. Some
removal of zinc at higher concentrations (ca. 5 to 10 mg/L) in ALDs has been reported by Nuttall
and Younger (2000).
Premature Failure of Two ALDs
Jennings
Construction of the ALD at the Jennings site was completed during April 1993.
Although the ALD successfully reduced the acidity of the mine water, the amount of flow
passing through it began to decrease after 4 to 5 months, as a small leak developed near the
beginning of the third ALD cell. The flow from this leak progressively increased until it
accounted for more than 80 percent of the total flow after 9 months of operation.
Analysis of the water quality and flow data provided insight into the possible
mechanisms of failure. Essentially 100 percent of the aluminum was retained within the ALD.
Most of the ferric iron, which accounted for about 10 percent of the total iron in the mine water,
was also retained in the ALD. Nearly 100 percent of these two species were retained with no
loss in efficiency, even as the flow deceased towards the end of 1993. Both of these species
form stable precipitates under the ambient conditions in the ALD, and are undoubtedly
responsible for the decreasing permeability and eventual clogging of the drain. In addition to the
constant removal of the two easily precipitated species, there is an initial retention of ferrous iron
during the first few months of operation, probably due to oxygen scavenging by the ferrous
species, adsorption on limestone surfaces or ion exchange on clay minerals in the limestone.
During construction, the air in the drain contains oxygen, which is available for reaction if it is
not flushed from the system prior to operation. Up to 40 percent of the iron retained in the drain
may have resulted from the oxidation of ferrous iron and the subsequent precipitation of ferric
hydroxide.
The total quantity of retained material was calculated at 581 kg of aluminum and 572 kg
of iron. Thus, a combination of both iron and aluminum could be responsible for clogging the
Jennings ALD. However, it might be argued that aluminum was more important for two reasons.

43
First, given the assumption that about 40 percent of the iron precipitate was caused by oxidation,
and some of the remainder by adsorption processes, it probably occurred throughout the ALD,
rather than in the one section, where the actual plug developed. Second, the portion of the ALD
where the clog was suspected was excavated, revealing the formation of a white gelatinous
substance, similar to aluminum precipitates seen elsewhere. Aluminum is thought to be the
major cause of failure at this site. In the absence of reducing conditions, such as those generated
in RAPS, the ferric iron in the influent may have also contributed to the problem.
REM
Construction of the REM ALD was completed in 1992. The ALD produced an average
of 54 mg/L of alkalinity with its 6.8 hours of detention time. About 6 years after construction,
water was noted leaking from the ALD. Effluent flow was significantly reduced after 9 years of
operation, and the volume of leakage increased. In year 10, the effluent stopped flowing
completely.
Only historical influent water quality exists for this site. Prior to construction of the
ALD, the mine discharge contained 4.5 mg/L of aluminum. Throughout the 10 years of
monitoring this ALD, aluminum floc was observed emanating from the effluent pipe. If the
effluent pipe was blocked (by hand) for 30 to 60 seconds, a slug of aluminum precipitates would
flow out of the effluent pipe. This indicates that aluminum solids were precipitating within the
ALD and were most probably the cause of its eventual failure.
Compost Wetlands
Mine water containing Fe
3+
, aluminium, or dissolved oxygen (DO) concentrations greater
than 1 mg/L has been treated with surface-flow compost wetlands. Compost wetlands generate
alkalinity through a combination of bacterial activity and limestone dissolution. The desired
sulfate-reducing bacteria requires a rich organic substrate which allow anoxic conditions to
develop. Limestone dissolution also occurs readily within this anoxic environment. A substance
commonly used in these wetlands is spent mushroom compost, a substrate that is readily
available in western Pennsylvania. However, any well-composted equivalent should serve as a
good bacterial substrate. Spent mushroom compost has a high CaCO
3
content (about 10 percent
dry weight), but mixing in more limestone may increase the alkalinity generated by CaCO
3
dissolution. Compost substrates that do not have a high CaCO
3
content should be supplemented
with limestone. The compost depth used in most wetlands is 30 to 45 cm. Typically, a ton of
compost will cover about 3.5 square yards about 45 cm thick. Cattails or other emergent
vegetation are planted in the substrate to stabilize it, and to provide additional organic matter to
“fuel” the sulfate reduction process. As a practical tip, cattail plant/rhizomes should be planted
deep into the substrate prior to flooding the wetland cell.
Compost wetlands, in which much or most of the water flows over the surface of the
compost, remove acidity (e.g., generate alkalinity) at rates of approximately 2 to 12 g d
&1
m
&2
.
This range in performance is largely a result of seasonal variation: acidity removal rates are
lower in winter than in summer (Hedin et al., 1991). Hedin et al. (1994a) recommended sizing
compost wetlands based on acidity removal rates of 3.5 to 7 g d
&1
m
&2
. Since the beneficial
reactions occur in the compost and limestone layers, and not in the surficially flowing water,
these systems must be quite large. For the past 5 years or so, reducing and alkalinity-producing
systems (RAPS) have been used to treat net acidic water containing ferric iron, aluminum, or DO
concentrations greater than 1 mg/L, instead of compost wetlands. However, at sites with
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

44
sufficient land area and/or minimal elevation difference between the mine discharge and the
stream, compost wetlands may be the most appropriate choice.
Reducing and Alkalinity-Producing Systems (RAPS)
RAPS is a generic term that describes the chemistry within a certain type of passive
treatment. In addition to producing alkalinity via dissolving limestone, these systems promote
reducing conditions by incorporating organic matter. The RAPS design directs water to flow
down through organic matter into the limestone. The reducing conditions facilitate sulfate
reduction, which generates alkalinity (reaction Y), and may also precipitate some metals to
sulfides. Ferric iron can be reduced to ferrous iron, eliminating the precipitation of ferric
hydroxide and subsequent clogging and armoring of the limestone.
2 CH
2
O = SO
4
2-
→H
2
S + 2 HCO
3
-
(Y)
This type of system was first implemented by Doug Kepler at the Howe Bridge site.
These systems were termed successive alkalinity-producing systems (SAPS), indicating that
more than one of these units could be used in series to treat very highly acidic water (Kepler and
McCleary 1994). Similar systems have also been referred to as vertical flow systems, vertical
flow ponds, or vertical flow wetlands. Chemically, biologically, and physically these systems
behave similarly, and will be referred to as RAPS in this manual. A layer of limestone (0.6 to
1.2 m thick) is placed on the bottom of an excavated area. A network of perforated pipes is
placed in the lower portion of this limestone layer. Organic material (0.15 to 0.61 m thick),
which typically has been composted, is placed above the limestone, and serves as the nutrient
source for the sulfate reducing bacteria. In Pennsylvania, spent mushroom compost has been the
organic material of choice. It roughly consists of composted horse manure (56 percent by
weight), hay (22 percent), straw (10 percent), chicken manure (10 percent), and gypsum (2
percent), but can differ between mushroom farms, since each uses its own recipe. Mine water
flows down through the system, encountering reducing conditions within the compost before
contacting the limestone. In the reducing environment, dissolved oxygen is removed, which
prevents ferrous iron oxidation, and any ferric iron already present is reduced to the ferrous state.
Thus, RAPS are appropriate for water containing ferric iron, which could armor the limestone in
an ALD.
It is thought that RAPS may also be more resistant to plugging by aluminum than ALDs
because of their larger cross sectional area and higher available head pressures (Watzlaf and
Hyman 1995). The oldest RAPS in this study (Howe Bridge) treated water for 11 years before
being replaced. After 11 years, it was still able to pass 50 percent of the influent water through
the compost and limestone layers. This system received less that 0.2 mg/L of aluminum. It
appeared that the progressive reduction in permeability was due to precipitation of iron
hydroxides on top of the compost layer, with an accumulation of iron sludge in excess of 15 cm
on top of the compost. Reduced permeability may also result from storm-mobilized silt and
other solids, as well as precipitation of metal sulfides within the organic layer. Thus, continued
monitoring of the actual performance of these systems is warranted.
In practice, RAPS, ALDs, settling ponds, and aerobic wetlands are used as unit
operations in a total remediation system. For example, RAPS are usually preceded by a settling
pond/wetland to settle iron and other solids, which could reduce permeability of the system.
RAPS and ALDs are followed by settling ponds and aerobic wetlands for oxidation,
precipitation, and settling of metals. After these ponds and wetlands, additional RAPS may be
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

45
used, each separated by a settling pond and wetland, to sequentially improve the water quality
when sufficient alkalinity cannot be introduced in the initial ALD or RAPS.
Site Descriptions
Howe Bridge
-
Water flows through a compost wetland (0.14 ha) prior to entering a
RAPS (0.14 ha). The RAPS contains a 0.4-m layer of limestone gravel covered by a 0.2-m layer
of spent mushroom compost and about 1.5 m depth of water. Perforated drainage pipes (black
plastic corrugated sewer pipe) are placed in a serpentine pattern in the bottom of the limestone
layer. These pipes only cover about one-half of the total surface area of the system (~0.07 ha).
Influent water is collected prior to the compost wetland.
Oven Run D (#1 and #2)
- This system treats discharges from reclaimed surface and
daylighted deep mines. Two RAPS are in series, each with a surface area of about 0.15 ha. Both
contain a 0.91-m thick layer of limestone and a 0.15-m thick layer of compost covered by 1.5 m
of water. a wetland precedes each RAPS with a surface area of 0.11 ha and a depth of 0.076 to
0.152 m of water. Influent water for each RAPS is sampled prior to the wetlands.
Oven Run E (#1 and #2)
- Abandoned deep mine drainage is piped to two RAPS in a
series. Each RAPS has a surface area of 0.26 ha, and the same thickness of limestone, compost
and water as the Over Run D RAPS, outlined above. RAPS #1 is preceded by a 1.8-m deep
pond (0.10 ha) and a wetland (0.12 ha). RAPS #2 is preceded by a pond (0.11 ha) and a wetland
(0.11 ha).
Jennings
- A system of perforated pipes was placed within a 0.31-m thick bed of inert
river gravel, which was wrapped with a geotextile fabric. Above the gravel layer is a mixture of
limestone and spent mushroom
compost that is 0.8 m thick. This mixture consists of 270 tonnes
of compost and 345 tonnes of limestone aggregate (9.5 mm x 1 mm (i.e., 3/8 in x 16 mesh)).
Influent water is sampled prior to entering the RAPS (Jennings Water Quality Improvement
Coalition 1999).
Results
While alkalinity is produced solely by limestone dissolution in ALDs, it is produced by
both limestone dissolution and sulfate reduction in RAPS. Table 13 and Table 14 present the
data obtained for six RAPS that have been monitored for up to 9 years. Shown in Table 13 are:
(1) the alkalinity produced by limestone dissolution (based on increases in calcium, where a 1
mg/L increase stoichiometrically yields 2.497 mg/L of alkalinity as CaCO
3
); (2) the alkalinity
produced by sulfate reduction (based on decreases in sulfate, where a 1 mg/L decrease
stoichiometrically yields 1.042 mg/L of alkalinity as CaCO
3
); (3) the measured total alkalinity
generated by the RAPS; and (4) the specific rate of generation of alkalinity calculated as grams
per day, per square meter of surface area, measured at the top of the compost layer.

46
Table 13. Construction Specifications and Quantification of Alkalinity Generation within RAPS
RAPS Site
Howe
Bridge
Oven Run
D #1
Oven Run
D #2
Oven Run
E #1
Oven Run
E #2
Jennings
Yr Built
1991
1995
1995
1997
1997
1997
Avg. Flow, L/min
70.9
342
323
408
413
61.4
Compost
Qty, tonnes or m3
272 t
140 m
3
140 m
3
248 m
3
248 m
3
270 t
td1, hr
8.8
1.7
1.8
2.5
2.5
24
Alkalinity by SO
4
reduction mg/L
as CaCO
3
92
25
41
61
21
57
Limestone
Qty, tonnes
454
1349
1349
2425
2425
345
Td1, hr
34
21
22
32
32
24
Alkalinity by limestone
dissolution, mg/L as CaCO3
120
70
4
130
30
419
Total Measured Alkalinity Generated
3, mg/L as CaCO3
212
97
29
149
58
424
Alkalinity Generation Rate, gd
-1
m
-2
18.2 - 36.3 53.5
17.1
40.3
16.2
60.4
1
td based on quantity of limestone or compost at construction and average flow rates using td = V/Q and
assuming void volumes of 49 percent for limestone and specific yields of 25 percent and 20 percent for
compost and compost/limestone mixture, respectively.
2
Jennings contained a compost and limestone mixture,
the 25-hr td is for the mixed layer.
3
Total alkalinity generated based on changes in measured net acidity
between the influent and effluent of RAPS.
The Howe Bridge RAPS produced approximately equal amounts of alkalinity from
sulfate reduction and limestone dissolution over the past 9 years. Much of the alkalinity that was
derived from sulfate reduction occurred in the first 2 to 3 summers of operation. (See Figure
11.). Seasonal trends in sulfate reduction were shown in the first few years of operation.
Although it is more difficult to see seasonal trends in more recent years because of lower
sampling frequency, it is apparent that the alkalinity production is not reaching the high levels
achieved in the first few years. Alkalinity generation rates were calculated as 19.6 g d
-1
m
-2
using the total surface area of the top of the compost. However, the perforated piping in the
limestone layer extended only about half way into the system, potentially causing it to perform
as if water actively flowed through only half of the RAPS. Taking this into account, actual
alkalinity generation rates are probably on the order of 39 g d
-1
m
-2
.

47
0
100
200
300
400
500
Sep-91 Sep-92 Sep-93 Sep-94 Sep-95 Sep-96 Sep-97 Sep-98 Sep-99 Sep-00
Date
Alkalinity, mg/L (as CaCO
3
)
Sulfate Reduction
LS Dissolution
Sum, SO4 + LS
Net Acidity Change
Figure 11. Alkalinity Generation in the Howe Bridge RAPS
Alkalinity from sulfate reduction and limestone (LS) dissolution were calculated from differences in sulfate and
calcium, respectively. The sum of these calculated alkalinities is also plotted with actual measured changes in
net acidity.
Both Oven Run sites consist of two RAPS in series. The rationale for this was twofold:
(1) one system could be put offline for maintenance and (2) during the design life of the two
RAPS, the first system was expected to contribute more alkalinity during the first half, and the
second system would contribute more alkalinity in the last half. At site D, the first RAPS
produced alkalinity at a rate of 57.4 gd
-1
m
-2
, and the second at a rate of 20.6 gd
-1
m
-2
over five
years of operation. Similarly, at site E, the first RAPS produced alkalinity at a rate of 42.7 gd
-
1
m
-2
, and the second at a rate of 15.6 gd
-1
m
-2
, over three years of operation. It was difficult to
determine any seasonal trends in the alkalinity production at either site because of the low
sampling density and extremely variable flow rates at each site (a very wet period, followed by
an extended drought period produced a greater than tenfold difference between high and low
flows).
For the Jennings RAPS, the compost and limestone were mixed together instead of
maintaining two distinct layers. This design was chosen because laboratory tests indicated that
the water at Jennings was capable of depleting the calcium carbonate within a 0.7-meter thick
layer of spent mushroom compost in about two years (Watzlaf 1997). After the calcium
carbonate was depleted in the laboratory tests, sulfate reduction virtually ceased, presumably
because of the lower pH environment. At this lower pH, fermentative bacteria, as well as
sulfate-reducing bacteria may not be as active. The fermentative bacteria breaks down complex
organics into simpler forms that the sulfate reducers can use. The rapid depletion of calcium
carbonate was caused by the production of acidity during aluminum precipitation (an aluminum
concentration of 23 mg/L will produce 128 mg/L of acidity upon hydrolysis). The Jennings
RAPS produced the greatest change in net acidity of any of the systems, attributing over 90
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

48
percent of the alkalinity production to limestone dissolution. As discussed above, some (128
mg/L) of this change can be attributed to aluminum precipitation. This RAPS displayed no clear
seasonal trends.
Table 14 shows changes in the major water quality parameters. The change in calcium
and sulfate concentrations were used to estimate the contributions of limestone dissolution and
sulfate reduction, respectively, as described above. Net acidity was determined using the
peroxide oxidation method, and the change in net acidity between the inlet and outlet is listed as
the total alkalinity generated in the second to last column in Table 13. Manganese, which is
expected to be conserved in these systems, was present in the influent and effluent at about the
same levels. Iron, and aluminum when present, were retained by the systems.
Table 14. Water Quality Before and After Contact with Reducing and Alkalinity Producing System
RAPS
Howe
Bridge
Oven Run
D #1
Oven Run
D #2
Oven Run
E #1
Oven Run
E #2
Jennings
Net Acidity, mg/L
In
314
99.8
6.38
212.
63.2
272
as CaCO3
Out
102
2.56
-23.0
63.2
5.3
-152
Alkalinity, mg/L as
In
31.3
1.5
N/A
0.0
9.4
0.0
CaCO
3
Out
57.8
29.0
31.6
9.4
25.5
204
Calcium, mg/L
In
190
300
327
149
201
109
Out
238
328
328
201
213
277
Iron, mg/L
In
189
40.6
1.69
18.6
9.21
68.5
Out
72.1
3.41
0.47
9.21
3.93
14.7
Manganese, mg/L
In
37.0
28.2
27.3
12.1
11.9
18.6
Out
35.7
27.4
22.0
11.9
11.3
17.6
Aluminum, mg/L
In
<0.2
1.45
1.21
16.4
9.40
24.1
Out
<0.2
0.82
0.32
9.40
4.36
0.84
Sulfate, mg/L
In
1186
1356
1340
932
873
799
Out
1098
1332
1301
873
853
744
The majority of these metals were presumably retained in the wetlands that precede the
RAPS, though significant levels of iron may have been removed on top of the compost in the
RAPS. However, the Jennings site has no such wetland and retains 85 percent of the iron and all
of the aluminum. In some cases, such as the Oven Run E sites, preventative maintenance is
performed by periodic high-flow flushing, during which the RAPS pond level is lowered. The
results of two flushes are reported below. Additional details of these and other flushes can be
found in Watzlaf et al. (2002).
Case Study: Flushing the DeSale II RAPS
The DeSale II site is located in Butler County, Pennsylvania within the headwaters of
Seaton Creek, a heavily impacted tributary in the Slippery Rock Creek Watershed. The system
consists of an equalization pond, two RAPS, an oxidation pond, wetlands, and a horizontal
limestone bed. Each RAPS is approximately 100 m long and 16 m wide and consists of (from
the bottom up) 15 cm of limestone for pipe bedding (AASHTO #57), the lower discharge/flush
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

49
pipes, 60 cm of limestone (AASHTO #1 is approximately 10 cm, 90 percent calcium carbonate),
the upper discharge/flush pipes, 60 cm of limestone (AASHTO #1), 15 cm of spent mushroom
compost, and 75 to 90 cm of water. Networks of piping drain four quadrants at two different
vertical levels (upper and lower) within the system. This more extensive underdrain system was
developed in an attempt to optimize both the distribution of flow during normal operation, and
the flushing of accumulated iron and aluminum solids. The underdrain was constructed of 10-
cm diameter Schedule 40 PVC pipe. Perforated laterals were placed on 1.4-m centers and
connected to an unperforated header with a sanitary-type tee. Perforations were hand-drilled
with two, 1.3-cm perforations, which were offset approximately 30
o
from the top of pipe. The
perforation spacing was equal to the lateral spacing (1.4 m). Four separate header pipes were
used for each underdrain level, thus dividing the surface area into approximately equal
quadrants. The upper and lower underdrain levels effectively divide each RAPS into eight
separate cells, four upper and four lower.
In the two years after its completion, the RAPS functioned very effectively, increasing
pH from 3.1 to 6.9, adding 370 gm/L of alkalinity, and decreasing iron and aluminum
concentrations from 27 to 5 mg/L, and 11 to 0.3 mg/L, respectively.
The right system was flushed nine months after it began treating water. Each of the eight
pipes was flushed sequentially at full volume (660 to 1360 L/min per pipe) for nine minutes. A
total of 69,700 L of water was removed (~ 4 percent of the total volume of water in the RAPS).
Samples were collected at 15, 30, and 45 seconds, then at 30 second intervals from 1 to 5
minutes, and then at 1-minute intervals from 5 to minutes. Dissolved oxygen, temperature, and
pH were monitored continuously and recorded at each sample interval. Flows were measured
periodically (~ every 1 to 3 minutes) using three different methods: horizontal pipe discharge
method, time volumetric method, and water level changes in RAPS. All three flow measurement
techniques were in fairly good agreement (within ~ 15 percent). Samples were not filtered and
consisted of an unacidified and acidified sample. Samples were analyzed in the laboratory for
concentrations of standard and trace mine drainage metals and sulfate.
The left system was flushed after 14 months of treatment. Based on the results of the
first flush, it was decided to flush the left system much more aggressively (i.e., to drain the
system completely). The pipes draining the four upper quadrants were opened at the same time.
After these pipes had drained for 11 hours, the flow had diminished to a trickle. After closing
the valves to the upper pipes, the lower flush pipes were opened and drained for an additional 4.5
hours. Flows were measured periodically (every 10 to 20 minutes) using the horizontal pipe
discharge method, which had compared favorably to the timed volumetric method and water
level changes method during the previous flush. A total of 1,430,000 L of water was removed
from the system. Unfiltered samples were collected from each pipe at 10 minute intervals.
Temperature, pH and flow measurements were taken between sample collection.
The two RAPS at the De Sale II site are essentially equivalent. They were constructed in
parallel, have the same dimensions, contain the same type and amount of media (limestone,
compost, etc.) and receive water from a common source.
From available monitoring data, including flow measurements and water quality
analyses, it was calculated that the right RAPS had accumulated 780 kg of iron and 312 kg of
aluminum during the first nine months of operation. The basic criterion used during this flush
was that the water should be allowed to flow until it ran clear. In practice, the flush was actually
continued for some additional time. During the flush, a total of 69,700 L of water was removed.
The maximum metal concentrations occurred in the first few minutes of flushing when the water

50
was visibly discolored. Both the visual observations and the lab analyses indicate that the initial
slug of material was removed from the system within two minutes. After seven minutes of
flushing, the iron and aluminum concentrations were the same as the concentration of dissolved
metal, indicating that no solid material was eluting.
Integration of the concentration versus time graphs indicated that only 1.4 kg of iron (0.2
percent of the iron retained since construction) and 0.9 kg of aluminum (0.3 percent of the
aluminum retained since construction) were flushed from the system. If one assumes that the
water flowed into the pipe uniformly from every direction, the range of influence of this flush
can be estimated from the pipe dimensions, the gallons flushed, and by assuming a limestone
porosity of about 50 percent (Hedin and Watzlaf 1994). It is estimated that the last water
through each pipe had been, on average, only 10 cm from the pipe before the flush began. Thus,
it is doubtful that much, if any of the metal oxyhydroxide-laden water actually entered the pipes
during this limited flush. Our conclusion is that “flushing until the water runs clear” is probably
not a sufficient criterion for effective flushing.
The left RAPS was flushed 14 months after it began treating water. From available
monitoring data, including flow measurements and water quality analyses, it was calculated that
the RAPS had accumulated 948 kg of iron and 499 kg of aluminum during these first 14 months
of operation. Because of the low amount of metals removed during the flushing of the right
RAPS, the criterion used during this flush was that the water should be allowed to flow as long
as possible (i.e., until the system was drained). In practice, the four pipes draining the upper
quadrants were flushed until the flow slowed to a trickle, then the pipes draining the lower four
quadrants were opened and allowed to flow until the system was totally drained. During the
flush, a total of 1,430,000 L of water was removed. To a first approximation, the flows in both
the upper and lower sections decrease going from quadrant 4 to quadrant 1. Qualitatively, this is
consistent with the pressure drop expected because of the increasing length of 10 cm diameter
pipe draining the quadrants. However, it would also be consistent with a clogging mechanism in
which the settling of suspended material, such as clays, predominated in the quadrants closest to
the RAPS inlet. The flows dropped slowly at first, and then more rapidly after the first 7 to 8
hours. At about 11 hours, the upper quadrants had drained and the valves to the lower quadrants
were opened. Flows were higher and longer for the upper quadrants than for the lower
quadrants,because these pipes drained, the standing water, the compost water, and the top
limestone layer (total of 1.5 m of head), whereas the lower quadrant pipes drained only the
bottom limestone layer (0.6 m of head).
The temperature and pH were monitored throughout the flush. The trends for both
parameters were the same; both decreased as the cooler, more acidic surface water penetrated the
lower strata faster than the chemical and thermal equilibration could occur. At about 7 hours the
values began to climb toward their earlier levels. This was at the same point at which the flow
sharply decreased (and residence time increased) indicating that the thermal and chemical
equilibration rates were now becoming competitive with the flow rate. At a little over 8 hours
(where the breaks in the upper quadrant trend lines occur), it was necessary to shut off the flow
due to darkness. The next morning, the temperature and pH continued to increase further to near
their initial values. The pH actually attained somewhat higher values, perhaps because of the
overnight stoppage of flow during which extended contact with the limestone occurred.
The maximum metal concentrations occurred in the first few minutes and corresponded
to visibly discolored water similar to what was seen for the right RAPS flush. In total, little
additional material was removed from the system even after prolonged flushing. Of the retained

51
948 kg of iron and 335 kg of aluminum, the flush removed 10.0 kg of iron (1.1 percent of the
iron retained since construction) and 6.53 kg of aluminum (1.3 percent of the aluminum retained
since construction).
Prior to the flush, no decrease in the permeability of the RAPS had been observed. Using
a hand level, there was no measurable (< 1.5 cm) difference in the elevation between the RAPS
water level and the level of the discharge pipe, indicating that very little head was necessary to
push the water though the RAPS. The system was probably maintaining permeability because
very little void volume had been lost up to that time. The 948 kg of retained iron corresponds to
1810 kg of Fe(OH)
3
. Our measurements indicate that a cubic centimeter of iron sludge contains
0.17 g of iron. Using this value, approximately 5.58 m
3
of iron sludge was retained in the RAPS.
Making similar assumptions for the aluminum sludge results in 2.92 m
3
of aluminum sludge, for
a total sludge volume of 8.50 m
3
. Assuming a 25 percent void in the compost and a 50 percent
void in the limestone, the RAPS contains about 595 m
3
of void space, with the precipitated
sludge occupying only about 1.4 percent of this void. Thus, it may be argued that too little
material had accumulated to be flushed effectively. Larger masses of material would present a
larger cross-sectionof the rapidly flowing water, more likely to be transported down-flow.
It is important to note that it is unlikely that the precipitates were distributed uniformly
throughout the available void volume. It is more likely that precipitation occurred in a band
(Watzlaf 1997). The width and position of the band would be determined by the pH gradient and
rates of precipitation and agglomeration. Therefore, the permeability of the RAPS could be
significantly reduced long before 100 percent of the void volume was occupied.
It is interesting to note that, although the clogging of RAPS is thought to be due to
aluminum precipitation, iron is being retained as well. The amounts of iron and aluminum being
retained in these systems are shown in Table 15.
Table 15. Total Amounts of Retained Iron and Aluminum Prior to Flushes at the DeSale II Site
RAPS
Fe Retained (kg)
Al Retained (kg)
Fe/Al Molar Ratio
Right
780
312
1.21
Left
948
499
0.92
Table 16. Amount of Water, Iron, and Aluminum Flushed for the Two RAPS at the DeSale II Site
Water Flushed
Iron Flushed
RAPS
Aluminum Flushed
L
% of total
kg
% of retained
kg
% of retained
Right
69,700
5
1.4
0.2
0.9
0.3
Left
1,430,000
100
10.0
1.1
6.5
1.3
Neither flush removed very much of the retained iron or aluminum. The most efficient
flush achieved only 1.1 percent removal of the incremental amount of metals accumulated since
the previous flush. (See Table 16.) None of the systems were experiencing any loss of
permeability prior to the flushes. In fact, only a very small percentage (1.1 to 1.4 percent) of the
void volumes were calculated to be filled with iron and aluminum precipitates. Lack of
efficiency has not yet led to failures of these systems and, in one case, efficiency may be
improving with time. However, the long-term prospects for these systems appear questionable at
best, if the current levels of metal removal via flushing continue.
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

52
Other Types of Water Treatment Systems
There are several other types of systems treating coal mine drainage, ranging from purely
passive, to semi-passive, to active. Table 17 lists many of these technologies with their
electricity requirement, the presence of moving parts, the required frequency of minor and major
maintenance, frequency of chemical addition, and estimated design life. Brief descriptions of
some of the widely used systems are given below (including the systems already discussed in
this paper). Further description and a listing of additional types of systems can be found
elsewhere (USEPA 1983, Younger et al. 2002, Brown et al. 2002, EPA 1983, PIRAMID
Consortium 2003).
Aerobic Wetlands
— Effective for the treatment of net alkaline mine drainage, aerobic
wetlands typically consist of an aeration structure (riprapped ditch, waterfall), a deep
unvegetated pond (1.2 to 2.4 m deep) and a shallow wetland (~0.15 m deep) that usually contains
cattails (typically
Typha latifolia
). The deeper ponds are designed to hold precipitated iron
oxides, while the cattail wetland is used to remove remaining dissolved and suspended iron.
Anoxic Limestone Drains
— Buried beds of limestone are designed to intercept mine
water in an anoxic state and add bicarbonate alkalinity. The presence of aluminum and ferric
iron will result in precipitation of these metal hydroxides within the ALD and could lead to
premature failure by limiting the reactivity of limestone and/or clogging with these precipitates.
Aerobic wetlands are used after these systems for the precipitation and collection of metal
precipitates.
Compost (anaerobic) Wetlands
— These systems typical contain limestone and composted
organic matter in a vegetated substrate. Typical vegetation includes cattails. Most flow is surficial.
Sulfate reduction and limestone dissolution occurs within these systems.
Limestone Beds
— Beds of limestone that are exposed at the surface. Water level within
the beds are below the exposed surface of the limestone. Water flows horizontally through these
beds. They are designed for use after iron is removed from the water and are intended to remove
manganese by encouraging conditions beneficial for biological manganese oxidation. Usually
placed at the end of treatment systems, and can also add additional alkalinity (Rose et al. 2003).
Limestone Ponds
— Constructed over upwellings of mine drainage; water flows upward
through the limestone. Function similarly to ALDs. Generally used when water has low DO and
contains low levels of aluminum and ferric iron.
Open Limestone Channels
— Channel or ditch lined with limestone. Usually placed on
a slope so the flowing water scours the limestone surface and voids to keep “clean”. This system
takes into account that limestone may armor with ferric hydroxides, but relies on the premise that
armored limestone will continue to dissolve at a slower rate. If a large settling pond is not used
at the end of these systems, metal precipitates can enter and damage the watershed (Ziemkiewicz
et al. 2003).
Diversion Wells
— Cylindrical structures in which a split of a contaminated stream water
flows upward through a bed of limestone at a velocity capable of fluidizing the bed. The
agitation of the limestone functions to keep the surfaces from armoring. In addition, the
limestone fines are generated and carried out of the system into the receiving stream, where they
may continue to dissolve and add alkalinity; hence a length of the stream may be sacrificed for
downstream improvement in water quality. This system must be filled with the proper size and
amount of stone, or bed will collapse and lose effectiveness. Stone must be replenished every
few days.
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

53
Limestone Sands
— This technology utilizes several, large, strategically placed piles of
fine limestone within streambeds of a contaminated watershed. Some alkalinity is added during
base flow conditions. During storm flow conditions the limestone is transported throughout the
watershed where it is incorporated in the stream sediments and dissolved to add alkalinity.
Again, a length of the stream may be sacrificed for downstream improvement in water quality.
The piles must be replaced at periodic intervals.
RAPS
— A reducing and alkalinity producing system . consists of a layer of limestone
overlain by a layer of composted organic matter. A drainage system is placed within the
limestone layer to force the mine water to flow downward through the compost and limestone.
The compost removes dissolved oxygen and reduces ferric iron to ferrous iron to minimize
armoring of the limestone. Typically the design includes 2 m of head between the water surface
in the RAPS, and the subsequent unit operation accounts for any loss of permeability. These
systems are commonly flushed periodically in an attempt to remove precipitated metals and
maintain permeability. RAPS are preceded with a pond to for allow precipitation of metals
within the pond and not in the RAPS. Aerobic wetlands are used after these systems for the
precipitation and collection of metal precipitates.
ReRAPS
— Recirculating RAPS, in which alkaline water produced by the RAPS (see
above) is mixed with the influent water in a pond to raise pH high enough to precipitate
aluminum outside of the RAPS. This water is then pumped to the RAPS. System has been used
to treat coal pile runoff where influent water flow is intermittent (Garrett et al. 2002).
Water-Powered Devices
— These devices use available head pressure at the site to move
some type of device (e.g., wheel, drum, tipping bucket) that meters out an alkaline material.
Aerobic wetlands are used after these systems for the precipitation and collection of metal
precipitates.
Windmills
— These are typically used for aeration by using the power generated from
the windmill to pump air through tubing into the mine water. They can also use power to meter
out alkaline materials as in water-powered devices.
Sodium Carbonate Briquettes
— Na
2
CO
3
pressed into briquettes, commonly referred to
as soda ash, are placed in a gravity dispenser (hopper) and allowed to dissolve in flowing mine
water. Ponds are used after these systems for the precipitation and collection of metal
precipitates.
Liquid Sodium Hydroxide
— A 20 to 50 percent solution of NaOH, sometimes referred
to as “caustic soda”, is stored in a large tank and gravity fed into the mine water. Ponds are used
after these systems for the precipitation and collection of metal precipitates.
Hydrated Lime
— Calcium hydroxide (Ca(OH)
2
), usually in powdered form, is added to
mine water via a screw feeder. Ponds are used after these systems for the precipitation and
collection of metal precipitates.
Quick Lime
— Calcium oxide (CaO), requires water to make up slurry (called milk of
lime) prior to adding to mine water. Ponds are used after these systems for the precipitation and
collection of metal precipitates.

 
54
Table 17. Techniques Used for Treating Coal Mine Drainage
Maintenance
Frequency
Technology
Electricity
Moving
Parts
Minor
Major
Frequency
of Chemical
Addition
Design
Life
(years)
Aerobic Wetlands
Anoxic Limestone Drains
Compost Wetlands
Limestone Beds
Limestone Ponds
Open Limestone Channels
N
N
monthly
none
anticipated
none
20 - 30
Diversion Wells
N
N
weekly
none
weekly
20 - 30
Limestone Sands
N
N
6 months
6 months
6 months
6 months
RAPS
N
N
monthly
6 months
none
20 - 30
RERAPS
Y
Y
monthly
none
none
20 - 30
Water-Powered Devices
Windmills
N
Y
weekly
weekly-yearly
none - monthly
5 - 10
Sodium Carbonate
Briquettes
Liquid Sodium Hydroxide
N
N
daily
weekly -
monthly
daily - monthly
5 - 10
Hydrated Lime
Quick Lime
Y
Y
daily
weekly -
monthly
daily - monthly
5 - 10
Designing Passive Treatment Systems
Characterizing Mine Drainage Discharges
In order to select the most effective passive treatment unit operations and to size them
properly, the untreated mine water must be well characterized. The quality and quantity of some
mine discharges are very consistent, while other discharges may vary by orders of magnitudes in
both contaminant concentrations and flow. At an absolute minimum, water quality and quantity
data should be collected during high and low flow periods. It is recommended that the discharge
be monitored periodically (e.g., monthly) for a complete water year. It is best to select the
monitoring dates in advance, and follow through on the monitoring regardless of weather. Both
the flow rate and chemical composition of a discharge can vary seasonally and in response to
storm events. If the passive treatment system is expected to operate during all weather
conditions, then the discharge flow rates and water quality should be measured in different
seasons and under representative weather conditions.
One of the most important measurements in sizing each passive unit is to obtain an
accurate measurement of the total flow of the mine discharge(s) or seep(s). Water samples
should be collected at the discharge or seepage point for chemical analysis, which should include
pH, alkalinity, iron, manganese, aluminum, and hot acidity (H
2
O
2
method) measurements. If an
anoxic limestone drain is being considered, and the pH is less than 5, iron concentrations should
be speciated into ferric and ferrous. At pH levels above 5, one can assume that all dissolved iron
is in the ferrous form. The samples should be analyzed for other ions that are usually present in
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

55
significant concentrations in coal mine drainage: calcium, magnesium, potassium, sodium,
cobalt, nickel, zinc, and sulfate. A cation/anion balance can be calculated to help verify the
laboratory analyses (see
Chemical Characteristics of Mine Drainage
on page 12).
In addition, pH, alkalinity, and dissolved oxygen should be measured in the field. If the
pH is greater than 4.5, a simple test for determining whether the water is net alkaline should be
performed: after recording the pH of the water, collect a sample , and add hydrogen peroxide ;
then stir or shake the sample and measure the pH again. If the pH drops below 4.5, the water is
net acidic. If the pH remains above 4.5, the water is net alkaline. Very inexpensive hydrogen
peroxide (3 percent solution), purchased at a pharmacy or grocery store, can be used. The
amount of hydrogen peroxide added to the sample is not critical, 5 to 10 mL per 100 mL of
sample is adequate.
Selecting Unit Operations
Water quality for a given discharge will determine the unit operations for designing the
most effective passive treatment systema. Table 18 shows three major classifications of mine
water quality. Each classification is appropriate for a particular unit operation. Examples of
class I (net alkaline) discharges are Penn Allegh, Brinkerton, and Scrubgrass; examples of class
II discharges are Elklick, Howe Bridge, and Morrison. (See Table 4.) Class III discharges are
Jennings, Schnepp, and REM-R. (See Table 11 and Table 12.)
Table 18. Classification of Mine Discharges
Water Quality Parameter*
Classification
I
II
III
pH
> 4.5
-
-
H
2
O
2
pH
> 4.5
< 4.5
< 4.5
Net Acidity
< 0
> 0
> 0
Ferric Iron
-
< 1
-
Aluminum
-
< 1
-
Dissolved Oxygen
-
< 1
-
Appropriate
Unit Operations
Aerobic
Pondsand
Wetlands
Anoxic
Limestone
Drains
Reducing and
Alkalinity
ProducingSystems
* pH in standard units, concentrations in mg/L, acidity in mg/L as CaCO
3
Currently, there are several types of unit operations for the treatment of coal mine
drainage; however, three of the most effective are aerobic ponds and wetlands, anoxic limestone
drains, and reducing and alkalinity-producing systems. In aerobic ponds and wetlands, oxidation
reactions occur and metals precipitate primarily as oxides and hydroxides. Most aerobic
wetlands contain cattails that grow in a clay or spoil substrate. However, plantless systems (i.e.,
ponds) have also been constructed and function similarly to those that contain plants. It is
recommended that net alkaline water be aerated to the maximum extent possible, conveyed to an
aerobic pond and polished with an aerobic wetland.
The ALD is a buried bed of limestone intended to add alkalinity to the mine water. The
limestone and mine water are kept anoxic so that dissolution can occur without ferric
oxyhydroxides armoring the limestone. ALDs are only intended to generate alkalinity, and must
be followed by an aerobic system in which metals are removed through oxidation and hydrolysis

56
reactions.
A RAPS consists of layers of limestone and compost. The water flows down through the
compost to remove oxygen and to reduce ferric iron to ferrous iron. The limestone adds
alkalinity. Most systems are designed to facilitate the periodic removal of aluminum and iron
precipitates by flushing water through the system.
Each of these three passive technologies is appropriate for a particular type of mine water
problem, but they are most effectively used in combination with each other. Figure 12 can be
used to determine the selection and sequence of unit operations for an effective passive treatment
system.
Figure 12. Selection of Passive Treatment Unit Operations
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

57
Sizing Passive Systems
The size of the passive treatment system depends on the loading rate of contaminants.
Calculate contaminant (iron, manganese, acidity) loads by multiplying contaminant
concentrations by the flow rate. If the concentrations are mg/L and flow rates are L/min, the
calculation is:
Load (g/day) = Flow (L/min) x Concentration (mg/L) x 1.44 (g min/mg day)
(11)
Aerobic Ponds and Wetlands
Sizing criteria for abandoned mined land (AML) uses 20 g d
&1
m
&2
for iron, and 1.0 g d
&1
m
&2
for manganese. These are intended to cost-effectively decrease contaminant concentrations
(Hedin et. Al. 1994a). In many situations at abandoned mined lands, the goal is to improve
water quality, not consistently achieve a specific effluent concentration. The AML sizing
criteria are based on measurements of contaminant removal by existing constructed wetlands.
Most of the removal rates were measured for treatment systems (or parts of treatment systems)
that did not consistently lower contaminants to federal effluent standards. In particular, the iron
sizing factor for alkaline mine water (20 g d
&1
m
&2
) is based on data from six sites, only one of
which lowers iron concentrations to compliance.
It is possible that iron removal rates are a function of iron concentration (i.e., as
concentrations decrease, the size of the system necessary to remove a unit of iron contamination
(e.g., 1 g/d) increases). To account for this possibility, we have provided a more conservative
sizing value for systems where the effluent must meet regulatory guidelines. (See Table 1.) We
refer to these as “compliance criteria.” The sizing value for iron (10 g d
&1
m
&2
) is in agreement
with the findings of Stark et al. (1990) for a constructed compost wetland in Ohio that receives
marginally acidic water. This rate is still larger, by a factor of 2, than the iron removal rate
reported by Brodie et al. (1991) for aerobic systems in southern Appalachia that are regularly in
compliance.
The manganese removal rate used for compliance (0.5 g d
&1
m
&2
) is based on the
performance of five treatment systems, three of which consistently lower manganese
concentrations to compliance levels. A higher removal value (1 g d
&1
m
&2
), is suggested for
AML sites. Because the toxic effects of manganese at moderate concentrations (< 50 mg/L) are
generally not significant, except in very soft water (Kleinmann and Watzlaf 1988), and the size
of wetland necessary to treat water that contains manganese is so large, AML sites with iron
problems should receive a higher priority than those with only manganese problems.
Net alkaline water contains enough alkalinity to buffer the acidity produced by metal
hydrolysis reactions. The metal contaminants (iron and manganese) will precipitate, given
enough time. The generation of additional alkalinity is unnecessary, so incorporation of
limestone or an organic substrate into the passive treatment system is also unnecessary. The goal
of the treatment system is to aerate the water and promote metal oxidation processes. In many
existing treatment systems where the water is net alkaline, the removal of iron appears to be
limited by dissolved oxygen concentrations and pH. Standard features that can aerate the
drainage, such as waterfalls or riprap ditches, should be followed by quiescent areas. Aeration
only provides enough dissolved oxygen to oxidize about 50 mg/L Fe
2%
. AML with higher
concentrations of Fe
2%
may require a series of aeration structures and wetland basins. The
wetland cells allow time for iron oxidation and hydrolysis to occur, and space in which the iron

58
floc can settle out of suspension. The entire system can be sized based on these iron removal
rates. If manganese removal is desired, base the system’s size on manganese removal rates.
Removal of iron and manganese occurs sequentially in passive systems; if both iron and
manganese removal are necessary, add the two wetland sizes together.
Anoxic Limestone Drains (ALD)
The primary chemical factor believed to limit the utility of an ALD is the presence of
ferric iron (Fe
3%
), aluminum (Al
3%
) and dissolved oxygen. When acidic water containing
any
Fe
3%
or Al
3%
contacts limestone, metal hydroxide particulates (FeOOH or Al(OH)
3
) will form.
No oxygen is necessary. Ferric hydroxide will precipitate on and around limestone, limiting
further dissolution. It has not been determined if precipitation of aluminum hydroxides limit
limestone dissolution. The buildup of both precipitates within the ALD can eventually decrease
the drain permeability and cause plugging. The presence of dissolved oxygen in mine water will
promote the oxidation of ferrous iron to ferric iron, and the precipitation of solids that may limit
limestone dissolution and reduce permeability in the ALD. While the short-term performance of
ALDs that receive water containing elevated levels of Fe
3%
, Al
3%
or DO can be spectacular (total
removal of the metals within the ALD) (Nairn et al. 1991), the long-term performance of these
ALDs is not good. (See
Premature Failure of Two ALDs
on page 43.)
Mine water that contains very low concentrations of DO, Fe
3%
and Al (all < 1 mg/L) is
ideally suited for pretreatment with an ALD. As concentrations of these parameters rise above 1
mg/L, the risk that the ALD will fail prematurely also increases. The length of time an ALD
operates before failing is a function of these contaminant concentrations; the amount of initial
void volume in the ALD, the cross-section of the ALD perpendicular to the flow, and the
dissolution rates of limestone (creation of new void volume).
In some cases, the suitability of using an ALD to treat mine water can be evaluated by the
type of discharge, and field measurements of pH. Net acidic mine waters that seep from spoils
and flooded underground mines and have a field pH ofgreater than 5, characteristically have
concentrations of DO, Fe
3%
, and Al that are all less than 1 mg/L. Such sites are generally
excellent candidates for treatment with an ALD. Mine waters that discharge from open drift
mines or have pH of less than 5 must be analyzed for Fe
3%
and Al. Mine waters with pH of
greater than 5 can contain dissolved Al and Fe
3%
. In northern Appalachia, for example, most
mine drainages with a pH of less than 3 contain significant concentrations of Fe
3%
and Al,
rendering them inappropriate for treatment with an ALD.
The mass of limestone required to neutralize a certain discharge for a specified period of
time (M
t
) can be readily calculated from the mine water flow rate and assumptions about the
alkalinity-generating performance of the ALD (equations 12 and 13). Research indicates that
approximately 15 hours of contact time between mine water and limestone in an ALD is
necessary to achieve a maximum concentration of alkalinity. In order to achieve 15 hours of
contact time within an ALD, 2,800 kg of limestone is required for each L/min of mine water
flow (equation 12). In equation 13, M
c
represents the mass of limestone consumed over a period
of time. For example, an ALD that discharges water with 300 mg/L of alkalinity (the maximum
sustained concentration thus far observed in an ALD effluent), dissolves 1,750 kg of limestone
(90 percent calcium carbonate) in ten years, per each L/min of mine water flow. Equations 12
and 13 must be summed to construct an ALD that contains sufficient limestone (90 percent
calcium carbonate) to ensure a 15-hour retention time throughout a 20-year period. Therefore, a
limestone bed should contain 6,200 kg of limestone for each L/min of flow, which is equivalent

59
to 26 tons of limestone for each gallon per minute of flow. The calculation assumes that the
ALD is constructed with 90 percent CaCO
3
limestone rock that has a porosity of 49 percent. The
calculation also assumes that original mine water does not contain ferric iron or aluminum. The
presence of these ions could result in faster rates of limestone dissolution through the generation
of acidity during hydrolysis. More importantly, they have the potential to limit limestone
dissolution and cause a significant reduction in permeability that could very well lead to failure
(as previously discussed). For a more detailed discussion of limestone dissolution rates, see
Cravotta and Watzlaf (2002).
M
t
= (flow x bulk density
LS
x t
d
) / void ratio
(12)
M
c
= (flow x alkalinity concentration x design lifetime) / CaCO
3
content
(13)
There are still some concerns about the ability of ALDs to maintain unchanneled flow for
a prolonged period of time, how much of the CaCO
3
content of the limestone can be expected to
dissolve, whether the ALDs will collapse after significant dissolution of the limestone, and
whether inputs of DO that are not generally detectable with standard field equipment (0 to
1 mg/L) might eventually result in armoring the limestone with ferric hydroxides. However, the
long-term effectiveness of several of the ALDs discussed here seems to indicate that the above
calculations are valid.
The anoxic limestone drain is just one component of a passive treatment system. When
an ALD operates ideally, its only effect on mine water chemistry is to raise (or keep) pH to (at)
circumneutral levels, and increase concentrations of calcium and alkalinity. Dissolved Fe
2%
and
manganese should be unaffected by flow through the ALD. The ALD must be followed by a
settling basin or wetland system in which metal oxidation, hydrolysis and precipitation can
occur. The type of post-ALD treatment system depends on the acidity of the mine water and the
amount of alkalinity generated by the ALD. If the ALD generates enough alkalinity to transform
the acid mine drainage to a net alkaline condition, then the ALD effluent can be treated with an
aerobic pond and wetland. If possible, the water should be aerated as soon as it exits the ALD
and directed into a settling pond. An aerobic wetland should follow the pond. The total post-
ALD system should be sized according to the criteria provided earlier for net alkaline mine
water. At this time, it appears that mine waters with acidity concentrations less than 150 mg/L
are readily treated with an ALD and aerobic wetland system.
If the mine water is contaminated with only Fe
2%
and manganese, and the acidity exceeds
300 mg/L, it is unlikely that an ALD constructed using current practices will discharge net
alkaline water. When this partially neutralized water is treated aerobically, the iron will
precipitate rapidly, but the absence of sufficient buffering can result in a discharge with low pH.
Building a second ALD to recharge the mine water with additional alkalinity after it flows out
of the aerobic system is currently not feasible because of the high dissolved oxygen content of
water flowing out of aerobic systems. If the treatment goal is to neutralize all of the acidity
passively, then a RAPS should be built to generate additional alkalinity. Such a treatment
system thus contains all three passive technologies. The mine water flows through an ALD, into
an aerobic pond and wetland, and then into a RAPS, followed by another pond and wetland.
If the mine water is contaminated with ferric iron (Fe
3%
) or aluminum, higher
concentrations of acidity can be treated with an ALD than when the water is contaminated with
only Fe
2%
and manganese. This enhanced performance results from a decrease in mineral acidity
due to the hydrolysis and precipitation of Fe
3%
and aluminum within the ALD. These metal-
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

60
removing reactions decrease the mineral acidity of the water. ALDs constructed to treat mine
water contaminated with Fe
3%
and aluminum and having acidity greater than 1,000 mg/L have
discharged net alkaline water. The long-term prognosis for these metal-retaining systems is not
good. However, even if calculations of system longevity (as described above) are inaccurate for
waters contaminated with Fe
3%
and aluminum, their treatment with an ALD maybe cost-effective
in some instances, when compared to chemical alternatives (Skousen and Faulkner 1992).
When a mine water is contaminated with Fe
2%
and manganese and has an acidity between
150 mg/L and 300 mg/L, the ability of an ALD to discharge net alkaline water will depend on
the concentration of alkalinity produced by the limestone system. The amount of alkalinity
generated by a properly constructed and sized ALD depends on the chemical characteristics of
the mine water. An experimental method has been developed that results in an accurate
assessment of the amount of alkalinity being generated when a particular mine water contacts a
particular limestone (Watzlaf and Hedin 1993). The method involves the anoxic incubation of
the mine water in a container (cubitainer) filled with limestone gravel. This cubitainer test may
be used in the design of passive systems, as outlined in Figure 12. The cubitainer test can
determine if the ALD will impart sufficient alkalinity to allow for the ALD effluent to be treated
with ponds and wetlands, or if the water needs additional treatment (RAPS) to add alkalinity. In
experiments at two sites, the concentration of alkalinity that developed in these containers after
48 hours correlated very well with the concentrations of alkalinity measured in the ALD
effluents.
Reducing and Alkalinity-Producing Systems (RAPS)
Based on the results of this study, RAPS were found to remove 40 g d
-1
m
-2
of acidity for
the initial system, and 15 g d
-1
m
-2
for the second RAPS in series. It is important to note that
these values were obtained from systems of similar construction, having compost layers about
0.2 m thick and limestone layers 0.4 to 0.9 m thick. If thinner layers were used, these surface
area-based acidity removal rates may not be applicable. It is reasonable to expect that alkalinity
production will be dependent on influent water quality. Jage et al. (2000) found that alkalinity
production in RAPS significantly correlated with detention time, influent total iron
concentrations, and non-manganese acidity concentrations. Rose and Dietz (2002) found
positive correlations between alkalinity production and influent iron and hydrogen ion
concentrations, and detention time in the compost. They also found acidity removal rates of 25 -
50 g d
-1
m
-2
for the 12 systems that they studied, and suggested using 25 g d
-1
m
-2
as a design
criteria for RAPS. Thomas and Romanek (2002) found alkalinity generation rates averaged 88
gd
-1
m
-2
in pilot scale studies using compost amended with fine-grained limestone (~1.2 mm).
Based on these findings, it is probably prudent to use a sizing criteria of 25 - 30 g d
-1
m
-2
for the
first RAPS in a series, and 15 g d
-1
m
-2
for a subsequent system. It is also recommended that the
limestone layer contain enough limestone to theoretically retain the water for 15 hours
throughout the design life of the system (6,200 kg of limestone per L/min of flow), the same
sizing criteria used for an ALD.
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

61
Constructing Passive Systems
Aerobic Ponds/Wetlands
A typical aerobic wetland is constructed by planting cattail rhizomes in soil or alkaline
spoil obtained onsite. Some systems have been planted simply by spreading cattail seeds, with
good plant growth after two years. The depth of the water in a typical aerobic system is 10 to 50
cm. Ideally, a cell should not be of a uniform depth, but should include shallow and deep marsh
areas and a few deep (1 to 2 m) spots. Note that common rooted aquatic vegetation cannot
tolerate water depths greater than 50 cm, and require shallower depths for propagation.
Typically, a pond is situated before the wetland to remove a majority of the iron
hydroxides. This pond is usually sized for an 8 to 24 hour retention time and is typically 1.5 to
2.5 m deep. To account for accumulations of iron, the value 0.17 g of iron per cm
3
can be used
so that the required detention time will be available for a predetermined time (i.e., its design
life). It is recommended that the freeboard of aerobic wetlands/ponds be constructed at about 1
m for the removal of iron. Observations of sludge accumulation in existing wetlands suggest
that a 1-m freeboard should be adequate to hold 20 to 25 years of FeOOH accumulation. We
have achieved good success when the pond and wetland have similar surface areas. This allows
for future removal of iron oxides from the pond without disturbing the vegetated wetland.
Recently iron oxides have been characterized for potential recycling (e.g., as pigments) (Kairies
et al. 2001, Hedin 2002).
Often, several wetland cells and/or ponds are connected by flow through a v-notch weir,
lined railroad tie steps, or down a ditch. Use of multiple cell/ponds can limit the amount of
short-circuiting, and aerates the water at each connection. If there are elevation differences
between the cells, the interconnection design should dissipate kinetic energy to avoid erosion
and/or the mobilization of precipitates. Spillways should be designed to pass the maximum
probable flow. Spillways should consist of wide cuts in the dike with side slopes no steeper than
2H:1V, be lined with non-biodegradable erosion control fabric and a coarse riprap, if high flows
are expected (Brodie 1991). Proper spillway design can preclude future maintenance costs
associated with erosion and/or failed dikes. If pipes are used, small diameter (< 30 cm) pipes
should be avoided, because they can plug with litter and FeOOH deposits. Pipes should be made
of PVC, PE or coated for long-term stability. More details on the construction of aerobic
wetland systems can be found in Hammer’s
Creating Freshwater Wetlands
, (1992) .
The floor of the wetland cell may be sloped up to a 3 percent grade. If a level cell floor is
used, then the water level and flow are controlled by the downstream dam spillway and/or
adjustable riser pipes.
As discussed previously, some of the aerobic systems that have been constructed to treat
alkaline mine water have little emergent plant growth and are better termed ponds than wetlands.
Metal removal rates in these plantless, aerobic systems appear to be similar to what is observed
in aerobic systems that contain plants. However, plants may provide value not reflected in
measurements of contaminant removal rates. For example, plants can facilitate the filtration of
particulates, prevent flow channelization and provide wildlife benefits that are valued by
regulatory and environmental groups.
ALDs
In an ALD, alkalinity is produced when the acidic water contacts the limestone in an
anoxic, closed environment. Limestone with higher CaCO
3
content (> 80 percent) has been
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

62
shown to dissolve faster than limestone with a higher MgCO
3
or CaMg(CO
3
)
2
content (~50
percent CaCO
3
) (Watzlaf and Hedin 1993). The limestone used in most successful ALDs have
80 to 95 percent CaCO
3
content. Most effective systems have used 5- to 20-cm-sized limestone.
Some systems constructed with fine and small gravel limestone have failed, apparently because
of plugging problems. The ALD must be sealed so that the inputs of atmospheric oxygen are
minimized, and the accumulation of carbon dioxide within the ALD is maximized. This is
usually accomplished by burying the ALD under 1 to 3 m of clay. Plastic is sometimes placed
between the limestone and clay as an additional gas barrier. In some cases, the ALD has been
completely wrapped in plastic before burial (Skousen and Faulkner 1992). This can also help
keep clay and dirt from getting into the pore volume from the bottom and sides of the
excavation. The ALD should be designed to inundate the limestone with water at all times.
Clay dikes within the ALD or riser pipes at the outflow of the ALD will help ensure inundation.
The dimensions of existing ALDs vary considerably. (See Table 7.) Narrower ALDs
have the advantage of minimizing short-circuiting, but present a small cross-section
perpendicular to the flow that may be more prone to clogging. Wider ALDs may be less likely
to suffer significant permeability reductions (clogging) but may allow short circuiting to occur.
In the end, however, site conditions will often dictate the dimensions of the ALD.
RAPS
RAPS are commonly constructed with a 1-m thick layer of limestone. A network of
perforated pipes is placed in the bottom of this limestone layer. On top of the limestone, a layer
of organic matter is placed that is typically 15 to 60 cm thick. Spent mushroom compost, which
is readily available and affordable in and around Pennsylvania, is an often used organic material.
Most spent mushroom compost consists of horse manure, hay, straw, chicken manure and
gypsum. Mine water flows down through the system, encountering the reducing environment of
the compost before contacting the limestone. The compost layer is intended to remove the
dissolved oxygen and convert any ferric iron to the ferrous state to avoid armoring of the
limestone. It is thought that RAPS may be less prone to aluminum plugging than ALDs because
of their larger cross-section (perpendicular to flow paths) and higher available head pressures.
The systems are generally constructed to allow for at least 2 m of head to be utilized, if needed,
to overcome losses in permeability. Alkalinity generation rates for these systems range from 40
to 60 g per day per m
2
of surface area for the first RAPS, and from 15 to 20 g per day per m
2
of
surface area for a second RAPS, when two RAPS are used in series (Watzlaf et al. 2000). Both
iron and aluminum are removed within these systems. Most are periodically flushed to extend
the life of these systems. No guidelines have yet been developed to guide the frequency,
duration, or intensity of the flushes.
A pond should be used to oxidize, precipitate and settle iron before the water enters the
RAPS to minimize the accumulation of iron precipitates (and other settleable solids) on top of
the compost layer in the RAPS. This pond will also serve as an equalization basin. The size of
this pond is site specific, but should be larger at sites where the pH of influent water is above 3.5.
Once the pH drops below ~ 3.0, iron is removed much more slowly from mine drainage.
Operation and Maintenance
Operational problems with passive treatment systems can be attributed to inadequate
design, unrealistic expectations, pests, inadequate construction methods, or natural problems. If
properly designed and constructed, a passive treatment system can be operated with a minimum

 
63
amount of attention and money.
Probably the most common maintenance problem is stability in the dike and spillway.
Reworking slopes, rebuilding spillways, and increasing freeboard can all be avoided by proper
design and construction using existing guidelines for such construction.
Pests can plague wetlands with operational problems. Muskrats will burrow into dikes,
causing leakage and potentially catastrophic failure problems, and can also uproot significant
amounts of cattails and other aquatic vegetation. Muskrats can be discouraged by lining dikes
and slopes with chain link fence or riprap to prevent burrowing (Brodie 1990). Beavers dams
cause water level disruptions and can seriously damage vegetation. They are very difficult to
control once established. Small diameter pipes traversing wide spillways (three-log structure)
and trapping have had limited success in beaver control. Large pipes with 90-degree elbows on
the upstream end have been used as discharge structures in beaver-prone areas (Brodie 1991).
Otherwise, shallow ponds with dikes and shallow slopes toward wide, riprapped spillways may
be the best design to deter beaver populations.
Insects, such as the armyworm, with their appetite for cattails, have devastated
monocultural wetlands (Hedin, et al. 1994a). The use of various plants in a system will
minimize such problems. Mosquitos can breed in wetlands where mine water is alkaline. In
southern Appalachia, mosquito fish (
Gambusia affinis
) have been introduced into alkaline-water
wetlands to control mosquito populations.
Conclusions
Characterization of influent water quality and quantity, including seasonal variation, is
important prior to the selection and development of a passive treatment system (Hyman and
Watzlaf 1995.). The presence or absence of periodic events, such as spring flushes of deposited
metal salts from within the mine area, may influence the selection and sizing of passive systems.
Aerobic ponds and wetlands can be very effective for the removal of iron from net
alkaline mine water. It appears that the original estimate of Hedin et al. (1994a) of 10 to 20 g d
-1
m
-2
remains a convenient pre-construction rule-of-thumb for estimating pond and wetlands
sizes. Recent studies have provided insight into the factors that control the overall processes,
and these approaches may be used to fine-tune sizing criteria. Modeling has concluded that
aeration to sparge carbon dioxide and increase pH can significantly increase iron oxidation rates,
thereby reducing the size of aerobic ponds and wetlands needed for iron removal.
ALDs can effectively treat net acidic mine water. The ideal influent water quality for an
ALD is net acidic water with a pH above 5.0. At this pH, neither ferric iron nor aluminum are
soluble in significant quantities. Intercepted ground water is typically low in dissolved oxygen,
and often contains partial pressures of carbon dioxide higher than atmospheric levels, which
allows for development of alkalinity concentrations greater than 100 mg/L as CaCO
3
. Near
maximum levels of alkalinity (usually between 150 and 300 mg/L) can be achieved with 15
hours or more of contact time. ALDs are tolerant of both ferrous iron and manganese, because
they remain soluble within the ALD. However, the presence of ferric iron, and particularly
aluminum, can reduce permeability of the ALD by precipitation of these metals within the voids
in the limestone. This has been documented in an ALD (Jennings) that received 21 mg/L of
aluminum and clogged within eight months. In the absence of ferric iron and aluminum, ALDs
have continued to perform well with no obvious seasonal variation or long-term reduction in
effectiveness.
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

64
Tracer studies indicated that while ALDs approximate plug-flow systems, some short
circuiting occurs, and dead areas do exist. Calculated detention times, using 49 percent porosity,
were in fairly good agreement with the median detention times of the tracer tests.
Water quality data determine the applicability of an ALD and flow data provide the basis
for sizing an effective ALD for the desired design life. At mine sites where the appropriate
water quality criteria were met and the ALD was sized properly, effective treatment of mine
drainage occurred, provided that the ALD was followed by ponds and/or wetlands for iron
oxidation, precipitation, and settling. At these sites, it is projected that the ALD will be effective
for the designed lifetime of 25 to 30 years and, in some cases, well beyond.
ALDs offer an effective means of introducing alkalinity into net acidic waters that
contain neither ferric iron nor aluminum. The presence of either of these ions will reduce
permeability of the ALD by precipitation, which will cause premature failure by clogging. In the
absence of these ions, ALDs have continued to perform well with no obvious seasonal variation
nor long-term degradation. Near maximum levels of alkalinity (usually between 150 and 300
mg/L) can be achieved with 15 hours or more of contact time. ALDs are tolerant of both ferrous
iron and manganese. ALDs must be viewed as a unit operation, not a standalone remediation
technique, and must be followed by a pond and wetland for iron oxidation, precipitation, and
settling.
Alkaline addition in a RAPS is dominated by the limestone dissolution pathway. The
acid neutralization potential afforded by a RAPS ranges from 35 to over 400 mg/L CaCO
3
.
Sulfate reduction contributed an average of 28 percent (with a range of 5 to 51 percent) of the
total alkalinity produced in the system. The rate of alkaline addition for a single RAPS is about
40 to 60 g d
-1
m
-2
. Rates for the second RAPS in a series fall off to about 1/2 to 1/3 of the rate of
the first system. Much of the variability in performance can be attributed to influent water
quality and detention time. As with ALDs, RAPS should be viewed as unit operations, not
stand-alone technologies. They must be preceded by a pond/wetland to precipitate iron and
other settleable solids. As with ALDs, RAPS must also be followed by a pond and wetland for
iron oxidation, precipitation, and settling.
Care should be taken to obtain sufficient water quality data of the target drainage,
including seasonal variation, before desiging and developing a passive treatment system. Site
and funding constraints may limit the applicability of passive techniques for some mine
drainages. However, for those drainages with appropriate water quality and land availability,
passive treatment systems continue to perform very well.
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

 
65
Abbreviations and Acronyms
ALD
anoxic limestone drain
AML
abandoned mined land
AASHTO
American Association of State Highway and Transportation Officials
atm
atmosphere
Eh
oxidation-reducing potential
Ha
hectare
ICAP-AES
inductively-coupled argon plasma –atomic emission spectroscopy
ISE
ion selective electrode
L
liter
L/min
liters per minute
mL
mililiter
mg/L
miligram per liter
RAPS
reducing and alkalinity-producing systems
SRB
sulfate reducing bacteria
s.u.
standard units
μm
micrometer

 
66
References
American Public Health Association. 1985.
Standard Methods for the Examination of Water
and Wastewater
. Washington, DC: American Public Health Association. 1,268 pp.
American Public Health Association. 1998.
Standard Methods for the Examination of Water
and Wastewater
. Washington, DC: American Public Health Association. 1,162 pp.
Barnes, H. L. and S. B. Romberger. 1968. Chemical Aspects of Acid Mine Drainage.
J. Water
Pollution Control Federation.
40:370-384.
Batty, L. C. and P. L. Younger. 2002. Critical Role of Macrophytes in Achieving Low Iron
Concentrations in Mine Water Treatment Wetlands.
Envir. Sci. Technol
. 36:3,997-4,002.
Beck, J. V. and D. G. Brown. 1968. Direct Sulfur Oxidation in the Solubilization of Sulfide
Ores by Thiobaccillus Ferrooxidans.
J. Bacteriology.
96:1,433-1,434.
Brodie, G. A., Britt, C. R., Tomaszewski, T. M., and H. N. Taylor. 1991. Use of Passive Anoxic
Limestone Drains to Enhance Performance of Acid Drainage Treatment Wetlands.
Proceedings of Reclamation 2000: Technologies for Success
(Durango, CO). pp. 211-222.
Brodie, G. A. 1990. Treatment of Acid Drainage Using Constructed Wetlands: Experiences of
the Tennessee Valley Authority.
Proceedings of the 1990 National Symposium on Mining
(Lexington, KY). pp. 77-83.
Brodie, G. A. 1991. Achieving Compliance with Staged Aerobic Constructed Wetlands to Treat
Acid Drainage.
Proceedings of the 1991 National Meeting of the American Society of
Surface Mining and Reclamation
(Princeton, WV). pp. 151-174.
Brown, M., Barley, B., and H. Wood. 2002.
Minewater Treatment: Technology, Application
and Policy
. Dorchester, United Kingdom: IWA Publishing. 453 pp.
Butler, J. N. 1991.
Carbon Dioxide Equilibria and Their Applications
. Chelsea, MI: Lewis
Publishers. 259 pp.
Cravotta, C. A. III and G. R. Watzlaf. 2002. Design and Performance of Limestone Drains to
Increase pH and Remove Metals from Acidic Mine Drainage. In:
Handbook of
Groundwater Remediation Using Permeable Reactive Barriers.
USA: Elsevier Science. pp.
19-66.
Davies, S. H. and J. J. Morgan. 1989. Manganese (II) Oxidation Kinetics on Metal Oxide
Surfaces.
J. Colloid and Interface Science.
129:63-77.
Dempsey, B. A., Roscoe, H. C., Ames, R., Hedin, R., and B.-H. Jeon. 2001. Ferrous Oxidation
Chemistry in Passive Abiotic Systems for Treatment of Mine Drainage.
Geochemistry:
Exploration, Environment and Analysis.
1(1):81-88.
Diehl, D. and W. Stumm. 1984. Is Dissolved Mn
2%
Being Oxidized by O
2
in Absence of Mn-
Bacteria or Surface Catalysts.
Geochemica et. Cosmochimica Acta.
48:1,571-1,573.
Dietz, J. M. and B. A. Dempsey. 2001. Treatment of Mine Drainage Ising Recirculated Iron
Oxides in a Complete Mix Reactor.
Proceedings of the 2001 Annual Conference of the
National Association of Abandoned Mine Land Programs
(Athens, Ohio). pp. 1-14.

67
Duncan, D., Landesman, J. and C. Walden. 1967. Role of Thiobaccillus Ferrooxidans in the
Oxidation of Sulfide Minerals.
Canadian J. Microbiology.
13:377-403.
Ehrlich, H. L.
Geomicrobiology
. New York: Dekker. 1981, 393 pp.
Fales, H. A. and F. Kenny. 1940.
Inorganic Quantitative Analysis
. New York: D. Appleton
Century Co. 713 pp.
Garrett, W. E., A. A. Bartolucci, Pitt, R. R. and M. E. Vermace. 2002. Recirculating –
Reducing and Alkalinity Producing System (RERAPS) for the Treatment of Acidic Coal Pile
Runoff.
Proceedings of the 2002 National Meeting of the American Society of Mining and
Reclamation
(Lexington, KY). pp. 539 - 557.
Gerber, D. W., Burris, J. E. and R. W. Stone. 1985. Removal of Dissolved Iron and Manganese
Ions from Water by a Sphagnum Moss System.
Proceedings of Wetlands and Water
Management on Mined Lands
(University Park, PA). pp. 365-372.
Girts, M. A., Kleinmann, R. L. P. and P. M. Erickson. 1987. Performance Data on Typha and
Sphagnum Wetlands Constructed to Treat Coal Mine Drainage.
Proceedings of the 8th
Annual Surface Mine Drainage Task Force Symposium
(Morgantown, WV).
Groudev, S. 1979. Mechanism of Bacterial Oxidation of Pyrite.
Mikrobiologija
. (16)75-87.
Hammack, R. W., Dvorak, D. H. and H. M. Edenborn. 1993. The Use of Biogenic Hydrogen
Sulfide to Selectively Recover Metals from a Severely Contaminated Mine Drainage. In:
Proceedings of the International Biohydrometallurgy Symposium
(Jackson Hole, WY).
Hammer, D. A. 1992.
Creating Freshwater Wetlands
. Boca Raton, FL: Lewis Publishers. 298
pp.
Hedin, R. S., Hyman, D. M. and R. W. Hammack. 1988.
Implications of Sulfate Reduction and
Pyrite Formation Processes for Water Quality in a Constructed Wetland: Preliminary
Observations
. U.S. Bureau of Mines IC 9183. pp. 382-388.
Hedin, R. S., Dvorak, D. H., Gustafson, S. L., Hyman, D. M., McIntire, P. E., Nairn, R. W.,
Neupert, R. C., Woods, A. C., and H. M. Edenborn. 1991.
Use of a Constructed Wetland for
the Treatment of Acid Mine Drainage at the Friendship Hill National Historic Site, Fayette
County, PA
. Interagency Agreement #4000-5-0010. 128 pp.
Hedin, R. S., Nairn, R. W. and R. L. P. Kleinmann. 1994a.
Passive Treatment of Coal Mine
Drainage
. U.S. Bureau of Mines IC 9389. 35 pp.
Hedin, R. S. and G. R. Watzlaf. 1994. The Effects of Anoxic Drains on Mine Water Chemistry.
Proceedings of the International Land Reclamation and Mine Drainage Conference and the
Third International Conference of the Abatement of Acidic Drainage
,
Volume 1
(Pittsburgh,
PA). U.S. Bureau of Mines Special Publication SP 06B-94. pp. 185-194.
Hedin, R. S., Watzlaf, G. R. and R. W. Nairn. 1994. Passive Treatment of Acid Mine Drainage
with Limestone.
J. Environmental Quality.
23(6):1,338-1,345.
Hedin, R.S. 2002. Recovery of Marketable Iron Oxide from Mine Drainage.
Proceedings of the
2002 National Meeting of the American Society of Mining and Reclamation
(Lexington,
KY). pp. 517 – 526.

68
Hem, J. D. 1985.
Study and Interpretation of the Chemical Characteristics of Natural Water
,
3rd Edition. U.S. Geological Survey Water Supply Paper 2254. 263 pp.
Huntsman, B. E., Solch, J. G. and M. D. Porter. 1978. Utilization of Sphagnum Species
Dominated Bog for Coal Acid Mine Drainage Abatement.
Abstracts from the Geological
Society of America 91st Annual Meeting
(Toronto, Ontario, Canada). p. 322.
Huntsman, B. E., Kleinmann, R. L. P. and T. O. Tiernan. 1985. Hydrologic and Geochemical
Considerations in Maintaining Man-Made Wetlands Constructed for Acid Mine Drainage
Abatement. Presented at Wetlands and Water Management on Mined Lands (Penn State
University, University Park, PA).
Hyman, D. M. and Watzlaf. 1995. Mine Drainage Characterization for the Successful Design
and Evaluation of Passive Treatment Systems.
Proceedings of the 17
th
Annual Association of
Abandoned Mine Lands Conference
. (French Lick, IN). pp. 170-185.
Jage, C. R., Zipper, C. E., and A. C. Hendricks. 2000. Factors Affecting Performance of
Successive Alkalinity-Producing Systems.
Proceedings of the 2000 National Meeting of the
American Society of Surface Mining and Reclamation
(Tampa, FL). pp. 451-458.
Jennings Water Quality Improvement Coalition. 1999. P
assive Treatment of Acid Mine
Drainage, Vertical Flow System
. Final Report for Project 18 of US EPA FY96 Section 319
Grant.
Kairies, C. L., Watzlaf, G. R., Hedin, R. S., and R. C. Capo. 2001. Characterization and
Resource Recovery Potential of Precipitates Associated with Abandoned Coal Mine
Drainage.
Proceedings of the 2001 National Meeting of the American Society for Surface
Mining and Reclamation
(Albuquerque, NM). pp. 278-279.
Kepler, D. A. and E. C. McCleary. 1994. Successive Alkalinity-Producing Systems (SAPS) for
the Treatment of Acid Mine Drainage. P
roceedings of the International Land Reclamation
and Mine Drainage Conference and the Third International Conference of the Abatement of
Acidic Drainage, Volume 1
(Pittsburgh, PA). U.S. Bureau of Mines Special Publication SP
06B-94. pp. 195-204.
Kepler, D. A. 1995. Successive Alkalinity-Producing Systems (SAPS).
Proceedings of the
West Virginia Surface Mine Drainage Task Force Symposium
(Morgantown, WV).
Kepler, D. A. and E. C. McCleary. 1997. Passive Aluminum Treatment Successes.
Proceedings of the West Virginia Surface Mine Drainage Task Force Symposium
(Morgantown, WV).
Kirby, C. S., and J. A. Elder Brady. 1998. Field Determination of Fe
2+
Oxidation Rates in Acid
Mine Drainage Using a Continuously Stirred Tank Reactor.
Applied Geochemistry.
13:509-
520.
Kirby, C. S., H. M. Thomas, G. Southman and R., Donald. 1999. Relative Contributions of
Abiotic and Biological Factors in Fe(II) Oxidation in Mine Drainage.
Applied Geochemistry.
14:511-530.
Kirby, C. S. 2002. Problems in acidity and alkalinity measurements in mine drainage.
Proceedings of the 2002 National Meeting of the American Society of Mining and
Reclamation
. (Lexington, KY), pp. 1068 - 1071.
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

69
Kleinmann, R. L. P. 1979. The Biogeochemistry of Acid Mine Drainage and a Method to
Control Acid Formation. PhD Thesis, Princeton University. 104 pp
Kleinmann, R. L. P. 1985.
Treatment of Acidic Mine Water by Wetlands
. U.S. Bureau of Mines
IC 9027. pp. 48-52.
Kleinmann, R. L. P., Crerar, D. A., and R. R. Pacelli. 1981. Biochemistry of Acid Mine
Drainage and a Method to Control Acid Formation.
Mining Engineering
. 33:300-306.
Kleinmann, R. L. P. and G. R. Watzlaf. 1988. Should the Effluent Limits for Manganese be
Modified? In:
Proceedings of the Mine Drainage and Surface Mine Reclamation
Conference, Volume II: Mine Reclamation, Abandoned Mine Lands and Policy Issues
(Pittsburgh, PA). U.S. Bureau of Mines IC 9184. pp. 305-310.
Kleinmann, R. L. P., Jones, J. R. and P. M. Erickson. 1988. An Assessment of the Coal Mine
Drainage Problem.
Proceedings of the 10th Annual Conference of the Association of
Abandoned Mine Land Programs
(Wilkes-Barre, PA). pp. 1-9.
Kleinmann, R. L. P., Tiernan, T. O., Solch, J. G. and R. L. Harris. 1983. A Low-Cost, Low-
Maintenance Treatment System for Acid Mine Drainage Ising Sphagnum Moss and
Limestone.
Proceedings of the 1983 Symposium on Surface Mining, Hydrology,
Sedimentology and Reclamation
(University of Kentucky, Lexington, KY). pp. 241-245.
Luther, G. W. III. 1987. Pyrite Oxidation and Reduction: Molecular Orbital Theory
Considerations.
Geochemica et Cosmochimica Acta.
53:3,193-3,200.
McIntire, P. E. and H. M. Edenborn. 1990. The Use of Bacterial Sulfate Reduction in the
Treatment of Drainage from Coal Mines.
Proceedings of the 1990 Mining and Reclamation
Conference and Exhibition
(West Virginia University, Morgantown, WV). pp. 409-415.
McKnight, D. M., B. A. Kimbell and R. L. Runkel. 2001. pH Dependence of Iron
Photoreduction in a Rocky Mountain Stream Affected by Acid Mine Drainage.
Hydrological
Processes.
15:1,979-1,992.
Nairn, R. W., Hedin, R. S. and G. R. Watzlaf. 1992. Generation of Alkalinity in an Anoxic
Limestone Drain.
Achieving Land Use Potential Through Reclamation
. (Proc. Conf. of the
9th National American Society for Surface Mining and Reclamation Meeting, Princeton,
WV). pp. 206-219.
Nairn, R. W., Hedin, R. S. and G. R. Watzlaf. 1991. A Preliminary Review of the Use of
Anoxic Limestone Drains in the Passive Treatment of Acid Mine Drainage.
Proceedings of
the 12th Annual West Virginia Surface Mine Drainage Task Force Symposium
(Morgantown,
WV).
Nawrot, J. 1990. Mine Drainage Treatment through Enhanced Biological Sulfate Reduction
within Wetlands.
Proceedings of the 1990 Mining and Reclamation Conference and
Exhibition
. Vol. II. pp. 353-364.
Nealson, K. H. 1983a. The Microbial Iron Cycle.
Microbial Geochemistry
, W. E. Krumbein,
Ed. Boston, MA: Blackwell Scientific. , pp. 159-190.
Nealson, K H. 1983b. The Microbial Manganese Cycle.
Microbial Geochemistry
. W. E.
Krumbein, Ed. Boston, MA: Blackwell Scientific. pp. 159-190.
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

70
Nordstrom, D. K. and J. W. Ball. 1986. The Geochemical Behavior of Aluminum in Acidified
Surface Waters.
Science.
232:54-56.
Nuttall, C. A, and P. L. Younger. 2000. Zinc Removal from Hard, Circum-Neutral Mine Waters
Using a Novel Closed-Bed Limestone Reactor.
Water Research
. 34:1,262-1,268.
PIRAMID Consortium. 2003.
Engineering Guidelines for the Passive Remediation of Acidic
and/or Metalliferous Mine Drainage and Similar Wastewaters.
Report European
Commission 5th Framework RTD Project No. EVK1-CT-1999-000021. University of
Newcastle Upon Tyne, Newcastle Upon Tyne UK. 166 pp.
Postgate, J. R. 1984.
The Sulphate-Reducing Bacteria
. 2nd d. NY: Cambridge University
Press. 208 pp.
Robbins E.I., Nord, G. L. Jr., Savela, C. E., Eddy, J. L., Livi, K. J., Gullett, C. D., Nordstrom, D.
K., Chou, I. M. and K. M. Briggs. 1996. Microbial and Mineralogical Analysis of
Aluminum-Rich Precipitates that Occlude Porosity in a Failed Anoxic Limestone Drain,
Monongalia County, West Virginia.
Proceedings of the 13th Annual International
Pittsburgh Coal Conf, Volume 2.
(Pittsburgh, PA). p. 761-765
Rose, A. W. and J. M. Dietz. 2002. Case Studies of Passive Treatment Systems: Vertical Flow
Systems.
Proceedings of the 2002 National Meeting of the American Society of Mining and
Reclamation
(Lexington, KY). pp. 776-797.
Rose, A. W., P. J. Shah and B. Means. 2003. Case Studies of Limestone-Bed Passive Systems
for Manganese Removal from Acid Mine Drainage.
Proceedings of the 2003 National
Meeting of the American Society for Mining and Reclamation
(Billings, MT). pp. 1,059-
1,078.
Sencindiver, J. C. and D. K. Bhumbla. 1988. Effects of Cattails (
Typha
) on Metal Removal
from Mine Drainage. 1
988 Mine Drainage and Surface Mine Reclamation, Volume I: Mine
Water and Mine Waste
. U.S. Bureau of Mines IC 9183. pp. 359-369.
Silverman, M. P. 1967. Mechanisms of Bacterial Pyrite Oxidation.
J. Bacteriology
. 94:1,046-
1,051.
Simmons, J., Ziemkiewicz, P. and D. C. Black. 2002. Use of Steel Slag Leach Beds for the
Treatment of Acid Mine Drainage.
Mine Water and the Environment.
21(2):91-99.
Singer, P. C. and W. Stumm. 1970a.
Oxygenation of Ferrous Iron
. Water Pollution Control
Research Series. Federal Water Quality Administration. 14010-06169. 183 pp.
Singer, P.C. and W. Stumm. 1970b. Acidic Mine Drainage: the Rate Determining Step.
Science.
167:1,121-1,123.
Skousen, J. and B. Faulkner. 1992. Preliminary Results of Acid Mine Drainage Treatment with
Anoxic Limestone Drains in West Virginia.
Proceedings of the 13th Annual West Virginia
Surface Mining Drainage Task Force Symposium
(Morgantown, WV).
Spratt, A. K. and R. K. Wieder. 1988. Growth Responses and Iron Uptake in Sphagnum Plants
and their Relations to Acid Mine Drainage (AMD) Treatment.
Mine Drainage and Surface
Mine Reclamation, Volume I: Mine Water and Mine Waste
. U.S. Bureau of Mines IC 9183.
pp. 279-286.

71
Stark, L., E. Stevens, H. Webster and W. Wenerick. 1990. Iron Loading, Efficiency, and Sizing
in a Constructed Wetland Receiving Mine Drainage. Paper in
Proceedings of the 1990
Mining and Reclamation Conference and Exhibition: Volume II
, J. Skousen, J. Sencindiver
and D. Samuel, Eds. pp. 393-401.
Stone, R. W and B. J. Pesavento. 1985. Micro- and Macro-Biological Characteristics of
Wetlands Removing Iron and Manganese.
Proceedings of Wetlands and Water Management
on Mined Lands
(Pennsylvania State University, University Park, PA). p. 373.
Stumm, W. and J. Morgan. 1981.
Aquatic Chemistry
.New York: Wiley-Interscience. 583 pp.
Stumm, W. and J. Morgan. 1996.
Aquatic Chemistry.
New York: Wiley & Sons. 1,022 pp.
Temple, K. G. and E. W. Delchamps. 1953. Autotrophic Bacteria and the Formation of Acid in
Bituminous Coal Mines.
Applied Microbiolog.
, 1:255-258.
Thomas, R. C. and C. S. Romanek. 2002. Acid Rock Drainage in a Vertical Flow Wetland I:
Acidity Neutralization and Alkalinity Generation.
Proceedings of the 2002 National Meeting
of the American Society of Mining and Reclamation
(Lexington, KY).
Turner, D. and D. McCoy. 1990. Anoxic Alkaline Drain Treatment System, a Low Cost Acid
Mine Drainage Treatment Alternative.
Proceedings of the National Symposium on Mining
.
(University of Kentucky, Lexington, KY). pp. 73-75.
United States Environmental Protection Agency. 1983.
Design Manual: Neutralization of Acid
Mine Drainage
. Office of Research and Development, Industrial Environmental Research
Laboratory. 231 pp.
Watzlaf, G. R. 1997. Passive Treatment of Acid Mine Drainage in Down-Flow Limestone
Systems.
Proceedings of the 1997 National Meeting of the American Society for Surface
Mining and Reclamation
(Austin, TX).
Watzlaf, G. R. and R.S. Hedin. 1993. A Method for Predicting the Alkalinity Generated by
Anoxic Limestone Drains.
Proceedings of Fourteenth Annual West Virginia Surface Mine
Drainage Task Force Symposium
(Morgantown, WV).
Watzlaf, G. R. and D. M. Hyman. 1995. Limitations of Passive Systems for the Treatment of
Mine Drainage.
Proceedings of the 17
th
Annual Association of Abandoned Mine Lands
Conference.
(French Lick, IN). p. 186-199.
Watzlaf, G. R., Schroeder, K. T. and C. L. Kairies. 2000. Long-Term Performance of Alkalinity-
Producing Passive Systems for the Treatment of Mine Drainage.
Proceedings of the 2000
National Meeting of the American Society for Surface Mining and Reclamation
(Tampa, FL).
pp. 262-274.
Watzlaf, G. R., Schroeder, K.T ., and C. L. Kairies. 2001.
Modeling of Iron Oxidation in a
Passive Treatment System.
Proceedings of the 2001 National Meeting of the American
Society for Surface Mining and Reclamation
(Albuquerque, NM). pp. 626-638.
Watzlaf, G. R., Kairies, C. L., and K. T. Schroeder. 2002. Quantitative Results from the Flushing
of Four Reducing and Alkalinity-Producing Systems.
Proceedings of the 23
rd
West Virginia
Surface Mine Drainage Task Force Symposium
(Morgantown, WV).
Wieder, R. K. 1989. A Survey of Constructed Wetlands for Acid Coal Mine Drainage Treatment

72
in the Eastern United States.
Wetlands.
9(2):299-315.
Wieder, R. K. 1994. Diel Changes in Iron (III)/Iron (II) in Effluent from Constructed Acid Mine
Drainage Treatment Wetlands.
Journal of Environmental Quality.
23(4):730-738.
Wieder, R. K. and G. E. Lang. 1982. Modification of Acid Mine Drainage in a Freshwater
Wetland.
Proceedings of the Symposium on Wetlands of the Unglaciated Appalachian
Region, West Virginia University, Morgantown, WV
, B. R. McDonald, Ed. pp. 43-53.
Wildeman T., Machener, S., Klusman, R., Cohen, R. and P. Lemke. 1990. Metal Removal
Efficiencies from Acid Mine Drainage in the Big Five Constructed Wetland.
Proceedings of
the 1990 Mining and Reclamation Conference and Exhibition, Vol. II
. pp. 425-433.
Wildeman T. R., Spotts, E., Schafer, W. and J. Gusek. 1994. Characterization, Management and
Treatment of Water for Metal-Mining Operations. Short Course at the 1994 Meeting of the
American Society for Surface Mining and Reclamation (Pittsburgh, PA), 1994.
Younger, P. L., Banwart, S. A., and R. S. Hedin. 2002.
Mine Water: Hydrology, Pollution,
Remediation
. Dordrecht: Kluwer Academic Publishers. 442 pp.
Ziemkiewicz, P. F., J. G. Skousen and J. Simmons. 2003. Long-Term Performance of Passive
Acid Mine Drainage Treatment Systems.
Mine Water and the Environment.
22:118-129.

Exhibit N:
Applications of Passive Treatment to Trace Metals Removal
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

KEVIN L. HOOVER
Sr. Environmental Scientist
TERRY A. RIGHTNOUR
Vice President
EES Division, Gannett Fleming, Inc.
Hyde, Pennsylvania
ROBERT COLLINS
CCB Coordinator
RICHARD HERD
Advisor
Allegheny Power
Greensburg, Pennsylvania
INTRODUCTION
Treatment of wastewater to meet regulatory discharge
standards is a significant cost for the electric utility industry,
and increasing competition is leading many companies to
search for more cost-effective treatment alternatives. One field
showing particular promise is
passive treatment technology
,
or the design of treatment systems based on processes that
cleanse water in the natural environment. These systems
generally have much lower operation and maintenance (O&M)
costs than conventional chemical treatment and can be
successfully applied to a variety of industrial discharges where
conditions are favorable to biological and geochemical
contaminant removal processes.
A major category of contaminants regulated under the National
Pollutant Discharge Elimination System (NPDES) is aqueous
metals, also referred to as
trace metals
. One source of metals-
bearing wastewater of particular importance to coal-fired utilities
is the leachate that can develop from coal combustion by-
product (CCB) sites. While generally non-acidic and free of
hydrocarbons, CCBs can leach trace metals to water passing
through landfills. Parameters that typically exceed compliance
limitations are iron and manganese. Depending on local
regulations, heavy metals can also be of concern.
To date, Allegheny Power (AP) has installed passive wetland
treatment systems to treat metals-bearing leachate at two of its
closed CCB facilities. Work was initiated in 1988 with
construction of a prototype treatment wetland at the Albright
closed CCB landfill in northern West Virginia. With positive
results from this system, in 1994 AP entered into a tailored
collaboration with the Electric Power Research Institute (EPRI)
to advance this cost-saving and environmentally-friendly
technology. This jointly-funded project centered on a full-
scale application of passive treatment at the Springdale closed
CCB landfill in western Pennsylvania and included a major
research and development component to evaluate existing and
experimental technologies for the treatment of CCB leachate.
The research and development team consisted of members
from AP, EPRI, the EES Division of Gannett-Fleming, Inc. (EES)
and the Pennsylvania State University. This paper has been
prepared to summarize the case histories of the Albright and
Springdale projects, and to present the findings of the
associated research programs with regard to future applications
of passive treatment within the utility industry.
HISTORY OF PROJECTS
Albright System
In 1986, the West Virginia Department of Natural Resources
(WVDNR) expressed concern that metals-contaminated leachate
emanating from the Albright closed CCB landfill was impacting
its receiving stream, and indicated that treatment of the
discharge would be necessary. Conventional chemical
treatment options were evaluated by AP, but were found not
to be cost-effective due to the site’s remote location, terrain
constraints and unmanned status. At the time, passive
technology was in its infancy, a promising approach to
wastewater compliance, but with no hard design standards
applicable to the treatment of CCB leachate. In search of a
more cost-effective means of treating these waters, AP retained
the services of EES to investigate the viability of using wetland
treatment for this site. The investigation and subsequent
design led to approval from the WVDNR for construction of
an R&D passive wetland treatment system at Albright.
The initial Albright system consisted of four small basins
formed by dikes in an existing drainageway and vegetated
with transplants from surrounding wetlands. Completed in
1988, this system proved successful in meeting NPDES
limitations for iron, but not for manganese. In the early 1990s,
work by the US Bureau of Mines (US BoM) indicated that
manganese removal rates are much lower than those for iron in
wetland environments, and that removal rates for both
parameters are largely a function of wetland surface area
1
. Two
additional basins were added to the system during 1992 to
provide additional surface area and, thereby, increase
manganese removal capacity. While showing significant
reductions in manganese discharge levels, the expanded system
was still unable to meet compliance for that parameter. In 1993,
pilot level modifications were made to evaluate preliminary
data by others on the ability of limestone beds to remove
manganese
2
. Based on these results and findings from the
Springdale system after its construction, the Albright system
was modified in 1996 to include three rock drains, reaching the
final configuration shown by Figure 1. Following a brief period
of inoculation for the manganese-oxidizing bacteria, almost
total removal was achieved for manganese at Albright, and
Applications of Passive Treatment to Trace Metals Removal
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

that system is now fully in compliance. Each major component
of the Albright system has been continually monitored for
influent and effluent water quality, and flow, for nearly 10 years.
Springdale System
Leachate from the Springdale landfill underdrain had been
discharging since the site was closed in 1975. In 1994, the
Pennsylvania Department of Environmental Protection
(PADEP) indicated that the existing NPDES permit for this
discharge would soon be revised to require more stringent
effluent limits on iron and manganese. Based on the success
at Albright, AP entered into a Consent Order and Agreement
with the PADEP to meet the expanded effluent criteria using
passive wetland treatment.
The new NPDES permit also included future compliance with
a number of other trace metals for which no passive design
standards were available at the time. In response to this need,
the AP/EPRI tailored collaboration project was designed with
dual purposes of: (1) using proven passive wetland
technologies to comply with existing NPDES limits for iron
and manganese and (2) designing and evaluating emerging
and experimental technologies aimed at achieving eventual
compliance with the additional parameters.
At Springdale, insufficient land area was available below the
discharge to construct a system to receive gravity flow,
necessitating a pumping facility to convey the leachate to a
more suitable site uphill. Based on the leachate chemistry, it
was determined that compliance with existing dissolved iron
limitations could be met by use of a simple oxidation/
precipitation basin, which would also equalize the intermittent
flow from the pumps before entering a wetland system. These
facilities were constructed in 1994 and achieved immediate
compliance for dissolved iron. In 1995, eight additional treatment
cells were added to the system in advance of issuance of the
new NPDES permit. These included four vegetated wetland
basins for iron polishing, two rock drains to culture manganese-
oxidizing bacteria, an organic upflow cell to promote sulfide
mineral formation, and an algal growth basin for vegetative
uptake of trace metals. The completed system, shown by
Figure 2, was immediately successful in meeting compliance
for all parameters except boron, which continues to be the
focus of additional efforts by AP to identify an effective passive
treatment mechanism.
Influent and effluent loadings were monitored at ten points
within the system for a period of two years following
construction to evaluate the treatment effectiveness of the
major components and technologies for a broad spectrum of
parameters. Of particular interest was development of design
criteria from the manganese-oxidizing rock drains, which were
later applied to achieve manganese compliance at the Albright
Figure 1. Albright System Layout
Figure 2. Springdale System Layout
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

site. Additional experiments in phytoremediation are
continuing in the on-site research facility, which has both
greenhouse-enclosed and exposed test cells to evaluate the
influence of climate on plant uptake rates.
SELECTED PASSIVE TECHNOLOGIES
The passive technologies employed at Albright and Springdale
have proven very effective for removing trace metals.
Monitoring results are summarized by Table 1, which shows
average influent and effluent parameter concentrations, percent
change, and degree of significant. The following provides a
brief summary of these technologies and guidelines for their
application.
Oxidation/Precipitation Basins
Oxidation/precipitation (O/P) basins are open water
impoundments designed to provide aeration for precipitation
of aqueous metals, detention time to settle precipitates, and
storage volume for accumulating precipitate sludge. They are
most effective for removing large-volume sludge formers and
are a key component in passive systems where iron is present
in quantity. Results from Springdale indicate that arsenic,
aluminum, and zinc will also tend to co-precipitate with iron.
Iron sludge consists primarily of the amorphous oxyhydroxide
limonite
(FeOOH
w
nH
2
O), formed by the process given below.
In the aeration step, oxygen is introduced passively by means
such as a splash plate or corrugated trough. Limonite sludge
forms quickly thereafter, but settles very slowly. A detention
time of at least 24 hours is recommended to produce a clear
water discharge, with additional storage capacity for
accumulated sludge usually maintaining the design detention
time at 40% of the total volume occupied.
2 Fe
2+
+ ½ O
2
+ n H
2
O = FeOOH
w
nH
2
O + 4H
+
O/P basins function best in the circumneutral pH range of 6 to
9 SU. A single passive aeration device can only introduce
enough oxygen to precipitate about 50 mg/L of iron
3
. For
higher loadings, a series of basins and aerators can be
employed. Oxidation of aqueous iron results in the generation
of acidity (H
+
), decreasing the pH of the wastewater. When
significant amounts of iron are being removed, measures may
be necessary to neutralize excess acidity with downstream
components. The rate of iron precipitation also begins to
diminish at a pH below 6 SU, with higher concentrations of
iron becoming stable despite the presence of oxygen.
Vegetated Wetlands
Vegetated wetlands used for treatment are typically constructed
as shallow basins with 1 to 2 feet of organic-rich planting
substrate. For optimum plant development, a substrate meeting
the classification of clay loam with at least 12% organic content
has been found to best duplicate conditions found in natural
wetlands
4
. The substrate is planted with species selected as
appropriate for the local climate. Cattails are generally the
hardiest plants for applications with high metals concentrations
or potential for sludge accumulation
5
. Flow within the basins
is best regulated at a depth of 0.1 foot or less
6
.
Vegetated wetlands function as both physical filters and sites
of biogeochemical activity to alter or fix contaminants in place,
and are effective against a broad spectrum of parameters.
Surface air contact creates an oxygen-rich,
aerobic
environment, which promotes the oxidation and precipitation
of aqueous metals. Below the surface, the organic planting
substrate consumes oxygen, creating an
anaerobic
environment that promotes sulfide mineral formation. Results
from Albright and Springdale show that vegetated wetlands
Table 1. Performance Results for Springdale and
Albright Passive Treatment Systems
Albright System
Springdale System
Parameter
Influent
Effluent
Influent
Effluent
Flow
GPM
20
40
pH
SU
6.50
15%
7.46
7.04
8%
7.61
Acid.
mg/L
4
-79%
1
23
-40%
14
Alk.
mg/L
34
110%
72
106
15%
121
Al
mg/L 0.628
-86%
0.089
0.891
-71%
0.260
Sb
mg/L
**
*
As
mg/L 0.003
-18%
0.002
0.061
-91%
0.005
Ba
mg/L
**
*
Be
mg/L
**
0.0007
-39%
0.0004
B
mg/L
**
15.92
-12%
14.03
Cd
mg/L
**
*
Cr
3+
mg/L
**
*
Cr
6+
mg/L
**
*
Co
mg/L
**
*
Cu
mg/L 0.001
51%
0.002
0.012
11%
0.013
Fe, Tot.
mg/L
2.45
-86%
0.33
12.46
-98%
0.27
Fe, Dis. mg/L
**
6.09
-98%
0.10
Pb
mg/L
0.002
-19%
##
0.001
*
Mn
mg/L
8.52
-87%
1.07
2.71
-92%
0.21
Hg
mg/L 0.0002
5%
##
0.0002
*
Mo
mg/L
**
0.296
-38%
0.183
Ni
mg/L 0.088
-74%
0.023
0.063
-48%
0.033
Se
mg/L 0.004
-48%
0.002
0.002
-15%
0.002
Ag
mg/L
**
0.0024
-85%
0.0004
Sr
mg/L
**
5.17
-8%
4.76
Tl
mg/L
**
*
Sn
mg/L
**
*
Ti
mg/L
**
*
V
mg/L
**
*
Zn
mg/L 0.083
-55%
0.037
0.089
-72%
0.024
SO4
mg/L
870
-17%
723
1334
-3%
1288
SS
mg/L
**
0.2
-46%
0.1
TDS
mg/L
1334
-13%
1164
1818
1%
1828
TSS
mg/L
8
-87%
1
25
-68%
8
† - Significant at 90%, ‡ - Highly Significant at 95%.
* Non-Detect ** Not Sampled
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

are effective for the removal of aluminum, arsenic, copper, iron,
manganese, nickel, and zinc. Other studies indicate vegetated
wetlands to be effective against cadmium
7
, cobalt
8
, and lead
7
,
and those at Springdale show some effect on beryllium and
molybdenum as well. Most other trace metals can be considered
candidates for removal in vegetated wetlands, but confirming
research is sparse. Boron, commonly associated with CCB
leachate, does not show significant removal in vegetated
wetlands. Compliance sizing criteria for vegetated wetlands
are available from the US BoM
1
for iron, manganese, and acidity
based on surface area, as follows:
Iron
10 grams/(meter
2
- day)
Manganese
0.5 grams/(meter
2
- day)
Acidity
3.5 grams/(meter
2
- day)
These values are additive, so a vegetated wetland should be
designed with sufficient area to remove each contaminant
separately. Preliminary findings from Albright indicate that
these criteria may not be sufficient for treatment of iron and
manganese to levels approaching 1 mg/L
5
. Vegetated wetlands
are limited in their capacity to accommodate large volumes of
iron sludge and should be placed after an O/P basin to limit
iron loading. Their biological processes will also diminish below
a pH of 4 SU. Periodic maintenance is necessary to eliminate
flow path short circuits, remove accumulated sludge, and
replace spent substrates. Control of internal flow velocities is
important for avoiding short-circuits or particle transport. As
a general rule, a minimum substrate surface width of 1 foot is
recommended for each gallon per minute of influent flow.
Manganese-Oxidizing Rock Drains
“Rock drains” are basins filled with loose stone or gravel that
provide substrates for the growth of bacteria which oxidize
aqueous manganese (Mn
2+
) as energy for their life processes.
These bacteria combine manganese and oxygen to form the
mineral
pyrolusite
(MnO
2
), the “black slime” coating commonly
found on river rocks. Manganese will not normally precipitate
below a pH of 9.5 SU
9
in chemical treatment, but in the presence
of bacteria it can be effectively removed in waters with a pH as
low as 6 SU and possibly as low as 5 SU
10
. The basic chemical
reaction for this can be summarized as follows:
Mn
2+
+ H
2
O + ½ O
2
= MnO
2
+ 2H
+
Detailed design criteria have not been published for rock drains.
However, both the Albright and Springdale systems show good
performance with basins having a total detention volume of
approximately 48 hours. The bacteria grow only on the surface
of the stones, so treatment efficiency is believed to also be a
function of stone surface area. Rock diameters of 1 to 6 inches
appear to produce a good ratio of surface growth area to void
space. Water levels within the basins are generally maintained
near the surface of the bed, and bacterial growth can occur
throughout the water column in the substrate. Multiple basins
with intermediate cascade aeration points have been found to
introduce the oxygen necessary for the bacterial activity.
Manganese-oxidizing bacteria are generally ubiquitous in the
environment and will normally colonize a completed rock drain
by natural growth within several months of construction.
Rock drains can be very effective against aqueous manganese,
showing almost total removal under ideal conditions. They do
not appear to function well with an influent iron concentration
of greater than 1 mg/L
11
, but the Albright application does
achieve very low iron and manganese discharge concentrations
with an average influent iron of 1.2 mg/L. When treating
wastewater containing both iron and manganese, O/P basins
and/or vegetated wetlands should be employed to remove iron
upstream of a rock drain. At Springdale, the rock drains show
some associative reduction of boron, molybdenum, and
strontium, while those at Albright show significant reductions
in aluminum, arsenic, copper, and nickel at low concentrations.
Organic Reduction Environments
A second bacterially-mediated process with potential for
removal of trace metals is sulfate reduction. Anaerobic bacteria
decompose organic matter in the presence of sulfates to
generate sulfide, a powerful reducing agent. Sulfide is capable
of joining with most aqueous metals to form sulfide minerals,
with M
2+
representing the metal in the following:
M
2+
+ 4CH
2
O + 2SO
4
2-
=
MS
2
+ H
2
+ 2HCO
3
-
+ 2CO
2
+ 2H
2
O
Organic reduction environments can be created in many forms.
One type used for acidity removal is a Sustained Alkalinity
Producing System (SAPS), which functions by downflow of
water through a layer of compost followed by a layer of
limestone. Horizontal migration of water through organic-rich
planting substrates will also result in sulfide generation in
vegetated wetlands. For the Springdale project, an experimental
cell was constructed using upflow through limestone and
compost. Although sulfide was produced in abundance by
this method, there were insufficient aqueous metals remaining
at that point in the system for any significant removal to occur.
In fact, some influent metals concentrations were so low that
additional amounts were leached from the compost. It is
concluded that this method of treatment would be more
effective against high concentrations of trace metals, and may
not be able to achieve extremely low effluent concentrations.
Phytoremediation
Growing plants must take in nutrients and minerals, including
small quantities of trace metals, from their surroundings to
produce new tissue. Once incorporated in plant tissue, trace
metals tend to be less mobile and are essentially removed from
the environment until the plant decays, or possibly longer.
A treatment method known as
phytoremediation
uses this

basic life process as a tool for removing contaminants from
wastewater. Plants do not uptake trace metals as a sufficient
percentage of their body mass to make this form of treatment
practical for high-concentration parameters, such as iron and
manganese. Even if a plant accumulates 1% of its mass in a
given metal, that still generates 100 pounds of plant matter for
every pound of metal removed. Instead, research is focused
on identifying
hyperaccumulators
, those plants that can store
exceptionally large amounts of trace metals in their tissues
without ill effect. These plants may be a practical treatment
method for removing low concentrations of trace metals, and it
is suspected that at least some of the trace metal removal
occurring at Albright and Springdale is a result of this process.
Research is also focusing on the emerging field of
transmigratory phytoremediation
, where plants modify a
contaminant to a benign form and pass it back to the
environment, rather than accumulating it in their tissues. This
eliminates the potential problem of disposing of large volumes
of plant matter. EPRI-supported research is being conducted
in conjunction with the Springdale project to examine plant
species that can volatilize selenium, continuously removing
that contaminant out of wastewater and releasing it to the
atmosphere as an innocuous methyl compound
12
.
Phased Element Removal Technology
One of the most important developments to come from the AP
research has been the recognition that each wastewater
contaminant has a preferred environment of removal. Passive
systems treating for multiple parameters may require more than
one internal treatment method, necessitating some form of
ordering protocol. To aid in the design of multi-environment
passive systems, EES has developed a set of guidelines known
as Phased Element Removal Technology (PERT
TM
)
13
, the
tenants of which are as follows:
w
Generally target contaminants in decreasing order of
concentration, as the parameter with the greatest loading
often controls the treatment efficiency of lesser
constituents.
w
Sequence treatment environments in order of increasing
sensitivity to chemical or physical loading.
w
Eliminate high-volume sludge formers as early as possible
in the system and provide sufficient storage volume for
the accumulated sludge.
w
Use narrow, elongated treatment cells to increase the
potential for separation of individual removal processes
within multiple-parameter treatment environments.
w
Identify limiting reagents and provide mechanisms for their
introduction.
w
Size components for flow capacity as well as chemical
loading capacity to avoid hydraulic overloads and
transport of incompatible contaminants to sensitive
downstream components.
w
Maximize influent contact with the effective treatment
substrate through close hydraulic control to prevent flow
path short-circuits.
w
Allow for ready access to treatment components and for
system maintenance, adjustment, and repair.
ECONOMIC ANALYSIS
An extensive cost analysis has been performed for the
Springdale passive treatment system
14
, and the methodology
later applied to the Albright system
5
. Comparisons were made
to four chemical treatment alternatives based on capital
construction costs and the present values of projected O&M
costs. Findings are summarized by Table 2, which compares
the results for the two passive systems to the least expensive
chemical alternative, a caustic soda drip-feed system, and the
most expensive, a hydrated lime silo system. The wastewaters
treated at Albright and Springdale are equivalent in chemical
loading, and similar construction methods would be required
for a chemical alternative on both sites, so this comparison is
reasonably accurate.
The Albright system is seen to have a significantly lower capital
construction cost than that of the chemical alternatives, while
the Springdale system has a comparable cost. The Springdale
system includes a number of experimental components that
Table 2. Summary of Estimated Costs for Passive and Chemical Treatment Alternatives
Passive Treatment
Chemical Treatment
Albright
Springdale
Low Cost
High Cost
System
System
Drip-Feed
Lime Silo
Capital Construction Cost
$231,965
$701,742
$619,740
$743,980
10 Year O&M Present Value
$115,290
$67,094
$95,344
$148,313
10 Year Total Present Value
$347,255
$768,836
$715,084
$892,293

elevate its cost compared to a strictly compliance application.
The largest capital cost factor for both forms of treatment is
basin construction. The relative requirements for basin
construction between passive and chemical alternatives would
be approximately equal for other wastewater applications, so
similar capital cost performance can be expected on other sites.
It is noted that passive systems may require a larger land surface
area to construct than chemical alternatives in some cases,
and for this reason may not be suited to applications where
construction space is severely limited. The opposite can also
be true, as the Albright system achieved compliance on a site
where a chemical alternative would be extremely difficult to
construct. Construction space evaluations and cost estimates
should be prepared from conceptual design layouts prior to
committing to a given treatment alternative.
In terms of future O&M costs, the Albright system is
comparable to the chemical alternatives within a 10 year
projection, while the Springdale system is considerably lower.
The reduced O&M cost for Springdale reflects less frequent
replacement of its treatment substrates, which are protected
from sludge accumulation by the equalization basin. Such a
basin was not possible in the construction area of the Albright
site. Both passive systems have lower operator supervision
time, mechanical maintenance, and consumable chemical costs.
Passive systems are self-regulating and require only cursory
operator supervision, as opposed to chemical systems, which
can require frequent or continuous operator presence.
Additional savings are realized by eliminating the costs of
chemical storage, reporting, and safety training. Longer term
projections of O&M costs indicate that both passive systems
represent the least expensive alternative as the costs for capital
replacement of mechanical chemical system components
become a consideration.
CONCLUSIONS
Passive treatment has proven to be a reliable and cost-effective
alternative to chemical treatment for the Albright and Springdale
CCB sites. Results from both projects have led to significant
advances in the understanding of passive removal processes
and the development of improved design standards. The
technologies employed are readily adaptable to other metals-
bearing wastewaters found within the utility industry, provided
attention is given to the individual limitations of each treatment
method. The cost savings observed for the AP projects are
inherent in the nature of passive treatment, and similar savings
can be expected for future applications where conditions are
appropriate to its use.
As a result of these experiences, passive wetland treatment is
now a major component of Allegheny Power’s Environmental
Management System for CCB facilities.
REFERENCES
1. Hedin, R. S. and Nairn, R. W., “Designing and Sizing
Passive Mine Drainage Treatment Systems,” Proceedings
of Thirteenth Annual West Virginia Surface Mine Drainage
Task Force Symposium, Morgantown, WV, April 8-9, 1992.
2. Eddy, David P., “Treatment of Acid Mine Water Using a
Sequential Cell Simulated Wetland: A Mesocosm
Experiment,” Thesis for The Pennsylvania State University
Intercollege Graduate Degree Program, 1994.
3. Hedin, R. S., Nairn, R. W., and Kleinmann, R. L. P., “Passive
Treatment of Coal Mine Drainage,” Bureau of Mines
Information Circular 9389, U.S. BoM, Pittsburgh, PA, 1994.
4. Bishel-Machung, L., Brooks, R. P., Yates, S. S., and Hoover,
K. L., “Soil Properties of Reference Wetlands and Wetland
Creation Projects in Pennsylvania,” Wetlands 16(4): 532-
541, 1996.
5. Rightnour, T. A., and Hoover, K. L., “Albright Wetland
Treatment System Final Report,” AP Internal Document,
1998.
6. Stark, L. R., Brooks, R. P., Webster, H. J., Unz, R. F., and
Ulrich, M., “Treatment of Mine Drainage by a Constructed
Multi-Cell Wetland: The Corsica Project.” Report Number
ER9411, Pennsylvania State University, Environmental
Resources Research Institute, University Park, PA, 1994.
7. US EPA, “Emerging Technology Summary: Handbook for
Constructed Wetlands Receiving Acid Mine Drainage,”
EPA/540/SR-93/523, 1993.
8. Eger, P., Melchert, G., Antonson, D., and Wagner, J., “The
Use of Wetland Treatment to Remove Trace Metals from
Mine Drainage,” Constructed Wetlands for Water Quality
Treatment, Lewis Publishers, Ann Arbor, MI, 1993.
9. Humenick, Michael J., Jr., Water and Wastewater
Treatment, Marcel Dekker, Inc., New York, NY, 1977.
10. Gordon, J. A., and Burr, J. L., “Treatment of Manganese
from Mining Seep Using Packed Columns,” Journal of
Environmental Engineering
,
115(2): 386-394, 1989.
11. Hedin, R. S., and Nairn, R. W., “Contaminant Removal
Capabilities of Wetlands Constructed to Treat Coal Mine
Drainage” Constructed Wetlands for Water Quality
Improvement, Lewis Publishers, Ann Arbor, MI, p. 187-
195, 1993.
12. Zayed, A. and Terry, N., “Selenium Volatilization in Roots
and Shoots: Effects of Shoot Removal and Sulfate Level,”
Journal of Plant Physiology, 143:8-14, 1994.
13. Rightnour, T. A., and Hoover, K. L., “An Application of
Phased Element Removal Technology for Passive
Treatment of Fly Ash Leachate Contaminants,”
Environmental Professional Journal, Vol. 19, Special Issue
No. 1, 201-208, 1997.
14. Rightnour, T. A., and Hoover, K. L., “An Analysis of Cost
Savings for a Constructed Wetland Treatment System,”
Electric Power Research Institute Interim Report, Draft,
July, 1997.
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Exhibit O:
Rapid Manganese Removal from Mine Waters Using an Aerated
Packed-Bed Bioreactor

Reproduced from Journal of Environmental Quality. Published by ASA, CSSA, and SSSA. All copyrights reserved.
Rapid Manganese Removal from Mine Waters Using an Aerated Packed-Bed Bioreactor
Karen L. Johnson* and Paul L. Younger
ABSTRACT
garded as a two-step process: in the first step, Mn
2
?
In the UK, the Environmental Quality Standard for manganese
is sorbed onto the manganese oxide or oxyhydroxide
has recently been lowered to 30
?
g/L (annual average), which is less
surface with concomitant partial oxidation of Mn
2
?
to
than the UK Drinking Water Inspectorate’s Maximum Permitted Con-
Mn
3
?
; in the second step, the disproportionation of Mn
3
?
centration Value (50
?
g/L). Current passive treatment systems for
to Mn
4
?
occurs (Morgan and Stumm, 1964). The rate
manganese removal operate as open-air gravel-bed filters, designed
of abiotic Mn oxidation has been summarized by Mor-
to maximize either influent light and/or dissolved oxygen. This re-
gan and Stumm (1964), who found it to be dependent
quires large areas of land. A novel enhanced bioremediation treatment
on both the concentration of Mn
2
?
ions and the quantity
asystempassivelyfor aermanganeseated subsurfaceremovalgravelhas bed.beenTdevelopedhe provisionthatof airconsistsat depthof
of Mn oxide present (Eq. [1]):
and the use of catalytic substrates help overcome the slow kinetics
?
[Mn
II
]
?
t
?
k
0
[Mn
II
]
?
k
1
[Mn
II
][MnO
2
]
usually associated with manganese oxidation. With a residence time
[1]
of only 8 h and an influent manganese concentration of approximately
20 mg/L,
?
95% of the manganese was removed. The treatment system
where
k
0
?
4
?
10
12
M
?
3
[O
2
.Aq][OH
?
]
2
and
k
1
?
also operates successfully at temperatures as low as 4
?
C and in total
10
18
[O
2
.Aq][OH
?
]
2
.
darkness. These observations have positive implications for manga-
There are numerous strains of bacteria that can in-
nese treatment using this technique in both colder climates and where
crease Mn oxidation rates by as much as five orders of
large areas of land are unavailable. Furthermore, as the operation of
magnitude (Nealson, 1983), the rate being dependent
oxyhydroxide,this
passive treatmentwhich
issystema powerfulcontinuallysorbentgeneratesfor most
freshpollutantmanganesemetals,
on which species of bacteria is involved (e.g., Zhang
it potentially has major ancillary benefits as a removal process for
et al., 2002). The degree of influence that bacteria have
other metals, such as zinc.
on Mn oxidation has proved difficult to determine, as
the majority of metabolic inhibitors that are used to
prevent biotic activity in control experiments also influ-
M
anganese is a common contaminant in many
ence Mn oxidation rates (Shiller, 2004). Despite these
mine waters and though not as ecotoxic as other
problems, it is widely assumed that Mn oxidation at
metals found in such waters (such as Fe, Al, and Zn),
circumneutral pH is biologically catalyzed (e.g., Zhang
it nevertheless has various undesirable properties, in-
et al., 2002).
cluding a propensity for precipitating in water distri-
There are many other catalysts for Mn oxidation, al-
bution pipe networks (eventually causing blockage of
though none are as effective as Mn oxides. Many authors
supply pipes), imparting an unpleasant “metallic” taste
have investigated the catalytic effects of clay minerals
to drinking water, and staining laundry. In the UK, the
on Mn oxidation (Blume and Schwertmann, 1969; Reddy
Environmental Quality Standard (EQS) for manganese
and Perkins, 1976; Wilson, 1980; Yavuz et al., 2003).
has recently been lowered to 30
?
g/L dissolved Mn (an-
Reddy and Perkins (1976) found that under alternate
nual average) to comply with European Directives. This
wetting and drying conditions, illitic clay was capable
new EQS is actually lower than both the UK Drinking
of fixing significant quantities of Mn either by physical
Water Inspectorate’s Maximum Permitted Concentration
entrapment or precipitation. Potter and Rossman (1979)
Value and the USEPA’s secondary maximum contami-
proposed that clay minerals (illite and montmorillonite)
nant level, both of which are 50
?
g/L in drinking water.
are necessary for the formation of certain Mn precipi-
For this reason, there has been renewed interest in cost-
tates, such as desert varnish.
effective Mn removal technologies.
Junta and Hochella (1994) characterized the role that
Manganese is notoriously difficult to remove using
mineral surfaces play in the heterogeneous oxidation of
(CreraractivationdoeseithernotactiveandreadilyenergyBarnes,or
passiveoccurrequired1974).withouttreatmentManganesefor
Mneitherbecauseoxidehighlyoxideprecipitationofformationoxidizingthe
high
appearscomposition,throughnificantMn
2
?
. Theyrolethatadsorptionshoweditinofiscontrollingthetheontoimmediategeometricthat“steps”thetheoxidationcharacter,ratesurfaceonofmineraloxidationthatofmoreMnplayssurfaces.
2
?
thanduringbeginsasig-theIt
and/or high pH (above pH 9) conditions (Sikora et al.,
the early stages of the reaction. After the initial oxida-
2000). Manganese oxidation is autocatalytic and is re-
tion of adsorbed Mn
2
?
at the mineral surface, the newly
formed site becomes the most reactive site for continua-
K.L. Johnson, School of Engineering, University of Durham, Durham,
tion of the adsorption–oxidation process.
DH1 3LE, UK. P.L. Younger, IRES, University of Newcastle, Newcas-
tle on Tyne, NE1 7RU, UK. Received 5 Aug. 2004. Technical Reports.
PASSIVE TREATMENT
*Corresponding author (karen.johnson@durham.ac.uk).
Passive treatment utilizes naturally available energy sources
Published in J. Environ. Qual. 34:987–993 (2005).
such as topographical gradient and microbial metabolic energy
doi:10.2134/jeq2004.0300
©
ASA, CSSA, SSSA
Abbreviations:
SEM, scanning electron microscopy; XRD, X-ray dif-
677 S. Segoe Rd., Madison, WI 53711 USA
fraction.
987
Published online May 11, 2005

Reproduced from Journal of Environmental Quality. Published by ASA, CSSA, and SSSA. All copyrights reserved.
988
J. ENVIRON. QUAL., VOL. 34, MAY–JUNE 2005
to treat contaminated water and requires regular but infre-
influent pipe near the bottom and an effluent pipe near the
quent maintenance to operate successfully over its design life
top on the opposite side (Fig. 1). This arrangement of flow was
(Younger et al., 2002). Because Mn oxidation is difficult in the
designed to limit the development of preferential flow-paths.
presence of ferrous iron (Nairn and Hedin, 1993), Mn removal
Three containers (A, B, and C) were filled to a depth of
systems are usually placed at the end of the treatment process
1 cm with bentonite, which was saturated with deionized water.
stream, so that they receive waters from which the majority
A 1-mm layer of MnO
2
powder was added to the hydrated
of the iron has already been removed (Fe
?
1 mg/L). Effective
bentonite surface. Finally the container was filled with clean
Mn removal passive treatment systems are essentially a form
single-size 20-mm-diameter dolomite clasts from the Raisby
of bioremediation, typically consisting of oxic “rock filters,”
quarry (Raisby formation) in northeastern England (UK Na-
hosting algal and/or bacterial consortia that create high-pH
tional Grid Reference NZ 311 337). A fourth reactor was set
microniches within which the precipitation of Mn oxyhydrox-
up as a “control,” containing only relatively inert quartzite gravel
ides and oxides occurs (e.g., Phillips et al., 1995; Hamilton et al.,
in place of the dolomite and MnO
2
powder (though still with
1999). For the algae in such systems to photosynthesize effec-
bentonite as in the other reactors). Mine water was pumped
tively, unobstructed daylight and low influent turbidities are
into and out of the system using two separate peristaltic pumps.
necessary. They are therefore subject to marked seasonal and
Aeration of the substrate was provided using an aeration
diurnal variations in performance efficiency.
pump. In the field, aeration of the substrate would be provided
In the United States, the “Pyrolusite Process” (Vail and
using a passive aeration system that uses modest heads of
Riley, 2000), a patented bioremediation method, has been
water to produce blasts of compressed air.
successfully used to treat manganiferous waters. In this system,
A key part of this work was to understand the role of aer-
a limestone gravel reactor is inoculated with manganese-oxi-
ation in enhancing Mn oxidation at various stages of biofilm
dizing bacteria, which are chosen site specifically. However,
development and under different environmental conditions.
recent work (Rose et al., 2003) suggests that special inocula-
Dissolved oxygen levels in the influent and effluent water were
tion may not be necessary.
measured on a weekly basis using a YSI (Yellow Springs, OH)
All of these existing systems require relatively large areas
Model 95 meter, which was calibrated using air as a standard
of land as they must be shallow to ensure sufficient infiltration
for 100% dissolved oxygen. However, dissolved oxygen in the
of light and/or dissolved oxygen. The new enhanced bioreme-
bioreactors themselves was not monitored. Reported dissolved
diation method discussed in this paper takes the form of a sub-
oxygen concentrations are accurate to
?
2%. The pH was mea-
surface flow gravel bed (Johnson, 2003a). The provision of air
sured on a weekly basis using a Camlab (Cambridge, UK) MY/6P
at depth is achieved using a passive aeration system and the
Ultrameter. Readings are accurate to
?
0.01 pH unit.
use of a catalytic substrate helps overcome the slow kinetics
The reactors were operated with and without aeration both
that are usually associated with Mn oxidation and, most impor-
at room temperature and at 4
?
C at different stages of biofilm
tantly, allows deeper systems to be built where large areas of
development. It should be noted that the quartzite reactor
land are not available.
was not operated without aeration during the start-up phase
or operated at 4
?
C due to the fact that only one control reactor
MATERIALS AND METHODS
could be accommodated. When the dolomite reactor was oper-
Static batch experiments (250 mL) were performed to iden-
ated at 4
?
C, the light was left switched on in the cold room
tify suitable substrates and conditions for Mn removal. The
to distinguish between temperature and light effects and sub-
results (Johnson, 2003b) directed us toward a dolomite sub-
sequently (once it was demonstrated that light was not a con-
strate with a bentonite and MnO
2
(Sigma-Aldrich, St. Louis,
trolling factor in the manganese removal process) the reactor
MO) basal layer, as this combination proved to be the most
was operated in complete darkness for the remaining period.
effective at promoting Mn oxidation.
Substrate and Mine Water Characterization
Experimental Design
Net-alkaline mine water from a recently closed (December
Continuous flow experiments were set up and operated
1998) fluorite mine (Frazer’s Grove Mine in the North Pen-
at room temperature and in natural light conditions (unless
nines, United Kingdom; Johnson and Younger, 2002) was used
otherwise stated) for a total of 277 d. The design of each re-
in the laboratory experiments. The authors felt it was impor-
actor consisted of a 5-L rectangular plastic container with an
tant to use real mine water as it is preferable to use in situ
Fig. 1. Sketch diagram (not to scale) of the small-scale continuous flow experiments.
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Reproduced from Journal of Environmental Quality. Published by ASA, CSSA, and SSSA. All copyrights reserved.
JOHNSON & YOUNGER: RAPID MANGANESE REMOVAL FROM MINE WATERS
989
microbial communities (Johnson, 2003), which are more repre-
RESULTS
sentative of real ecosystems than synthetic systems. The water
had a pH of 7. The geochemistry of the mine water varied with
Dissolved oxygen levels in both the influent and efflu-
time due to ground water rebound in the area of the mine [see
ent waters were always
?
95% (
?
2%); this is likely to
Johnson and Younger (2002) for more detailed geochemical
be due to the fact that both influent and effluent vessels
analysis]. During the experimental period, we measured bicar-
were open to the atmosphere. Effluent pH was between
bonate alkalinity values between 120 to 160 mg/L as CaCO
3
8.00 and 8.20 in all reactors for the duration of the
equivalent, and 15 to 30 mg/L Mn, 5 to 10 mg/L Fe, and 5 to
experimental period.
10 mg/L Zn. The water was stored for 72 h before use to allow
The results for the aerated dolomite and quartzite
the iron to precipitate out to ensure that ferrous iron could not
reactors can be categorized into two phases: an initial
interfere with the Mn oxidation process. Metal concentrations
“start-up” period lasting approximately 2 mo, during
were measured after this storage period using a Unicam (Cam-
which 60% of the influent Mn and 85% of the influent
bridge, UK) 929 atomic absorption spectrophotometer. Ana-
Zn were removed in the dolomite reactors and 40% of
lytical precision was
?
0.1 mg/L. Effluent water samples were
the influent Mn and 83% of the influent Zn were removed
taken from the reactors on a daily basis, filtered (using 0.45-
?
m
in the quartzite reactor; and a second “established” part
filters), and acidified. There was no significant difference in
metal concentrations between filtered and nonfiltered samples
of the experiment, when a black biofilm became evident
except for samples taken from the control reactor during the
on the substrate surfaces, during which 97% of the influ-
period of reaeration.
ent Mn and 93% of the influent Zn were removed in
Flow rates were measured using a graduated container and
the dolomite reactors and 90% of the influent Mn and
stop-watch, and nominal residence times were calculated ac-
90% of the influent Zn were removed in the quartzite
cordingly, taking into account the porosity of the bentonite
reactor.
and dolomite system (determined to be approximately 50%).
The addition of the disinfectant Virkon to Reactor B
Flows were generally adjusted to ensure a residence time of
on Day 148 resulted in the breakdown of the black man-
around 8 h (approximately 5 mL/min flow rate) as the initial
ganese oxyhydroxide biofilm, which was partly washed
batch experiments had indicated that the majority of Mn and
away. The effluent remained acidic and effervescent for
Zn would be removed in this time (Johnson, unpublished data).
several weeks afterward. Percentage Mn removal de-
All percentage removals quoted for Mn and Zn are for an
creased to approximately 29% and percentage Zn re-
8-h residence time.
moval to approximately 66% after the disinfectant had
No microbiological analyses were performed on the sub-
been added. There was no change in either Mn or Zn
workstratestooorktheplace.waterHowevdue
toer,resourceReactor
Blimitationswas poisonedat
the(ustimeing thethe
percentage removal with or without aeration.
disinfectant Virkon [DuPont, Wilmington, DE]) on Day 148 to
Figure 2 shows the effect of aeration on Mn removal
gain some understanding of the degree of abiotic versus abiotic
in the dolomite reactors during the start-up phase. When
manganese oxidation. At the end of the experimental period,
aeration was removed from Reactor B, effluent Mn
biofilm-covered dolomite clasts were removed from Reactor
concentration increased and the reactor subsequently
C and compared with fresh dolomite using X-ray diffraction
took longer to become “established.” It is interesting to
(XRD), scanning electron microscopy (SEM), and elemen-
note that the nature of the start-up phase is very differ-
tal analysis.
ent in the dolomite and quartzite reactors. The dolomite
Fig. 2. Effluent manganese concentration in the quartzite (control) and dolomite reactors (Reactor B and average of Reactors A and C) showing
the effect of aeration during the start-up phase.
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Reproduced from Journal of Environmental Quality. Published by ASA, CSSA, and SSSA. All copyrights reserved.
990
J. ENVIRON. QUAL., VOL. 34, MAY–JUNE 2005
Fig. 3. Effluent manganese concentration in the quartzite (control) and dolomite reactors (average of Reactors B and C) showing the effect of
aeration during the established phase.
reactors all removed approximately 60% of influent Mn
quartzite reactors is most clear during the reintroduction
during the start-up phase and there was a step increase
of aeration. Figure 4 shows that when aeration was rein-
to approximately 97% Mn removal during the estab-
troduced, Mn effluent concentrations returned to previ-
lished phase. In comparison, Mn removal in the quartz-
ous (aerated) levels in the dolomite reactor whereas
ite reactor increased gradually during the start-up phase
they did not return to previous (aerated) levels in the
until it reached approximately 90% removal in the es-
quartzite reactor. It was also noted that Mn oxyhydrox-
tablished phase.
ide deposits were dislodged during the reaeration pro-
Figure 3 shows the effect of aeration on Mn removal
cess in the quartzite reactor whereas the attachment
in the dolomite and quartzite reactors during the estab-
was undisturbed in the dolomite reactors.
lished phase. When aeration was removed from the do-
Figure 5 shows both Mn and Zn removal performance
lomite (B and C) and the quartzite (control) reactors,
in the dolomite reactor (A) at 4
?
C. It is clear that Mn
effluent Mn concentration increased more in the quartz-
and Zn removal is inhibited at 4
?
C without aeration.
ite reactor than in the dolomite reactors. The difference
There is no explanation for the precipitous drop in efflu-
in Mn removal performance between the dolomite and
ent Mn concentrations around Day 100 but it does dem-
Fig. 4. Effluent manganese concentration in the quartzite (control) and dolomite reactor (Reactor C) showing the effect of the reintroduction
of aeration in the established phase.
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Reproduced from Journal of Environmental Quality. Published by ASA, CSSA, and SSSA. All copyrights reserved.
JOHNSON & YOUNGER: RAPID MANGANESE REMOVAL FROM MINE WATERS
991
Fig. 5. Effluent manganese and zinc concentrations in the dolomite reactor (Reactor A) showing the effect of aeration during operation at 4
?
C.
onstrate the large variability in the data. With aeration,
the dolomite underneath. However, Tipping et al. (1984)
at 4
?
C, Mn and Zn removal returned to previous (unaer-
found Ca to be abundant in the Mn oxyhydroxides pre-
ated at 20
?
C) levels. When aeration was moved from
cipitated from natural lake waters in northwestern En-
the substrate (as is shown in Fig. 1) to the influent water
gland. Aluminium and silica are also present in the anal-
reservoir there was a decrease in Mn removal equivalent
ysis of the black deposit suggesting that a clay mineral
to there being no aeration (data not shown on graph).
(possibly bentonite) may be associated with the black
Figure 6A shows that there were some Mn deposits
deposit.
present on the surface of the dolomite before its use in
the reactors. It is clear from Fig. 6B that after use in the
reactors the Mn deposit is quite extensive in its cover-
DISCUSSION
age. The XRD peak count was very low due to the largely
All of the reactors took approximately 8 wk to become
amorphous nature of the deposit but four of the five great-
“established.” This appears to be a common “start-up”
est peaks characteristic of nsutite (
?
MnOOH) were pres-
period for microbial communities engaged in Fe and Mn
ent in the diffractogram (d-spacings: 4.00
x
, 1.64
x
, 2.42
7
,
oxidation (Bourgine et al., 1994). The percent metal re-
and 2.13
5
) and it is likely that this is the only crystalline
moved was greater during the established phase than
phase present.
during the start-up phase in both the dolomite and the
Elemental analysis suggests that the black precipitate
quartzite reactors. The start of the established phase co-
is an oxide with a Mn to Zn relative abundance ratio
incided with the development of a black precipitate (bio-
of 3:1, which is consistent with the molar ratios of Mn to
film) on the substrate surface. Although no attempt was
Zn removed in the continuous flow experiments. There
made to identify the species of manganese-oxidizing bac-
were also significant quantities of calcium present in the
teria present, the addition of the disinfectant Virkon re-
precipitate although it is not clear whether the Ca is part
sulted in a dramatic decrease in percentage Mn and Zn
of the black precipitate or if we are “seeing” the Ca from
removed. The decrease to approximately 29% Mn re-
moval after the addition of disinfectant may give an in-
dication of the amount of manganese removal that was
taking place without microorganisms. However, the facts
that dead cells may still provide more sorption sites than
the inorganic substrate alone and the lower pH environ-
ment created by the addition of Virkon may have affected
the percentage Mn removed. Therefore, 29% may not
be an accurate estimate of abiotic percentage Mn re-
moved but it does indicate that Mn removal was at least
in part a biotic process. It is interesting to note that in the
acidic environment, Zn removal is relatively higher than
Fig.and6. Secondaryafter
(B) useelectronin the continuousimage
of theflowdolomiteexperimentsbothshowingbefore (A)Mn
Mn removal in comparison with percentage removals
deposit (light areas) on the dolomite (dark areas). The scale bar
under circumneutral conditions. This may be due to pref-
shows 100
?
m.
erential adsorption of Zn over Mn onto manganese oxy-

Reproduced from Journal of Environmental Quality. Published by ASA, CSSA, and SSSA. All copyrights reserved.
992
J. ENVIRON. QUAL., VOL. 34, MAY–JUNE 2005
hydroxides at low pH (whereas at circumneutral pH Mn
second phase of aeration. This suggests that the dolo-
is preferentially adsorbed) (Nicholson and Eley, 1997).
mite substrate provides a better surface for the attach-
The black precipitate recovered from the surface of the
ment of Mn oxyhydroxide precipitates than the quartz-
dolomite was nsutite,
?
MnOOH, which was identified us-
ite gravel. The reasons for the superiority of the dolomite
ing XRD. The smaller peaks for nsutite were not present
surface over the quartzite surface could be due to the
but this is likely to be due to the amorphous nature of
physical characteristics of the substrates such as surface
much of the precipitate; Giovanoli (1980) specifically
roughness (SEM photographs of the dolomite and quartz-
points out the difficulty in definitively identifying nsut-
ite surfaces show that the dolomite surface is much rougher
ite using XRD. Elemental analysis using SEM showed
than the quartzite surface; Johnson 2003b) or geochemi-
that there was Zn present, presumably as a sorbed phase,
cal properties. The authors have started a new research
on the surface of the nsutite. The relative abundance of
project to address this question and are currently run-
Mn to Zn of was 3:1, which was also the approximate
ning bioreactors in triplicate using the pure minerals
molar ratio of Mn to Zn in the influent water for the
calcite, dolomite, magnesite, and quartzite. In this work,
majority of the time. There is a positive correlation
environmental scanning electron microscopy (E-SEM)
(
R
2
?
0.74) between percent Mn removed and percent
will be used to assess the spatial distribution and compo-
Zn removed in the dolomite reactors. However, there
sition of the manganese removing biofilms and the asso-
is a statistically insignificant relationship (
R
2
?
0.20)
ciated manganese oxyhydroxides that accumulated on
between the Mn and Zn removed in the quartzite reac-
the substrate surface. Biofilm community development
tor. This is probably due to the greater variability in
and composition will be monitored over time using de-
both Mn and Zn removal in the quartzite reactor.
naturing gradient gel electrophoresis (DGGE). Pilot-
During the start-up phase (with aeration), approx-
scale reactors have also been established in the field at
imately 40 and 60% of Mn were removed in the quartzite
three different sites in northeastern England.
and dolomite reactors, respectively. During the estab-
Since approximately 0.15 to 0.30 mg/L of dissolved oxy-
lished phase (with aeration), the difference between the
gen are required to oxidize 1 mg/L Mn
2
?
, either partially
two types of reactor was less noticeable with approx-
to Mn
3
?
or fully to Mn
4
?
(Sikora et al., 2000), there is
imately 90 and 97% of Mn removed in the quartzite and
more than sufficient oxygen present in fully saturated
dolomite reactors, respectively. However, the variation
waters (which typically contain approximately 10 mg/L
in percent Mn removed in both the start-up and the es-
dissolved oxygen) to oxidize the approximately 20 mg/L
tablished phases was greater in the quartzite reactor than
of dissolved Mn
2
?
in the influent water. Since we know
in the dolomite reactors, suggesting that the dolomite/
that the influent water was nearly saturated in dissolved
MnO
2
combination provides a better substrate surface
oxygen, this suggests that it was not the extra oxygen pro-
for sustaining high percentage manganese removal.
vided by the aeration that increased Mn removal but
The effects of aeration were examined during both
the actual aeration process. This hypothesis is supported
the start-up and established phases in the dolomite re-
by the fact that when the influent water reservoir was
actor and it is clear that aeration is important at both
aerated directly (rather than as shown in Fig. 1) while
stages. The dolomite reactors removed approximately
the reactor was in the cold room, there was a subsequent
60% of the influent Mn in the start-up phase with aera-
decrease in percentage Mn removed equivalent to no
tion compared to approximately 22% without aera-
aeration being present. The generation of bubbles at the
tion. During the established phase, the difference was
substrate surface probably increased the potential mass
less noticeable in the dolomite reactor with approx-
transfer of oxygen (Baylar and Emiroglu, 2004) to the
imately 97% of Mn removed with aeration and approx-
reactive surfaces, thereby making oxygen more readily
imately 91% Mn removed without aeration. The quartz-
available to take part in the oxidation process and sup-
ite gravel reactor was more dependent on aeration to
port other essential microbiological support systems.
sustain high percentage Mn removal. In the established
The importance of this aeration was highlighted when
phase, approximately 90% of the Mn was removed with
the reactors were exposed to stressful environmental
aeration in the quartzite gravel reactor, dropping to
conditions such as low temperatures. Without aeration,
approximately 60% Mn removal without aeration. This
Mn removal in the dolomite reactor fell dramatically
suggests that during the established phase, maintaining
from approximately 97% to approximately 38%. When
high percentage manganese removal in the dolomite
the reactor was reaerated, manganese removal increased
reactor was less dependent on aeration than the quartz-
back to approximately 97% within 24 h. It seems plausi-
ite reactor.
ble that the group of microorganisms that were success-
The differences in Mn removal performance between
fully aiding Mn removal without aeration at room tem-
the quartzite and dolomite reactors are also noticeable
perature were not able to continue doing so at 4
?
C; with
during the reaeration phase. With the reintroduction of
their subsequent drop in activity and the reintroduction
aeration, the quartzite reactor only recovers to approx-
of aeration, another group of microorganisms were able
imately 67% Mn removal whereas the dolomite reactors
to take advantage of the environment and enhance Mn
recover completely to approximately 97% Mn removal.
removal once again.
It was noted that the black precipitate that had coated
Metal removal rates quoted for wetland-type passive
the quartzite gravel was significantly dislodged by the
treatment systems are usually quoted in units of g/m
2
/d
reintroduction of aeration, whereas the precipitate on
(e.g., Hedin et al., 1994), where the m
2
term refers to
the dolomite substrate remained attached during the
land area. This allows one to compare the land require-
Electronic Filing - Received, Clerk's Office, July 2, 2009
* * * * * PCB 2010-003 * * * * *

Reproduced from Journal of Environmental Quality. Published by ASA, CSSA, and SSSA. All copyrights reserved.
JOHNSON & YOUNGER: RAPID MANGANESE REMOVAL FROM MINE WATERS
993
ments of various treatment options. However, this is
Johnson, K.L. 2003a. The importance of aeration in passive treatment
clearly not appropriate for a volume-based treatment
schemestion
11(2):205–212.for
manganese removal. Land Contamination Reclama-
system. Calculations were performed to determine the
Johnson, K.L. 2003b. Oxidative passive treatment to remove Mn
2
?
amount of Mn removed in the small-scale continuous
from mine waters: What is the best substrate? p. 85–97.
In
8th Int.
flow experiments both in g/m
3
/d and in g/m
2
/d (assuming
Mine Water Association Congress, Johannesburg, South Africa.
a subsurface treatment system depth of 1 m). With an
IMWA, Johannesburg.
8-h residence time, the Mn removal rate was calculated
Johnson,chemicalK.L.,consequencesand P.L. Younger.of
the abandonment2002.
Hydrogeologicalof
Frazer’s Groveand
geo-car-
as 60 g/m
2
/d, which is an order of magnitude greater than
bonate hosted Pb/Zn fluorspar mine, North Pennines, UK.
those quoted by Nairn and Hedin (1993) and demon-
p. 347–364.
In
P.L. Younger and N.S. Robins (ed.) Mine water
strates the ability of the treatment system to overcome
hydrogeology and geochemistry. Spec. Publ. 198. Geol. Soc., London.
the slow oxidation kinetics usually associated with Mn
Junta,mineralJ.L.,surfaces:and M.F.A
microscopicHochella.
1994.and
spectroscopicManganese
(II)study.oxidationGeochim.at
oxidation. This has major implications for the treatment
Cosmochim. Acta 58:4985–4999.
of manganiferous waters in areas where there is little
Morgan, J.J., and W. Stumm. 1964. Colloid-chemical properties of
land available. Furthermore, as the operation of this pas-
manganese dioxide. J. Colloid Sci. 19:347–359.
sive treatment system continually generates fresh Mn
Nairn, B., and R.S. Hedin. 1993. Contaminant removal capabilities
oxyhydroxide, which is a powerful sorbent for most pol-
of
In
G.A.wetlandsMoshiriconstructed(ed.) Constructedto
treat coalwetlandsmine drainage.for
water
p.quality187–195.im-
lutant metals (Jenne, 1968), it potentially has major an-
provement. Lewis Publ., Boca Raton, FL.
cillary benefits as a removal process for other metals
Nealson, K.H. 1983. The microbial manganese cycle.
In
W.E. Krum-
such as zinc.
bein (ed.) Microbial geochemistry. Blackwell Sci. Publ., Oxford.
Nicholson, K., and M. Eley. 1997. Geochemistry of manganese oxides:
Metal adsorption in freshwater and marine environments. p. 309–326.
ACKNOWLEDGMENTS
In
K. Nicholson, J.R. Hein, B. Buhn, and S. Dasgupta (ed.) Manga-
This work was completed as part of a Ph.D. thesis that was
nese mineralisation: Geochemistry and mineralogy of terrestrial
funded by the Engineering and Physical Sciences Research
and marine deposits. Spec. Publ. 119. Geol. Soc., London.
Council (Grant no. 98316317). The authors would also like to
Phillips,1995.
ManganeseP.,
J. Bender,removalR. Simms,from S.acidRodriguez-Eaton,coal-mine
drainageandbyC.a
Britt.pond
thank Mr. Paul Allison of Durham Industrial Minerals Ltd.
containing green algae and microbial mat. Water Sci. Technol.
for supplying the dolomite that was used in the experiments.
31(12):161–170.
In addition, thanks must go to Professor David Manning at New-
Potter, R.M., and G.R. Rossman. 1979. Mineralogy of manganese
castle University for his useful comments on the manuscript.
dendrites and coatings. Am. Mineral. 64:1219–1226.
Reddy, M.R., and H.F. Perkins. 1976. Fixation of manganese by clay
REFERENCES
minerals. Soil Sci. 121:21–24.
Rose, A.W., P.J. Shah, and B. Means. 2003. Case studies of limestone-
Baylar, A., and M.E. Emiroglu. 2004. An experimental study of air
bed passive systems for manganese removal from acid mine drain-
entrainment and oxygen transfer at a water jet from a nozzle with
age. p. 1059–1078.
In
Proc. of the 20th Conf. for the American
air holes. Water Environ. Res. 76(3):231–237.
Society of Mining and Reclamation, Billings, MT. ASMR, Lexing-
Blume, H.P., and U. Schwertmann. 1969. Genetic evaluation of profile
ton, KY.
distribution of Al, Fe and Mn oxides. Soil Sci. Soc. Am. Proc.
Shiller, A.M. 2004. Temperature dependence of microbial Mn oxida-
33:438–444.
tion as a control on seasonal concentration variability of dissolved
Bourgine, F.P., M. Gennery, J.I. Chapman, H. Kerai, J.G. Green, R.J.
metals in flood-plain rivers.
In
Proc. of the 227th National American
Rap, S. Ellis, C. Gaumard, C. Stretton, J. Chandler, K. Kaur, C.
Chemical Society Spring Meeting, Anaheim, CA. 28 Mar.–1 Apr. 2004
Harwood, A.A. Redhead, P. Howsam, J. Beechey, D. Osbourne,
[Online]. Available at http://membership.acs.org/g/geoc/Anaheim_
and B. Giles. 1994. Biological processes at Saints Hill water-treat-
2004_Abstracts.pdf (verified 31 Jan. 2005). Am. Chem. Soc., Wash-
ment plant, Kent—Discussion. J. Inst. Water Environ. Manage.
ington, DC.
8(4):379–392.
Sikora, F.J., L.L. Behrends, G.A. Brodie, and H.N. Taylor. 2000.
Crerar, D.A., and H.L. Barnes. 1974. Deposition of deep-sea manga-
Design criteria and required chemistry for removing manganese in
nese nodules. Geochim. Cosmochim. Acta 38:270–300.
acid mine drainage using subsurface flow wetlands. Water Environ.
Giovanoli, R. 1980. On natural and synthetic manganese nodules.
Res. 72(5):536–544.
p. 159–202.
In
I.M. Varentsov (ed.) Geology and geochemistry of
Tipping, E., D.W. Thompson, and W. Davison. 1984. Oxidation prod-
manganese. 2nd Int. Symp., Sydney, Australia. Akademiai Kiado,
ucts of Mn(II) in lake waters. Chem. Geol. 44:359–383.
Stuttgart, Germany.
Vail, W.J., and R.K. Riley. 2000. The Pyrolusite Process: A bio-
Hamilton, Q.U.I., H.M. Lamb, C. Hallett, and J.A. Proctor. 1999.
remediation process for the abatements of acid mine drainage.
Passive treatment systems for the remediation of acid mine drain-
Green Lands 30(4):40–46.
age at Wheal Jane, Cornwall. Water Environ. Manage. 13:93–103.
Wilson, D.E. 1980. Surface and complexation effects on the rate of
Hedin, R.S., R.W. Nairn, and R.L.P. Kleinmann. 1994. Passive treat-
Mn (II) oxidation in natural waters. Geochim. Cosmochim. Acta
ment of polluted coal mine drainage. Bureau of Mines Info. Circ.
44:1311–1317.
9389. United States Dep. of Interior, Washington, DC.
Yavuz, O., Y. Altunkaynak, and F. Guzel. 2003. Removal of copper,
Jenne, E.A. 1968. Controls on Mn, Fe, Co, Ni, Cu and Zn concentra-
nickel, cobalt and manganese from aqueous solution by kaolinite.
tions in soils and water: The significant role of hydrous Mn and
Water Res. 37:948–952.
Fe oxides. p. 337–387.
In
Trace inorganics in water. Advanced
Younger, P.L., S.A. Banwart, and R.S. Hedin. 2002. Mine water:
Chemistry Series 73. Am. Chem. Soc., Washington, DC.
Hydrology, pollution, remediation. Kluwer, Amsterdam.
Johnson, D.B. 2003. Chemical and microbiological characteristics of
Zhang, J., L.W. Lion, Y.M. Nelson, M.L. Shuler, and W.C. Ghiorse.
mineral spoils and drainage waters at abandoned coal and metal
2002. Kinetics of Mn(II) oxidation by Leptothrix discophora SS1.
mines. Water Air Soil Pollut. 3:47–66.
Geochim. Cosmochim. Acta 65(5):773–781.

Back to top