PUBLICATION 460-133

Powell RiveR PRoject

Reclamation Guidelines foR suRface mined land

Passive Treatment of Acid-Mine Drainage

Authored by C. Zipper, Extension Specialist, Crop and Soil Environmental Sciences, Virginia Tech;
J. Skousen, Extension Specialist, Land Reclamation, West Virginia University, and
C. Jage, former graduate student, Crop and Soil Environmental Sciences, Virginia Tech

Introduction mining areas, the most common of these minerals is

iron pyrite (FeS2). Using FeS2 as an example, AMD for-
Acidic mine drainage (AMD; also called “acid rock mation can be represented as follows:
drainage” or “acid drainage”) is an environmental pol-
lutant that impairs water resources in mining regions FeS2(s) + 3.5 O2 + H2O → Fe+2 + 2 SO4-2 + H+ (eq. 1)
throughout the world. Where such treatment is required
legally, treatment must be efficient and continual. Fe+2 + 0.25 O2 + H+ → Fe+3 + 0.5 H2O (eq. 2)
Treatment methods are commonly divided into either
“active,” meaning reliance on the addition of alkaline Fe+3 + 2 H2O → FeOOH (s) + 3 H+ (eq. 3)
chemicals to neutralize the acidity, or “passive.” The
term “passive treatment” means reliance on biologi-
cal, geochemical, and gravitational processes. Passive
treatment does not require constant care or the chemi-
cal reagents that characterize “active” AMD treatment.

This publication presents guidance for design of pas-

sive treatment systems for AMD. Our emphasis is to
describe clearly the mechanisms governing these sys-
tems’ treatment effectiveness and performance. Parties
intending to construct passive treatment systems may
refer to other sources that include more detailed design
and construction guidance, including those listed as
“Design Guidelines” in the references below. Articles
by Skousen and Ziemkiewicz (2005), and Skousen
(1996), contain photographs that illustrate many of
these concepts.
Figure 1. Acid drainage from an abandoned coal mine has
impacted this southwestern Virginia stream. The orange
Acid Mine Drainage coloration is due to the iron (Fe) that has been mobilized
by the sulfur oxidation process, which causes acid mine
Acid drainage occurs when minerals containing reduced drainage to occur. When the acid-forming minerals do not
forms of sulfur (S) oxidize via exposure to oxygen and contain significant quantities of iron, the acidic waters do
water during earth disturbances (Figure 1). In coal- not have the orange coloration pictured here.

www.ext.vt.edu
Produced by Virginia Cooperative Extension, Virginia Tech, 2018
Virginia Cooperative Extension programs and employment are open to all, regardless of age, color, disability, gender, gender identity, gender expression, national origin, political affiliation, race, religion, sexual orientation, genetic informa-
tion, veteran status, or any other basis protected by law. An equal opportunity/affirmative action employer. Issued in furtherance of Cooperative Extension work, Virginia Polytechnic Institute and State University, Virginia State University,
and the U.S. Department of Agriculture cooperating. Edwin J. Jones, Director, Virginia Cooperative Extension, Virginia Tech, Blacksburg; M. Ray McKinnie, Administrator, 1890 Extension Program, Virginia State University, Petersburg.

VT/0218/460-133 (CSES-216P)
The process is initiated with pyrite oxidation and release acid-drainage treatment, the water’s acidity is an indi-
of ferrous iron (Fe+2), sulfate (SO4 2-), and hydrogen (H+) cator of the amount of treatment that must be applied to
(eq. 1). The sulfur-oxidation process is accelerated by neutralize both the H+ present in solution (the solution
the presence of Thiobacillus and Ferroplasma bacteria. pH), and the H+ that will be generated as soluble met-
Ferrous iron undergoes oxidation forming ferric iron als oxidize and/or hydrolyze, and precipitate. Acidity is
(Fe+3) (eq. 2). Finally, Fe+3 is hydrolyzed (reacts with usually expressed on a CaCO3-equivalent basis, recog-
H2O) to form ferric hydroxide (FeOOH), an insoluble nizing that 50 mg of CaCO3 can neutralize 1 mg of H+.
orange-colored precipitant, and release additional acid-
ity (eq. 3). The rate of FeOOH formation is pH-depen- Fe in solution is a source of acidity because the chemi-
dent; it occurs rapidly when pH > 4. cal reactions that cause it to change from liquid to
solid phase (to precipitate) and removed from solution
While most of the acid drainage in the Appalachian release H+ (see equation 3). Similarly, soluble Al also
region occurs with pyrites associated with coals, many contributes to acidity because it hydrolyzes as OH- ions
metals other than Fe occur as reduced-S mineral forms. are added and pH is raised; Al eventually precipitates
Many ores that are mined and processed to produce met- as a solid-phase chemical form, releasing H+ in the
als occur as reduced-S mineral forms (e.g., CdS, CuS, process:
HgS, NiS, PbS, ZnS); therefore, acid mine drainage is
associated with both coal and non-coal mining, and the Al +3 + 3 H2O → Al(OH)3 (s) + 3 H+ (eq. 4)
FeS2 oxidation process described above is illustrative
of a more general process of reduced-S-mineral oxida- Another common source of acidity in mine drainage
tion which releases SO42- , H+, and associated metals. is Mn. There are several pathways by which Mn can
be neutralized, all of which generate H+, including that
Acid drainage can have dramatic impact on water represented below which occurs when O2 is present:
quality. When large quantities of reduced-S mineral
are exposed to air and water, large quantities of H+ are Mn2+ + 0.25 O2 + 1.5 H2O → MnOOH (s) + 2H+ (eq. 5)
released causing very low water pH – sometimes < 3.
Like the FeOOH formation reaction described above
Generally, the extreme acidity mobilizes (produces
(eq. 3), Mn precipitation reaction rates are influenced
soluble forms which are carried by drainage waters)
by pH and occur more rapidly as pH is raised, but Mn
metals that are released from the sulfide minerals that
precipitation generally occurs more slowly and requires
oxidize and from associated minerals. This mobiliza-
higher solution pH’s than other metals. The presence of
tion occurs because a number of metals – including Al,
acid-soluble metals such as Fe, Al, and Mn also tends
Cu, Fe, Hg, Ni, Pb, Zn – become more soluble in water
to hinder, or buffer, changes in pH that would otherwise
as pH declines and are said to be “acid soluble.” Acid
occur in response to OH- additions because their pre-
drainage also produces waters that contain high con-
cipitation generates additional H+.
centrations of dissolved mineral salts (“total dissolved
solids,” or TDS), including the sulfates produced by If a solution’s chemistry is known, acidity can be esti-
mineral oxidation, mobilized acid-soluble metals, and mated using an equation that considers the molecular
other mineral components that become soluble due to weight of each acid-soluble metal in solution at an
mineral dissolution caused by the extreme acidity. In environmentally significant concentration, the number
sum: acid drainage is an acidic, sulfate rich, high TDS, of protons (H+) to be generated per mole as those met-
metal-bearing solution that harms stream ecology. als precipitate, and the solution pH. A commonly used
form of that equation is as follows:
Acidity and Alkalinity A = 50*[ 2*(Fe+2/56) + 3*(Fe+3/56) +
In this bulletin, passive system performance measures 2*(Mn/55) + 3*(Al/27) + 1000*(10-pH)] (eq. 6)
are expressed as removal of acidity from the treat-
ment waters. Acidity is a measure a solution’s capacity In this equation, A = Acidity, expressed as mg/L
to neutralize additions of an alkaline reagent, such as CaCO3 equivalent (50 is the initial multiplication fac-
NaOH. One way to determine acidity is to add NaOH tor because 50 mg of CaCO3 can neutralize 1 mg of
to the solution in an amount sufficient to raise pH to H+); and Fe+2, Fe+3, Mn, and Al are concentrations in
8.3; the amount of acid-neutralization (OH-) added is solution, as mg/L. When other acid-soluble metals are
converted to an equivalent acidity concentration. In present in known concentrations (Zn, Cu, Ni, etc.),

www.ext.vt.edu
2
similar terms are added to the equation to estimate the Aerobic Wetlands
acidity contributions by these other metals. The acidity
Aerobic wetlands are the simplest type of passive treat-
calculated using equation 6 is an estimate of the true
ment system but are limited in the types of waters they
acidity, because acid-soluble metals can react along
can treat effectively. Aerobic wetlands are used to treat
several reaction paths, depending on solution chem-
mildly acidic or net-alkaline waters containing elevated
istry. However, several studies (Watzlaf et al. 2004,
Fe concentrations. They have limited capacity to neu-
Kirby and Cravotta 2005) demonstrate that calculated
tralize acidity. These systems’ primary function is to
acidities are generally close to measured acidities for
allow aeration to the mine waters flowing among the
acid drainage solutions collected from the Appalachian
vegetation, allowing dissolved Fe to oxidize, and to
coalfields, and that pH alone is generally a poor estima-
provide residence time where the water is slowed for
tor for acidity.
Fe oxide products to precipitate. Because Fe precipita-
A solution’s alkalinity is measured by titrating with a tion generates H+, the water leaving the aerobic wetland
strong acid, usually H2SO4, to pH of 4.5. Thus, a solu- may be lower in pH than that entering, even if Fe con-
tion with pH > 4.5 but < 7 can have both alkalinity and centrations are less.
acidity. In this bulletin, the term “alkalinity” is used
Where influent waters are net-alkaline and Fe is not in
generally to describe a solution’s potential to neutralize
solution at significant concentrations, aerobic wetlands
acidity. The net acidity of a solution can be determined
are also capable of removing manganese (Mn). How-
by subtracting measured alkalinity from measured or
ever, Mn-removal effectiveness is limited by several
calculated acidity; and the net alkalinity can be deter-
factors. Mn removal occurs via mechanisms similar to
mined by subtracting acidity from alkalinity. See Kirby
the removal of Fe, but more slowly. Dissolved Mn can
and Cravotta (2005) for a thorough review of acidity
oxidize to solid-phase forms, such as MnO2 (manga-
and alkalinity. Throughout this bulletin, acidity and
nese dioxide), but the process is very slow when pH < 8
alkalinity concentrations are expressed on a CaCO3-
and, like Fe oxidation, it generates acidity. As with Fe,
equivalent basis.
Mn oxidation occurs both chemically and microbially.
However, Fe oxidation is a preferential process so Mn
Passive Treatment of AMD oxidation does not occur as a significant process until
Fe oxidation is nearly complete. For aerobic wetlands
Passive treatment systems for acid drainage are intended to remove Mn successfully, large areas or additional
to renovate and improve the quality of waters that pass wetland cells are required.
through them. These systems are modeled after wet-
lands and other natural processes, with modifications A typical aerobic wetland system is a shallow, surface-
directed toward meeting specific treatment goals. Early flow wetland planted with cattails (Typha sp.) (Figure
research included investigations of natural wetlands 2). The depression that holds the wetland may or may
that were receiving AMD (Weider and Lang 1982) that not be lined with a synthetic or clay barrier. Depend-
raised pH and reduced Fe concentrations without vis- ing on landscape conditions, the lining can be intended
ible deterioration.

A critical step in designing passive treatment is to char-

acterize the waters to be treated. This can be done by
measuring the discharge or flow of those waters and
the concentrations of water-quality constituents of con-
cern over an extended period–ideally at least a year—to
determine how these quantities vary seasonally. Based Figure 2. Simplified cross-sectional view of an aerobic
on this information plus knowledge of the system’s pur- wetland. These systems can treat mildly acidic,
pose (whether it will stand alone or be used in combina- circumneutral, or alkaline waters that contain dissolved
tion with another treatment process, or is intended for metals such as Fe (iron) which can be transformed to the
regulatory compliance), the elemental concentrations solid phase via oxidation. Aquatic plants such as cattails
translocate O2 to the subsurface through their roots,
and flow volume to be treated are determined – these
which aids metal oxidation. The aquatic plants also help to
are the “design conditions.” Site characteristics, espe- prevent channelization of the waters flowing through the
cially land availability, also influence passive treatment wetland, slowing water velocities and aiding solid-phase
system selection and design. metal removal via sedimentation.

www.ext.vt.edu
3
to either to keep treatment waters from draining Table 1. Suggested design criteria for aerobic
out through the depression’s base or to prevent wetlands (g/m2/day of surface area). (Hedin and
environmental waters from moving into the system and Watzlaf 1994a, 2002).
thus diluting the waters to be treated. Fe Mn

Substrates vary from natural soils to composted For regulatory compliance 10 1

organic matter. Shallow water levels (10-30 cm) are
For other purposes 20 2
generally maintained for the aerobic conditions and
to enable growth of aquatic plants which aid wetland Data from Skousen and Ziemkiewicz (2005)
performance. Constructing the system with variable demonstrate that aerobic wetland treatment systems
do not always perform in an equivalent manner to the
water depths within the 10-30 cm range encourages
design criteria (Figure 3). On a total acidity basis, the
plant community diversity. Plants serve as physical
design criteria for non-regulatory compliance (20 g/m2/
obstacles which help to prevent “channelized flow,” day of Fe) is equivalent to about 36 g/m2/day of total
a condition that occurs when flowing waters are acidity, assuming that most of the soluble iron is in the
concentrated within the shortest distance between the Fe+2 form, which would be expected at pH = ~7. For
entrance and exit; thus, plants can help to disperse non-regulatory compliance purposes, Fe-based design
water flows throughout the wetland for more effective criteria would be roughly half that level (meaning
treatment. Dispersed flow causes the waters to move that the wetland as designed to be twice as large).
more slowly, allowing more time for the oxidation and The aerobic wetland systems demonstrate variable
aiding in physical filtration and sedimentation of small performance, with only 2 of the 6 performing within the
range of design criteria for regulatory compliance and
particles. Species such as cattails translocate O2 from
2 additional approximating non-regulatory compliance
the atmosphere through their roots, which also aids criteria.
oxidation. Although most aquatic plants will remove
some metals from the water column, their capacity
is quite limited relative to overall metal loadings that
these systems usually receive. Therefore, metal uptake
by plants plays only a minor role in water treatment by
these systems.

Aerobic wetlands are often designed in series with a

small sedimentation basin that contains no plants.
Waters to be treated may flow through the sedimentation
basin prior to entering the aerobic wetland which allows
suspended sediments and easily hydrolyzable Fe+3 to
settle out, The pre-treatment basin is designed to extend
the aerobic wetland’s useful life by accumulating
Figure 3. Data compiled by Skousen and Ziemkiewicz
particles that would otherwise settle in the wetland
(2005), which document performance of six aerobic
system. wetland systems. All values charted are averages of
multiple samples collected over extended time periods.
Commonly cited design criteria are based on the
total surface area required to treat anticipated Fe and Anaerobic Wetlands
Mn loadings (Table 1). Because performance varies
Modifications of the aerobic wetland design allow
with factors such as weather and streamflow, more
these systems to add alkalinity and effectively treat net
conservative design criteria are recommended when acid waters (Figure 4). These include the addition of
systems are intended to achieve regulatory compliance a bed of limestone beneath or mixed with an organic
because the penalties for even occasional failure to substrate, which encourages generation of alkalinity as
perform as expected can be significant. bicarbonate (HCO3-). Sulfate reduction is a microbial

www.ext.vt.edu

4
process that occurs under anoxic (low O2) conditions as MgCO3 (dolomitic limestones). The limestone is
when sulfates and biodegradable organics are present. placed so that waters must move through organic
Sulfate-reducing bacteria utilize the O that enters the substrate prior to contacting it, which allows bacteria
anoxic environment as a component of sulfate (SO42-) in the organic material to remove O2 from the
for metabolic processing of biodegradable organics, percolating waters. This process helps to prevent
transforming the associated S to either hydrogen sulfide armoring of the limestone. The term “armoring” refers
gas (H2S) or to a solid-phase sulfide. The most common to coating of Fe on limestone surfaces, a process that
form of sulfide reduction generates H2S and bicarbonate renders those surfaces less reactive.
alkalinity:
Anaerobic wetlands are capable of removing acid-solu-
SO4-2 + 2 CH2O → H2S + 2 HCO3- (eq. 7) ble metals, especially Fe and Al, and producing alkalin-
ity. However, their effectiveness is limited by the slow
Sulfate reduction is common in both natural and treat- mixing of the alkaline substrate water with acidic waters
ment wetlands where its occurrence can often be near the surface. Thus, these systems commonly require
detected as visible bubbles emerging from the substrate large surface areas and long retention times. Like other
accompanied by the “rotten egg” odor of H2S gas. When passive treatment systems, their effectiveness at remov-
acid-soluble metals are in solution, sulfate reduction ing Mn is limited unless very large areas are used.
can form solid-phase metal sulfides as an alternative
end product, which removes metals from solution and Research reported by Skousen demonstrates that sub-
deposits them in the substrate. In the following, “M” strate processes – alkalinity generation to stimulate
represents a sulfide-forming metal and “MS” represents oxidation and hydrolysis, and metal sulfide formation
a metal sulfide. – are the primary drivers of water quality renovation
over the long term. When the systems are first con-
M + SO4-2 + CH2O → MS + HCO3- (eq. 8) structed, mechanisms such as plant uptake and sorption
by organic materials can contribute to metal removal,
The other alkalinity generating process is dissolution of but the capacity of these mechanisms to remove metals
the limestone within or below the organic substrate: is soon exhausted as absorption sites are filled.

CaCO3 + H+ → Ca2+ + HCO3- (eq. 9) General guidelines for anaerobic wetland construction
advocate use of a 30–60 cm layer of organic matter
The bicarbonate (HCO3-) is a source of alkalinity, and over 15–30 cm bed of limestone, or placing a mixture
can neutralize H+ and/or raise pH to enhance precipita- of organic matter and limestone to a depth of 50–100
tion of acid-soluble metals: cm. The organic matter must be water permeable and
biodegradable. Materials such as spent mushroom
HCO3- + H+ → H2O + CO2 (aq) (eq. 10)
compost have been used successfully at a number of
sites in northern Appalachia. Cattails (Typha sp.) or
other aquatic vegetation may be planted throughout
the wetland to supply additional organic matter for the
O2-consuming bacteria and to promote metal oxida-
tion with the release of oxygen from their root system.
If sediments or easily hydrolysable Fe are present in
Figure 4. Simplified cross-sectional view of an anaerobic the waters to be treated, a pretreatment pond can be
wetland. These systems are also called “composted included in the system design.
wetlands” because the substrate layer generates alkalinity.
The substrate can be comprised of biodegradable organic The design depth of water over the organic/limestone
materials that are placed over limestone, as represented, mixture varies. Some anaerobic wetland designs
or as an organic-limestone mixture. The circular arrows maintain water depths of 10–30 cm to encourage
represent diffusion of treatment waters through the aquatic vegetation that prevents flow channelization
substrate which generates alkalinity.
and adds fresh organic material to the substrate. Other
High-Ca limestones, with greater than 90% CaCO3, designers use deeper waters and do not encourage
are preferred for passive treatment use because they vegetation, reasoning that the translocation of oxygen
are more soluble than impure limestones or those to substrates through the roots of species, such as
that contain a larger proportion of total carbonates cattails, hinders substrate functions which are critical

www.ext.vt.edu

5
to performance and require anoxic conditions. When
plants are not present, physical barriers can be installed
to prevent flow channelization and additional organic
materials can be added manually (Figure 5).

Available guidelines for system sizing recommend

planning for acidity removal rates of 3.5 g/m2/day Figure 7. Simplified cross-sectional view of an anoxic
when the system is intended for regulatory compli- limestone drain. These systems are used to add alkalinity
ance and 7 g/m2/day otherwise. However, performance to anoxic (low O2) waters that are low in constituents that
data (Skousen and Ziemkiewicz 2005; Figure 6) dem- can easily precipitate under low-O2 conditions (i.e. Al,
onstrate that anaerobic wetland performance is highly Fe+3). The standpipe exit is elevated, and other measures
are taken to exclude O2 from the drain so as to prevent
variable and that these systems tend to neutralize acidity
precipitation of metals within the gravel pores.
more effectively when receiving higher influent acidity
concentrations and loadings. Of the 16 systems whose
performance was documented, eight were removing
acidity at rates of ~10 g/m2/day or greater.

Anoxic Limestone Drains

Anoxic limestone drains (ALDs) can also be used to
treat acidic waters (Figure 7). ALDs are limestone-
filled trenches through which acidic water is directed
so the limestone can produce bicarbonate alkalinity
via dissolution (eq. 9). ALDs are capped with clay or
compacted soil to prevent AMD contact with oxygen.
The effluent is held in a settling pond to allow pH
adjustment and metal precipitation prior to being
discharged to natural water courses. Where water
quality is suitable for an ALD, an ALD can be used to
pretreat the AMD prior to forcing the waters through
subsequent passive treatment units.

Figure 6. Graphs prepared from data

Figure 5. Because poor water quality and low nutrient availability compiled by Skousen and Ziemkiewicz
hindered plant growth in this aerobic wetland, haybales were placed to (2005), which document performance of
prevent flow channelization and to add biodegradable organic materials 11 anaerobic wetland systems. All values
that help stimulate alkalinity generation via sulfate reduction. The inset charted are averages of multiple samples
photo shows iron floc accumulating on the biodegradable organic collected over extended time periods, and
material. Water draining from an abandoned deep mine enters the system the systems represent a variety of operating
to the right of the inset photo, and flows through the system toward an conditions.
exit at the lower right.

www.ext.vt.edu
6
When working as intended, ALDs can renovate acidic result, standard design guidelines for ALDs generally
waters more cost effectively than wetland-based recommend that drains be constructed to achieve
systems. ALDs, however, are not capable of treating 15 hour retention times over their design lifetimes.
all AMD waters. Significant concentrations of O2, Limestone will most likely dissolve over that lifetime,
Al or Fe+3 in the water will cause an ALD to clog so the initial construction achieves a longer retention
with metal-hydroxides once a pH of 4.5 or above is time. An equation that is commonly used in ALD
reached. When excess Fe+3 is present in the AMD or design is as follows:
is allowed to form from Fe+2 due to the presence of
O2, solid-phase Fe can precipitate within the ALD (eq. M = [Q * ρ * t / V] + [Q * C * T / x] (eq. 10)
3), while Al precipitation can occur as pH increases
even when O2 is excluded (eq. 4). If significant metal In this equation, M represents the mass of limestone to
precipitation occurs within the ALD, the precipitant be used to construct the drain. The first bracketed term
floc (a gel comprised of hydrolyzed solid-phase metal represents the volume of limestone required to achieve
precipitants) clogs the ALD’s pores, hindering the the design retention time, while the second bracketed
waters from moving through the system and impairing term represents the volume of limestone expected to
dissolve over the design lifetime. Adding the two quan-
its function. Once an ALD becomes clogged with
tities is intended to assure that the mass of limestone
precipitants, it becomes non-functional and must be
remaining in the drain throughout its design lifetime is
replaced, repaired, or abandoned.
adequate to achieve the design retention time. To cal-
To maximize the probability that an ALD will not clog, culate the first bracketed term: Q = water flow rate; ρ
Fe+3, Al, and dissolved O2 concentrations of the influent = limestone density; t = design retention time; and V
waters should all be below 1 mg/L. However, Skousen = bulk void volume of the limestone gravel, expressed
and others (2000) recommend that ALDs can be used as a percent of total volume. To calculate the second
successfully for AMD with dissolved O2 concentrations bracketed term: Q = influent flow rate; C = expected
of up to 2 mg/L and Al concentrations of up to 25 mg/L, rate of alkalinity generation, expressed on an mg-
when less than 10 percent of total Fe in the Fe+3 form. CaCO3/L per unit time basis; T = the design lifetime;
Although such ALDs can be expected to clog eventu- and x = the CaCO3 content of the limestone, for which
ally, they can still offer cost-effective water treatment 90% or greater is recommended.
compared to other passive system processes during
their time of operation. ALDs are far less costly to con-
struct than anaerobic or vertical flow wetland treatment
systems and can render less costly treatment on a life-
cycle basis, even if periodic but infrequent repair and
replacement is required.

The term “retention time” means the amount of time

the average quantity of water flowing through the drain
spends within the drain structure, and is calculated as
[total pore volume]/[Average Rate of Flow Through the
Drain]. A common recommended design criterion is that
ALDs should achieve at least 15 hour retention times.
Figure 8. Data from Watzlaf, Schroeder, and Kairies
This criterion is based on research conducted by Watzlaf (2000) collected using water monitoring ports in four
and others (2000) who established series of sampling anoxic limestone drains demonstrated that alkalinity
ports in several ALDs; they were able to withdraw concentrations approached a maximum after 14 to
water samples at several points so as to observe how 23 hours in drains. As a result, design guidelines for
water quality changed as it moved through the ALDs. ALDs generally recommend that drains be constructed
They observed that alkalinity generation reached to achieve 15 hour retention times over their design
maximum levels, which was within the range of 150 lifetimes. Average influent acidities (mg/L as CaCO3) were
– 300 mg/L (as CaCO3) after 14 to 23 hours retention 52 at Elklick, 382 at Morrison, and 472 and 411 at Howe
time. As dissolved Ca+2 and HCO3- concentrations Bridge 1 and 2, respectively. (Figure from Watzlaf et al.
2004).
approached saturation, continued limestone dissolution
and alkalinity generation was hindered (Figure 8). As a

www.ext.vt.edu
7
ALD performance data compiled by Skousen and Ziem- flows, as the abrasive action of fast-moving water and
kiewicz (2005) demonstrates that, like other passive suspended particles tends to dislodge the Fe armoring,
treatment system, ALD performance is highly variable refreshing the surface for treatment effectiveness. OLCs
(Figure 9). Twenty of the 36 ALDs had estimated aver- can be effective as one element of a passive treatment
age influent acidities of 440 mg/L or less, close to the system (Figure 10), but typically are not relied upon for
range of systems used to generate the 15-hour retention stand-alone AMD treatment (Ziemkiewicz et al. 1997;
time design guideline. All generated alkalinities of 400 Skousen, Sexstone, and Ziemkiwicz 2000).
mg/L or less, and most were close to or within the 150
to 300 mg/L alkalinity target range. However, 14 of the
ALDs documented were treating waters with influent
acidities of 591 mg/L or greater, and most of these sys-
tems were generating alkalinities considerably above
the commonly cited 150-to-300 mg/L target range;
among this group, those with longer retention times
tended to generate the most alkalinity.

Figure 10. Performance of 16 open limestone channels,

as documented by Skousen and Ziemkiewicz (2005),
plotted by influent pH. All values charted are averages of
multiple samples collected over extended time periods.

Vertical Flow Systems

Vertical flow passive-treatment systems combine the
treatment mechanisms of anaerobic wetlands and ALDs
Figure 9. Data compiled by Skousen and Ziemkiewicz
(2005) document performance of 36 anoxic limestone in an attempt to compensate for the limitations of both
drains. All values charted are averages of multiple samples (Hendricks 1991; Duddleston et al. 1992; Kepler and
collected over extended time periods. All but 1 of the McCleary 1994) (Figure 11). Vertical-flow systems
systems with influent acidities ≤ 440 mg/L generated have also been called SAPS (for “successive alkalin-
< 400 mg/L alkalinity, regardless of retention time. Six ity producing systems,” Kepler and McCleary 1994)
of the systems with higher influent acidities, primarily and RAPS (for “reducing and alkalinity producing sys-
those with longer retention times demonstrated tems,” Watzlaf et al. 2000).
greater alkalinity generation performance. Three of the
documented systems failed to generate alkalinity and are The basic elements of these systems are similar to the
not represented. anaerobic wetland, but a drainage system is added to
force the AMD into direct contact with the alkalinity-
Open Limestone Channels producing substrate. The three major system elements
An open-air analog to the ALD is the open limestone are the drainage system, a limestone layer, and an
channel (OLC). These systems are employed where AMD organic layer (Figure 12). The system is constructed
must be conveyed over some distance prior to or during within a water-tight basin, and the drainage system is
treatment. OLCs are simple to construct and operate, if constructed with a standpipe to control water depths
the terrain is favorable: an open channel conveying the and ensure that the organic and limestone layers
AMD is lined with high-Ca limestone. Even though remain submerged. As the AMD waters flow down-
the limestone typically becomes armored with Fe, the ward through the organic layer, essential functions are
armored limestone retains some treatment effectiveness. performed: dissolved oxygen is removed by aerobic
OLCs are most effective when placed on slopes of greater bacteria utilizing biodegradable organic compounds as
than 20 percent or when they receive periodic high energy sources, and sulfate-reducing bacteria generate

www.ext.vt.edu
8
alkalinity and sequester metals as sulfides (eq. 7-9). An Two major limitations to the long term performance of
organic layer capable of lowering DO concentrations vertical flow systems are accumulation of metal floc, pri-
to < 1 mg/L is essential to prevent limestone armoring marily Fe and Al in the limestone layer, and degradation
and for sulfate reduction. In the limestone layer, CaCO3 of the organic layer. In order to postpone floc accumula-
is dissolved by the acidic, anoxic waters moving down tion and eventual clogging, a valved flushing pipe is typi-
to the drainage system, producing additional alkalin- cally included as a part of the drainage system (Kepler and
ity. The final effluent is discharged into a settling pond McCleary 1997; Figure 13). When opened, this valved
for acid neutralization and metal precipitation prior to drain discharges at a lower elevation than the standpipe.
ultimate discharge. For influents containing significant Head pressure moves waters through the system rapidly,
quantities of Fe+3 and/or sediment, vertical-flow sys- flushing the gel-like flocs of Al and Fe (“floc”) that tend
tems should be preceded by either a settling pond or an to accumulate in the drainage pipes and limestone pores.
aerobic wetland so as to limit accumulation of solids Opening this valve periodically is intended to remove
on the organic layer surface. For treating highly acidic loose floc from the limestone layer, discharging it into the
discharges, several vertical flow cells can be placed in settling pond. In order to assure adequate head pressure
sequence, separated by settling ponds. for flushing, water depths above the organic layer of 1
to 2 m or greater are generally recommended. However,
as demonstrated by Watzlaf, Kairies, and Schroeder
(2002), only a small proportion of the total accumulated
floc is removed by typical flushing operations. Because
floc accumulates within the limestone layer, many verti-
cal flow systems are designed with intricate drainage con-
figurations to aid floc removal by the flushing operation.

Generally, systems are built with high-calcium (> 90%)

limestone in the 4-to-6 inch size range with a limestone
layer thickness in the range of 60 to 100 cm. In order
to assure that the volume of limestone is adequate to
assure long-term performance, the criteria used to size
Figure 11. The upper surface of a vertical flow ALDs (eq. 10) can also be applied to calculate a volume
system constructed in 1990 to receive effluent from of limestone.
Westmoreland Coal Company’s Pine Branch refuse pile
(Duddleston et al. 1992). The vertical plastic piping visible The organic layer is the other major system element
in the background was installed to enable “clean out” of that is critical to long-term performance. To operate
the drainage system, should that prove necessary, prior to properly, the organic layer must be sufficiently
the development of the flushing valve concept depicted
in Figure 13.

Figure 12. Simplified cross-sectional view of a simple

vertical flow treatment system. Waters enter the system
on the left, and it must pass vertically through the organic
matter and limestone gravel to enter the perforated Figure 13. An underdrain flushing system can extend
piping within the limestone layer to exit the system. The the useful lifetime of vertical flow treatment systems by
drainage system is outfitted by a standpipe that maintains preventing accumulation of metal hydroxide floc in the
water levels so as to assure the organic matter and limestone layer. During normal operation, the system
limestone remain submerged. Alkalinity is generated by discharges through the standpipe. Periodically, floc is
sulfate reduction and limestone dissolution. Some metals removed from the underdrain by opening the flushing
are deposited within the system, but most metal removal valve so waters can exit rapidly, carrying loose floc into
occurs through precipitation in the settling pond that the settling pond. Once the flow ceases, the flushing valve
receives discharge waters. is closed and the system resumes normal operation.

www.ext.vt.edu
9
biodegradable to allow essential functions – but it systems than in anaerobic wetlands, so vertical-flow
also must be permeable, so water can move through systems generally require shorter residence times
it into the limestone. Organic functionality can be and smaller surface areas. Skousen and Ziemkiewicz
expected to degrade over time, due to microbial (2005) describe a system design guideline that sizes
biodegradation and floc accumulation in pore spaces. vertical flow systems to treat 20 g/m2/day with a
Materials that have been used successfully include 15-hour retention time, assuming the system can attain
spent mushroom compost, composted manure, and a 20-year useful life. As a general guideline, Watzlaf et
mixtures of composted materials with less-expensive al. (2004) recommend that the initial vertical-flow cell
organic sources such as rotting hay. Published design in a sequence can be sized assuming it will generate
guidelines recommend organic layer depths of 15–60 40–60 g/m2/day of alkalinity, and that subsequent
cm. Carefully install the organic layer to assure that the cells can be expected to generate 15–20 g/m2/day.
material is well mixed and applied at a uniform depth Analysis of performance data gathered by Skousen and
across the limestone-layer surface. Once the organic Ziemkiewicz (2005) and by Jage, Zipper, and Noble
layer is in place, any activity causing compaction or (2001) demonstrate that influent acidity concentrations
physical disturbance should be avoided to prevent
creation of zones of preferential vertical flow that that
will “short circuit” the system and decrease treatment
effectiveness (Figure 14)

Vertical flow systems can neutralize acidity and promote

metal precipitation in difficult treatment situations. Due
to the forced contact of the AMD with the limestone,
acid neutralization is more rapid in vertical flow

Figure 14. Top-view representation of the organic layer

and underlying drain configuration (black bars, with
arrows representing drainage water flow) of a vertical flow
system in Pennsylvania, with colors representing depth
of oxidation. The dark blue colors represent oxidizing Figure 15. Performance data for 18 vertical
conditions occurring > 45cm below the organic layer flow systems from Jage, Zipper, and Noble
surface, meaning that water is flowing downward rapidly (2001), and for 15 vertical flow systems from
through these areas, while the red colors represent Skousen and Ziemkiewicz (2005). One S&Z
depths of oxidation < 15 cm, meaning that the opposite point is off-scale in the right plot: (x,y) = (668,
is occurring; yellows and greens represent intermediate 843). Both data sets indicate an influence by
conditions. (Figure from Demchek and Skousen 2000). influent acidity on system performance.

www.ext.vt.edu
10
and loadings influence performance (Figure 15). Of water quality, NMA is as in equation 11, and tr = average
the 33 vertical-flow systems whose performance was residence time in the limestone layer expressed in hours.
documented by two studies, 19 generated alkalinity at
> 20 g/m2/day and 13 generated alkalinity at > 40 g/m2/ Equation 12 was developed by analyzing data from 18
day. Generally, systems receiving higher acid loadings vertical flow systems receiving influent waters with Fe
generated more alkalinity, on both a concentration and < 300 mg/L, Al < 60 mg/L, and NMA < 500 mg/L (Jage
a per unit surface area basis, than those receiving lower et al. 2001). Equation 12 is not expected or intended to
loadings. give precise results, but it can be used to provide design
guidance. Figure 15 shows the relationship of observed
Jage et al. (2001) have described a more detailed sys- alkalinity generation, averaged over periods exceeding
tem for vertical flow system sizing. Their guidelines are one year, to equation 12 predictions for 18 vertical flow
based on the observation that the primary factor govern- cells at 5 locations. All values plotted are system aver-
ing alkalinity generation is the rate at which limestone ages over periods exceeding one year.
dissolves, which is affected by solution chemistry. Res-
idence-time in the limestone layer is one factor govern-
ing the limestone dissolution rate, as limestone can be
expected to dissolve most rapidly during the first few
hours of AMD contact. As the waters in contact with
the limestone become saturated with dissolved Ca2+ and
HCO3-, the rate of limestone dissolution slows consid-
erably. Another factor governing the rate at which lime-
stone dissolves is pH; at lower pH’s, CaCO3 dissolves
more rapidly. Based on these observations, they defined
a method for sizing the limestone layer of a vertical
flow system. The first step is to calculate a quantity
which they termed as non-manganese acidity (NMA)
for the influent waters:

NMA = A - 100 * Mn / 55 (eq. 11)

where A = acidity (mg/L as CaCO3) and Mn = Mn con- Figure 16. Observed alkalinity generation, as system
centration (mg/L) of the influent waters; NMA = non- averages for multiple observations over >1 year, and
Mn acidity, that component of acidity derived from H+ predicted alkalinity generation predicted by Equation 12
and from Al and Fe, expressed as mg/L CaCO3. In other for 18 vertical flow cells. Alkalinity generation, as plotted
words, NMA is total acidity, as estimated using equa- here, is equivalent to “Acidity Treated,” as charted above.
tion 6, minus its Mn-derived component. The rationale Equation 12 can be applied to illustrate the logic
for using NMA as a basis for sizing is that Mn is not for constructing several vertical flow systems in
removed effectively by most vertical flow systems series, separated by settling ponds, as an alternative
because its rapid oxidation requires pH be greater than to construction of a single large system. Consider a
the practical maximum (from pH 7.0 to 7.5) that most discharge with Fe = 60 mg/l, Mn = 20 mg/l, and total
passive treatment systems are able to achieve. This prac- acidity = 300 mg/l as CaCO3 (Figure 17). Equation 12
tical maximum occurs because CaCO3 dissolution rates predicts that a limestone residence time on the order of
decline as solutions become increasingly alkaline. 350 hours would be required in a single cell to generate
300 mg/L acidity as needed to neutralize 260 mg/L of
Once the design rate of alkalinity generation has
influent non-Mn acidity. As an alternative, construction
been determined, the limestone residence time can be
of two cells in series, each with a 15-hour residence
estimated using the equation below:
time and separated by a settling pond, are predicted by
Alk = 99.3 * log 10(tr) + 0.76 * Fe + 0.23 * equation 12 as being capable of generating a comparable
NMA - 58.02 (eq. 12) amount of alkalinity (150 mg/L each, 300 mg/L total).

where Alk = alkalinity to be generated (mg/L as CaCO3);

Other Passive Treatment System Types
Fe = total iron concentration (mg/L) of the design influent

www.ext.vt.edu
11
 The above are the most widely used types of passive
treatment systems for acid drainage. Several other
types of systems are also available, generally for use in
specialized situations. Such methods include limestone
ponds, limestone sand beds, limestone diversion
wells, methods that utilize biodegradable organic
materials under anoxic conditions to stimulate sulfate
reduction (“bioreactors”), and other methods. Several
of these other system types are described in the Passive
Treatment Design Guideline publications listed as
references below.

Developing a Passive Treatment Strategy

The design of all passive treatment systems starts with
characterization of influent AMD chemistry and flow.
Regular sampling over at least a 12-month period
Figure 17. Alkalinity generation vs. limestone residence should account for the variations that may occur or in
time predicted by equation 12 for influent water for which
response to seasonal changes or storms. At a minimum,
Fe = 60 mg/l, Mn = 20 mg/l, and total acidity = 300 mg/l.
The solid line represents the predicted performance, while all water samples should be analyzed for pH, total
the dashed lines above and below represent the 95% Fe, Mn, Al, SO42-, total alkalinity and total acidity.
confidence derived from the regression that generated Additional analyses, including Fe+2, Fe+3, and dissolved
equation 12 (Jage et al. 2001). O2 are necessary if ALD treatment is being considered.
If the AMD discharge is intended to achieve regulatory
compliance, sampling
personnel should assure that
worst-case conditions, however
defined, are represented by the
sampling data.

The selection of a passive

treatment system is governed
by both influent water quality
and site characteristics. The
diagram in Figure 18 illustrates
a decision process for selecting
an appropriate system for
a given discharge. For net
alkaline discharges containing
elevated concentrations of
Fe, no additional alkalinity
additions are needed. The
only conditions necessary to
complete treatment are an
oxidizing environment and
sufficient residence time to
Figure 18. Passive system design flow chart. More detailed versions of the flow chart
can be accessed in the other sources, including the passive treatment system design allow for metal oxidation and
guidelines referenced by this publication. precipitation. These conditions
can be provided by a settling

www.ext.vt.edu
12
pond and, if sufficient area is available, the settling system performance varies significantly among con-
pond may be followed by an aerobic wetland for structed systems, due to differences in site conditions.
effluent polishing (more complete removal of settleable One factor that influences the performance of systems
contaminants). that utilize limestone is influent acidity, as systems
receiving higher acidity are often able to neutralize
The treatment of net acidic drainage can be handled in a acidity more effectively. Passive treatment systems
number of ways depending on influent chemistry. If the are effective at renovating waters with low pH, high
influent quality is suitable for an ALD, an ALD can be acidity, and high concentrations of acid soluble metals
employed as a pretreatment method. A post-ALD set- including Fe and Al, but their effectiveness at removing
tling pond or aerobic wetland is required to allow for Mn is generally limited unless large treatment areas are
the oxidation and precipitation of metals. available.
Acidic waters that are not suitable for ALDs can be
treated with either an anaerobic wetland or a vertical Acknowledgments
flow system. Due to the potentially large demands on
Thanks to Art Rose (Pennsylvania State University)
land area of anaerobic wetlands, they are usually only
and George Watzlaf (U.S. Department of Energy) for
practical for low-flow situations. For systems that
their assistance to the research that supports these
receive water that has a pH greater than 4, settling ponds
recommendations.
may precede an anaerobic wetland cell to remove sig-
nificant quantities of Fe. The remaining discharges can
be treated using a vertical flow system. References and Literature Cited
Where terrain is suitable, an open limestone channel
Passive Treatment Design Guidelines
can be used to carry the treatment waters while adding
additional alkalinity in the process. Hedin, R. S., R. W. Nairn, and R. L. P. Kleinmann.
1994a. Passive Treatment of Coal Mine Drainage.
Another factor governing system selection will be the Bureau of Mines Inf. Circ. IC9389. Washington,
cost of available treatment options. Skousen and Ziem- D.C.: U.S. Bureau of Mines.
kiewicz (2005) have documented cost-effectiveness for
a number of different passive system installations by Hedin, R. 2002. Passive treatment of polluted mine
estimating construction costs from system dimensions, waters. In Mine Water Hydrology, Pollution, Reme-
assuming 20 year lifetimes, and calculating the cost per diation. Chapter 5. Ed. Younger, P.L., S. Barnhart,
ton of acid neutralization using measured performance and R. Hedin. Norwell, Mass.: Kluwer Academic
data. Per-unit acidity treatment costs vary dramati- Publishers.
cally for all system types. In general, anoxic limestone
drains render the most cost-effective passive treatment Skousen, J. 1996. Overview of Passive Systems for
while open limestone channels were also able to render Treating Acid Mine Drainage. Morgantown: West
treatment that was less costly, on average than either Virginia University Extension Service.
type of wetland system and vertical flow systems. Of
Skousen, J., A. Sexstone, and P. Ziemkiewicz. 2000.
course, ALDs have very specific influent water qual-
Acid mine drainage treatment and control. In Rec-
ity requirements and are not an option for many AMD
lamation of Drastically Disturbed Lands. Ed. R.
discharges, while OLCs can only operate within water
Barnhisel, W. Daniels, and R. Darmody Madison,
conveyances. Generally speaking, the systems that are
Wis.: American Society of Agronomy. 131-168.
able to treat the most problematic waters–highly acidic
and oxidized–are also the most costly. Skovran G. A. and C. R. Clouser. 1998. Design Consid-
erations and Construction Techniques for Succes-
sive Alkalinity Producing Systems. In Proceedings
Conclusions of the 1998 Annual Meeting of the American Soci-
Passive treatment systems can be a component of an ety for Surface Mining and Reclamation. Princeton,
AMD treatment strategy. They can function as either W.Va.: ASSMR.
stand-alone treatment strategies or as pre-treatment to
reduce the cost of active treatment. Passive treatment Watzlaf, G., K. Schroeder, R. Kleinmann, C. Kairies, R.

www.ext.vt.edu
13
 Nairn. 2004. The Passive Treatment of Coal Mine Journal of Environmental Quality 30: 1015-1022.
Drainage. DOE/NETL-2004/1202. Washington,
D.C.: U.S. Department of Energy. Kepler, D. A., and E. C. McCleary. 1997. Passive Alu-
minum Treatment Successes. In Proceedings of the
National Association of Abandoned Mine Lands
Programs. Davis, W.Va.: NAAMLP.
Other References Kepler, D. A., and E. C. McCleary. 1994. Successive
Demchak, J., J. Skousen. 2000. Treatment of Acid Mine Alkalinity-Producing Systems (SAPS) for the
Drainage by Four Vertical Flow Wetlands in Penn- Treatment of Acidic Mine Drainage. In Proceed-
sylvania. Morgantown: West Virginia University ings of the International Land Reclamation and
Cooperative Extension Service. Mine Drainage Conference. Washington, D.C.:
U.S. Bureau of Mines. 195-204.
Duddleston, K. N., E. Fritz, A. C. Hendricks, and K.
Roddenberry. 1992. Anoxic Cattail Wetland for Kirby, C., and C. Cravotta. 2005. Net alkalinity and
Treatment of Water Associated With Coal Mining net acidity I. Theoretical considerations, and II.
Activities. p. 249-254. In Proceedings of the 1992 Practical considerations. Applied Geochemistry 20:
Annual Meeting of the American Society for Sur- 1920-1964.
face Mining and Reclamation. Princeton, W.Va.:
ASSMR. 249-254. Rose, W.A., 2006. Long-term performance of vertical
flow ponds – an update. In Proceedings of the Sev-
Demchek, J., J. Skousen, and T. Morrow. 2001. Treat- enth International Conference on Acid Rock Drain-
ment of Acid Mine Drainage by Four Vertical Flow age. Lexington, Ky.: ASMR. 1704–1716.
Wetlands in Pennsylvania. Morgantown: West Vir-
ginia University Cooperative Extension Service. Skousen, J., and P. Ziemkiewicz. 2005. Performance
of 116 Passive Treatment Systems for Acid Mine
Hedin, R. S. and G. R. Watzlaf. 1994. The Effects of Drainage. In Proceedings, National Meeting of the
Anoxic Limestone Drains on Mine Water Chemistry. American Society of Mining and Reclamation. Lex-
In Proceedings of the International Land Reclama- ington, Ky.: ASMR.
tion and Mine Drainage Conference. Washington,
D.C.: U.S. Bureau of Mines. 185-194. Watzlaf, G. R., Kairies, C. L., and K. T. Schroeder.
2002. Quantitative Results from the Flushing
Hedin, R. S., G. R. Watzlaf, and R. W. Nairn. 1994b. of Four Reducing and Alkalinity-Producing
Passive Treatment of Acid Mine Drainage with Systems. In Proceedings of the 23rd West Virginia
Limestone. Journal of Environmental Quality 23: Surface Mine Drainage Task Force Symposium.
1338-1345. Morgantown: West Virginia University.

Hendricks, A. C. 1991. The use of an Artificial Wet- Watzlaf, G., K. Schroeder, and C. Kairies. 2000. Long-
land to Treat Acid Mine Drainage. In Proceedings term performance of anoxic limestone drains. Mine
of the International Conference on the Abatement Water and Environment 19:98-110.
of Acidic Drainage. Montreal, Canada.
Wieder, R K., and G. E. Lang. 1982. Modification of
Jage, C., C.E. Zipper, and R. Noble. 2001. Factors Acid Mine Drainage in a Freshwater Wetland. In
affecting alkalinity generation by successive Proceedings of the Symposium on Wetlands of the
alkalinity producing systems: Regression analysis. Unglaciated Appalachian Region. Morgantown:

www.ext.vt.edu
14