USDA
EPA
ui
III
« £
« o 5P
o.n *
o o) <£
United States
Department of
Agriculture
Northeast Watershed Research
Center
University Park PA 16802
United States
Environmental Protection
Agency
Office of Environmental
Processes and Effects Research
Washington DC 20460
EPA-600/7-84-044
March 1984
Research and Development
Hydrology and
Water Quality on
Stripmined Lands
Interagency
Energy/Environment
R&D Program
Report
-------
HYDROLOGY AND WATER QUALITY ON STRIPMINED LANDS
by
A. S. Rogowski and H. B. Pionke
U.S. Department of Agriculture, ARS
Northeast Watershed Research Center
University Park, Pennsylvania 16802
EPA-IAG-D5-E763
Project Officer
Clinton W. Hall
Office of Energy, Minerals and Industry
Washington, D.C. 20250
Office of Research and Development
U.S. Environmental Protection Agency
Washington, D. C. 20250
U.S. Environmental Protection Agen
k'.-Tion 5, Library (5PL-16)
2JO S. Dearborn Stt-eet, Room 1670
Chicago, -IL 60604
-------
DISCLAIMER
This report has been reviewed by the Office of Energy, Minesoils and
Industry, U.S. Environmental Protection Agency, and approved for
publication. Approval does not signify that the contents necessarily
reflect the views and policies of the U.S. Environmental Protection Agency,
nor does mention of trade names or commercial products constitute endorse-
ment or recommendation for use.
-------
FOREWORD
The Federal Water Pollution Control Act Amendments of 1972, in part, stress
the control of nonpoint source pollution. Sections 102 (C-l), 208 (b-2,F) and
304(e) authorize basin scale development of water quality control plans and
provide for area-wide waste treatment management. The act and the amendments
include, when warranted, waters from agriculturally and silviculturally related
nonpoint sources, and requires the issuance of guidelines for both identifying
and evaluating the nature and extent of nonpoint source pollutants and the
methods to control these sources. Research program at the Northeast Watershed
Research Center contributes to the aforementioned goals. The major objectives
of the Center are to:
. study the major hydrologic and water-quality associated problems
of the Northeastern U.S. and
. develop hydrologic and water quality simulation capability useful
for land-use planning. Initial emphasis is on the hydrologically
most severe land uses of the Northeast.
Within the context of the Center's objectives, stripmining for coal ranks
as a major and hydrologically severe land use. In addition, once the site is
reclaimed and the conditions of the mining permit are met, stripmined areas
revert legally from point to nonpoint sources. As a result, the hydrologic,
iii
-------
physical, and chemical' behavior of the reclaimed land needs to be understood
directly and in terms of control practices before the goals of Sections 102,
208 and 304 can be fully met.
Signed:
^,B.
H. B. Pionke
Director
Northeast Watershed
Research Center
-------
ABSTRACT
Studies were conducted to evaluate physical properties of spoils resulting
from surface-coal mining and reclamation operations in Clearfield County,
Pennsylvania. Bulk density, evapotranspiration, water retention, infiltration,
and hydraulic conductivity values were determined at 10 sites randomly located
within a 4-ha experimental area. Average bulk density of the surface 0.5-m
layer of minesoil was 1,763 kg/m while specific surface at most sites averaged
2
31 m /g. Microlysimeter data indicated that evapotranspiration (ET) on
minesoil could be approximated by class-A pan evaporation or by model results.
A large amount of spatial variation was observed in infiltration, water retention
and hydraulic conductivity values. In the uppermost 0.75 m of the profile most
minesoils on the average retained 35 mm of water, between 10 and 1,500 kPa,
compared to 136 mm for the adjoining soils. When water was available ET
approached potential, however, hydraulic properties of the minesoil would likely
lead to droughty conditions and extended periods of plant stress.
Two, 3.6-m deep, 2.4-m diamecer caissons were used co measure water movement
in reconstituted spoil profile. In here we present details of instrumentation
and propose methods for determining spoil-water-retentivity curve and hydraulic
conductivity as a function of water content and/or tensiometer pressure including
a correction for coarse fragment content. This correction may be significant
when modeling flows in spoil materials. During water application we found that
on spoil alone, water infiltrated faster and less of it was retained in the
profile after drainage than on topsoiled spoil. More sediment moved with
-------
infiltrating water in the topsoiled profile, and oxygen concentrations after
application of water were lower than on spoil alone. Similar situations may
prevail in the field. During the initial wetting, settlement was considerable
and changes in temperature and bulk density were good indicators of water
movement within the spoil and topsoiled spoil profiles. However, flow through
larger channels did contribute significant amounts of water to the water table
in the spoil without being detected by radioactive probes, temperature probes,
or tensiometers.
Total acidity provides a reasonable estimate of other major chemical
parameters contained in spoil drainage. In caisson percolate, good linear
correlations existed between total acidity and Al, total soluble Fe, Mg, and
SO, concentrations. These exclude the lowest concentrations (<10 meq/L)
which normally do not represent the important acid contributing spoils. The
+2
relationship between total acidity and Ca, Fe , Mn and pH was much less
precise.
A partially weathered spoil, containing both pyrite and acid products,
exerts the primary control on the concentration of acid products in percolate.
Normal variations in hydrology appear less important. This was apparent from
analyses of data taken over several caisson runs as well as within one run.
Except for extreme hydrologic conditions or the "first flush" of total acidity
observed initially, the maximum and stabilized total acidity in percolate
increased with spoil depth or as other chemically related spoil properties
increased.
The total acidity of percolate appeared related to both the spoil depth in
the caisson and to the total acidity of spoil extract. If the total acidity
from a spoil layer extract, compared to that of the overlying spoil layers, is
substantially greater (about 1 order of magnitude), this spoil layer will
VI
-------
distinctly and identifiably influence percolate quality. Where total acidity is
not substantially greater, the spatial and hydrologic variability will tend to
mask the contributions from each layer into what appears as a total acidity
increase with depth. A substantial increase of total sulfate and pyrite content
in the spoil profile (about 1 order of magnitude) corresponded with a similar
increase in the total acidity of percolate. Extreme total pyrite or SO, con-
centrations in the spoil profile can be used to delineate the primary total
acidity contributing zones.
The concentrations of Cd, Cr, total soluble Fe, Hg, Mn and Zn in spoil_
extracts exceeded EPA water quality drinking standards from one (Cr) to all
(Mn) spoil layers. The standards for Pb and Cu were not exceeded. Generally,
trace metal concentrations in the spoil layer extracts agree closely with those
observed in the corresponding spoil percolate. The Cu and Zn concentrations in
the percolate were considerably higher. In caisson 2, the trace metal concen-
trations in the spoil layer extract, caisson percolate, and the caisson well
agree closely. This suggested that reduction in trace metal concentrations
caused by dilution or other processes occurs primarily after entry into the
groundwater or stream system rather than in the spoil profile. Concentration
of Cu, Zn and Mn in spoil extract could be approximated from the SO
concentration.
Potentially there are several chemical and hydrologic problems associated
with placement of acid spoil materials. The rational for a deep placement
well below the soil surface, and preferably below a water table, is to prevent
or minimize oxidation of pyrite to sulfuric acid and associated salts by
reducing the supply of oxygen. If, however, substantial sulfuric acid or
associated salts are already contained within the spoil because of present or
previous mining, handling and reclamation operations (or if large supplies of
-------
indigenous salts exist, placement below a water table) may actually increase the
rate of acid and salt leaching. Specific placement of acid- and salt-containing
spoil should be aimed at preventing contact with percolating water or rising
water tables. We recommend placement based on chemical and physical spoil
properties that may affect water percolation and 0,., diffusion rates in the
profile. Both the deeper placement of acid spoil and coarser particle size can
substantially reduce the amount of acid drainage. Placement above the water
table with emphasis on percolate control may be better for high sulfate spoils,
while placement below the non-fluctuating water table may be better for pyrite
spoils.
viii
-------
CONCLUSIONS
Federal and State surface coal mining reclamation laws have set water
quality effluent standards both explicitly and implicitly. To achieve some of
these explicit standards, the emphasis is on effluent treatment. In contrast,
the groundwater quality standards are not explicit, but instead require that
practices or technologies be applied that minimize the degradation in ground-
water quality. Essentially, this calls for the proper application of an often
nonexistent technology to achieve minimum degradation. From the economic and
aesthetic viewpoint the onsite control methods often appear most attractive.
If necessary, when used in conjunction with the more expensive offsite effluent
treatment processes, onsite controls may help to achieve effluent standards at
a substantially reduced cost.
Selecting the onsite control technology to optimize relatively long-term
effluent and groundwater quality while minimizing short-term erosion may prove
complex and very difficult. By law or necessity, these decisions and their
continual updating fall to the appropriate State or Federal agencies, who must
designate acceptable technologies or guidelines for selecting these
technologies. Part of the problem is that the processes controlling acid
production and salt or trace metal losses are not sufficiently well understood
on a large enough scale for adequate design of appropriate control technologies.
Control methods such as those described here are based on assumptions, and they
contain limitations that often make them inappropriate for a particular site.
Hence, there is a great need for improved understanding of the systems, and for
-------
innovative application of this understanding to better and more efficiently
reclaimed surface coal mines.
Studies dealing with the physical and hydraulic properties of minesoils
indicated that surface mining and reclamation operations have resulted in
2 3
minesoils at 31 m /g specific surface and 1,763 kg/m bulk density. On mine-
soils studied, the bulk density first decreased and then increased with depth.
Between 10 and 1,500 kPa tensiometer pressure, most minesoils retained about
one-fourth as much water as did natural soils (35 vs. 136 mm). Evapotranspira-
tion from the minesoils could be approximated by class-A pan evaporation
results or modeled using an ET model.
Single ring infiltrometer data results and saturated hydraulic conductivity
of minesoils were variable. Exposure to the elements for 4 months lowered the
mean value of hydraulic conductivity for the minesoils at the site but increased
the variability.
The findings of this study have the following hydrologic implications.
Where geology is similar to that of the study area, surface mining will
generally result in a minesoil with low infiltration rates (high density and
low porosity). While evapotranspiration may approach potential rates when
water is available, hydraulic properties of minesoil may result in droughty
conditions and periods of plant stress. However, hydrologic changes resulting
from mining can be highly variable from site to site and will also vary over
time as a result of weathering processes.
Investigations were conducted to compare changes in soil morphology and
chemistry before and after surface coal mining and reclamation operations in
Clearfield County, Pennsylvania. Four minesoil pits located within a disturbed
area and four natural soil pits located in adjacent undisturbed areas were
described and sampled. The minesoils were classified as Udorthents, three of
-------
the natural soils were classified as Typic Dyst}rochrepts and one as an Aquic
Fragiudult. The most prominent feature of the minesoils was their high degree
of coarseness and high rock fragment content; their roots tended to concentrate
along soil coarse fragment interface. In general, the chemical constituents of
the minesoils were similar to those of the natural soils. However, as a result
of extended weathering more total bases have been leached from the natural
soils and significantly more extractable aluminum was found in them than in the
minesoils. The high content of carboniferous shale and coal fragments in the
minesoils affected organic carbon and nitrogen determinations, while comparison
of mineralogy suggested that the minesoils studied were derived from the same
materials from which the natural soils had originally developed.
Preliminary results of instrumenting two large caissons filled with strip
mine-spoil and applying water indicated that on the spoil alone, water initially
infiltrated faster and carried much less sediment with it than it did on top-
soiled profile. As we applied water to the caissons we also observed the
pressure buildup within the profiles. After wetting, oxygen concentrations
recovered faster in spoil alone than in topsoiled spoil. In both caissons,
infiltrating water reached the water table faster than expected from observing
the movement of the wetting front, this was attributed to flow through larger
channels.
Two-probe gamma density probe was well suited to measuring the rate of
wetting front advance in the spoil, while response time of porous cup tensiome-
ters was delayed. When there was a difference in temperature between the
applied water and spoil, thermocouple output registered temperature changes
within the profile caused by the wetting-front passage. Water redistribution
profiles indicated fast drainage on spoil alone and delayed drainage in the
topsoiled profile. Water table data showed that in both caissons more than 50
percent of applied water drained from the profiles studied within 24 h.
xi
-------
Spoil hydraulic properties were obtained experimentally in the laboratory
and successfully compared with water content and tensiometer pressure values
measured in the caissons. Large corrections appear necessary to adjust spoil
retentivity curves for the coarse fragment contents present in most
Pennsylvania spoils. This could be significant in simulation and model
building.
In a larger context of reclamation and management of spoils this study
suggests that total reliance on radioactive probes to monitor water content
and movement in the spoil materials may not be warranted. Rapid drainage
through large channels can easily go undetected unless water table behavior
is monitored closely at the same time.
A number of spoil properties need to be considered before deciding where
and how to place acid spoils. Generally these properties relate to and often
control percolation and 0 diffusion rates into the spoil bank. These include
porosity, water retention, and particle size distribution of the topsoil, the
layer below the topsoil, and the worst acid spoil layer. Depth below surface
and differences in particle size distributions can substantially reduce the
amount of percolate or slow the amount of 0 delivered. Analysis of both SO,
and pyrite concentrations in spoil can identify the worst acid spoil layer for
subsequent isolation and can indicate whether placement should be above or
below the water table. Placement above the water table with emphasis on
percolate control may be a better method for highly SO -contaminated spoil,
whereas placement below nonfluctuating water table may be best for highly
pyritic spoils low in SO,. In addition, the rate of acid production and loss
from spoil can potentially be controlled by increasing the spoil particle size.
We have concluded that the spoil percolate quality is generally stable.
The maximum or stabilized total acidity in caisson percolate increased with
XI1
-------
spoil depth or other chemically related spoil parameters except for extreme
hydrologic conditions. More normal variations in hydrology do not appear to
dominate percolate quality.
The spoil percolate quality appears related to both the spoil depth in the
caisson, and the quality of the spoil or the spoil extract. If the total
acidity of a spoil layer extract is substantially greater than that from over-
lying layers, this spoil layer will exert an identifiable and dominating
influence on percolate quality. Where not substantially greater, the spatial
and hydrologic variability will tend to mask the contributions from each layer
into what appears as a total acidity increase with depth. A substantial
increase of total- sulfate and pyrite content of the spoil matrix (about or
above 1 order of magnitude) corresponded to similar increases in the total
acidity of spoil extract and caisson percolate. Soil layers with extremely
high total pyrite or SO, matrix concentrations can also be used to delineate
the primary total acidity contributing zones in the spoil profile.
We found that the concentration of Cd, Cr, total soluble Fe, Hg, Mn and Zn
in spoil extracts exceeded EPA water quality drinking standards from one (Hg)
to all (Mn) spoil layers. The standards for Pb and Cu were not exceeded.
Generally, the trace metal concentration in the spoil extracts agreed closely
with those observed in the corresponding spoil percolate. The Cu and Zn con-
centrations in the percolate were considerably higher. In caisson 2, the
trace metal concentrations in the spoil extract, caisson percolate and the
caisson well agreed closely. This suggests that reduction in trace metal con-
centrations achieved by dilution or other processes occurs primarily after
entry into the groundwater or stream system rather than within the spoil
profile.
xxn
-------
The major chemical controls on acid products loss to groundwater from the
spoil particle are the acid products diffusion and the acid production rates.
The waters percolating to the groundwater table dissolve and remove acid
products from the surface of the spoil particle. The concentration of acid
products at that surface is likely controlled by either the diffusion rate of
acid products from interior oxidation or storage sites, or the acid production
rate controlled by either the kinetics of pyrite oxidation or the diffusion of
0- to pyrite oxidation sites within the particle.
Pyrite oxidation in the spoil particle, potentially the most important
process, appears primarily controlled by either the pyrite oxidation or 0
diffusion rate. For new, unweathered spoil, the experimentally determined
pyrite oxidation rate compared reasonably well with published values. The
rate of 0.16 mg SO generated/g pyrite/hour remained undiminished following
approximately a 1500 hour incubation period in 20 percent 0 . Thus, pyrite
oxidation rather than the 0 diffusion rates appeared to be controlling. As
weathering removes the shallow pyrites, the 0? diffusion or acid product
diffusion rate appears to be controlling.
Because the spoil particles used contain substantial concentrations of
acid products (3.1%) as well as pyrite (3.8%), the initial acid product losses
to groundwater can be very large, even in the absence of 0,.,, especially where
submerged or frequently flushed. As particle leaching progresses, diffusion
rather than percolation or groundwater flow rates becomes potentially control-
ling because of the very low diffusion coefficient of these acid products.
The acid product contributing zone was 0.022 cm depth for a 0.5 cm diameter
assumed cylindrical spoil particle which amounted to 17 percent of the total
particle volume. Thus, 83 percent of the spoil particle volume was
noncontributing.
xiv
-------
In cases where diffusion is limiting, the basic spoil particle size is
important in controlling the rate of salt and acid loss to groundwaters since
this reduces the acid contributing volumes relative to the total spoil mass.
Total acidity provides a reasonable estimate of other major chemical
parameters contained in spoil drainage. In caisson percolate, good linear
correlations existed between total acidity and Al, total.soluble Fe, Mg, and
SO concentrations. These exclude the lowest concentrations (<10 meq/1) which
normally do not represent the important acid contributing spoils. The relation-
ship of total acidity with Ca, ferrous Fe, Mn and pH was much less precise.
Concentration of Cu, Zn and Mn in spoil extract could be better approximated
from the sulfate concentration than from the pH values.
xv
-------
-------
CONTENTS
Foreword ±±±
Abstract v
Conclusions ±K
Figures xviii
Tables xxiii
1. Introduction 1
References 12
2. Site Description 19
Chemical Properties 19
Natural Soils 22
Minesoils 25
Root Density and Particle Size Distributions 26
Chemical Properties 28
Organic Carbon 31
Clay Mineralogy 33
Sulfur (S) 33
Spectrometric Analyses 34
Spectrographic Analyses 35
Physical Properties 36
Specific Surface 39
Bulk Density 42
Minesoil Water 45
Moisture characteristic 45
Water retention 47
Hydraulic conductivity 50
Infiltration 51
Evapotranspiration 54
References 55
3. Caisson Studies 61
Monitoring Water Movement Through Strip Mine
Spoil Profiles 61
Instrumentation 65
Methods 67
Results and Discussion 69
Modeling Water Flux on Strip-Mined Land 79
The Caissons 80
Flow Models 83
Model 1 83
Model 2 85
Results and Discussion 87
xvi
-------
CONTENTS (continued)
Chemical Considerations 93
Analysis, Sampling and Storage Methods 93
Chemical Properties of the Spoil 96
Chemical Properties of the Spoil Percolate ...... 100
Chemical Interrelationships of Spoil and
Spoil Percolate 104
Processes Controlling Acid Product and Acid Losses
from the Spoil Particle 109
Laboratory Setup and Procedure Ill
Calculation Method 115
Experimental Results 122
Diffusion of Acid Products to the Particle Surface
or 02 inco the Particle as the Controlling
Process 124
Rate of Pyrite Oxidation as the Controlling
Process 126
Interpretation of Results 127
Chemical Parameter Interrelationships 128
References 135
4. Field Studies 140
Physical and Hydrologic Setting 140
Chemical Setting 146
Water Quality 148
Site Response in Time 148
Controlling Processes 156
Site Response in Space 159
Mean values 159
Conclusions 162
References 164
5. Recommendations 165
Introduction 165
What Was Done 167
Subsidence of Materials and Changes in Density and
Porosity 168
Water Movement in the Profile 170
Chemical Effects of Atmospheric and Hydrologic
Isolation 173
Intraparticle Diffusion Processes Affecting Acid
Production and Loss 178
References 182
xvxi
-------
FIGURES
Number Page
2.1 Schematic diagram of study area and location of experimental
pits P1-P4 for the soils and P5-P8 for the minesoils 19
2.2 Cation exchange capacity on the <2 mm material (clear) and
total volume (shaded), base saturation, pH and total
acidity for the natural soils and the minesoils 29
2.3 Study area and location of minesoil pits, lysimeter sites,
and infiltration sites 39
2.4 Determination of bulk density by a modified excavation
technique (Bertram, 1973) 43
2.5 Water retentivity curve (gravimetric) for a minesoil
corrected and not corrected for coarse fragments 47
2.6 A microlysimeter site 51
2.7 Cumulative ET values during the study as determined by model
results, changes in lysimeter weight, and evaporation from
the class A pan 55
3.1 Backhoe excavates successive layers of spoil at Kylertown .... 62
3.2 Research staff assembles boxes for transport of Kylertown spoil
by truck to the experimental facility 62
3.3 Boxes are unloaded at the experimental research facility 63
3.4 Caissons at the experimental research facility to be filled
with Kylertown spoil 63
3.5 Numbered boxes are positioned near the appropriate caisson
preparatory to reconstituting the field profile 64
3.6 Inside of caisson 1 showing the 2.5' depth of spoil in place,
access tubes for measuring density and moisture and a
lysimeter for collecting the effluent, ladder was used by
personnel who rebuilt the field profile layer by layer 64
3.7 Caisson instrumentation description 65
XVlll
-------
FIGURES (continued)
Number
3.8 Caisson profile density before (original) and after (new)
water application 70
3.9 Water redistribution profiles for caissons 1 and 2 for
indicated times 74
3.10 Comparison among tensiometer pressures at 0.3-, 0.6- and 0..9-m
depths on caisson 1 and 2 for 27 and 15 days after water was
applied 75
3.11 (a) Water retentivity curves averaged for all depths for spoil
profile corrected and uncorrected for coarse fragments and
specific retention; (b) matched (Green and Corey, 1971)
hydraulic conductivity values as a function of water,
corrected for coarse fragments and specific retention;
(c) hydraulic conductivity as a function of tensiometer
pressure; curves labeled la refer to caisson 1, they were
matched at K = 0.13 m/hr, curves labeled 2 were for spoil
material in caisson 2 and were matched at K =1.89 m/hr. . . 76
sat
3.12 Schematic representation of the two caissons 81
3.13 Moisture characteristic (a) and hydraulic conductivity (b) as
a function of tensiometer pressure 82
3.14 Moisture and pressure profiles simulated with model 1 on a
hypothetical topsoiled spoil (a) , natural soil (b), and
nontopsoiled spoil (c). The numbers on the curves
indicate time in minutes from the beginning of water
application at the surface 88
3.15 Comparison of experimental (caisson 1) water contents obtained
on a topsoiled minesoil during the run with a moisture
profile at 0.8 h predicted by model 1 (3.15a); comparison of
experimental water contents taken 1 day (3.15b) and 1 month
after the run (3.15c) with moisture profiles at 1 day and at
1 month predicted by model 1. Solid circles (•) are
experimental (caisson 1) values at indicated times; open
circles (O) in (3.15c) show profile water content 44 days
after the run 89
3.16 Comparison of experimental water contents obtained on non-
topsoiled spoil profiles 2 h (•) and 30 days (±) after
the water application ceased with simulated profiles 91
xix
-------
FIGURES (continued)
Number Page
3.17 Comparison of experimental water contents on nontopsoiled spoil
for two intermittent rain periods (• and *) 41 h apart with
water content profiles simulated using model 2 92
3.18 Relation of total acidity from lysimeter drainage with spoil
and spoil extract analyses for corresponding spoil layers. . . 104
3.19 Relation of total acidity from lysimeter drainage with spoil
and spoil extract analyses for corresponding spoil layers. . . 105
3.20 Total acidity as related to flow duration on caissons 1
(7/18/78) and 2 (8/23/78) 107
3.21 Total acidity as related to quantity of leachate under
saturated conditions for spoil 21 109
3.22 Relationship of total salts (TS) with electrical conductivity
(EC) for selected column leachates 113
3.23 Relationship of total salts (TS) with electrical conductivity
(EC) for selected column leachates (expanded) 114
3.24 Relationship of SO/ with electrical conductivity (EC) for
selected column leachates 114
3.25 Relationship of SO. with electrical conductivity (EC) for
selected column leachates (expanded) 115
3.26 Assumed geometry and labelling of individual spoil particle. . . 117
3.27 Example of column leaching sequence following NL incubation. . . 117
3.28 Relationship between C^ availability, incubation time and the
acid product concentration in leachate (extrapolated to
zero dilution) 119
3.29 Expanded scale relationship between 0,, availability, incubation
time and the acid product concentration in leachate (extra-
polated to zero dilution) 120
3.30 Relationship of acid production rate for different incubation
periods under air and N? atmospheres 122
3.31 Relationship between two measures of total acidity 129
3.32 Relationship between total acidity and sulfate concentration . . 130
3.33 Relationship of calcium, magnesium, manganese and aluminum
with total acidity 131
xx
-------
FIGURES (continued)
Number Page
3.34 Relationship between total soluble iron and total acidity .... 131
3.35 Relationship of log Zn, Cu and Mn concentrations with pH 132
3.36 Relationship of Zn, Cu and Mn concentrations with sulfate
concentrations 134
4.1 Schematic diagram of the experimental site 140
4.2 Vegetative cover density (T/ha) at the experimental site 142
4.3 Relative distribution of infiltration at 60 minutes for a 65 mm
Summer storm (200 = 43 mm, 150 = 38 mm, 100 = 31 mm,
50 = 23 mm) 143
4.4 Water content at the surface (a) and at OJ [m (b) and water
table evaluations during a monitoring period in the
summer, site scales are in meters 144
4.5 Characteristic distribution pattern of solar radiation (a) and
of precipitation (b) at the experimental site 145
4.6 Site response to a spring storm 145
4.7 Typical oxygen profiles in August (a), and October (b) at the
experimental site 148
4.8a Overall response of the experimental site during the study
period: Water levels 149
4.8b Overall response of the experimental site during the study
period: Iron (Fe) concentration 150
4.8c Overall response of the experimental site during the study
period: Aluminum (Al) concentration 151
4.8d Overall response of the experimental site during study
period: Sulfate (SO.) concentration 152
4.9 Water table elevations (solid line) and SO, concentrations
(dashed line) during 500-700 day time segment, points
represent experimental values of SO, concentration 159
4.10 Average changes in pH (units), and iron (Fe), manganese (Mn),
magnesium (Mg), aluminum (Al), and sulfate (SO,) in mg/1,
outside scale is in meters 160
xxx
-------
FIGURES (continued)
Number Page
5.1 Water profiles in the topsoiled spoil prior to and following
ponding 170
5.2 Water profiles (a) in nontopsoiled spoil following wetting,
and (b) corresponding elevations of water table and amounts
of water applied 172
5.3 Total acidity as a function of (a) leachate, and (b) percolate
quantity at 1.2 m (•) , 2m (A), and 2.5 m (o) below the
spoil surface 176
5.4 Assumed structure of cylindrical spoil particle 179
5.5 Contributing spoil volume as a function of cylindrical
particle radius 180
xxii
-------
-------
TABLES
Number Page
2.1 Morphological Properties of Soils and Minesoils 23
2.2 Roots and Particle Size Distributions for Soils (Pit 1-4)
and Minesoils (Pit 5-8) 27
2.3 Iron Oxide and Extractable Aluminum in Minesoils (Pit 5-8)
and Contiguous Natural Soils (Pit 1-4) 30
2.4 Percentage of Organic Carbon, Nitrogen, C/N Ratio and
Clay Mineral Content of Minesoils (Pit 5-8) and
Contiguous Natural Soils (Pit 1-4) 32
2.5 Total Sulfate, Pyritic and Organic Sulfur Contents of
Selected Horizons 34
2.6 Quantitative Spectrometric Analysis of Selected Horizons
(Presented in Percentages) 35
2.7 Semiquantitative Spectrographic Analysis of Selected
Horizons (Presented in PPM) 36
2.8 Coarse Fragment Content (By Weight) of the Minesoil Horizons. . . 41
2.9 Organic Matter (OM) , 1500 kPa Water Content (W) and Specific
Surface (SS) of Minesoils 41
2.10 Surface Bulk Density (Mean), and Coefficient of Variation
(C.V.) of Minesoils Determined by Various Methods 44
2.11 Depth Bulk Density Distributions on the Minesoil, and the
Site Mean, Standard Deviation (SD) and Coefficient of
Variation (C.V.) 45
2.12 Minesoil Moisture Characteristics 46
2.13 Water Content of Minesoils (Sample) After 2 Days of Drainage,
of Clods Desorbed at 30 kPa, and Moisture Characteristic
Values (MC) at 10 kPa for Surface Horizons 48
2.14 Water Retained between 10 and 1500 kPa Tensiometer Pressure at
Four Minesoil Sites and at One Natural Soil Site 49
xxiii
-------
TABLES (continued)
Number page
2.15 Saturated Hydraulic Conductivity at 10 Lysimeter Sites (L-l
to L-10) Before (Initial) and After 4 Months of Field
Exposure (Final) Corrected to 20°C 52
2.16 Infiltration Velocities on Natural Soil, Topsoiled, and
Nontopsoiled Minesoil 53
3.1 Location of Instruments Caisson 1 and 2 66
3.2 Sediment Concentration in the Well and in the Effluent from
Lysimeters at Different Times After Water was Applied 70
3.3 Oxygen Concentration at Selected Depths After Water
Application 71
3.4 Depth of Water Infiltration with Time After Application 72
3.5 Distribution of Total Available Pore Space (TPS) and of Spoil
Water in Caissons with Depth 78
3.6 Water Depth, Seepage Flux, and Contained Volume in the Saturated
Zone After Water was Applied to Caissons 79
3.7 Selected Average Physical and Hydrologic Properties of
Caisson Materials 82
3.8 Experimental Values of Seepage Flux and Antecedent Moisture
Content (6) 86
3.9 Simulated Amounts, Duration and Intensities of Applied Rain
and Antecedent Water Content 86
3.10 Experimental and Simulated Profile Water Content Values,
Matching Factors and Their Reciprocals 91
3.11 Spoil Pyrite Analysis 97
3.12 Values of pH, Electrical Conductivity (EC), Total Acidity, Ca,
Mg, and SO, for Twenty-One Layers of Reconstructed Spoil
Profile. 98
3.13 Trace Metal Concentration of Spoil Extract 99
3.14 Chemical Characteristics of Spoil Drainage at Selected
Depths within Caissons 1 and 2 101
3.15 Trace Metal Contamination of Caisson Percolate and Well
Waters Taken from Caisson 1 and 2 for Run 1 103
xxiv
-------
TABLES (continued)
Number
Page
3.16 Incubation Sequences, Time and Composition of Eluant Gas
by Column 112
3.17 The Weathered Volume (Contributing) and Depth (Ar) as Related
to Different Diffusion Coefficients 125
3.18 Effect of Spoil Particle on Contributing Volume 125
3.19 Relationship of the Total Acidity Determined at pH 8.2 (y)
to Different Chemical Parameters (x) 129
4.1 Time of Occurrence of Major Storms and Associated Well
Elevation Minima and Maxima 154
4.2 Controlling Processes Affecting Groundwater Quality on
Stripmines 157
4.3 Acidity, pH, and Concentration of Selected Chemical Constituents
in Well Water on a Mine and Reclaimed Site Measured During
High and Low Water Levels 161
4.4 Average Water Levels, Acidity and Concentrations of SO,, Fe and
Al at High and Low Water Levels for Grouped Wells 163
5.1 The Sulfate, Sulfide and Organic Sulfur Content for Unmined
Coal Overburden 174
xxv
-------
-------
SECTION 1
INTRODUCTION
Current literature contains numerous studies of acid generation, neutralization,
and transformation in strip-mine spoil materials, yet we do not understand well the
rate-determining processes on the field scale. We know even less about the spoil
water flow, oxygen diffusion, surface runoff, erosion, evapotranspiration, and
temperature distributions within the spoil banks.
Prevailing economics, and the concerted national effort towards energy self-
sufficiency, virtually ensure strip mining of our shallow coal deposits. This
activity will cause considerable changes in the amount, distribution, and quality
of water in mined areas, and change the temperature and evapotranspiration (ET)
regimes. The oxidation of pyrite, and subsequent flushing of oxidation products,
could result in acid discharges from some areas. Modern methods of strip mining
and reclamation (Grim and Hill, 1974) are likely to alleviate many of the problems
by judicious placement and isolation of acid-producing materials, topsoiling, liming,
and rapid revegetation. Nevertheless, at present we have no single technique avail-
able to assess beforehand the degree of impact and the long-range effects of strip
mining on a given area. Extensive bibliographies, with abstracts on the subjects
of strip mining, mine drainage reclamation, and interstate water compacts are readily
available (NTIS, 1975; Bituminous Coal Research, Inc., 1964 through 1975; Maloney,
1975; Frawley, 1971). For clues to changes expected in soil profiles before and
after strip mining, we should look closely at the status of nitrogen, weatherable
minerals, and available water (Ahmad, 1971). Some conventional hydrologic studies
(Collier et al., 1970) showed, as expected, that past strip-mining activities
-------
significantly increased acidity and mineralization of surface water and ground
water. In other studies, Corbett (1965) and Agnew (1971) pointed out beneficial
effects, such as increased infiltration and water-storage capacity, as well as
delayed runoff and extended base flow, particularly during the dry periods.
Corbett and Agnew (1968) identified the "flushout" phenomenon common to many
strip-mine locations. Flushouts result from intense rains during periods of
low flow and drought, and may account for a large proportion of acid-loading of
streams (Corbett, 1969). Smith et al. (1974) also pointed out the positive
effects of strip mining, as in the control of slope and grade, and the possible
management of root-zone depth and weatherable mineral content. However, at the
same time, temporary organic matter deficiencies and weak soil structure will
probably prevail. If we know the properties of overburden in advance, we may
reconstruct soils by altering their physical and chemical characteristics to suit
better the growth of plants. Destruction of compact soil layers, apt to impede
movement of water and growth of roots (McCormack, 1974), has sometimes led to
improvement of soil environment for plant growth. Nevertheless, the continuous
long-term and recurring problem of pyrite weathering, and concurrent transport of
oxidation products still defies solution. The transport, in general, is
accomplished by percolating, seeping, or flowing water that feeds streams, drains
the area, or recharges the ground water. Although modern methods in strip mining
(Grim and Hill, 1974; USEPA, 1973) can minimize these effects, long-term impact
and legal aspects and implications (Goldberg, 1971; Goldberg and Power, 1973) are
still largely unknown. Smith et al. (1971), in studies of 70- to 130-year-old
weathered iron-ore spoil, suggested that reclaimed sites may have a higher density,
lower porosity, continuing lack of structure, and lower N and organic matter
contents, as well as poorer infiltration rates, than adjacent undisturbed soils.
However, their results also showed that properly placed mine soils provide
-------
increased water-holding capacity, greater root depth, and a more weatherable
mineral matrix.
Lysimeter studies (Lowry and Finney, 1962) have shed some light on the oxida-
tion of pyrite and leaching rates of the oxidation products within a small portion
of the profile under vertical flow conditions. Recently, Smith et al. (1974)
adapted selected chemical, physical, and mineralogical measurements, used to
classify natural soils, to coal overburden sections and strip-mine spoil in
West Virginia. Improved classification of mine soils, based on observable
properties of mine soil profiles, will most certainly provide the needed basis
for more precise management.
The actual' acid production and the quality of spoil water are due to modi-
fication of the potential acid production by neutralizing or the rate-limiting
factors. We could estimate potential acid production by leaching spoil samples,
by the analysis of total pyrite (VonDemfange and Warner, 1975), or by classify-
ing spoils to estimate the available or total pyrite. Analogously, we can
determine the neutralization potential by mineralogically based estimates of the
quantity of CO. materials.
Leaching study provides directly an estimate of readily available pyrite.
It is, however, a laborious site specific procedure. The total pyrite analysis
estimates potential acid production, but does not give a measure of availability.
Some use it directly as input to equations for determining the acid or
neutralization potential (Caruccio, 1973; VonDemfange and Warner, 1975). The
surface area exposed affects pyrite availability (OSU, 1970). Classification of
potential acid production, according to pyrite particle-size and amount associated
with selected strata, is currently under active investigation (Caruccio, 1973).
We can classify the dominant chemical processes according to acid generation,
neutralization, and transformation reactions. According to Singer and Stumm
-------
(1968), the summary reactions for the acid generation by oxidation of pyrite
minerals are:
FeS2(s) + 3.5 02 + H242~ + Fe2+ + 2H+, (1.1)
Fe2+ + 0.25 02 + H+ = Fe3+ + 0.5 H20, (1.2)
Fe3+ + 3 H20 = Fe(OH)34- + 3H+, and (1.3)
•*eS2(s) + 14 Fe3+ + 8 H20 = 15 Fe2+ (1.4)
+ 2SO 2" + 16H+.
4
In these reactions, four moles of H are being generated for each mole of FeS_
4- - 2-
consumed, two moles of H each eventually resulting from oxidation of 2S ->• 2SO.
and Fe ->• Fe The challenge is to relate the pertinent characteristics of
these reactions to acid production on a spoil bank scale.
Under certain conditions, laboratory tests show that reaction rate [1] depends
on water concentration, pH, texture of pyritic materials, and temperature.
Apparently, the concentration build-up of oxidation products, which would slow
acid production, does not occur when humidity is near 100%. Condensation of
gaseous H_0 on the reactant sites causes solubilization, and then reaction product
removal by dripping (Shumate et al., 1971). Laboratory-derived, simple relation-
ships between the reaction rates and pH, temperature, dissolved 0,, and relative
humidity are referenced in OSU (1971) , and specifically described in Smith et al.
(1968), OSU (1970), and Clark (1966).
Estimating acid production by reaction (1.4), which is directly dependent on
reaction (1.2), is much more complex. According to the literature, this requires
3+ 3+ 2+
an estimate of Fe concentrations and the Fe /Fe ratio as input to reaction
-------
3+ 2+
(1.4) (Smith et al., 1968). Under conditions of low Fe /Fe ratios (<0.3) and
high CL (15% in atmosphere), pyrite oxidation by reaction (1.1) greatly exceeds
the pyrite oxidation rate by reaction (1.4) (Shumate et al., 1971), and the
practical assumption that acid production depends directly on the dissolved CL
3+ 2+
concentration is reasonable. At low Fe /Fe ratios (20% ferric or lower),
experimental data have also shown (OSU, 1971) that microbially induced oxidation
is small and, consequently, acid drainage produced is small, as compared with
that produced independently of microbial activity. Several authors have argued
that, in the presence of 0«, acid generation is related to CL consumption, regardr
less of whether chemically or microbially induced oxidation dominates (Harvard
University, 1970; Shumate et al., 1971; Smith, 1974).
Additional simplification is given by Morth , who predicted acid production
reasonably well by relating the experimental rate of acid generation to the CL
concentration (in terms of partial pressure), rather than dissolved CL. If this
can be done, then the calculated or observed CL in the spoil atmosphere, and its
distribution down through the spoil bank, can relate directly to acid production.
In a laboratory study (NUS Corporation, 1971), acid, Fe, and conductivity result-
ing from pyrite oxidation were all noted as a function of 0. concentration (to
40%) in a carrier gas, indicating that partial pressure of CL, rather than
dissolved 0-, could be used directly to predict acid production.
The amount, reactivity, and distribution of CO. minerals affects the extent
of neutralization of acid effluent from the pyritic reaction sites. The CO-
A. H. Morth. 1971. Acid mine drainage: a mathematical model. Ph.D.
Thesis. Ohio State University, Columbus, Ohio.
-------
minerals can be calcite (Ca), dolomite (Ca, Mg), or siderite (Fe). However, if
siderite breaks down, the amount of H from Fe will be greater than the HCO
from siderite. The question remains to be answered whether siderite can in fact
produce significant alkalinity. Drainage waters will generally increase in pH
2~f* 2+ 2+
with a corresponding increase in Ca , Mg , or possibly Fe content, unless
3
2+ 2+
Ca concentration exceeds the solubility of CaSO,, or unless Fe is oxidized
and exceeds the solubility of the appropriate Fe compounds.
"Spoil" water behaves similarly to soil water to the extent that a given spoil
material corresponds to original soil. For example, for Pottsville series and
Lower Kittanning bed of the Allegheny series, the percentage of material <2 mm
(the soil-size fraction) varies between 17-64% in spoils, as compared with 35-95%
in the original soils (Plass and Vogel, 1973). Data of Ciolkosz et al. (1983) also
show wide variations in textures of mine soils. When larger sizes predominate,
spoil is no longer analogous to soil, and may have to be treated as rockfill (Leps,
1973, p. 91). Frequently, mine spoils have many air pockets (large voids), these
may contribute to the non-Darcy behavior of the spoil. Under those circumstances
Darcy's Law would not hold, and for saturated conditions turbulent flows could
prevail. Although Coleman does not explicitly indicate the amount of >25-min
materials in the spoils that he tested for infiltration, close examination of
his Figure 13 suggests that these materials compose about 60% of the total.
2
Coleman also found that the dark carbonaceous spoil banks in central
Pennsylvania had an infiltration rate about 10 times higher than the forest soil
2
G. B. Coleman. 1951. A study of water infiltration into spoil banks in
central Pennsylvania. M.S. Thesis. The Pennsylvania State University,
University Park, Pennsylvania.
-------
nearby. However, spoil material, derived from the yellow thick-bedded shales and
sandstones, had infiltration rates four to five times lower than wooded soil. In
strip-mining literature, soil is commonly considered as material <2 mm in diameter.
2
Coleman concluded that higher soil contents (39%) decrease the air space in the
yellow shale and mixed spoils, causing a decreased rate of infiltration, while
the lower soil content (20%) of the dark shales had the opposite effect. He also
showed that frost, compaction, and erosion decreased infiltration rates on the
spoils.
Brown obtained analytical flow nets for the three vertically separated flow
systems in Clearfield County, PA. He found an "upper" system present in the in-
terval between the topographic highs and the roofs of the deep mines. This system
controlled the amount of water entering the deep mines through the roof and along
the stripped outcrop. Vertical permeabilities were generally increased by
fracturing and collapse of the mine roofs.
o
Brown also found that the "middle" flow system was the most complex, with
discharge occurring to both the stream valley and into the top of the Connoquenes-
sing Sandstone formation. This flow system seemed to reflect local topographic
features. Finally, he found that the "lower" flow system was more regional in
extent and, generally, ignored local topography. He believed discharge occurred
along the deeply incised streams. In some areas, ground water flow seemed to be
3
controlled predominantly by fractures and joints. Brown concluded that by
defining the density, orientation, and permeabilities of these fractures for each
3
R. L. Brown. 1971. Shallow groundwater flow systems beneath strip and deep
coal mines at two sites, Clearfield County, PA. Ph.D. Thesis. The Pennsylvania
State University, University Park, Pennsylvania.
-------
rock type, and by relating the occurrence of fractures to structural features,
ground water flow systems could be defined, and unique flow solutions could be
obtained. Emrich and Merritt (1969) concluded that acid drainage from coal mines
may often pollute underlying aquifers through joints, fractures, and abandoned oil
and gas wells. Moving down the hydraulic gradient, it may discharge as Fe-rich
•
springs or flowing wells, or may pollute domestic or municipal water supplies.
Thames et al. (1974) assumed Darcian flow to model water transfer in
Arizona's Black Mesa spoil. Considering the heterogeneity of our spoil materials,
this approach may be open to question. Other than the work of Davis et al. (1964),
ElBoushi and Davis (1969), and ElBoushi (1975) on infiltration into rock piles,
not specifically related to strip mines, there seems to be no information in the
literature on infiltration and unsaturated flow in strip-mine spoil. The problem
is further complicated by the current practice of spreading topsoil over the
coarser spoil material. Thus, in any consideration of these flows, the fingering
process, due to the wetting front instability (Hill and Parlange, 1972), may need
to be taken into account.
Two sources of heat exist in spoil materials, one at the surface due to solar
radiation, the other within the spoil bank as a result of internal exothermic
chemical reactions. These include oxidation of pyrite, condensation of H70, and
possibly solubilization of acid-generated by-products. Solar heat can increase
the surface temperature to as much as 80°C on the dark organic shale and bituminous
coal (Deely and Borden, 1969). The high surface temperatures (particularly above
50°C) in spoils hinder the establishment of vegetative cover. Mulching with
organic materials has been somewhat successful in lowering the temperature and
conserving moisture (Kusko and Hutnik, 1974). The coarseness of the spoil materials
acts like a gravel mulch, tending to minimize the temperature fluctuations. Ahmad
and Ghosh (1971) found that diurnal fluctuations of temperature in the spoil at a
8
-------
60-cm depth were insignificant. From the standpoint of evapotranspiration (ET),
tops of plants growing on the spoil can experience extreme moisture stress, while
roots remained reasonably cool. In the literature there is no mention of an El-
model developed for, and operating on, strip-mined soils. Possibly, the
Blackland ET-model (Ritchie, 1972; Richardson and Ritchie, 1973) can be adapted
for this purpose.
Lovering (1948) first studied geothermal gradients. He showed that pyrite
oxidizing above a water table in rocks undisturbed by mining operations caused
noticeable inflections in the temperature gradient at a 200-m depth. Ahmad and
Ghosh (1971), used the differences in temperature to map acid-producing areas
in Ohio, and concluded that maximum ground temperatures at 60-cm depth corre-
sponded to a high percentage of pyrite content and low pH. The ground
temperature gradient seemed higher in acid-producing areas relative to nonacid-
producing areas. Their observations suggest a qualitative partitioning of a
strip-mine watershed based on detailed temperature surveys into acid-contributing
and acid noncontributing zones.
The term mode ling, when used in strip-mining literature, has different mean-
ings for different individuals. For example, Caruccio and Ferm (1974) discussed
a conceptual sedimentary model for coal. They concluded that, by identifying the
environment of deposition for a particular stratum, they could forecast the
occurrence and the mode of distribution of S in coal and associated rocks.
Miller and Thompson (1974) examined the effects of coal barrier width, using
the work of Cedergren (1967) on application of flow nets to seepage flows. They
concluded that compared to a 6-m coal barrier, a 15-m barrier decreased the
restricted flow 53%, and a 30-m barrier decreased it 75%.
Herricks et al. (1975) used the Stanford Watershed Model (SWM) to generate
streamflow sequences for areas with no recorded historical record, based on the
-------
sequences for an instrumented subwatershed. He also used simulated SO. concen-
4
trations in a conservative stream-loading model to provide an indication of the
amount and intensity of acid-mine drainage, and in addition obtained an indica-
tion of the erosive action of selected storm discharges through the use of sedi-
ment transport theory.
Sternberg and Agnew (1968) suggested a general analytical model of surface-
mine drainage. They obtained solutions for changes in ground water level and
flow that would occur in response to a uniform rate of seepage.
By far the most practical approach to modeling has come from Ohio State
University, and is best summarized in Shumate et al. (1971), Morth , Morth et al.
(1972), and Ricca and Chow (1974). In Shumate et al. (1971), a conceptual model
for pyrite oxidation was proposed, physical, chemical, and biological systems
involved were discussed, and some abatement techniques were considered.
Morth used Shumate's conceptual model to produce a mathematical model of
acid-mine drainage for a drift mine. The key assumption of this model, based on
a review of pertinent laboratory data, was that the oxidation rate was propor-
tional to the 0? concentration. Thus, Morth developed a program to predict 0?
concentration within small channels of coal or shale. The output of the program
was a set of 0- gradients for a given set of channel lengths and atmospheric
pressures. The hydrologic aspects of this program are somewhat less sophisticated.
The program is based on a simple material balance (In-Out = Accumulation) using
precipitation amount and estimates of infiltration capacity and evapotranspiration.
It is the hydrologic system of essentially undisturbed overburden, with water
seeping through "cracks and crannies" to reappear in an established water table,
that may at times intersect parts of mined-out areas. In strip mines, however,
the graded overburden is in a form of rubble and the acid-forming material may be
either buried deeply, distributed randomly, or placed otherwise through the spoil.
10
-------
The current literature survey suggests that much work has been done on the
basic chemical reactions of acid generation, neutralization, and transformation;
yet, the rate-determining chemical processes on the field scale are not well
defined. Even less has been done on the subject of water flow, air and CL
diffusion, evapotranspiration, temperature distribution, and possible surface
»
runoff. It is the purpose of this study to remedy the situation.
We plan in what follows to examine the impact of strip mining on the hydrology
of a mined area. In doing so we will examine experimental results obtained from
our laboratory and field studies and attempt to draw general conclusions applicable
to other areas.
'Ricca and Chow (1974) combined North's Acid Drainage Model with the Ohio
State version of Stanford Watershed Model (SWM) to produce daily mine-water
outflow and daily acid load from a watershed, on which the experimental mine
modeled by Morth and Morth et al. (1972) was located. They found that, generally,
the agreement between simulated and observed values was quite satisfactory. The
concept of using SWM to predict the impact of strip mining on an area has apparent-
ly not yet been tested. Unmined watersheds, however, have been modeled extensively,
and now some mined watersheds are being modeled. Because the calibration of SWM
requires abundant data, extension of other areas generally is based on the assump-
tion of similarity, which may not always be valid.
11
-------
-------
REFERENCES
1. Agnew, A. F. Coal Mining Hydrology and the Environment, Or Give the Devil His
Due. In: AIME Environmental Quality Conference, June 7-9, Am. Inst. Min.,
Metall. Pet. Eng., New York, N.Y., 1971. pp. 157-164.
2. Ahmad, M. U. Summary of the Proceedings of the Acid Mine Drainage Workshop.
Acid Mine Drainage Workshop, Athens, Ohio, August 2-6. Department of Geology,
Ohio University, Athens, Ohio, 1971. pp. 1-6.
3. Ahmad, M. U., and B. A. Ghosh. Temperature Survey of Coal Mines Producing
Acid Water. In: Proceedings, Seventh International Symposium on Remote
Sensing of Environment, 2:1109-1154. Willow Run Laboratory, University of
Michigan, Ann Arbor, Mich., 1971.
4. Bituminous Coal Research, Inc. Mine Drainage Abstracts, A Bibliography.
Bituminous Coal Research, Inc., Monroeville, Pa., 1964.
5. Bituminous Coal Research, Inc. 1965-1974 Supplements. Mine Drainage Abstracts,
A Bibliography. Bituminous Coal Research, Inc., Monroeville, Pa.,
6. Bituminous Coal Research, Inc. Reclamation of Coal-Mined Land, A Bibliography
with Abstracts. Bituminous Coal Research, Inc., Monroeville, Pa., 1975.
7. Caruccio, F. T. Characterization of Strip-Mine Drainage by Pyrite Grain Size
and Chemical Quality of Existing Groundwater. In: Ecology and Reclamation of
Devastated Land, Volume 1, R. J. Hutnik and C. David, eds. Gordon and Breach
Publishers, New York, 1973. pp. 193-226.
8. Caruccio, F. T., and J. C. Perm. Paleoenvironment Predictor of Acid Mine
Drainage Problems. In: Fifth Symposium on Coal Mine Drainage Research,
Louisville, Ky. National Coal Association, Washington, D.C., 1974. pp. 5-10.
9. Cedergren, H. T. Seepage, Drainage, and Flow Nets. John Wiley and Sons, Inc.,
New York, 1967.
12
-------
10. Ciolkosz, E. J., R. L. Cunningham, G. W. Petersen, and R. C. Cronce.
Characteristics Interpretations and Uses of Pennsylvania Minesoils.
Progress Report 381. The Pennsylvania State University, Agricultural
Experiment Station, University Park, Pa., 1983. 88 pp.
11. Clark. C. S. Oxidation of Coal Mine Pyrite. J. Sanit. Eng. Div., Am.
Soc. Civ. Eng. 92:127-145, 1966.
12. Collier, C. R., R. J. Pickering, and J. J. Munsen, eds. Influences of
Strip Mining on the Hydrologic Environment of Parts of Beaver Creek
Basin, Kentucky, 1955-66. Geological Survey Prof. Paper No. 427C.
U.S. Government Printing Office, Washington, D.C., 1970. pp. 1-58.
13. Corbett, D. M. Runoff Contributions from Cast Overburdens of Surface
Mining Operations for Coal, Pike County, Indiana. Water Resources
Research Center, Rep. No. 1 (1965). Indiana University, Bloomington,
Ind., 1965. pp. 1-67.
14. Corbett, D. M. Acid Mine Drainage Problems of the Patoka River Watershed.
Southwestern Indiana Water Resources Research Center, Rep. of Invest.
No. 4. Indiana University, Bloomington, Ind., 1969. pp. 1-173.
15. Corbett, D. M., and A. F. Agnew. Coal Mining Effect on Busseron Creek
Watershed, Sullivan County, Indiana. Water Resources Research Center,
Rep. of Invest. No. 2. Indiana University, Bloomington, Ind., 1968.
pp. 1-187.
16. Davis, S. N., R. D. Bernard, M. K. Botz, and C. N. Cullimare. Some
Hydrogeologic Factors in Water Contamination by Nuclear Cratering.
U.S. Atomic Energy Comm. HNS-1229-49, Las Vegas, Nev., 1964.
17. Deely, D. J., and F. Y. Borden. High Surface Temperatures on Stripmine
Spoils. In: Ecology and Reclamation of Devastated Land, R. J. Hutnik
and G. Davis, eds. Gordon and Breach Publishers, New York, 1969.
pp. 69-79.
13
-------
18. ElBoushi, I. M. Amount of Water Needed to Initiate Flow in Rubbly Rock
Particles. J. Hydro!., 27:275-289, 1975.
&•
19. ElBoushi, I. M., and S. N. Davis. Water Retention Characteristics of
Coarse Rock Particles. J. Hydrol., 8:431-441, 1969.
20. Emrich, C. H., and G. L. Merritt. Effects of Mine Drainage on Groundwater.
Ground Water, 7(3):27-32, 1969.
21. Frawley, M. L. Surface Mined Areas: Control and Reclamation of Environ-
mental Damage, A Bibliography. PB-203-448. U.S. Department of Interior,
Washington, B.C., 1971. 67 pp.
22. Goldberg, E. F. The Legal Framework of Acid Mine Drainage Control. Acid
Mine Drainage Workshop, August 2-6, Athens, Ohio. Department of Geology,
Ohio University, Athens, Ohio, 1971. pp. 76-85.
23. Goldberg, E. F., and G. Power. Legal Problems of Coal Mine Reclamation.
Water Pollution Control Research Series. EPA-14010 FZU 03/72, U.S.
Environmental Protection Agency, Raleigh, NC, 1973.
24. Grim, E. C., and R. D. Hill. Environmental Protection in Surface Mining
of Coal. EPA-670/2-74-093, U.S. Environmental Protection Agency,
Cincinnati, Ohio, 1974. pp. 1-277.
25. Harvard University. Oxygenation of Ferrous Iron. 14010-06169, Contract
PH 36-66-107, Federal Water Quality Administration, U.S. Department of
Interior, Washington, D.C., 1970.
26. Herricks, E. E., V. 0. Shanholtz, and D. N. Contractor. Models to Predict
Environmental Impact of Mine Drainage on Streams. Trans. ASAE 18(4) :
657-663, 1975.
27. Hill, D. E., and J.-Y. Parlange. Wetting Front Instability in Layered
Soils. Soil Sci. Soc. Am. Proc., 36(5):697-702, 1972.
14
-------
28. Kusko, M., Jr., and R. J. Hutnik. Effect of Mulches and Amendments on
the Survival and Growth of Vegetation Planted on Anthracite
Processing Wastes. Special Research Report SR-101. Pennsylvania State
University, University Park, Pa., 1974.
29. Leps, T. M. Flow through Rock fill. In: Embankment-Dam Engineering.
Casagrande Volume, R. C. Hirshfield and S. J. Poules, eds. John
Wiley and Sons, Inc., New York, 1973. pp. 87-107.
30. Lovering, T. S. Geothennal Gradients, Recent Climatic Changes, and Rate
of Sulfide Oxidation in the San Manuel District, Arizona. Econ. Geol.,
43(1):1-20, 1948.
31. Lowry, G. L., and J. H. Finney. A Lysimeter for Studying the Physical
and Chemical Changes in Weathering Coal Spoils. Ohio Agricultural
Experiment Station Research Circular 113, Wooster, Ohio, 1962.
32. McCormack, D. E. Soil Reconstruction: For the Best Soil After Mining.
In: Second Res. and Appl. Technol. Symposium on Mined Land
Reclamation, October 22-24, Louisville, Ky. National Coal Association,
Washington, D.C., 1974. pp. 150-162.
33. Maloney, F. E., ed. Interstate Water Compacts, A Bibliography. Water
Resources Sci. Inf. Center, Office of Water Research and Technol.
OWRT/WRSIC 75-205. U.S. Department of Interior, Washington, B.C.,
1975.
34. Miller, J. T., and D. R. Thompson. Seepage and Mine Barrier Width.
In: Fifth Symposium on Coal Mine Drainage Research, October 22-24,
Louisville, Ky. National Coal Association, Washington, D.C., 1974.
pp. 103-127.
15
-------
35. Morth, A. H., E. E. Smith, and K. S. Shumate. Pyritic Systems: A Mathematical
Model. EPA Technol. Series 14010 EAR, Contract No. 14-12-589, Report No.
665771. U.S. Environmental Protection Agency, Washington, D.C., 1972. 171 pp.
36. National Technical Information Service. Strip Mining, A Bibliography with
Abstracts. U.S. Department of Commerce, Springfield, Va., 1975.
37. NUS Corporation. The Effects of Various Gas Atmospheres on the Oxidation of
Coal Mine Pyrites. Water Pollution Control Research Series 14010 ECC-08/71.
U.S. Environmental Protection Agency, Washington, B.C., 1971.
38. Ohio State University Research Foundation. Sulfide to Sulfate Reaction
Mechanism. Water Pollution Control Research Series 14010 FPS-02/70.
Federal Water.Quality Administration, U.S. Department of Interior,
Washington, D.C., 1970.
39. Ohio State University Research Foundation. Pilot Scale Study of Acid Mine
Drainage. Water Pollution Control Research Series 14010 EXA-03/71.
U.S. Environmental Protection Agency, Washington, D.C., 1971. 84 pp.
40. Plass, W. T., and W. G. Vogel. Chemical Properties and Particle Size
Distribution of 39 Surface-Mine Spoils in Southern West Virginia. USDA
Forest Service Research Paper NE-273. Northeastern Forest Experiment
. Station, Upper Darby, Pa., 1973.
41. Ricca, V. T., and K. Chow. Acid Mine Drainage Quantity and Quality
Generation Model. Trans. Soc. Min. Eng., Amer. Institute of Mining
Eng., 256(4):328-336, 1974.
42. Richardson, C. W., and J. T. Ritchie. Soil Water Balance for Small
Watershed. Trans. ASAE, 16:72-77, 1973.
43. Ritchie, J. T. Model for Predicting Evaporation from a Row Crop with
Incomplete Cover. Water Resour. Res., 8(5):1204-1213, 1972.
16
-------
44. Shumate, K, S., E. E. Smith, P. R. Dugan, R. A. Brant, and C. I. Randies.
Acid Mine Drainage Formation and Abatement. Water Pollution Control
Research Series Program 14010 FPR. U.S. Environmental Protection Agency,
Washington, D.C., 1971.
45. Singer, P. C., and W. Stumm. Kinetics of the Oxidation of Ferrous Iron.
In: Proceedings, Second Symposium on Coal Mine Drainage Research,
May 14-15, Pittsburgh, Pa. Coal Industry Advisory Comm. to Ohio River
Valley Sanitation Commission, 1968. pp. 12-34.
46. Smith, E. E., K. Svanks, and K. Shumate. Sulfide to Sulfate Reaction Studies.
In: Proceedings, Second Symposium on Coal Mine Drainage Research,
May 14-15, Pittsburgh, Pa. Coal Industry Advisory Comm. to Ohio River Valley
Sanitation Commission, 1968. pp. 1-11.
47. Smith, J. J. Acid Production in Mine Drainage Systems. In: Extraction of
Minerals and Energy: Today's Dilemma, R. A. Deju, ed. Ann Arbor Science
Publishers, Inc., Ann Arbor, Mich., 1974. pp. 57-75.
48. Smith, R. M., W. E. Grube, Jr., T. Arkle, Jr., and A. Sobek. Mine Spoil
Potentials for Soil and Water Quality. EPA-670/2-74-070, Environmental
Protection Technol. Series, U.S. Environmental Protection Agency,
Cincinnati, Ohio, 1974. 303 pp.
49. Smith, R. M., E. H. Tryon, and E. H. Tyner. Soil Development on Mine Spoil.
Bull. 604T. West Virginia University, Agricultural Experiment Station,
Morgantown, W. Va., 1971.
50. Sternberg, Y. M., and A. F. Agnew. Hydrology of Surface Mining, A Case
Study. Water Resour. Res., 4(2):363-368, 1968.
17
-------
51: Thames, J. L., E. J. Crompton, and R. T. Patten. Hydrologic Study of a
Reclaimed Surface Mined Area on the Black Mesa. In: Second Research
and Appl. Technol. Symposium on Mined Land Reclamation, October 22-24,
Louisville, Ky. National Coal Association, Washington, B.C., 1974.
52. U.S. Environmental Protection Agency. Processes, Procedures and Methods
to Control Pollution from Mining Activities. EPA-430/0-73-011, U.S.
Environmental Protection Agency, Washington, B.C., 1973. pp. 1-390.
53. VonDemfange, W. C., and B. L. Warner. Vertical Bistribution of Sulfur
Forms in Surface Coal Mine Spoils. In: Third Symposium on Surface
Mining and Reclamation, October 21-23, Louisville, Ky. National Coal
Association, Washington, B.C., 1975. 1:135-147.
18
-------
-------
SECTION 2
SITE DESCRIPTION
CHEMICAL PROPERTIES
The experimental site (Figure 2.1) selected was a reclaimed 4 ha strip mine
adjoining undisturbed land and located near Kylertown in Clearfield County,
Pennsylvania, within the Pittsburgh Plateau section of the Appalachian Plateau
Province. The geologic system (Pennsylvanian) was characterized by cyclic
Location Mop
UNDISTURBED
Figure 2.1. Schematic diagram of study area and location of experimental
pits P1-P4 for the soils and P5-P8 for the minesoils.
19
-------
sequences of sandstone, shale, coal and clay, and underlain by flat-lying to gently
folded reddish, yellowish and brownish clay shale and yellowish brownish sandstone,
shale and siltstone. The coal strata included the middle and lower Kittanning
seams (C and B) of the Allegheny group, Millstone Run and Mineral Creek formations.
In general, sandstones overlay the middle Kittanning C-seam, with siltstones and
shales predominating above the lower Kittanning B-seam. The general topography of
the area was broad with rounded divides separated by narrow V-shaped valleys
(Glass, 1972). The mean elevation of the site was about 450 m with a northern to
northeastern aspect.
Before stripping, the vegetation of the area was hardwood forest, grasses,
legumes and trees have since become established on the site following reclamation.
The lower portion of the experimental site (B-spoil in Figure 2.1) was mined and
reclaimed in 1969 before passage of the Pennsylvania legislation requiring topsoil-
ing and burial of acid materials. Although the area has been contoured, topsoil
was not spread over the spoil. Beech, pine, and aspen have been planted on this
area. The upper portion of the area (C-spoil in Figure 2.1) was mined and reclaimed
in 1974 after passage of the stripmine legislation. As a direct consequence, 15 to
20 cm of topsoil was spread over the spoil, the site was limed and planted to grass
and legumes. The eastern half of the area (pit 8) was also heavily manured resulting
in a good stand of clover. On the mined area, the disturbed material extends to a
depth of about 10 m, and small pockets and larger tracts of original undisturbed soil
surround the area. Pit sampling was done in 1976, 7 years (B-spoil) and 2 years
(C-spoil) after reclamation.
Four pits (pit 1-4) about 2 m deep, 3 m long and 1.5 m wide were sited around
the periphery of the mined land on soils originally present in the area, and four
pits (pit 5-8) were randomly located within the reclaimed area (Figure 2.1).
Detailed morphological descriptions of the profiles were made according to the Soil
20
-------
Survey Manual (Soil Survey Staff, 1951), and the pedons were classified according to
the recent system of soil taxonomy (Soil Survey Staff, 1975). Root distribution
descriptions were made using the procedure outlined in Appendix II of Soil Taxonomy
(Soil Survey Staff, 1975). Four square decimeters outlined on an acetate sheet was
placed over the horizons and the number of roots were counted and averaged.
Bulk samples (~5 kg) were obtained from each horizon of the natural soils and
minesoils. Coarse fragments (>19 mm diameter) were sieved from the bulk samples
and weighed in the field. The material <19 mm was air dried and retained for
further analyses. Particle-size analyses of the samples from each horizon were
made by sieving and by the pipette method after dispersion with sodium metaphosphate
(NaPOJ and sodium carbonate (Na-CO-) (Kilmer and Alexander, 1949). Chemical analy-
ses were made on that fraction which passed through a 2-mm sieve. Soil reaction (pH)
was determined with a pH meter and glass electrode on three 1:1 (by weight) soil
solution suspensions, with distilled water, IN KC1, and 0.01 M CaCl2, respectively.
The basic cations (Ca, Mg, K and Na) were extracted with IN NH.OAc solution at pH 7
(Peech et al., 1947). Ca and Mg levels were determined by atomic absorption
spectrophotometry, and K and Na contents were determined by flame emission
spectrophotometry. Extractable acidity was determined by titration of a BaCl--
triethanolamine soil extract at pH 8.1 (Yuan, 1959). The Aluminon method (Yuan,
1959) was used to determine Al. Free iron oxides, extracted by Na dithionite-
citrate-bicarbonate, were analyzed for using orthophenanthroline. Organic carbon
content was determined by ignition with a Fischer induction furnace (Young and
Lindbeck, 1964) and by the Walkley-Black method (U.S. Salinity Laboratory, 1954).
Total nitrogen was determined by the Kjeldahl method (Jackson, 1958).
Quantitative spectrometric and semiquantitative spectrographic analyses, as
well as sulfur analyses on samples from representative horizons of soils and
minesoils were performed by the Mineral Constitution Laboratory, The Pennsylvania
21
-------
State University, University Park, Pennsylvania. Total sulfur was determined using
the high temperature combustion method (LEGO furnace). When total sulfur exceeded
0.05% of the sample by weight, pyritic, sulfate, and organic forms were analyzed
separately. Depending on the amount present, pyritic sulfur was determined either
by atomic absorption spectroscopy or by the Eschka method (D271-70), and sulfate
sulfur was obtained by precipitation with Bad (D2492-68) (ASTM, 1971).
Samples from each horizon were prepared for X-ray diffraction by removing free
iron oxides and organic matter (Anderson, 1963). Clay material (<2u) was removed
by centrifugation and two subsamples were obtained. One subsample was saturated
with Mg and the other with K. X-ray diffraction patterns were run on the room
temperature (25 C) Mg saturated slides and again after solvation with ethylene
glycol at 80 C. Patterns were also obtained from the room temperature (25 C) K
saturated slides, and after 2 hours of heat treatment at both 300 and 500 C.
The undisturbed Hazleton and Dekalb soils, at pits 1 and 2, respectively, are
within a forested area. Trees had been removed from the site on which the Cookport
and Hazleton soils were located (pits 3 and 4). The pedogenetic features common to
well developed soils were expressed in the soils at these four locations. In
contrast, soils at pits 5 through 8 are located within the area which has been
stripmined and reclaimed, and consequently exhibit characteristics of minimal
pedogenetic development.
Natural Soils
Pit 1. Structural development and horizon differentiation were evident in this
soil (Table 2. .1) . Clay films have accumulated in the B22t and B23t horizons due to
pedogenetic processes, however, the clay content does not qualify either of these as
diagnostic argillic horizons.
22
-------
TABLE 2.1. MORPHOLOGICAL PROPERTIES OF SOILS AND MINESOILS
Pll
)
2
3
4
5
6
7
8
Horizon
Al
A:
B21
B22t
B23i
C
AJ
B21
B22
B23
C
A)
A2
Bl
Bxli
Bx2l
Bx3l
Bx4i
Al
A2
B2H
B22l
B23t
IJC]
DC2
Ap
Cl
C2
C3
C4
o
Ap
AC
a
C2
Ap]
Ap2
a
C2
C3
C-s
AC
a
C2
c?
C4
C5
Depth (cm)
0-3
300
10-41
41-74
74-109
J09-176
0-10
10-33
33-51
51-71
71-74
0-15
15-13
43-66
66-91
91-117
117-135
135-167+
0-8
8-20
20-48
48-74
74-94
94-125
1 25-183 +
0-20
20-36
36-56
56-74
74-114
1)4-152
0-13
13-28
28-53
53-84
0-10
10-31
31-51
51-66
«~127
127-154
0-18
18-36
36-58
58-102
102-145
145-1834
Moist Color
Matrix Monies
JO YR 3/2
10 YR 5/4
10 YR 5/4
7.5 YR 5/4
& 4/4
7.5 YR 5/8
gw
CW
gw
gw
CW
cw
cw
e
p
aw
ai
aw
gw
gw
ds
CW
aw
gw
cv.
p
ab
ab
gw
cw
c*
as
gw
pw
gw
C)
•sil = sili loam. l = loajn. cnl = channcr\ loam, cnsil = channcry sih loajn, vcnsil=vcr>- channery sih loam. vcn) = ver>- channcry
loam. cl = clay loam, vcnsl=vcr>' channen sandy loam. vcnls = ven' channen,' loamy sand, vshl = v<:ry shaJy lo^m. vshsl = ver>
shajy s^ndy loam. vshb = vcr> shaJy lo^my sand, li — lo.
•cu=ciear wav> . cs-cicar smooth, ps ="pn.dual smooth, fu=graduaJ uavy, di = difTubC urcpular. j^craduaJ irrcjrular.
aw=abrup; uav\. aj-abrupi incnjlar. di = difTu->c smooth. ab = abnjpi broken. a> = aorup; smooih. ci = clear irrepular.
23
-------
Pit 2. This soil is similar to that of pit 1 except that it is shallower in
depth to bedrock. This pedon is well drained and moderately deep. Both pedons
had characteristics common to the Hazleton and Dekalb series (loamy-skeletal,
mixed, mesic). In both, the particle size control sections (25-100 cm) contain
less than 35% rock fragments by volume, and consequently the pedons do not fit
the modal concept of their respective series and are considered taxadjunct to those
series. They are both classified as coarse loamy, mixed, mesic Typic Dystrochrepts.
Pit__3. The Al horizon consisted of depositional material that had eroded from
adjacent stripmined areas. The very firm, thick, platy structure showed evidence of
compaction. The soil in pit 3 was also the only one with a fragipan (Bx horizons).
Clay films were evident in the Bxlt through the Bx4t horizons, and the Bxlt horizon
had sufficient illuvial clay to qualify as an argillic horizon. The mottles in the
Bx2t through Bx4t horizons was indicative of the impeded drainage. The fragipan
extended from 66 cm to the bottom of the pit at 167 cm.
This soil is moderately well drained. It differs slightly from the Cookport
series in having less clay in the particle size control section (25-100 cm) . The
pedon is classified a coarse-loamy, mixed, mesic, Aquic Fragiudult.
Pit 4. The surface horizon has received erosional deposits resulting from
stripmining and reclamation of adjacent areas. Beneath this horizon the soil was
undisturbed.
The soil was similar to that encountered in pit 1, except for a higher percent-
age of channers in the particle size control section (coarse-loamy, mixed, mesic
Typic Dystrochrept). In addition, a weathered coal seam (coal blossom) was found at
a 94 cm depth (IIC1) in this profile overlying a mottled underclay layer (IIIC2).
As in pit 1, this pedon contained less than 35% rock fragments by volume in the
particle-size-control section (25-100 cm) and did not fit the modal concept of
the Hazleton series.
24
-------
Minesoils
Pit 5. The surface horizon had moderate medium platy structure parting to
weak very fine subangular blocky structure indicating that soil forming processes
have occurred. The abrupt color change from the Ap horizon to the C horizons is
due to the topsoiled nature of this profile.
Pit 6. The Ap horizon had granular and weakly developed subangular blocky
structure. Vegetation has exerted its effect enhancing this development. The
transitional (AC) horizon had subangular blocky structure grading to structureless,
massive soil material. The C3 horizon was composed almost entirely of rock frag-
ments with about 15 to 20% void space.
Pit 7. The surface horizons, Apl and Ap2, had weakly developed fine subangu-
lar blocky structure grading to massive material at about 30 cm. The color of Ap
horizons in this profile was lithochromic. In the C3 horizon, no fine material
was present. The horizon was composed mainly of large rock fragments with 20%
void space.
Pit 8. A weak fine granular and a weak very fine subangular blocky structure
was present in the Ap horizon. Bands of different colored soil were present in
the C4 and C5 horizons. The C5 horizon (145 cm) was composed primarily of rock
fragments.
The minesoils in pits 5 through 8 are loamy skeletal, mixed, mesic members of
the Udorthent great group. The minesoils do not fit into any of the established
Udorthent subgroups. Some modifications in Soil Taxonomy criteria are necessary
in order to be applied to minesoils. The epipedons of the minesoils were the
only horizons that exhibited any pedogenetic development. The C horizons were
always structureless and massive with wavy horizon boundaries. Pits 5, 6, and
8 were located in C-spoil material. Consequently, their A to C horizon color
changes were primarily due to the soil material spread over the spoil during
25
-------
reclamation. Minesoil at pit 7 was located on the B-spoil, it did not have abrupt
color change, since no topsoil had been spread over this area during reclamation.
Root Density and Particle Size Distributions
Sieve analysis data (Table 2.2) show that a majority of materials in minesoils
(>75 percent on the average by weight) were coarse fragments. About 30% of the
coarse fragments in minesoils were larger than 76 mm (Pedersen et al., 1978). In
their fine earth fraction (<2 mm), most minesoils had significantly less silt and
more sand than the natural soils. The clay content, however, for both soils and
minesoils did not vary significantly from horizon to horizon nor from pit to pit.
Root penetration in soils (pit 1-4) was limited either by bedrock or a
fragipan. In pit 3, roots were numerous to 43 cm, below which they significantly
decreased in number and size to the fragipan (Bx) horizon. Few roots penetrated
the fragipan, some seemed to grow laterally along the upper boundary, and those
that penetrated the fragipan were located along prism walls. In pit 4, most of the
roots were located in the A horizon, and roots did not penetrate beneath the coal
blossom horizon (94-125 cm) . The forested profiles (pit 1 and 2) contained roots
throughout the solum.
In the minesoils, many very fine grass roots were present in the topsoiled
portion of the profile. These roots, at times tended to grow parallel to the
soil surface or spread on plates and ped surfaces. As the spoil became coarser,
fine fibrous roots were flattened on coarse fragments and between structural units.
Tap-roots were contorted, and there were fewer roots on the shale than on the
sandstone fragments. Few roots penetrated the massive C-horizons of the minesoils.
Thus, even though the minesoil profiles generally offer a deeper medium for root
proliferation, their coarse, structureless nature and resulting rapid drainage, and
low water retention capacity may limit root proliferation drastically.
26
-------
TABLE 2.2. ROOTS AND PARTICLE SIZE DISTRIBUTIONS FOR SOILS (PIT 1-4) AND MINESOILS (PIT 5-8)
I'arlicle Size
Distribution (mm)
Depth (cm)
I'll 1
0
3
10
41
7-1
109
I'll 2
0
10
33
51
71
Pil 3
0
15
43
66
91
117
135
I'll 4
I)
8
20
48
74
94
125
Values for 1
Horizon
Al
A2
1121
1)221
1)231
C
Al
1)21
1)22
1123
C
Al
A2
1)1
lixlt
I)x2l
I)x3l
Ux4l
Al
A2
l)2lt
!J22t
l)23l
IICI
IIIC2
Rools*
Many
Many
Many
Few
None
Many
Many
1'CW
None
Many
Common
Few
Many
Many
Few
None
'its 1-4: Average
Standard Deviation
% Coarse
Fiagmenut
23.2
23.1
322
33.9
14.3
57.4
41.7
37.7
38 3
40 5
45.8
55.7
38 5
44.7
49.6
36 5
29.0
49 5
34 1
35.5
34.2
38.8
34.5
31.2
35 1
37.4
9.K
Sand
2.0-
0.05
34.8
37.3
43.9
45 6
58.7
62.5
39.1
47 3
47.9
46 2
48.4
47 1
42.7
45.1
41 4
384
326
38.2
25.7
25.6
27.8
35.4
38.0
42 5
21.4
40.5
9.7
Sill
0.05-
0.002
(% <2 nun)
..55.6
51.1
42.0
40.3
29.7
23 4
46.7
38.8
39.0
40.9
41.4
37 5
44 0
42 2
44.3
43.4
47.4
44.4
58 0
57.1
54.3
46.7
47.2
47 4
53 7
44.7
8.0
Clay
<
0.002
9.6
It. 6
14.1
14.1
It. 7
14 1
14.1
13.8
13 1
12.9
10.2
15.4
13.3
12.7
14.3
18.2
200
176
16.3
17.3
17.9
17.8
14 8
10. 1
24.9
14.8
3.4
Depth (cm)
I'll 5
0
20
36
56
74
114
I'll 6
0
13
28
53
I'D 7
0
10
31
51
66
127
I'il 8
0
18
36
58
102
145
Values for 1
Hoi i/on
Ap
Cl
C2
C3
C4
C5
Ap
AC
Cl
C2
A pi
Ap2
Cl
C2
C3
C4
AC
Cl
C2
C3
C4
C5
Rools*
Many
Common/Few
None
Many
Many/Common
Few
Many
Common
None
Many
Few
None
Few
'Us 5-8: Average
Slaiulaul Deviation
% Coarse
Fragments!
64.4
84.6
79.7
79 4
804
90.0
73.3
71.1
79.3
92.8
82.2
54.1
864
88.9
660
K0.9
67.8
76.5
81.0
81.4
71.6
69.9
77.3
9.3
I'ailiclc Size
Distribution (mm)
Sand Sill
2.0- 0.05-
0.05 0.002
(% <2 mm)
45.5 383
61.4 26.6
62.3 27.2
62.1 25.4
62.3 24.3
64.6 22.2
47.6 35.7
53.0 310
58 8 26 1
66.5 20.3
51.7 31.9
53.6 28.9
53.5 327
61.4 28.1
65.7 25.5
64.6 26.1
55.1 35.1
54.8 33.1
50.6 37.1
50.0 38 9
48.0 41.7
41.1 46.5
56 1 310
7,3 6.7
Clay
<
0 (102
163
120
10.5
12.4
13.3
13.2
16.7
16.0
15 2
13.2
16.4
17.5
13.8
10.5
8.8
9.3
9.8
12.1
12.3
II. 1
10.3
12.4
12.9
2.6
*None = 0, few = (< I.OOO/m'), common = 1.000- lO.OOO/m*; many = (2*10,000/111*).
tFiagmenls>2 mm.
-------
Chemical Properties
Cation exchange capacity (CEC), cation exchange capacity corrected for coarse
fragment content, base saturation, pH, and total acidity for different horizons of
soils (pit 1-4) and minesoils (pit 5-8) are presented in Figure 2.2. The CEC
ranged from 8.5 to 61.2 meq/100 g in the natural soils and from 7.3 to 22.4 meq/100 g
in the minesoils. The high absorptive capacity of the carbonaceous matter (coal
blossom) in the IIC1 horizon accounted for the high CEC value of this horizon.
Whereas, the high CEC's of the Al horizons of pits 1 and 2 are attributed to the high
organic carbon contents. Most of the exchange sites on the soils studied were
+ +3
occupied by H and Al ions. Although the average CEC of the minesoils (17 meq/100
g) was higher than in the natural soils (14 meq/100 g), the differences were not
significant at the 5% level. However, when the CEC values in component horizons were
corrected for the coarse fragment content the differences became pronounced. A
3 meq/100 g nutrient content of corrected minesoils was low when compared to that for
the A and B horizons of the adjacent contiguous soils (13 meq/100 g). Therefore, on
coarse slowly weathering spoils nutrients availability may limit vegetative growth in
the absence of supplemental fertilization.
The pH values of the natural soils and minesoils appeared similar. Total
acidity of the natural soils was generally higher due to leaching of bases over
time. The high total acidities of the Al horizons of pits 1 and 2 are attributed
to the high organic matter content of these horizons and the effects of an over-
lying leaf litter, whose decomposition causes acidic conditions. The large dif-
ferences in total acidity for the AC horizon of minesoil in pit 8 resulted from
heavy liming and manuring of that part of the site.
In the natural soils, total acidity generally decreased with depth, whereas in
most minesoils (pits 6, 7, 8) it increased. On the average the natural soils and
minesoil in pit 7 had a greater percentage of iron oxides and extractable Al
28
-------
PIT HORIZON
DEPTH
icm)
CATION EXCHANGE CAPACITY
Im«q/io0gl
BASE SATURATION
1:1. soil: water)
TOTAL ACIDITY
imeq/ioogl
10
20
30
A I
A2
821
B22t
631
C
A I
B2i
B22
B23
C
A I
A2
8 I
B xit
B «2*
8.3)
B x4i
A i
A2
B 2lt
B 22*
B 23i
1 Ci
mc2
0-3
3-10
10-41
74-109
109-176
0-10
10-33
33-51
5l-7l
71-74
O-IS
15-43
43-66
66-91
117-135
135-167
0-8
8-20
20-48
48-74
74-9.4
94-125
125-163
a£3i&£&a_
10 20 30 2 4 6
I I I I I I I I
ttE
0 10 20 30 40
till
66.3
I
DO
NATURAL SOILS
PIT HORIZON
DEPTH
icml
CATION EXCHANGE CAPACITY
imeq/iOOgl
SASE SATURATION pH
in) u • i. soil: ~ater)
TOTAL ACIDITY
imtq/iOOgl
8
Ap
Cl
C2
C3
C4
cs
(AP
!•
Apt
AP2
C i
C2
C3
C4
AC
C I
C2
C3
C4
C5
0 10 20 30 4
1 1 1 1 1
0-20
20-36
36-56
56-74
74-1 14
P — r
! 1
M4-I52 1 !
0-13
13-28
28-53
53-84
0-10
10-31
31-51
51-66
66-127
127-157
0-18
18-36
36-58
58-102
102-145
145-183
fa I
*-~M I
LJ1
czzq
"
•A 1
I !
i — i
t^
fy^ \
1 .. I L .J. 1 I J . I I !
MINESOILS
Figure 2.2. Cation exchange capacity on the <2 mm material (clear) and total
volume (shaded), base saturation, pH and total acidity for the
natural soils and the minesoils.
29
-------
(Table 2.3). Presence of iron oxides and extractable Al is related to the weathering.
It appears that the natural soil horizons, because of their more weathered conditions,
contained significantly more iron oxides and extractable Al than the minesoils in
pits 5, 6, and 8. Minesoil in pit 7 which has been weathering for only seven years
also followed the same trend. The observed levels (Table 2.3) of Al could be toxic
to some plant species (Al > 1 to 2 meq/100 g).
TABLE 2.3. IRON OXIDE AND EXTRACTABLE ALUMINUM
IN MINESOILS (PIT 5-8) AND CONTIGUOUS
NATURAL SOILS (PIT 1-4)
Horizon
cm
Pit 1
0 Al
3 A2
10 B21
41 B22t
74 B23t
309 C
Pit 2
0 Al
10 B21
33 B22
51 B23
71 C
Pit 3
0 Al
15 A2
43 Bl
66 Bxlt
91 Bx2t
117 Bx3t
135 Bx4t
Pit 4
0 Al
8 A2
20 B2H
48 B22t
74 B23t
94 IJC1
125 II1C2
Average
Standard
Deviation
Iron
Oxide
%
3.3
2.8
3.1
3.2
2.6
2.2
2.4
2.1
1.9
1.8
1.9
4.1
2.2
2.7
3.0
3.3
3.7
4.5
3.0
3.3
3.4
3.8
4.0
1.7
4.0
3.0
0.8
hxtractable
Aluminum
meq/100 g
6.2
4.2
5.0
5.1
3.6
3.8
6.2
2.8
2.8
3.5
3.0
4.2
2.8
1.0
1.1
2.6
3.0
3.5
3.0
3.8
5.0
5.9
5.4
7.1
3.5
3.9
1.5
Horizon
cur
Pit 5
0
20
36
56
74
114
Pit 6
0
13
28
53
Pit 7
0
10
31
51
66
127
Pit 8
0
18
36
58
102
145
Ap
Cl
C2
C3
C4
C5
Ap
AC
Cl
C2
Apl
Ap2
Cl
C2
C3
C4
AC
Cl
C2
a
C4
C5
Iron
Oxide
%
1.7
1.9
2.1
1.8
1.8
1.8
1.2
:l-7
1.6
2.2
2.3
2.2
2.3
1.9
2.4
3.1
1.2
1.6
1.8
1.8
1.8
1.9
1.9
0.4
Extractable
Aluminum
meq/100 g
3.5
1.4
1.0
1.3
1.3
0.7
0.1
1.7
2.6
1.3
3.6
3.1
2.6
2.8
2.1
1.8
0.1
1.9
2.6
2.8
3.6
5.1
2.1
1.2
30
-------
Organic Carbon
In general, organic carbon content decreased with depth in the natural soils
and increased with depth in the minesoils (Table 2.4). Organic carbon contents
were highest in the partially decomposed coal seam (IIC1) of pit 4 and in the
horizons above B23t, and below (IIIC2) this seam. Whereas, in natural soils the
Al horizons in pits 1 and 2 have accumulated most organic matter from the forest
vegetation. The minesoil in pit 7 which was derived from carboniferous shale also
had a very high organic carbon content. Thus, the high carbon values of minesoils
are attributed to the presence of partly weathered coal or carboniferous shale
fragments rather than to vegetation remains.
The Walkley-Black method seemed to agree in general with the ignition method
values for the surface horizons of the natural soils, but gave lower values for
subsurface horizons. The values obtained by the Walkley-Black method for the
minesoils did not agree with those determined by ignition. The values obtained by
ignition were 1/3 to 6 times higher than those obtained by the Walkley-Black
method; however, this relationship was not constant and the AC horizon of pit 8
had higher organic C when determined by Walkley-Black method.
Apparently the organic C in minesoils cannot be accurately determined by
either of these methods when shale and coal fragments are present. Alternate
methods should be sought, which could give more reliable determinations.
The C/N ratios for three of the natural soil profiles decreased with depth
(pits 1, 2, and 3). However, high values of C in the coal blossom (IIC1) and
underclay (IIIC2) were responsible for higher C/N ratios in pit 4. The C/N
ratios obtained on the minesoils were very erratic because of high C values, and
were generally higher than could be attributed to either plant or animal
residues.
31
-------
TABLE 2.4. PERCENTAGE OF ORGANIC CARBON , NITROGEN, C/N RATIO AND CLAY MINERAL
CONTENT OF MINESOILS (PIT 5-8) AND CONTIGUOUS NATURAL SOILS (PIT 1-4)
U)
tsi
Clay Minerals
Horizon (cm)
Pit 1
0 Al
3 A2
10 U2I
41 11221
74 D23l
109 C
Pil 2
0 Al
10 1)21
33 D22
5 1 B23
71 C
Pit 3
0 Al
15 A2
43 1)1
66 Dxlt
91 Hx2t
117 I)x3t
135 13x41
Pil 4
0 Al
8 A2
20 1)211
48 0221
74 D23I
94 IICI
125 I1IC2
Average
Standaid
Deviation
C N
6.27 0.346
2.13 0.064
0.42 0.040
0.32 0.037
0.10 0.048
0.09 0.026
3.64 0.170
0.56 0.037
0.43 0.030
0.22 0.021
0.17 0.029
1.53 0.123
1.52 0.057
0.23 0.030
0.24 0.022
0.20 0.029
0.34 0.027
0.35 0.037
6.25 0.101
0.85 0.058
0.67 0.058
0.34 0.047
1.51 0057
12.59 0.347
2.03 0.059
1.72 0.076
2.85 0.088
•Hy ignition method.
C**Oiganic
carbon
C/N
18.1
33.3
10.5
8.7
2.1
3.5
21.4
15.1
14.3
10.5
5.9
12.4
26.7
7.7
10.9
6.9
12.6
9.5
61.9
14.7
26.2
7.2
26.5
36.3
34.4
—
M
5
15
5
tr
tr
tr
10
10
10
10
15
20
15
15
15
10
10
10
5
5
II
4
V
45
40
30
20
15
50
55
55
50
35
15
20
20
15
15
15
15
25
25
25
20
10
5
27
15
1 K
15 40
20 35
25 35
30 35
35 45
20 30
15 30
15 23
20 30
30 35
35 40
25 45
30 40
30 45
35 35
30 35
30 40
20 40
20 40
25 40
30 40
40 40
40 50
40 55
27 39
8 7
Q
ir
5
5
Ir
ir
Ir
5
tr
ir
ir
ir
ir
ir
ir
ir
ir
Ir
Ir
tr
5
Horizon (cm)
Pil 5
0 Ap
20 Cl
36 C2
56 C3
74 C4
114 C5
Pit 6
0 Ap
13 AC
28 Cl
53 C2
Pit 7
0 Apl
10 Ap2
31 Cl
51 C2
66 C3
127 C4
Pit8
0 AC
18 Cl
36 C2
58 C3
102 C4
145 C5
Average
Standard
0 Deviation
M = Monlmorillonite K =
V»Vermiculite
Q"
Kaolinilc
Quariz
C N
0.39 0.037
.81 0.068
.68 0.062
.87 0.049
.51 0053
.43 0.052
0.78 0.058
1.80 0.048
1.64 0.059
0.95 0.026
4.37 0.110
2.49 0.059
6.88 0.185
11.40 0.285
19.90 0.576
22.83 0.644
0.69 0.050
0.53 0.070
2.85 0.112
2.76 0.075
3.00 0.101
2.74 0.112
4.29 0.131
6.06 0.166
C/N
10.5
26.6
27.1
38.2
28.5
27.5
13.5
37.5
27.8
36.5
39.7
42.2
37.2
40.0
34.6
70.0
13.8
7.6
25.5
36.8
29.7
24.5
—
—
M
10
3
5
5
10
tr
Ir
5
ir
5
5
5
5
5
10
5
5
10
6
2
Clay
V
15
10
10
10
5
5
15
10
5
10
10
10
5
5
10
5
30
10
15
15
20
15
11
6
Minerals
1
35
45
40
45
40
45
30
40
50
40
40
40
45
40
40
40
20
45
35
35
35
30
39
7
K Q
40 tr
40 5
45 5
40 ir
45 5
40 5
40 3
45 5
45 tr
50
50 tr
45 If
45 5
45 5
40 5
40 10
35 10
40
40 Ir
40 5
35 5
40 5
42 6.
4 2
N^Nilrogcn
-------
Clay Mineralogy
The conversion of illite to vermiculite (Ciolkosz, 1978) is one of the
principal clay weathering transformations in Pennsylvania soils. Vermiculite
and kaolinite were the dominant clay minerals in the natural soils (pit 1-4),
while illite and kaolinite were dominant in the minesoils (pit 5-8)(Table 2.4).
The illite in the natural soils has been weathered to vermiculite while in the
minesoils much illite remains. Mineralogical data indicates that minesoils are
similar to the contiguous natural soils. However, the minesoils studied, all
weathering for less than 10 years, have not been exposed to as intensive
weathering as the natural soils.
The large amount of montmorillonite in pit 3 is attributed to the impeded
drainage (fragipan) in that profile, while topographic position of pit 4, and
to a certain extent 8, is probably responsible for the higher montmorillonite
content.
Sulfur (S)
The discussion covering S and the following two sections on spectrometric and
spectrographic analyses may be regarded as somewhat speculative in nature. We do
not know' the extent of natural variation associated with the elements considered.
Quite possibly some random degree of stratification in minesoils could be
responsible for higher or lower contents of certain elements as compared to natural
soils. Consequently, considerable caution should be exercised when ascribing any
of the reported results directly to the soil forming processes.
The total S content (Table 2.5) of natural soils was not high enough to warrant
component analyses into sulfate S, pyritic S, and organic S. The high sulfate S
content of the C2 horizon of pit 7 (0.16%) may be a result of the longer weathering
of this profile (7 years compared to 3 years for other minesoils) or due to the
33
-------
TABLE 2.5. TOTAL SULFATE, PYRITIC AND ORGANIC
SULFUR CONTENTS OF SELECTED HORIZONS
Pi!
1
2
3
4
5
6
7
8
Depth (cm)
3-10
41-74
0-10
33-51
15-43
46-66
117-135
0-20
4S-74
94-125
0-10
74-114
0-13
58-84
10-31
51-66
0-18
58-J02
HoniOD
A2
B22t
A3
B22
A2
Bl
Bx3t
A2
B22i
I1C1
Ap
C4
Ap
C2
Ap2
a
AC
a
Total'
0.03
0.03
0.04
0.03
0.04
0.03
0.04
0.03
0.03
0.19
0.03
0.13
0.03
0.13
0.10
0.38
0.04
0.08
Sulfate' Pyriiic Organic
0.01 0.03 0.15
0.05 0.07 0.01
0.05 0.08 0.00
0.05 0.02 0.03
0.16 0.07 0.15
0.04 0.04 0.00
Total sulfur*0.055c, no further analysis.
random stratification of the spoil material. Greater amounts of organic S forms
were found in the IIIC1 horizon of pit 4 and C2 horizon of pit 7. Pyritic S, the
principal component in the formation of acid mine drainage, was low in all of the
minesoil horizons yet higher than in the natural soils. Since highly acid
materials have been buried well below the sampling depth considered in this study,
they are not considered to be a major problem at this site.
Spectrometric Analyses
Quantitative spectrometric analyses (Table 2.6) reveal some trends, however,
little difference is noted between soil and minesoil.
The SiO« contents of the natural soils and minesoils are similar. The higher
values of SiO~ in the C4 horizon of pit 5 indicated that weathering at this depth
possibly has not exerted as much an affect as in the overlying horizons or may be
34
-------
TABLE 2.6. QUANTITATIVE SPECTROMETRIC ANALYSIS OF SELECTED
HORIZONS (PRESENTED IN PERCENTAGES)
PII Depth (cm)
]
2
3
4
Average
Standard
5
6
7
8
Average
Standard
3-10
41-74
0-10
33-51
15-43
43-66
117-135
8-20
48-74
94-125
Deviation
0-10
74-114
0-13
53-84
10-31
51-66
0-18
58-102
Deviation
Horizon
A2
B22t
Al
B22
A2
Bl
Bx3l
A2
B22l
J1C1
Ap
C4
Ap
C2
Ap2
C2
AC
C3
Wi. Ash
94.03
94.81
90.05
96.12
95.22
96.47
95.76
94.35
94.10
70.12
92.10
7.93
95.43
94.27
95.03
94.07
92.96
77.09
95.54
91.80
92.02
6.17
SiO;
79.0
76.5
84.0
86.0
84.0
83.0
81.5
78.5
74.0
71.0
79.75
4.84
82.0
90.5
82.0
79.0
78.5
73.0
86.0
77.5
81.06
5.40
AJ.O,
12.0
13.8
9.0
8.7
9.3
10.0
12.1
14.1
15.8
18.9
12.37
3.33
12.6
11.7
11.2
12.6
J3.3
15.5
8.3
14 J
12.42
2.15
TiO:
0.92
0.89
0.80
0.74
0.79
0.79
0.79
1.04
1.02
1.19
0.40
0.14
0.87
0.63
Q.S2
0.69
0.77
0.92
0.64
0.91
0.78
0.12
Fe,O3
4.75
5.75
2.92
2.71
2.65
3.30
4.58
4.55
5.95
5.30
4.25
1.26
4.22
5.62
3.93
6.80
5.23
8.75
3.85
6.55
5.62
1.70
MpO
0.78
0.61
0.76
0.39
0.39
0.49
0.75
0.69
0.91
0.68
0.65
0.17
0.53
0.49
0.48
0.54
0.51
0.72
0.37
0 69
0.54
0.11
CaO
0.11
0.14
0.12
0.12
0.09
0.17
0.17
0.11
0.08
0.32
0.14
.07
0.18
0.11
040
0.15
0.09
0.19
0.27
0.10
0.19
0.12
MnO
0.061
0.076
. 0.039
0.039
0.039
0.049
0.075
0.069
0 046
0067
0.056
0015
0.046
0.058
0.074
0.055
0088
0.074
0.064
o.io:
0.070
0.018
Na-O
0.33
0.27
0.29
0.24
0.25
0.28
0.47
0.37
0.40
0.03
0.24
0.12
0.26
0.12
0.24
0.12
0.14
0.19
0.20
0.30
0.20
0.07
K.O
1.96
2.24
1.43
1.46.
1.63
1.85
2.25
2.35
2.67
3.10
2.09
0.54
2.09
2.23
1.87
2.39
2.40
2.66
1.37
2.53
2.19
0.42
Total
99.9
100.3
99.4
100.4
99.1
99.9
102.7
101.8
100.9
100.9
102.8
101.5
101.0
102.4
101.0
102.0
101.1
102.9
due again to the random stratification of spoil material. Lower SiO values in C2
horizon of pit 7 indicate lower content: of sandstone, in the mixture of sandstone
and shale of the B-spoil. The higher values of Fe?0 for the deeper horizons of
<£ -)
minesoils may have resulted from the dissolution and redistribution of iron
sulfides. The greater proportion of K 0 in the minesoils as compared with the
natural soil horizons may indicate a more active weathering of illite.
Spectrographic Analyses
Semiquantitative Spectrographic analysis of select horizon material was con-
firmatory showing little difference between trace elements of soils and minesoils
(Table 2.7).
35
-------
TABLE 2.7. SEMIQUANTITATIVE SPECTROGRAPHIC ANALYSIS OF SELECTED
HORIZONS* (PRESENTED IN PPM)
Pn Depth (cm) Horizon Ba
]
2
3
3-10
41-74
0-10
33-51
15-43
43-66
117-135
4
Average
8-20
48-74
94-125
A2
B22l
Al
B22
A2
Bl
Bx3t
A2
B22I
1JC1
Siandard Deviation
5
6
7
8
Average
0-10
74-114
0-13
53-84
10-31
51-66
0-18
58-102
Ap
C4
Ap
C2
Ap2
C2
AC
C3
Standard Deviation
350
380
390
310
380
320
350
410
430
300
362
43
340
300
340
270
390
440
290
430
350
64
Be
<3
<3
<3
<3
<3
<3
<3
<3
<3
4
<3
<3
<3
<3
<3
<3
<3
<3
<3
<3
Ce
120
<100
<100
<100
<100
-OOO
<100
<100
ooo
120
<104
100
<100
120
<100
<100
'• 110
<100
<100
<104
Co
<20
<20
<20
<20
<20
<20
<20
<20
<20
39
<22
<20
<20
<20
<20
50
<20
<20
43
<27
Cr
74
100
74
56
70
66
84
100
100
120
84
20
76
80
86
60
96
110
50
100
82
20
Cu
23
31
16
16
15
18
24
20
32
54
25
12
22
26
21
30
35
50
12
50
31
14
Ga La
23 40
25 32
15 33
18 30
19 30
18 31
21 38
25 48
28 52
39 60
23 39
7 11
22 43
23 34
23 50
22 43
26 43
26 44
16 29
29 40
23 41
4 7
Ni
25
32'
<25
<25
26
<25
31
25
33
74
<32
<25
44
<25
27
27
40
<25
46
<32
Sc
12.0
15.0
6.6
8.0
8.4
10.0
12.0
12.0
17.0
25.0
13.6
5.6
13.0
13.0
12.0
13.0
14.0
18.0
5.4
16.0
13.0
3.6
Sr
41
47
35
30
50
43
50
64
86
160
61
38
42
46
45
45
64
120
41
80
60
28
V
96
100
66
62
72
.74
100
110
130
150
96
29
110
100
100
100
110
120
120
120
110
9
Y Zr
31 420
33 390
33 440
26 290
.26 340
36 340
29 '340
35 340
40 410
60 330
35 364
10 48
32 340
22 250
33 490
35 310
30 240
37 370
30 380
39 310
32 336
5 80
*Noi detected in all samples: Ag. Bi, Ge, Mo, Kb, Pb, Sn, Yb.
Where coarse shale fragments predominated in the deeper horizons of minesoils
(pits 7 and 8), the Cu content was slightly higher than that of natural soils,
whereas the Co content of the Ap2 of pit 7 and C3 of pit 8, attributed to higher
shale content in these profiles, exceeded the 1-40 ppm range reported for soils
(Baker and Chesnin, 1975). The Ba, Cr, V, Be, and Ni contents of the minesoils and
associated soils were within ranges expected for natural soils which indicates that
heavy metal toxicities would not be encountered on these minesoils.
PHYSICAL PROPERTIES
Typically, minesoils in Appalachia are low in organic matter and high in coarse
fragments, with weakly developed structure. Their surface horizons contain the
36
-------
greatest amounts of fine soil material, <2 mm in diameter. Large boulders and frag-
mented rock pieces are common. During the mining operation, overburden rock is
shattered by blasting, and as a result many "fines" essentially consist of pulverized
unweathered rock materials (Rogowski et al., 1977).
Here we will discuss the effects of surface mining and reclamation operations on
some physical properties of resultant minesoil. The objectives were to determine the
moisture characteristics and specific surface of minesoils, to establish the range
and variation in bulk density, to determine the hydraulic conductivity, and to
estimate initial evapotranspiration at the reclaimed site.
Bulk densities of minesoils especially at the surface are usually greater
than those of undisturbed soils because of their compacted state, lack of
structure, immature pedogenetic nature, and higher coarse fragment content
(Ciolkosz et al., 1983). Natural soils tend to be more porous with an intri-
cately developed system of cracks and fissures. Although, in general, the total
porosity is lower, pores in minesoils are typically larger.
Physical and morphological properties of both natural soils and minesoils
influence water retention and movement within a profile. Water retention
depends mainly on the amount of the <2 mm fraction and on structure. The degree
of tension with which the water is held determines its availability to plants.
Specific retention decreases with an increase in particle size (Farmer and
Richardson, 1976), resulting from a decrease in surface area and an increase in
pore size; however, it does not seem to be changed significantly by the shape
or type of fragments, nor by the time allowed for drainage (ElBoushi, 1966)1.
I. M. ElBoushi. 1966. Geologic interpretations of recharge through coarse
gravel and broken rock. Ph.D. Thesis. Stanford University, Stanford, California.
37
-------
It is difficult to obtain undisturbed samples of minesoils for determination of
hydraulic properties because of the large percentage of coarse fragments and weak
structural development of these materials. Sieved samples, particularly when cor-
rected for coarse-fragments content, may however give a useful estimate of hydraulic
properties, like moisture characteristic and hydraulic conductivity. Bruce (1972)
concluded that sieving of coarse-textured, organic-matter deficient materials did
not significantly modify their water retention properties. In general, research
results (Bruce, 1972; Unger, 1975) have indicated that if natural structure is
present, any type of sample disturbance will disrupt it, modifying pore-size
distribution and pore volume. However, in materials with little or no structural
development, like minesoils, this effect should either be minor or absent.
In the Appalachian minesoils we expect the storage of plant available water to
be significantly reduced because of their coarser texture, greater rock fragment
content and larger pores. Rates of infiltration and hydraulic conductivity usually
increase as the materials become coarser textured. The coarser texture, however,
restricts unsaturated moisture movement within the profile due to a decrease in
area for flow and increase in tortuosity (Mehuys et al., 1975). At times minesoils
may occupy up to 25% greater volume than the natural material (Van Voast, 1974),
frequently including randomly distributed large channels which allow rapid
drainage. Therefore, values for water movements into and through minesoils are
highly variable. ElBoushi (1966) has shown that water infiltrating through loose
granular material (similar to subsurface horizons of many minesoils) tends to
concentrate into discrete paths. Similar flows, but due to wetting front
instability, have been discussed by Hill and Parlange (1972), and more recently by
Raats (1973) and Philip (1975).
The experimental site (Figure 2.3) shows location of instrumentation, and
measurement sites. When sampling minesoils at pits 5, 6, 7, and 8 we could only
38
-------
UNDISTURBED
Symbol
itolion
Q Pin
'9 M.I S.I.
O Inl.l,..,.„„ P10
Ji Co-toil
I Bulk D.".Hr Sp°'' B«-"<».,
Figure 2.3. Study area and location of minesoil
pits, lysimeter sites, and infil-
tration sites.
obtain a few clods near the surface at pits 5 and 7, and none at all below the
0.2 m depth and at the other two pits. In addition to clods, bulk samples
(about 5 kg) were collected from each horizon down to a depth of 1.5 to 2.0 m.
Specific Surface
Specific surface area (SS) of the mineral fraction of minesoils in meters
squared per gram was calculated, using an equation of Young and Onstad (1976)
SS = - 2.36 + 7.96 (W) - 4.49 (OM) (2.1)
39
-------
In equation (2.1), W is the percent gravimetric water content at 1,500 kPa corrected
for coarse fragments and OM is the percentage of organic matter estimated by
multiplying percentage C (Walkley-Black) by a factor of 2.5, as suggested by
Broadbent (1953) for subsoil materials also corrected for coarse fragments.
In general, the correction for coarse fragments of properties such as organic
matter involved multiplying the value obtained by the percentage of material <2 mm,
while correction of water retention values such as 1,500 kPa values (W) involved
multiplying them by percentage of material <2 mm and adding a correction factor.
Correction factors were in turn obtained fay determining experimentally (ElBoushi,
1966) the amount of water retained by coarse fragments and multiplying these
values by percentage of material >2 mm. This amount of water retained by sandstones
(pits 5, 6, and 8) was 0.0282 g/g and for shale (pit 7) it was 0.0687 g/g. Thus,
the convections for sandstone (CSD) and shale (CSH) coarse fragments may be written
CSD = U (1-B) + 0.0282B (2.2)
CSH = U (1-B) + 0.0687B
where U is the uncorrected value of water retention determined in the laboratory
on the <2 mm material and B are the amounts of coarse fragments >2 mm present in
the respective horizon (Table 2.8) of the minesoil.
Table 2.9 lists organic matter, 1,500-kPa water content, and specific surface
values for the minesoils. Relatively high organic matter values reported for the
C horizons of minesoils may in part be due to inclusion of coal fragments (pit 7),
or to mixing in of plant debris during reclamation (pit 5, 6, and 8). Originally
the area was forested. While the tree trunks were removed during land preparation
before mining commenced, much of the rest was mixed with the material later
comprising upper layers discussed here. Organic matter contents exhibited much
larger variation among sites and horizons as measured by CV values than did 1,500
40
-------
TABLE 2.8. COARSE FRAGMENT CONTENT (BY WEIGHT) OF THE MINESOIL HORIZONS
Horizon
Ap
Cl
C2
C3
Ap
AC
Cl
C2
Coarse fragmencs
8/g
Pic 5
0.644
0.846
0.797
0.794
Pic 6
0.733
0.711
0.793
0.928
Horizon
Apl
AP2
Cl
C2
AC
Cl
C2
C3
Coarse fragmencs
g/g
Pic 7
0.822
0.541
0.864
0.889
Pit 8
0.678
0.765
0.812
0.814
TABLE 2.9.
ORGANIC MATTER (OM)*, 1,500 kPa WATER CONTENT (W)f,
AND SPECIFIC SURFACE (SS)t OF MINESOILS
Horizon
Ap
Cl
C2
C3
Ap
AC
Cl
C2
Apl
AP2
Cl
C2
AC
Cl
C2
C3
Mean
C.V.
OM
Pic 5
0.16
0.22
0.30
0.29
Pic 6
0.48
0.70
0.42
0.08
Pic 7
0.90
1.40
0.73
0.65
Pic 8
0.75
0.69
0.61
0.51
0.56
58.931
w
4.38
3.29
3.45
3.52
4.09
3.90
3.47
3.00
6.83
6.94
6.86
6.91
3.66
3. 70
3.58
3.56
4.45
33.331
SS
31.8
22.3
26.4
25.2
28.0
25.5
23.3
21.2
47.9
46.6
49.0
49.7
23.4
24.0
23.4
23.7
30.7
35. OX
*Walkley-Black OM - 2.5C, correcced for coarse fragmencs
fGravimecric, correcced for coarse fragmencs,
fSS from equacion (4.1).
41
-------
kPa values. The values of specific surface obtained for the minesoils appear to be
typical of the materials low in montmorillonite. Comparison of the A horizons of
the minesoils with the A horizons of the adjoining natural soils (not listed in
2
Table 2.9) showed them to be somewhat less (33.9 vs. 38.3 m /g), while a comparison
of G horizons suggested values on minesoils to be about two-third of those on
adjoining natural soils.
Pit 7 values were the highest primarily because of a higher water content at
1,500 kPa. This results since more water was found experimentally to be retained
on shale (0.0687 g/g) than on sandstone fragments (0.0282 g/g) used in the cor-
rection factor (equations (2.2) and (2.3)) computations. On the average, values
for other pits were about the same. Both the organic matter values and the
1,500-kPa values were corrected for the coarse fragment content. Thus, the results
represent specific surface area of the minesoil as a whole. Somewhat higher values
2
for minesoils (36.7 vs. 31.0 m /g) would be reported if this correction has not
been applied. Young and Onstad (1976) concluded that equation (2.1) when compared
with measured values of specific surface could explain about 86% of the variation.
Bulk Density
Bulk density near the minesoil surface was measured at 10 lysimeter sites
randomly located within the reclaimed area (Figure 2.3). Five replicate readings
around each site were made with a direct transmission gamma probe (Blake, 1965;
Troxler, 1970), at 0.15- and 0.30-m depth. Bulk density was determined on clods
collected near lysimeter sites and bulk density was also determined gravimetrical-
ly using a modified excavation technique (Bertram, 1973). In this technique, we
placed a steel ring (1 m in diam.) on the spoil surface at each of the 10 sites
and removed loose stones and excessive vegetation (Figure 2.4). A pliable rubber
liner was then fitted inside the ring and the volume of water needed to fill the
ring was determined. The minesoil within the ring was then removed to about the
42
-------
Figure 2.4. Determination of bulk density
by a modified excavation
technique (Bertram, 1973).
0.5-m depth and weighed. Small grab samples were taken to determine gravimetric
water content. The rubber liner was then replaced in the excavation and water
added to the same level as before. From volume differences and soil weight, cor-
rected for water content, dry bulk density was determined.
In addition, minesoil density to 1.5 m was determined at seven locations
(Figure 2.3) with the depth density gamma probe and corrected for water content
obtained with neutron scatter equipment. Access tube installation (Figure 2.3)
in the stony minesoil materials followed .Rawitz (1959).
43
-------
The average surface bulk densities obtained by various methods (Table 2.10) were
not significantly different from each other at the 5% level. The excavation technique
had the highest coefficient of variation, whereas the clods had the lowest. Results
also showed that the means of the sites did not vary significantly from one another.
It appears that the direct transmission gamma probe method is well suited to non-
destructive measurement of minesoil bulk density. It is by far the most convenient
and rapid of the methods tried. However, where very large fragments prevail, the
excavation method could prove superior.
TABLE 2.10. SURFACE BULK DENSITY (MEAN), AND
COEFFICIENT OF VARIATION (C.V.) OF MINESOILS
DETERMINED BY VARIOUS METHODS
Method Mean C.V.
kg/m2 %
Excavation 1.810 13.3
Gamma probe 1.700 11.2
Clods* 1.780 6.7
*Desaturated at 30-kPa tensiometer pressure.
In Table 2.11 depth bulk density measurements of minesoil at seven sites are
given. On the average the density in the 2-m profiles increases with depth and the
degree of variation (as measured by the coefficient of variation) appears to
decrease. Low density values (<1,000 kg/m ) are probably indicative of large spaces
and cavities within the spoil, while higher densities may imply proximity to solid
shale or sandstone fragments, or degree of equipment compaction at shallow depth.
Despite the diversity of locations, there was little apparent difference in bulk
density between the sites both on the topsoiled and nontopsoiled profiles when low
values were excluded.
44
-------
TABLE 2.11. DEPTH BULK DENSITY DISTRIBUTIONS ON THE MINESOIL, AND THE SITE MEAN,
STANDARD DEVIATION (SD) AND COEFFICIENT OF VARIATION (C.V.)
Depth Dl D2 D3 D4 D5 D6 D7 Mean 3D C.V.
0 k.g/m2 Z
0.305
0.610
0.914
1.219
1.524
1.829
2.134
1.730
1.670
1.650
1.720
1.060*
640*
1.790
1.360
1.210
1.610
1.660
1.660
1.620
1.630
1.240
890*
1.600
1.510
1.510
1.530
1.590
1.740
1.680
1.130
670*
730*
760*
1.590
1.390
1.590
1.600
1.530
1.580
1.580
1.630
1.400
1.590
1.610
1.580
1.600
1.550
1.540
1.630
1.450
1.600
1.530
1.650
1.650
1.570
1.499
1.532
1.543
1.588
1.600
1.586
1.620
199
178
183
84
54
44
81
13
12
12
5
3
3
5
*Value noc included in compucacion of mean.
Minesoil Water—
Moisture characteristic—Gravimetric moisture characteristics of sieved (<2 mm)
materials from each horizon of the minesoils was obtained experimentally in tripli-
cate by desorption (Richards, 1965). Subsequently, these values were corrected for
coarse fragments. Since clods could not be obtained at the pits on all but two
horizons of the minesoils, we have no measure of profile bulk density other than
gamma probe readings at different sites. The moisture data are therefore presented
on the gravimetric basis. Indeed, because of high coarse fragment content, with
many fragments being of boulder proportions, we may question advisability of
presenting the moisture data on other than gravimetric basis. Under conditions of
abundanc rainfall minesoils can supply ample water to plants (Plass and Vogel, 1973)
At times it even appears that minesoils have greater amounts of available water in
comparison to natural soils (Verma and Thames, 1975). This apparent increase in
availability often results because water in minesoils is normally held at lower
tensions than in corresponding natural soils.
The textural differences and organic matter deficiencies of minesoils account
in general for their lower water holding capacities, and when corrected for coarse
fragments, the minesoils actually retain substantially less water than comparable
natural soil horizons. Table 2.12 gives moisture characteristic data (gravimetric)
at four pits corrected and uncorrected for coarse fragments. Figure 2.5 shows
45
-------
TABLE 2.12. MINESOIL MOISTURE CHARACTERISTICS
Oepch Horizon 0
at
0
0.20
0.36
0.56
0
0.13
0.28
0.53
0
0.10
0.31
0.51
0
0.18
0.36
0.58
fCorrected
Ap
Cl
C2
C3
Ap
AC
Cl
C2
Apl
Ap2
a
C2
AC
Cl
C2
C3
for
0.4205*
0.1678t
0.3271
0.0734
0.3267
0.8888
0.3483
0.0931
0.4250
0.1341
0.3457
0.1200
0.3409
0.0929
0.2921
0.0472
0.3087
0.1114
0.3309
0.1890
0.3815
0.1112
0.3680
0.1019
0.4277*
0.1843t
0.3912
0.1445
0.3969
0.1310
0.3779
0.1262
coarse fragoe
Tensiomecer pressure (kPa)
1
0.3494
0.1425
0.2856
0.0678
0.2786
0.0790
0.2828
0.0798
0.3777
0.1215
0.2900
0.1039
0.2852
0.0814
0.2270
0.0425
0.2668
0.1039
0.2954
0.1727
0 . 3092
0.1014
0.3034
0.0947
0.3954
0.1739
0.3558
0.1362
0.3621
0.1244
0.3503
0.1211
tnC3 .
2
0.3337
0.1369
0.2727
0.0658
0.2650
0.0762
0.2696
0.0771
0.3577
0.1162
0.2716
0.0985
0.2301
0.0803
0.2255
0.0424
0.2568
0.1022
0.2779
0.1647
0.2972
0.0998
0.2970
0.0940
0.3523
0.1600
0.3261
0.1292
0.3276
0.1179
0.3206
0.1155
4
0.2975
0.1241
0.2337
0.0598
0.2249
0.0681
0.2289
0.0688
0.3115
0.1038
0.2343
0.0878
0.2564
0.0754
0.1930
0.0400
0.2328
0.0979
0.2498
0.1513
0.2525
0.0937
0 . 2509
0.0889
0.3102
0.1465
0.2899
0.1207
0.2890
0.1105
0.2835
0.1086
6
0.2766
0.1166
0.2208
0.0579
0.2158
0.0663
0.2141
0.0658
0.2892
0.0979
0.2081
0.0802
0.2193
0.0677
0.1690
0.0383
0.2007
0.0922
0.2134
0.1351
0.2244
0.0899
0.2156
0.0850
0.2921
0.1406
0.2676
0.1155
0.2708
0.1071
0.2692
0.1060
8 10
.in ... a f IT
Pit 5
0.2650 0.2569
0.1125 0.1096
0.2128 0.2098
0.0566 0.0562
0.2017 0.1989
0.0634 0.0628
0.1999
0.0629
Pic 6
0.2687
0.0924
0.1983 0.1930
0.0774 0.0758
0.2084 0.1999
0.0647 0.0637
0.1492 0.1426
0.0369 0.0364
Pit 7
0.1930 0.1876
0.0908 0.0898
0.2005 0.1946
0.1292 0.1265
0.2084 0.1999
0.0877 0.0865
0.1924 0.1875
0.0824 0.0819
Pie 3
0.2757 0.2559
0.1354 0.1290
0.2515 0.2349
0.1117 0.1078
0.2592 0.2386
0.1048 0.1009
0.2569 0.2421
0.1037 0.1009
15
0.2265
0.0988
0.1890
0.0530
0.1799
0.0590
0.1792
0.0593
0.2331
0.0829
0.1816
0.0725
0.1780
0.0592
0.1312
0.0356
0.1786
0.0882
0.1900
0.1244
0.1865
0.0847
0.1774
0.0807
0.2403
0.1240
0.2085
0.1016
0.2192
0.0973
0.2278
0.0983
24
0.2122
0.0937
0.1715
0.0503
0.1686
0.0567
0.1772
0.0589
0.2273
0.0813
0.1678
0.0685
0.1483
0.0530
0.1201
0.0348
0.1677
0.0863
0.1815
0.1205
0.1749
0.0831
0.1587
0.0787
0.2313
0.1211
0.1993
0.0994
0.2025
0.0941
0.2212
0.0970
33
0.1797
0.0821
0.1452
0.0462
0.1396
0.0508
0.1684
0.0570
0.2155
0.0782
0.1632
0.0672
0.1290
0.0491
0.1133
0.0343
0.1661
0.0860
0.1772
0.1185
0.1769
0.0834
0.1573
0.0785
0.2209
O.U77
0.1612
0 . 0900
0.1927
0.0922
0.2078
0.0946
100
0.1667
0.0775
0.1370
0.0449
0.1338
0.0496
0.1621
0.0557
0.1956
0.0729
0.1524
0.0641
0.1207
0.0473
0.1017
0.0335
0.1544
0.0839
0.1641
0.1125
0.1615
0.0813
0.1440
0.0770
0.2053
0.1127
0.1538
0.0887
0.1771
0.0892
0.1876
0 . 0908
1,500
0.0719
0.0438
0.0590
0.0329
0.0592
0.0345
0.0671
0.0362
0.0759
0.0409
0.0657
0.0390
0.0594
0.0347
0.0528
0.0300
0.0662
0.0683
0.0776
0.0694
0.0661
0.0686
0.0721
0.0691
0.0542
0.0366
0.0655
0.0370
0.0681
0.0358
0.0684
0.0356
typical retentivity curves of one horizon, corrected and uncorrected for coarse
fragments. While uncorrected data both in Table 2.12 and Figure 2.5 are similar to
that for a fine-grained soil, corrected data shows what can be expected on the mine-
soil when coarse fragment content is included. In reality, some sorting of coarse
fragments and fines does probably take place within the tninesoil profiles. Pockets
of fines could well exhibit moisture characteristics much like the uncorrected
curves shown, while aggregations of coarse fragments from which the fines have been
washed out would essentially be unaffected by internal gradients other than gravity.
46
-------
C2 (Pit 8)
o.i 0.3 0.3 a.4
Giovimelrtc Wot«r Content
Figure 2.5. Water retentivity curve (gravimetric)
for a minesoil corrected and not '
corrected for coarse fragments.
The corrected curves may not be actually realized anywhere within the spoil profile,
they do however represent an average between the two extremes discussed above.
Water retention—Unlike natural soils which may drain slowly in response to
gravity and matrix gradients within the structured profiles, coarse textured,
structureless, minesoils of Appalachia often drain rapidly through discrete large
pores or channels in response to gravity alone. Some water may be retained in the
form of films on the surface of the particles and some at the contact points.
Alternatively, if weathering of surface materials or deposition of fines forms a
47
-------
slowly permeable crust, infiltration and redistribution become greatly inhibited
(Crosby et al., 1977). At times, if a small head (0.1 m) is imposed on such
slowly permeable areas, piping results and new channels open up.
Water retention on the minesoils was estimated in two ways. First, water reten-
tion was studied in the field. Two days after a 38-mm rainfall, which effectively
saturated the minesoil, gravimetric samples were obtained from the surface and sub-
surface in close proximity to each of the lysimeter sites. Water content was deter-
mined on the dried and weighed samples and water retention was calculated. Second,
clods sampled from the vicinity of lysimeter sites were desorbed at 30 kPa dried,
weighed, and moisture content was calculated.
In Table 2.13 the amount of water retained by minesoils 2 days after a 38-mm
rainfall is compared with that retained by clods desorbed at 30-kPa tensiometer
pressure. If field-gravity drainage were to correspond to 30-kPa desorption, the
values for clods and samples would be the same, however, the clods in all cases
retained less water than the field samples. Comparing this data with corrected
values of moisture characteristic suggests a limit at about 10 kPa for these
minesoils should be used rather than the conventional 30 kPa value for estimating
TABLE 2.13. WATER CONTENT OF MINESOILS (SAMPLE) AFTER
2 DAYS OF DRAINAGE, OF CLODS DESORJSED AT 30 kPa,
AND MOISTURE CHARACTERISTIC VALUES (MC)
AT 10 k.Pa FOR SURFACE HORIZONS
Wacer concenc
Sice
L-l
L-2
L-3
L-4
L-5
L-6
L-7
L-8
L-9
L-1Q
!4ein
C.V.
Sample
0
.106
30 kPa clods
0.
3/8
.084
10 kPa MC
0
.110
0.088
0
0
0.
0
0
0.
0,
0.
0.
IS.
.082
.082
.091
.092
.105
,079
,062
.121
091
.3
0.
0.
0.
0,
0.
0.
0.
19.
_
.058
.062
,066
.068
,053
047
-
063
1
0
0.
0.
0.
17.
_
..
.092
_
.090
129
105
3
48
-------
water availability. Taking 10 kPa as the lower limit, and assuming average profile
density shown in Table 2.11 the amount of water available to plants in spoil
profiles between 10 kPa and 1500 kPa is estimated in Table 2.14.
TABLE 2.14. WATER RETAINED BETWEEN 10 AND 1500 kPa
TENSIOMETER PRESSURE AT FOUR MINESOIL
SITES AND AT ONE NATURAL SOIL SITE
Sit*
HoriiOD
Ap
Cl
C2
C3
Total '
Ap
AC
Cl
C2
Total
Apl
Ap2
Cl
C2
Total
AC
Cl
C2
C3
Total
Al
A2
B2)
B22t
Total
Thick-
ness.
mm
200
160
200
180
740
130
150
250
310
S40
100
210
200
150
660
180
180
220
440
1.020
30
70
310
330
740
Water retained
Unco rrec ted
-Ug
0.185
0.151
0.140
0.133
0.193
0.127
0.141
0.090
0.121
0.117
0.134
0.115
0.207
0.169
0.171
0.174
>
0476
0.226
0.135
0.141
mm
PitS
56.0
37.0
43.1
37.4
1 73.5(23.5) f
Plt6
37.9
29.2
54.5
43.9
365.5(19.7)
Pit?
18.3
37.7
41.5
26.S
124.4(1 8.9)
PitS
.55.1
46.6
58.3
121.6
282.0127.7)
12.6
22.0
74.5
89.3
398.2(26.8)
Corrected*
wt.8
0.066
0.023
0.028
0.027
0.052
0.036
0.025
0.006
0.022
0.057
0.018
0.013
0.092
0.071
0.065
0.065
0.366
0.174
0.092
0.093
HUB
20.0
5.7
8.7
7.5
41.9(5.7)
10.1
8.5
11-2
3.1
32.9(3.9)
3-3
18.4
5.5
3.0
30.2(4.6)
25.3
19.6
22.3
45.8
1130(11.1)
9.7
16.9
50.5
58.9
136.008.41
*Con«:ted for co*rs* fragments
"j"V'a)oes in brackets are expressed is ptrcentaff of u>uU thickne
(23.5) = 100(373.5/740).
49
-------
Compared to values on natural soil and on minesoil uncorrected for coarse
fragments, retention on corrected minesoils (except at pit 8) is very low
(Table 2.14). Topsoiled profiles (pits 5 and 6) retain most plant-available water
near the surface. On a nontopsoiled profile (pit 7) higher retention in the Ap2
may indicate a zone of illuviation, fines having been washed down from the mine-
soil surface. Compared with other profiles, the minesoil at pit 8 shows retention
similar to that of the natural soil. This pit is situated on a lower slope ad-
joining nontopsoiled materials, and contains more montmorillonite and vermiculite
than minesoils at other pits. Very low amounts of water retained in spoil profiles
in general will lead to low water availability and limited plant growth.
Retention values for natural soils are included for comparison only. Since
the analysis was performed on sieved <2-mm material and the profile does exhibit
structural development the results should be treated with caution (Bruce, 1972;
Unger, 1975).
Hydraulic conductivity—Using the spoil material packed into nicrolysineters
(Figure 2.6), saturated hydraulic conductivity of minesoil before and after an
ET study was determined in the laboratory by a constant head method (Klute, 1965).
Field exposure of microlysimeters for 4 months resulted in some fines being lost
through piping and some scaling of the surface.
Results of the hydraulic conductivity experiment performed before and after 4
months of field exposure are presented in Table 2.15. Large variability (as
measured by CV values) existed between the lysimeters, particularly during the
second run. High values of conductivity obtained for the second run (L-6 and L-8)
were due to washing out of fine material from the lysimeters. Such occurrences may
be quite representative of field conditions, when piping and internal erosion are
common. The conductivities generally were lower after microlysimeters were exposed
to the elements. Compared to initial results, field exposure seems to have
50
-------
Figure 2.6. A microlysimeter site.
decreased the magnitude and increased the variability of the average hydraulic
conductivity values. The values obtained represent unconsolidated spoil material.
No attempt was made to plug any channels which developed when fines were washed
out during drainage and subsequent exposure. The results may be indicative of
minesoil conductivity immediately following reclamation.
Infiltration—In contrast to the hydraulic conductivity measurement on
disturbed and repacked minesoils, infiltration was determined in the field on
51
-------
TABLE 2.15. SATURATED HYDRAULIC CONDUCTIVITY AT JO LYSIMETER
SITES (L-l TO L-10) BEFORE (INITIAL) AND AFTER 4 MONTHS
OF FIELD EXPOSURE (FINAL) CORRECTED TO 20'C
Sauirauec hyQraubc conducuvir}
L-ysimeLer siLe DO
L-)
L-2
L-3
L-4
U5
L-6
L-7
US
L-9
L-10
Mean*
C.V_ •>
]aiuaJ
39
303
Hi
126
7.<25
1.707
1.S2S
3.303
1.60S
667
640
p
FU-JL)
mra/bour
4
20
28
22
2.024
43.590
3.045
jo 39]
1<2
153
295
80
•Logarithmic mean, and coefficient of variation (C.V.).
9
undisturbed sites using a 0.5 m diam cylinder infiltrometer (Coleman, 1951) .
Half of the infiltrometer (0.1 m) was imbedded in the spoil and the opening made
by the cylinder walls was sealed with bentonite clay. Five-liter increments
(25 mm) of water were added to the cylinder repeatedly and allowed to infiltrate.
During the first addition we measured the initial infiltration rate as calculated
from the lowering of water level, while in the subsequent runs we looked for the
rate to stabilize and eventually decline. Field determined final infiltration
capacities were expected to be less than the hydraulic conductivity values
determined on the microlysimeters in the laboratory (Talsma, 1969). Less
disturbance, larger area, and possibly some surface crusting could be responsible
for the decreased rates. But even these lower rates were likely to overestimate
2
G. B. Coleman. 1951. A study of water infiltration into spoil banks in
central Pennsylvania. M.S. Thesis. The Pennsylvania State University,
University Park, Pennsylvania.
52
-------
the true infiltration capacity by about 14% (Tricker, 1978). Infiltration rates
were also measured on a 2.5 by 2.5 m infiltration plot cased to the depth of 1 m
and situated in the area that has not been topsoiled.
Table 2.16 shows results of the preliminary infiltration study on topsoiled
and nontopsoiled minesoil, on adjacent natural soil, and on the infiltration
plot. The infiltration rates generally were lower than the hydraulic conductivity
values given in Table 2.15. This is because of differences in methods of measure-
ment and basic hydraulic properties of materials. In general, hydraulic conduc-
tivity values are expected to be higher than the final infiltration rate (Talsma,
1969). Values determined using a constant head method (Klute, 1965) are generally
obtained on fully saturated disturbed and repacked materials. In contrast,
ponding of water during the infiltration measurements does not necessarily result
in a fully saturated flow; because of surface crusting slower unsaturated flows and
fingering are often more likely (Hill and Parlange, 1972; Raats, 1973; Philip,
1975). Comparison of results for the infiltrometers and infiltration plot shows
the values also to decrease with increase in size.
TABLE 2.16. INFILTRATION VELOCITIES ON
NATURAL SOIL, TOPSOILED, AND
NONTOPSOILED MINESOIL
Site
Natural soil
1-1
1-2
Topsoiled material
1-4
1-6
Nontopsoiled
materials
1-7
1-8
1-9
MO
Infiltration plot
Water
applied
mm
75
450
75
75
75
25
75
50
140
7
mm/hour
53
421
11
28
17
33
14
7.3
2.0
c.v.
%
4
4
34
11
17
42
12
22
47
Initial
140
-
190
343
286
21
10
7
Steady
mm/hour
40
-
19
19
19
19
6
2
Final
27
-
3
5
3
11
3
1
53
-------
Hydraulic properties such as moisture characteristic will also materially
influence the rate of water transfer within the minesoil profiles. The steepness
of the corrected retentivity curves (Figure 2.5) will be complemented (Rogowski and
Jacoby, 1979) by a corresponding steepness in the hydraulic conductivity—water
content curves (not shown). Thus a change in water content of as little as three
percentage points (from 0.20 to 0.17 g/g) may result in a hydraulic conductivity
3 0
change of three orders of magnitude (from 10 to 10 mm/hr). While hydraulic
conductivity values (Table 2.15) depict an unusual situation of complete saturation,
using final field infiltration rate values likely will lead to a better first
estimate of minesoil conductivity.
Infiltration values on minesoils also appear lower than on natural soils. This
is despite the fact that at 1-1 the surface horizon had been eroded and at site 1-2
the litter layer has been removed. Both these sites are in an area where trees have
been recently removed. Consequently, large amount of decaying roots and old root
channels are present which could be responsible for higher rates.
Evapotranspiration— At each lysimeter site some of the excavated minesoil
was repacked into small microlysimeters (0.2 by 0.3 m) (Rogowski and Jacoby,
1977) to approximate measured field bulk density. The microlysimeters (Figure 2.6)
were seeded with ryegrass (LoUwn perenne L.) and set on collecting pans in the
excavations. Acrylic liners were placed around each microlysimeter, and the
minesoil was returned to its original level. Vegetation was reestablished at
each site, and an evapotranspiration (ET) study was initiated. The amount of
water seeping through the minesoil and the weight of each cylinder was recorded
twice weekly. Rainfall, evaporation, maximum-minimum temperature, and solar
radiation data were also monitored during this time.
Figure 2.7 shows preliminary cumulative ET values during the study as deter-
mined by model results, changes in lysimeter weight, and evaporation from the class
54
-------
NOV. I
5OO
TIME (hrs)
IOOO
Figure 2.7.
Cumulative ET values during the study as
determined by model results, changes in
lysimeter weight, and evaporation from
the class A pan.
A pan. Although the model data appear lower, the results showed that ET from mine-
soil in the microlysimeters (2.4 mm/day average) could be approximated either by
class A pan evaporation (1.8 mm/day) or the model (Ritchie, 1972) (1.3 mm/day).
2
Linear regression of the lysimeter and the A pan data gave an r value of 0.984;
2
comparison with model data an r value of 0.950; and comparison of the model data
2
with the class A pan data an r value of 0.987. Under the conditions of the
experiment, lysimeters retained more water than would be expected under normal
conditions since downward seepage was prevented by a liner and contact with
underlying minesoil was not maintained. Other studies (Rogowski, 1978; Rogowski
and Jacoby, 1979) and careful consideration of hydraulic properties suggests that
droughty conditions can easily occur. Small changes in moisture content
(Figure 2.5) lead to large changes in the tensiometer pressure. Consequently,
little water is readily available (Table 2.14) for plant use on most minesoils
as compared to natural soils.
55
-------
-------
REFERENCES
1. Anderson, J. L. An Improved Pretreatment for Mlneralogical Analysis of
Samples Containing Organic Matter. Clays and Clay Minerals, 10:380-388,
1963.
2. ASTM. Standard Methods of Coal Analysis. American Society for Testing
Materials, Philadelphia, Pa., 1971.
3. Baker, D. E., and L. Chesnin. Chemical Monitoring of Soils for Environmental
Quality and Animal and Human Health. Adv. Agron., 27:305-374, 1975.
4. Bertram, G. E. Field Tests for Compacted Rockfill. In: Embankment-dam
Engineering, Casagrande Volume, R. C. Hirshfield and S. J. Poules, eds.
John Wiley and Sons, Inc., New York, 1973. pp. 1-20.
5. Blake, G. R. Bulk Density. In: Methods of Soil Analysis, Part 1, C. A.
Black, ed. American Society of Agronomy, Madison, Wis. Agronomy,
9:374-390, 1965.
6. Broadbent, F. E. The Soil Organic Fraction. Adv. Agron., 5:153-183, 1953.
7. Bruce, R. R. Hydraulic Conductivity Evaluation of the Soil Profile from
Soil Water Retention Relations. Soil Sci. Soc. Am. Proc., 36:555-561, 1972.
8. Ciolkosz, E. J., G. W. Petersen, R. L. Cunningham, and R. P. Matelski. Soils
Developed from Colluvium in the Ridge and Valley Area of Pennsylvania. In:
Proceedings of the International Geobotany Conference, University of
Tennessee Press, Knoxville, Tenn., 1978.
9. Ciolkosz, E. J., R. L. Cunningham, G, W. Petersen, and R. C. Cronce. Character-
istics Interpretations and Uses of Pennsylvania Minesoils. Progress Report 381.
The Pennsylvania State University, Agricultural Experiment Station, University
Park, Pa., 1983. 88 pp.
56
-------
10. Crosby, E. C., D. E. Overton, and R. A. Minear. A Simulation Model of Spoil
Bank Hydrology. In: Surface Mining and Fish/Wildlife Needs in the Eastern
U.S., D. E. Samuel, J. R. Stauffer, C. H. Hocutt, and W. T. Mason, Jr., eds.
Proceedings, Fifth Symposium on Surface Mining and Reclamation, NCA/BCR Coal
Conference and Expo IV, October 18-20, Louisville, Ky., 1977. pp. 28-31.
11. Farmer, E. E., and B. Z. Richardson. Hydrologic and Soil Properties of Coal
Mine Overburden Piles in Southeastern Montana. In: Proceedings, Fourth
Symposium on Surface Mining and Reclamation, NCA/BCR Coal Conference and
Expo III, October 19-21, Louisville, Ky., 1976. pp. 120-130.
12. Glass, G. B. Geology and Mineral Resources of the Philipsburg 7-1/2 Minute
Quadrangle, Centre and Clearfield Counties, Pennsylvania. Pennsylvania
Geological Survey, 4th Series, Department of Environmental Resources,
Harrisburg, Pa., 1972.
13. Hill, D. E., and J.-Y. Parlange. Wetting Front Instability in Layered Soils.
Soil Sci. Soc. Am. Proc., 36(5):697-702, 1972.
14. Jackson, M. L. Soil Chemical Analysis. Prentice Hall, Englewood Cliffs,
New Jersey, 1958.
15. Kilmer, V. J., and L. T. Alexander. Methods of Making Mechanical Analyses
of Soils. Soil Sci., 68:15-24, 1949.
16. Klute, A. Laboratory Measurement of Hydraulic Conductivity of Saturated
Soil. In: Methods of Soil Analysis, Part 1, C. A. Black, ed. American
Society of Agronomy, Madison, Wis. Agronomy, 9:214-215, 1965.
17. Mehuys, G. R., L. H. Stolzy, J. Letey, and L. V. Weeks. Effect of Stones
on the Hydraulic Conductivity of Relatively Dry Desert Soils. Soil Sci.
Soc. Am. Proc., 39:37-42, 1975.
57
-------
18. Pedersen, T. A., A. S. Rogowski, and R. Pennock, Jr. Comparison of Some
Properties of Minesoils and Contiguous Natural Soils. EPA-600/7-78-162,
U.S. Environmental Protection Agency, Research and Development Series,
Cincinnati, Ohio, 1978. 141 pp. ,
19. Peech, M. L., L. T. Alexander, L. A. Dean, and J. F. Reed. Methods of Soil
Analysis for Soil Fertility Investigations. U.S. Department of Agriculture
Circular 757, Government Printing Office, Washington, D.C., 1947.
20. Philip, J. R. Stability Analysis of Infiltration. Soil Sci. Soc. Am.
Proc., 39:1042-1049, 1975.
21. Plass, W. T., and W. G. Vogel. Chemical Properties and Particle Size
Distribution of 39 Surface-Mine Spoils in Southern West Virginia.
Northeast Forest Experiment Station Research Paper NE-276, 1973.
22. Raats, P. A. C. Unstable Wetting Fronts in Uniform and Nonuniform Soils.
Soil Sci. Soc. Am. Proc., 37:681-685, 1973.
23. Rawitz, E. Installation and Field Calibration of Neutron Scatter Equipment
for Hydrologic Research in Heterogeneous and Stony Soils. Water Resour.
Res., 5(2):519-523, 1969.
24. Richards, L. A. Physical Conditions of Water in Soil. In: Methods of
Soil Analysis, Part 1, C. A. Black, ed. American Society of Agronomy,
Madison, Wis. Agronomy, 9:128-152, 1965.
25. Ritchie, J. R. Model for Predicting Evaporation from a Row Crop with In-
complete Cover. Water Resour. Res., 8(5):1204-1213.
26. Rogowski, A. S. Water Regime in Stripmine Spoil. In: Surface Mining and
Fish/Wildlife Needs in the Eastern U.S., D. E. Samuel, J. R. Stauffer,
C. H. Hocutt, and W. T. Mason, Jr., eds., Proceedings of a Symposium,
December 3-6, Morgantown, West Va., 1978. pp. 137-145.
27. Rogowski, A. S., and E. L. Jacoby, Jr. Assessment of Water Loss Patterns
with Microlysimeters. Agron. J., 69:419-424, 1977.
58
-------
28. Rogowski, A. S., and E. L. Jacoby, Jr. Monitoring Water Movement through
Strip Mine Spoil Profiles. Trans. ASAE 22:104-109, 114, 1979.
29. Soil Survey Staff. Soil Survey Manual. U.S. Department of Agriculture
Handbook 18, U.S. Government Printing Office, Washington, D.C., 1951.
30. Soil Survey Staff. Soil Taxonomy: A Basic System of Classification for
Making and Interpreting Soil Surveys. U.S. Department of Agriculture
Handbook 436, U.S. Government Printing Office, Washington, D.C., 1975.
31. Talsma, T. In Situ Measurement of Sorptivity. Aust. J. Soil Res.,
7:269-276, 1969.
32. Tricker, A. S. The Infiltration Cylinder, Some Comments on its Use.
J. Hydrol., 36:383-391, 1978.
33. Troxler Electronic Laboratories, Inc. 2400 Series COMPAC, Surface
Moisture-Density Gauge Manual. Raleigh, N.C., 1970.
34. Unger, P. W. Water Retention by Core and Sieved Samples. Soil Sci. Soc.
Am. Proc., 39:1197-1200, 1975.
35. U.S. Salinity Laboratory. Diagnosis and Improvement of Saline and Alkali
Soils. U.S. Department of Agriculture Handbook 60, U.S. Government
Printing Office, Washington, D.C., 1954.
36. Van Voast, W. A. Hydrologic Effects of Strip Coal Mining in Southeastern
Montana—Emphasis: One Year Mining Near Decker. Montana College of
Mineral Science and Technology, Butte, Montana, 1974.
37. Veraa, T. R., and J. L. Thames. Rehabilitation of Land Disturbed by Surface
Mining Coal in Arizona. J. Soil Water Conserv., 30:129-132, 1975.
38. Young, J. L., and M. R. Lindbeck. Carbon Determination in Soils and Organic
Materials with a High Frequency Induction Furnace. Soil Sci. Soc. Am.
Proc., 28:377-381, 1964.
59
-------
39. Young, R. A., and C. A. Onstad. Predicting Particle-Size Composition of
Eroded Soil. Trans. ASAE, 19:1071-1075, 1976.
40. Yuan, T. L. Determination of Exchangeable Hydrogen in Soils by Titration
Method. Soil Sci. , 88:164-167, 1959.
60
-------
-------
SECTION 3
CAISSON STUDIES
MONITORING WATER MOVEMENT THROUGH STRIP MINE SPOIL PROFILES
The methods of measuring and simulating water flow on disturbed land are scarce.
The purpose of this study was to develop suitable instrumentation and adapt existing
methodology to the assessment of water flow on stripmined lands. The experimental
studies, described in this and the following section, are confined to the measurement
and modeling of water flow in reconstructed spoil profiles.
The spoil profile chosen for study and described in detail in Section 2 was
situated on a site near Kylertown, Pa. This site had been stripmined seven years
before for Lower Kittanning coal (B-seam), and two years before for Middle
Kittanning coal (C-seam). The regraded spoil, underlain by an undisturbed clay
layer, consisted of a mixture of grey shale and light colored sandstone covered in
part (C-spoil only) with topsoil material (loamy skeletal, mixed, mesic, Udorthents).
The area was planted in trees, grass and legumes and part of it had been limed,
fertilized and manured. The site was surrounded by wooded and cultivated patches of
original soils.
A 6-m spoil profile was excavated (Figures 3.1, 3.2) with a large backhoe, in
20 almost equal depth increments, transported to a research facility, and placed
by hand in two 2.4 m diameter caissons (Figures 3.3, 3.4, 3.5, 3.6) duplicating
the field order. A 15 kg sample from each box was used for analyses of particle
size, physical and chemical properties. The caissons, constructed from 14 gage,
corrugated, galvanized steel culverts, were placed under roof on a concrete pad,
and were precoated with 12 mil layer of NEXON—a commercially available
thermoplastic, corrosion resistant, coal-tar-based resin.
61
-------
Figure 3.1. Backhoe excavates successive layers
of spoil at Kylertown.
Figure 3.2. Research staff assembles boxes for
transport of Kylertown spoil by
truck to the experimental facility.
62
-------
Figure 3.3. Boxes are unloaded at the experimental
research facility.
Figure 3.4. Caissons at the experimental research facility
to be filled with Kylertown spoil.
63
-------
Figure 3.5. Numbered boxes are positioned near the
appropriate caisson preparatory to
reconstituting the field profile.
Figure 3.6. Inside of caisson 1 showing the 2.5' depth of spoil in
place, access tubes for measuring density and moisture
and a lysimeter for collecting the effluent, ladder
was used by personnel who rebuilt the field profile
layer by layer.
64
-------
To ensure uniform drainage, 0.6 m of inert quartz sand of a compatible particle-
size distribution with the spoil was placed at the bottom of each caisson. The
reassembled spoil profile was covered in caisson 1 with 0.5 m of topsoil, while in
caisson 2 a dark acid shale layer (0.4 m) taken from just above the B-coal seam was
placed below the reassembled profile and on top of the sand.
Instrumentation
Figure 3.7 shows the cross section of an instrumented caisson, and Table 3.1
gives the location of different types of instrumentation in both caissons.
A central well for measurement of water levels consisted of a rigid PVC (poly-
vinyl chloride) pipe with a 0.6 m long stainless steel screen at the bottom. Four
aluminum access tubes, for measurement of density and water content with gamma and
neutron probes, respectively, were coated with a corrosion resistant paint. The
access tubes were placed 0.3 m apart to accommodate a two-probe density probe
besides a standard neutron moisture gage.
r, P,
Figure 3.7. Caisson instrumentation description.
65
-------
TABLE 3.1. LOCATION OF INSTRUMENTS* CAISSON 1 AND 2
Caisson
0
18
30
61
90
122
183
244
274
274
274
335
Location
1 Caisson 2
(cm)
0
31
38
69
98
130
191
252
282
282
282
343
Type
Old surface
New surface
L, Tr, D, TC(2)
L, Tg, TC(2)
Tr
L, Tp, D, TC(2)
L, Tp, TC(1)
L, Tp, D, TC(2)
L, TC(1)
Bottom of spoil
Top of screen and sand
Bottom, well casing, access tubes
*L = lysimeter; Tr = mercury tensiometer; D = diffusion chamber;
TC = thermocouple; Tg = gage tensiometer; Tp = pencil
tensiometer.
Changes in moisture content and density were measured using neutron and gamma
probes, respectively. Organic C was determined by a Walkley Black method (Allison,
1965). Total S was obtained by a high temperature combustion, while standard tests
(ASTM, 1971) were used to evaluate sulfate, pyritic and organic S in each layer of
the spoil.
We measured leaching potential of the reassembled profile on a 10-kg subsample
of each excavated layer by placing it in a 20 cm diameter PVC cylinder and leaching
with 1000 ml of water. The leachate was filcered and on the filtrate pH, total
acidity, electrical conductivity, Ca, Mg, and SO, were determined. The results of
these analyses are discussed in detail elsewhere in this report.
Rainwater needed for application to the caissons was collected from the roof
of the research facility and stored in large tanks.
66
-------
Using values of bulk density, we also calculated the total available pore
space (IPS) from the equation
IPS = 1 - (Bulk Density/2650)
where 2650 was the assumed particle density in kg/m .
In addition to the 4 access tubes for measurement of density and water we
filled two other access tubes with ethylene glycol, and placed copper-constantan
thermocouples inside at selected depths to measure small changes in temperature
during water application. The method was similar to that used in borehole
geophysics (Keys and MacCary, 1971), except we used stationary thermocouples,
instead of a mobile probe, and added ethylene glycol to water to prevent freezing.
We installed porous cup tensiometers, by excavating a hole in the spoil and
filling the bottom part around the porous cup with a fine spoil slurry.
We also installed tension free lysimeters (Jordan, 1966), made a PVC and
stainless steel to prevent corrosion, at selected depths at a 10° slope while the
caissons were being filled and connected each one to its own manometer-outflow
assembly. The manometer-outflow assembly served a dual purpose: air pressure
changes could be read as we applied the water at the surface and we could
measure and sample water intercepted by the lysimeters for subsequent chemical
analysis.
Oxygen diffusion chambers similar to those described by Raney (1949), allowed
ambient spoil air to diffuse into a space previously swept clear with nitrogen.
Diffused oxygen was measured with an oxygen analyzer, and oxygen concentration in
the spoil was computed using appropriate calibration procedures.
Methods
In topsoiled caisson 1, we applied the stored rainwater at the average rate of
300 mm/h (1.4 m /h) for 92 min, using a pump and sprinkler system. The pump was
67
-------
operated during 0 to 70 min and 102 to 124 min time intervals. This rate caused
some surface ponding. In caisson 2, we applied the stored rainwater at the rate
of 1890 mm/h (8.8m /h) for 16 min (following the initial 5 min application at
the same rate as in caisson 1). The pump was operated during 0 to 5 min and 47 to
63 min time intervals. Even at the high application rate no noticeable ponding
was evident. The total amounts applied were 455 and 483 mm of water (2115 and
2246 L) to caissons 1 and 2, respectively. After water application we left caisson
2 uncovered but covered caisson 1 with a plastic sheet and allowed it to drain for
17 days. We then uncovered it, and on the 23rd day after water application raked
the surface to break the crust.
To determine a water retentivity curve of the spoil profiles we used the
following system. We cleaned and de-aired 80 mm diameter, 25 kPa (250 mb)
fast-flow retainer ceramic plates. Rings, the same size as the plates and 40
mm high were placed on the retainer plates and filled with <2-mm fraction of
spoil material. We used the <2-mm fraction in accordance with soil taxonomy
textural gropings for soils (Soil Survey Staff, 1975). The ring and retainer
assemblies were then placed on an extractor plate (25 kPa ceramic) in the
pressure cooker apparatus. Water was added and samples were allowed to
saturate overnight. The next day excess water was wiped off; samples were
weighed, and their volume measured. Subsequently, samples were desaturated at
1, 4, 8, 15 and 24 kPa (10, 40, 80, 150 and 240 mb) dried and weighed. To ex-
press water content on a total volume basis, calculated values for a water
retentivity curve were multiplied by a factor equal to the fractional amount of
spoil material <2 mm in a given sample. Specific retention for spoil particles
>2 mm was then measured, adjusted for the amount of >2-mm material present, and
added as a constant to the recalculated values. The procedure for measuring
specific retention, adapted from ElBoushi (1966), consisted of wetting a known
68
-------
volume of particles, allowing it to drain briefly (30 min), and computing water
retained from weight changes.
The Green and Corey (1971) model with Millington and Quirk (1960) pore size
interaction factor was used to calculate the hydraulic conductivity as a function of
tensiometer pressure and soil water content from corrected and uncorrected values of
water-retentivity curve. The factor of 4/3 was more suitable for the coarse materials
(Rogowski, 1972) than other factors commonly used in the basic Marshall (1958) model.
The model was matched at the saturated conductivity corresponding to average conduc-
tivity measured on the topsoiled and nontopsoiled portions of the experimental site.
Results and Discussion
We observed considerable settling, 0.18 m (6%) on caisson 1 and 0.31 m (10%) on
caisson 2, during the water application. The settling corresponded closely to the
field behavior of the freshly reclaimed spoil.
The results showed that the profile density increased in both caissons after
water was applied and spoil had settled (Figure 3.8). The density in caisson 1
increased most near the surface and progressively less with depth, while in caisson
2, except for the acid shale layer, the increase in density was least near the
surface becoming progressively more with depth. The acid shale layer changed little
in density after water application, probably because of the many large rock fragments.
Spoil settling and changes in density following water application may indicate
internal sediment movement. Table 3.2 shows sediment concentration values in the
effluent from lysimeters for different times after water was applied to each caisson.
Higher concentrations of sediment in effluent from caisson 1 compared with those from
caisson 2 suggests a possibility of greater internal erosion (piping) on topsoiled
areas. Such mass movement of soil may also transport into the profile iron- and
sulfur-oxidizing bacteria associated with the sediment.
69
-------
Ovnsi'y (Kq/m )
1000 2OOO
0
0«ns»» (Kg/m )
lOOO 2000
COisson I
Caisson 2
3.5
Figure 3.8. Caisson profile density before (original)
and after (new) water application.
TABLE 3.2. SEDIMENT CONCENTRATION IN THE WELL AND IN THE EFFLUENT
FROM LYSIMETERS AT DIFFERENT TIMES AFTER WATER WAS APPLIED
Time
Caisson 1
Depth*
Sediment
Time
Caisson 2
Depth
Sediment
min
m
ppm
min
m
*Depth below original surface.
ppm
34
46
46
47
62
90
111
128
130
133
0.30
0.61
1.83
1.22
2.44
0.61
1.83
1.22
0.30
2.44
90138
50604
6952
6746
3578
15702
2370
3090
16062
758
60
60
60
60
60
60
75
75
75
125
125
155
185
0.38
0.69
1.30
1.91
2.52
2.82
1.30
2.52
2.82
2.52
2.82
3.43
2.82
7828
1160
7504
2212
2292
1960
1388
888
592
548
332
108
128
70
-------
During water application, despite the central well open to the atmosphere, the
air pressure buildup in the profile was considerable (between 0.4 and 10.4 kPa).
The pressure buildup in the caisson 2 spoil was greater than in topsoiled caisson 1
since infiltrating water moved faster in the spoil alone and entrapped air could not
exit. An analogous situation existed with regards to oxygen concentrations. In
Table 3.3 we compare oxygen concentrations at three depths in the two caissons.
Preliminary data showed that oxygen concentrations in topsoiled caisson 1 did not
recover to the initial concentration as did those on spoil alone in caisson 2. The
data, however, were not sufficient to conclude that the topsoil layer impeded oxygen
movement. Oxygen concentration in the profile after caisson 1 was uncovered (17
days) and raked (23 days) showed only a slight'increase over previous readings.
TABLE 3.3. OXYGEN CONCENTRATION AT SELECTED DEPTHS AFTER WATER .APPLICATION
Time
Caisson 1, depth, m
0.30
1.22
2.44
Time
Caisson 2, depth, m
0.38
1.30
2.52
days
percent oxygen
days
percent oxygen
0
1
2
3
6
7
8
9
13
15
16
17
20
21
22
23
25
27
14.8
1.9
1.9
1.1
2.1
1.2
1.9
1.7
4.7
4.1
4.6
5.0
3.2
5.0
3.6
5.0
5.1
8.1
8.4
5.7
4.8
3.1
4.0
3.5
2.8
2.7
4.5
4.5
6.9
7.1
4.6
3.4
4.1
3.9
8.3
7.7
12.9
6.7
7.6
7.1
11.4
8.1
7.6
8.4
10.5
9.5
11.8
12.7
11.5
6.0
6.1
9.9
9.6
4.2
0
1
4
5
6
7
8
12
14
17.1
19.8
17.6
16.9
18.1
19.1
18.0
20.7
20.8
9.8
19.0
19.0
17.7
19.0
18.9
18.8
19.9
19.6
11.3
water
water
water
water
water
17.7
20.2
20.5
71
-------
Infiltration and redistribution of applied water and wetting-front passage in
the spoil profiles were monitored with the neutron moisture probe and gamma two-
probe density probe (Table 3.4) supplemented by thermocouple readouts. Since
neutron probe measures water content in a relatively large spherical volume of
material changes in water content associated with a wetting-front passage in
caisson 1 (Table 3.4) were not particularly informative. However, rapid changes
TABLE 3.4. DEPTH OF WATER INFILTRATION WITH TIME AFTER APPLICATION
Depth*
m
0.30
0.61
0.91
1.22
1.83
Timet
min
22
31
35
49
56
Caisson 1
Water
m
Before
0.1227
0.0924
0.1089
0.1007
0.1099
content!"
3. 3
/m
After
0.1223
0.2032
0.1372
0.1336
0.1085
Caisson 2
Depth
m
0.30
2.44
Time
min
1
54
Water
m
Before
0.0842
0.1019
content!
3/m3
After
0.3942
0.2899
Wet bulk density
kg/m
Before After
1253 1563
1499 1687
*Below original surface.
tTime to reach given depth; beginning at the time water was first
applied, pump operated during 0-70 min and 102-124 min time
intervals for caisson 1 and during 0-5 min and 47-63 time intervals
for caisson 2.
j_
tBased on neutron moisture probe readings before and after the front
passage.
§Computed from changes in wet bulk density and in initial water
content using gamma density probe and neutron moisture probe
readings.
72
-------
in wet bulk density measured with a dual gamma probe after a wetting-front passage in
caisson 2 indicated a higher water content just behind the front than at a later time
(2 hr later). Because of a very rapid movement of water in caisson 2, we could
monitor only two depths whereas in caisson 1 where infiltration was slower, we
obtained more readings. In general, the two-probe gamma density probe was better for
monitoring the wetting front movement than was the neutron moisture probe, since with
the former at any time only a small volume of spoil (25 mm slice) was being examined.
Similar conditions are likely to prevail in the field. Direct transmission dual
gamma probe could ideally be used above and below a topsoil-spoil interface to
monitor the rate of seepage into the underlying profile.
Since the applied water was at a different temperature than caisson spoil,
we could follow its movement by observing temperature changes. For shallow
depths, particularly in caisson 1, rapid temperature changes accurately
indicated the wetting front passage. However, infiltrating water appeared to
reach the bottom of caissons considerably earlier than any noticeable changes
in temperature at deeper depths could be observed. This indicates that prefer-
ential water flows do occur through select channels and larger pores of spoil
profiles. Single measurements are unlikely to intercept the flow channels in
the field. Consequently, it would seem prudent to use simultaneously a com-
bination of measurements including moisture, density, temperature and water
table elevation at a sufficiently large number of measuring sites.
Figure 3.9 shows water redistribution profiles based on neutron probe
measurements. Redistribution profile 2 on topsoiled caisson 1 represents
water content values immediately after the wetting front has passed a given
depth. Redistribution profile 3 on caisson 1 measured a day later corresponds
more closely to what may be expected under field conditions following a heavy
rain. While the topsoil in caisson 1 was draining, the water content of the
spoil below the 1-m depth remained relatively constant.
73
-------
Water Content (m3 / m3)
.10 .20 .30
Wqrer Content (m3 / m3)
.10 .20 .30
.40
I Initial
2 Weftmg
3 I days
5 8 •
Figure 3.9. Water redistribution profiles for caissons
1 and 2 for indicated times.
For caisson 2 on spoil alone redistribution profiles showed the water content
to increase gradually with depth up to 1.80 m. Below that depth, and in an acid
shale layer of lower density and higher porosity, the water content increased more
rapidly (redistribution profiles 3, 4 and 5) after a slower initial wetting
(redistribution profile 2). Coarse spoil layers in the field may initially act as
impeding layers and may not begin to fill with water until overlying spoil becomes
saturated.
In caisson 1 the redistribution was slow. After initial drainage (profiles 2
to 4), water redistribution took place over a period of one month with little change
in water content below the 1 m depth (profiles 4 to 7). At the same time we
observed considerable drying of topsoil. In caisson 2, however, water continued to
drain freely and relatively fast from the coarse grained spoil profile that has not
been topsoiled. This could be ascribed to smaller retentivity and lower air entry
value of spoil compared with topsoil.
74
-------
Redistribution profiles on caisson 1 and 2 in Figure 3.9 cover the same time
period as the tensiometer readouts in Figure 3.10. We obtained our best readings
IS 20 23
Tim* (Days)
Figure 3.10.
Comparison among tensiometer pressures
at 0.3-, 0.6- and 0.9-m depths on
caissons 1 and 2 for 27 and 15^ days
after water was applied.
with individual porous cup tensiometers at the 0.3-, 0.6- and 0.9-m depths, inserted
into the spoil profile from the surface. Tension in caisson 2 began to decrease in
the 0.3-m depth tensiometers 5 min after water was applied, and in the 0.6-m depth
tensiometers 10 min after application. After a pause (42 min), and 8 min after
restarting water application, tensiometers at 0.9-m depth began to change. In
contrast, two-probe data on caisson 2 indicated the wetting front passing the 0.3-
and 2.4-m depths at 1 and 7 min after water application was started and restarted
respectively, suggesting a several minute tensiometer response delay in the spoil.
In caisson 1 tension increased gradually until the surface was uncovered and raked,
it then registered a sharp increase at the shallower depths. However, in caisson 2
75
-------
marked increases in shallower values of tension could already be observed shortly
after water application ceased. This suggests droughty environment for plants
growing on nontopsoiled spoils in the field.
Figure 3.11 shows hydraulic properties of spoil material determined in the
laboratory. Large differences between water retentivity curves (Figure 3.11a)
corrected and uncorrected for coarse fragments may pose a problem when water flow
in strip mine spoils is studied. Materials used in topsoiling are likely to have
a water retentivity curve similar to the uncorrected curve in Figure S.lla, while
underlying spoil materials may have curves similar to the corrected curve. The
M
g
£
*
10°
10'
I I I I I
a.
I I I I
10'
I04
CL
s
0 JO .20 .30 .40 .50 60
Water Content (m-3/m3)
IU'
IOP
4
X
E
~IO''
3
u
i
i
"5
o
I
iri"
1 1
b.
Caisson it
Caisson 2
'
.
"
\
aisson lo
1 I
0 10 .20 .30
Water Content (m3/m3)
10'
10°
10*
Id3
10*
10"
Tensiom«r«r Pressure (mb)
I02 10' 10°
Tensiometer Pressure (UPa)
Figure 3.11.
(a) Water retentivity curves averaged for all depths for spoil
profile corrected and uncorrected for coarse fragments and
specific retention; (b) matched (Green and Corey, 1971)
hydraulic conductivity values as a function of water, corrected
for coarse fragments and specific retention; (c) hydraulic con-
ductivity as a function of tensiometer pressure; curves labeled
la refer to caisson 1, they were matched at K = 0.13 m/hr,
curves labeled 2 were for spoil material in caisson 2 and were
matched at K =1.89 m/hr.
sat
76
-------
measured water contents and tensiometer pressures in the topsoiled layer in caisson
1 (Figures 3.9, 3.10) are somewhat larger than those expected from the uncorrected
water retentivity curve, however, values for the spoil follow those for the
corrected curve. In caisson 2 measured values of water content below 1 m depth
appear larger than expected from the corrected water retentivity curve in Figure
S.lla, but a rapid decrease in tensiometer pressure above 1 m depth corresponds to
what would be expected from the corrected water retentivity curve. The curves in
Figure 3.lib show that small changes in water content can result in large changes
in hydraulic conductivity of the spoil. This means that even when water content of
the spoil varies little from place to place faster flows may occur through wetter
zones. Compared to natural soils under the field situations larger hydraulic
conductivity values (both saturated and unsaturated) could be expected at lower
water contents and higher values of tensiometer pressure on minesoils derived from
materials similar to those studied.
Table 3.5 shows the distribution of water in caissons with depth and accounts
for both the residual, and applied water. The water content values are based on
neutron moisture probe measurements, before and 1 day after water was applied to
the caissons. Since we could not integrate the values continuously throughout the
profile, we selected the depth intervals on the basis of similar porosity (TPS).
After the water was applied, we could account for both residual and applied water
in caissons 1 and 2 within 2- to 4% (98- and 104%). In caissons 1, the topsoiled
layer, except for the saturated zone, retained the most water, percentage of
saturation decreasing with depth. In caisson 2, the percentage of saturation
increased with depth. Such contrasting behavior is likely to prevail on field
spoils with and without topsoil. Surface crusting on some spoil materials as a
result of weathering may lead to results that are even more extreme than those
shown in Figure 3.9 and listed in Table 3.5 for topsoiled caisson 1. More of the
77
-------
TABLE 3.5. DISTRIBUTION OF TOTAL AVAILABLE PORE SPACE (T?S) AND OF SPOIL UATER IN CAISSONS WITH DEPTH
Caisson
1
Wacer concenCj_L
Depth
m
0.0-0.5
0.5-0.8
0.8-1.8
1.8-2.5
2.5-2.6
2.6-3.2
Applied
Total
TPS, L
478
487
1783
1320
242
1126
-
5646
Original
252
146
491
386
0
0
2115
3390
1st day
451
336
766
444
79
1126$
-
3317
Saturation*
:
94
69
43
34
33
100
-
58
Depth
m
0.0-0.5
0.5-0.8
0.8-1.8
1.8-2.1
2.1-2.5
2.5-3.1
Applied
Tocal
TPS, L
355
527
1585
476
778
1374
-
5095
Caisson
2
Water content. L
Original
196
130
552
179
189
22
2246
3514
1st day
12 3-1
229
952
363
613
1374
-
3654
Saturation
%
35
43
60
76
79
100
-
72
*(lst day/TPS) x 100.
tLess 0.13 n on caisson 1, and 0.31 m on caisson 2 on 1st day data, as result of settling.
TCorrecced for estimated 354 L that leaked out overnight.
total pore space is filled with water in caisson 2 than in caisson 1 (Table 3.4)
because in caisson 2 on the average water is distributed differently and different
amounts of total pore space are available.
Table 3.6 shows water table elevation, seepage flux, and volumes of water
percolating to the saturated zone at different times. Calculations for caisson 2
indicate that 27 min after water application ceased, 38% of applied water had
reached the water table, and by the next morning 59% had reached the water table.
In caisson 1 where water was ponded for 140 min, initial drainage accounted for
15% of the water applied, while overnight drainage accounted for 53%, which re-
sembled overnight drainage in caisson 2 (59%). Very rapid initial seepage flux
at the water table in both caissons slowed down considerably by the following
morning. Since water arrived at the water table fast and before the profiles
in caisson 1 and 2 were saturated, it is likely that infiltrating water may have
reached the water table through a limited number of larger channels.
78
-------
TABLE 3.6. WATER DEPTH*, SEEPAGE FLUXf, AND CONTAINED VOLUMEf IN
THE SATURATED ZONE AFTER WATER WAS APPLIED TO CAISSONS
Time§
min
72#
77
80
1280
60
67
90
150
1160
5480
Water depth
m
0
0.163
0.223
0.576**
0
0.205
0.436
0.464
0.671
0.711
Seepage flux
m/day
Caisson 1
_
19.7
36.4
0.18
Caisson 2
_
18.6
6.4
0.30
0.08
0.01
Volume
%
_
15
21
53
_
18
38
41
59
63
*Water depth above caisson bottom.
tlncoming seepage flux at the water table, corrected for porosity.
TVolume accrued as percentage of applied.
§0n caisson 1 2115 L water applied for 92 min, ponded for 140 min;
on caisson 2 2246 L water applied 0 to 5- and 47- to 63 min.
//Time for the first arrival of water at the bottom of each
caisson.
**This value would be 0.370 m if we excluded the amount that leaked
out of caisson.
MODELING WATER FLUX ON STRIP-MINED LAND
Although a slow, uniform, downward movement of a wetting front is modeled, it
is not always a typical infiltration mechanism on all soil materials. In some
soils infiltrating water appears to move down very rapidly through a system
of macropores (Quisenberry and Phillips, 1976; Dixon and Peterson, 1971). Such
behavior is particularly true in coarse textured stony materials and in soils
that have a well defined system of cracks and fissures (Ritchie et al., 1972).
In other soils, when a finer layer overlies a coarse textured one, instability
79
-------
of the wetting front at the interface may lead to fingering (Hill and Parlange,
1972; Raats, 1973; Philip, 1975). In Appalachian minesoils, because of their
coarse texture, some combination of these mechanisms can prevail.
The objective of this study was to model numerically infiltration and re-
distribution of water in a reclaimed topsoiled and nontopsoiled Appalachian
minesoil.
Appalachian minesoils are unlike natural agricultural soils for which the
numerical modeling techniques are relatively well developed and flow processes
are moderately well known. In these minesoils large size fractions predominate.
Particles larger than 25 mm frequently constitute more than 80% of the total
(Ciolkosz et al., 1983) and boulders, large rock fragments, channels and
fissures are common. Under saturated conditions, turbulent (Leps, 1973) or
transitional (Gill, 1976) flows may occur, and under unsaturated conditions,
water often infiltrates rapidly between large fragments and along discrete
channels down to the water table without noticeably wetting the rest of a
profile. In such circumstances modeling flow is very difficult. We have
attempted to adapt the techniques used in numerical modeling of water flow in
soils to the water flow in minesoils and then to verify model results with
experimental data.
The Caissons
Figure 3.12 is a schematic diagram of the two systems modeled. It repre-
sents two 2.4-m diameter, 3.0-m high caissons assumed to be packed with twenty
identical layers of spoil over sand. Layers 1 through 10 of caisson 1 are
referred to as spoil 1 and layers 11 to 20 of caisson 2 as spoil 2. Caisson 1
represents a system topsoiled with 0.5 m of fine textured soil. Caisson 2
represents a system without topsoil and having a 0.4-m layer very coarse acid
80
-------
Topsoiled System
Rain
Nontopsoiled System
Rain
iJ
3
Soil
Spoil 1
2
3
4
5
6
7
8
9
10
Sand
Spoil 1 1
13
15
16
18
20
— Well —
Acid Shale
Sand
Caisson 1
Caisson 2
Figure 3.12. Schematic representation of
• . the two caissons.
shale above the sand. Each caisson is thus essentially a two layer system of finer
material overlying coarser material. In both cases a water table at a depth of 3 m,
at the sand layer, is the lower boundary.
The actual caissons that these systems attempted to model were used to obtain
experimental data on water flow and on acid generation in spoil and are described
in detail in previous section. They were made from corrugated culvert pipe and
were packed with two ten-layer consecutive portions of a reclaimed field profile
6-m in depth. To approximate field conditions, the first ten layers (caisson 1)
were topsoiled and the second ten (caisson 2) were placed over layer of coarse acid
shale. The materials in the caissons were assembled layer by layer from sequential
excavations transported from a field site. Each caisson was equipped with a
central well and four access tubes for monitoring the water table and profile water.
Selected average properties are listed in Table 3.7 and are shown in Figure 3.13.
Although the moisture characteristic curves (Figure 3.13) do not show it, comparison
81
-------
TABLE 3.7. SELECTED AVERAGE PHYSICAL AND HYDROLOGIC
PROPERTIES OF CAISSON MATERIALS
Caisson
1
Soil
Spoil 1
Sand
2
Spoil 2
Acid shale
Sand
Coarse
fragments*
7
62
73
58
77
82
58
Total
pore
space
0.329
0.363
0.419
0.376
0.450
0.487
Specific
retentiont
3, 3
0.227
0.071
0.138
0.068
0.043
0.138
10 kPa
retention?
0.285
0.146
0.017
0.131
0.064
0.017
Saturated
conduc-
tivity
0.04
0.52
0.15
1.05
2.10
0.15
*Fragments > 2 mm.
tWater retained by the coarse fragments (ElBoushi, 1966) after drainage.
TWater retained by fines at kPa, corrected for coarse fragments.
ikPol
Figure 3.13.
Moisture characteristic (a)
and hydraulic conductivity
(b) as a function of
tensiometer pressure.
82
-------
of the values for total pore space in Table 3.7 with anticipated water content at
saturation in Figure 3.13 showed that large jumps in water content and hydraulic
conductivity at saturation could be expected because of anticipated channel flow
in both the spoil (1 and 2) and the acid shale. Values in Figure 3.13 were used
as input to models in numerical simulation of infiltration and redistribution of
water in caisson profiles.
The moisture characteristics were determined by desorption in the laboratory on
the material smaller than 2 mm and were corrected for the coarse fragment content.
Using these moisture characteristics and the Green and Corey (1971) model, unsatu-
rated hydraulic conductivity values were computed and matched at the saturated
hydraulic conductivity (K). Initially (Table 3.7), K values were assumed equal to
the average field values in surface layers of topsoiled and nontopsoiled spoil where
the caisson material originated (Pedersen et al., 1980). Subsequently, hydraulic
conductivity values deduced from travel times to the water table in the caissons
were also used. The ratio of hydraulic conductivities (K) in the topsoiled system
(K .,:K .,) was 1:14; the ratio in the nontopsoiled system (K .- :K . , ) was
^ soil spoil spoil shale
1:2. Because the experimental moisture characteristics (Figure 3.13) represented a
desorption phase; experimental values were divided by 1.6 to determine the wetting
or absorption phase required for modeling infiltration (Mein, 1971, p. 49).
Flow Models
Model 1—
Using the geometry of topsoiled and spoil systems shown in Figure 3.12, a one
dimensional, nonlinear, second order partial differential moisture flow equation
(equation (3)) was solved numerically for infiltration and redistribution phases
under variable rainfall conditions
36/3t = V • (KV<}>) (3.1)
83
-------
8 is the volumetric moisture content, t is time, K is the hydraulic conductivity,
V is the operator and is the total potential, tj> = fy + z with ip the tensiometer
pressure and z the depth measured vertically downwards. Numerical solution based
on a Crank-Nicholson finite difference scheme followed Mein (1971) with some
modifications and additions. These included adapting the solution to layered
systems, solving for variable intensity rain, and solving for both the infiltra-
tion and redistribution phases after water application ceased. Small depth
increments were taken close to the surface (Smith and Woolhiser, 1971) and at the
soil-spoil and spoil-acid shale interfaces, and these increments became larger
with depth. Some problems were encountered with convergence because of the low
initial moisture content in the spoil, the type of moisture characteristics used
(Figure 3.13), and incorporated changes in rain intensity. The program worked
best if the initial water content corresponded to water content at 300 kPa or
less. Under certain circumstances when convergence problems were encountered we
found it useful to punch out the output on cards every 100 time steps and restart
with new initial data.
In this manner moisture movement into and through the minesoils was modeled
essentially in a conventional way, the output consisting of moisture profiles,
wetting front positions, and pressure distributions with time. Although this
approach simulated adequately the conditions expected in the natural fine grained
soils adjoining the mined area and in the topsoiled minesoil represented by the
geometry of caisson 1, it posed several problems on nontopsoiled spoil, which
often contained as much as 80% coarse fragments. In the latter case if we used
the moisture characteristic values uncorrected for coarse fragments, the program
output showed unrealistically high profile water content. On the other hand, if
we used the moisture characteristics corrected for coarse fragments, the program
showed apparent saturation at much too low a. value.
84
-------
Model 2—
It could be argued that in coarse textured nontopsoiled spoil, water moved prin-
cipally through a few larger channels at the rate equal to the average percolation
rate, but independently of the matric potential gradient. In fact, any pressure
gradients that existed were primarily local (rather than continuous), and tended to
pull water from the relatively few larger channels through which it flowed to the
water table.
To model this type of flow we used the following simple procedure. First,
hydraulic conductivity values in Table 3.7 obtained in the laboratory were compared
with the seepage flux rates in Table 3.8. Since the seepage rates, calculated from
travel time through a 3-m profile, appeared to be similar in magnitude to the
hydraulic conductivity, they were taken to represent flux rates for different runs.
The profile was then arbitrarily divided into approximately sixty 0.05-m-thick
layers. Their properties such as percent coarse fragments, surface retention on
coarse fragments, total pore space and water retention at 10 kPa by fines were
interpolated from the available data for each of the 11 thicknesses (11 through 21)
in Figure 3.12. Water application at a known rate (Table 3.9) was then assumed to
start and continue for a prescribed time (Table 3.8). We further assumed that as
water went from layer to layer enough was pulled out by each layer to satisfy
surface retention on coarse fragments as well as the 10-kPa retention by fines and
that the rest moved on to the next layer, and so on until it reached the water
table. The water table then started to rise based on the total available pore
space and antecedent water content of the profile. Comparison with experimental
profiles could suggest through what fraction of the total area the water moved.
Model 2 is strictly an infiltration model, since both specific retention on coarse
fragments and retention by fines at 10 kPa are treated as constant during
85
-------
TABLE 3.8. EXPERIMENTAL VALUES OF SEEPAGE FLUX AND
ANTECEDENT MOISTURE CONTENT (0)
System
Soil, topsoiled (caisson 1)
and nontopsoiled (caisson 2)
Nontopsoiled (caisson 2)
Nontopsoiled (caisson 2)
Seepage
flux*
nrai/s
0.78
0.96
1.32
0.96
Antecedent
3, 3
m /m
0.1029
0.0926
0.1633
0.1962
Illustrated
in figure
5.2.3 and
5.2.4
5.2.5
5.2.6
5.2.6
*Computed as the depth to the water table divided by time required for water to
arrive at the water table.
TABLE 3.9. SIMULATED AMOUNTS, DURATION AND INTENSITIES OF
APPLIED RAIN AND ANTECEDENT WATER CONTENT
Simulated rain
System
Soil
Topsoiled
Nontopsoiled
Topsoiled
Nontopsoiled
Nontopsoiled
Amount
mm
459
459
201
459
15
0
510
135
0
139
0
211
0
122
Duration
min
92
92
40
92
5
39
16
112
33
72
2456
146
35
11
Intensity
mm/s
0.083
0.083
0.083
0.083
0.054
0.000
0.525
0.020
0.000
0.032
0.000
0.024
0.000
0.177
Antecedent
water
content
% by vol
0.20
0.20/0.10
0.20
0.20/0.10
0.095*
—
0.100
0.169
—
0.198
—
0.254
—
0.338
Illustrated
in figure
5.2.3
5.2.3
5.2.3
5.2.4
5.2.5
5.2.5
5.2.5
5.2.6
5.2.6
5.2.6
5.2.6
5.2.6
5.2.6
5.2.6
*Values on this and following lines represent average profile water content.
86
-------
infiltration. For redistribution, the time behavior of specific retention and
retention by fines at 10 kPa also could be investigated.
Results and Discussion
Figure 3.14 illustrates a set of simulated moisture and pressure profiles using
techniques of model 1 and data from Table 3.7 and Figure 3.13. Three types of
systems were analyzed. Figure 3.14a shows a topsoiled system, Figure 3.14c a non-
topsoiled system, and Figure 3.14b, for comparison, is a natural soil profile. The
total pore space in a soil layer was adequate for simulating water profile develop-
ment in a topsoiled layer, but the only way to make the incoming water move through
spoil to the depth of 3 m fast enough to approximate experimental conditions was to
set the total pore space to the value corrected for coarse fragments (Figure 3.13).
Values almost twice that (Table 3.7) prevailed in the actual profile. If the total
pore space values as given in Table 3.7 were used, the front advance was much too
slow to describe the actual situation.
Figure 3.15 compares model 1 simulation with the caisson 1 experimental data.
In Figure 3.15a water contents, measured during the run immediately after the
wetting front had passed, are compared with a predicted (model 1) profile midway
through the run. Except for the experimental point close to the surface, agreement
appears good. Figures 3.15b and 3.15c compare simulated moisture profiles with
caisson 1 experimental data taken the following day and the following month,
respectively. Model 1 simulations agree fairly well with caisson 1 experimental
results except in the soil layer near the surface. Thus, we believe that model 1
predictions of infiltration and redistribution in the profile below a 0.5-m depth
can be realistic enough on a topsoiled minesoil. The low experimental value of
water content near the soil surface in Figure 3.15a may be either an error or it
may represent a location that did not wet well during water application. Bulk
87
-------
a.
b.
c.
V
0.50
~ 1.00
r
a LSD
u
Q
2.00
2.50
n
• So.
- _- -_-i
f
?
c
i
(N
>
0
Us.
L, L
ft n 1 «
jjnm'
JiU2*
^£J5-
Spoi
•9 n i
J
?A"i
-
-
i
-
Soil
[
i
?
s^'
S-^J
*^Qs
^
.
-
^\
-
.
>
^
^x
Spoil
-
-j
I
1
-
k i
Water Content (m3/m3)
2.50
— 1,000 —10
-0.1 -1,000 -10 -0.1 -1,000 -10 -0.1
Tensiometer Pressure IkPal
Figure 3.14. Moisture and pressure profiles simulated
with model 1 on a hypothetical topsoiled
spoil (a), natural soil (b), and non-
topsoiled spoil (c). The numbers on the
curves indicate time in minutes from the
beginning of water application at the
surface.
88
-------
Spoil
Spoil
W»t*r Content
Figure 3.15.
Comparison of experimental (caisson 1) water contents
obtained on a topsoiled minesoil during the run with
a moisture profile at 0.8 h predicted by model 1
(3.15a); comparison of experimental water contents
taken 1 day (3.15b) and 1 month after the run
(3.15c) with moisture profiles at 1 day and at 1
month predicted by model 1. Solid circles (•) are
experimental (caisson 1) values at indicated times;
open circles (o) in (3.15c) show profile water
content 44 days after the run.
density data for caisson 1 (not shown) suggested that at the depth of 0.3 m a layer
of somewhat lower density was present (80% of the density above and below it).
Since the model assumed homogeneous conditions, this could possibly account for the
discrepancy. Higher caisson 1 experimental values near the surface than predicted
in Figure 3.15b may be due to a similar measurement error or may reflect higher
retention than the model can account for during the redistribution phase. In
Figure 3.15c the lower experimental value near the surface at 1 month may reflect
evaporation losses that occurred during that month and were not accounted for by
model.
89
-------
In Figure 3.16 we compared simulated results for model 1 and 2 with caisson 2
experimental values taken immediately after the first run. For model 1 the agree-
ment between the experimental (caisson 2) and simulated values got progressively
worse with depth. Although model 2 in general approximated the shape of experi-
mental profile reasonably well, it seemed to over- or underestimate caisson 2
experimental values. Table 3.10 lists matching factors necessary to correct model
2 predictions to the values observed after rain application and their reciprocals.
The discrepancy between the observed (caisson 2) and simulated values may be due to
errors. However, the reciprocals may also represent the fraction of an area through
which the flow takes place. In addition, Figure 3.16 shows experimental (caisson 2)
values at 1 month following the first run to demonstrate that in the spoil materials
studied relatively little redistribution took place after the initial infiltration
ceased. Consequently, expressing specific retention and retention of 10 kPa during
simulation as constants may not introduce excessive error.
Next we used model 2 methodology to illustrate a specific situation: a filling
up of the mined out area. Figure 3.17 illustrates progressive stages of this
process. The first dashed curve on the left shows initial water content distribution
in the profile in caisson 2. After the two bursts of rain (Table 3.9) experimental
water contents were determined and expected moisture profile was simulated. The
profile discussed here (caisson 2) is shorter (by 0.31 m) than the profile of
caisson 2 shown in Figure 3.16. This is because of the settling that occurred
following initial application of rain during the first run.
Simulation results (next solid curve) down to 2 m appear to reflect the experi-
mental values well. However, below the 2 m depth simulated values seem to exceed
the experimental. The reason may be that at higher initial profile water contents
more water is retained by the profile than the model allows, and consequently the
water table does not rise as high as projected.
90
-------
00 010 0 20 0 10 0 40 0 SO 0 <0
Figure 3.16.
Comparison of experimental water contents
obtained on nontopsoiled spoil profiles
2 h (•) and 30 days (A) after the water
application ceased with simulated
profiles.
TABLE 3.10. EXPERIMENTAL AND SIMULATED* PROFILE WATER CONTENT VALUES,
MATCHING FACTORS AND THEIR RECIPROCALS
Depth
Water content
Experimental
Simulated
Match factor
Reciprocal
m
3, 3
m /m
0.6
0.9
1.2
1.8
2.4
3.0
0.175
0.209
0.224
0.269
0.174
0.514
0.186
0.174
0.199
0.192
0.161
0.464
0.94
1.20
1.12
1.40
1.08
1.11
-
0.83
0.89
0.71
0.93
0.90
*Model 2.
91
-------
90 fl. 10 0 20 0 30 0 40 0 SO 0 CO
Water Content (m3/mj)
Figure 3.17.
Comparison of experimental water
contents on nontopsoiled spoil
for two intermittent rain periods
(• and *) 41 h apart with water
content profiles simulated using
model 2.
On the right hand side of Figure 3.17 a curve and (•*>) data points reflect the
last two bursts of rain (Table 3.9) following a 41 h pause (Table 3.9). The curve
describes simulated values, the points experimental (caisson 2) measurements.
Simulated values correspond to the distribution of total available pore space.
Experimental values (caisson 2) are again generally somewhat lower in the 0- to
0.5-m depth and below 1 m but higher above the acid shale, suggesting a possible
discrepancy between estimation of the total available pore space based in four
access holes and actual pore space as reflected by caisson 2 experimental values.
It is also possible, as mentioned previously, that the amount of water retained by
coarse fragments is larger at higher soil water contents.
Model 2 could probably be extended to larger areas. Necessary profile
properties, such as percent coarse fragments (> 2 mm) and total pore space, as
well as water retention by fines and coarse fragments could be ascertained
from test borings. Continuous well and rainfall records might provide adequate
92
-------
data to compute seepage flux. For example, we have found that values of seep-
age flux obtained at our field site (Pedersen et al., 1980) compare well with
those listed in Table 3.8. Subsequently, using methods outlined in Rogowski
(1980), contour maps of pertinent properties over an area and with depth can be
developed. If the rain and infiltration capacity distributions at a given site
are known, model 2 could be used to predict profile water contents and water
table accretion for a rain event.
CHEMICAL CONSIDERATIONS
In here we present and discuss the chemical properties of the spoil, spoil
extracts and percolate observed in the caissons and in the laboratory, if
closely coordinated and related to the caisson study. A main laboratory study
included here, examined the processes potentially controlling salt and acid
losses to percolate or groundwater on the spoil particle scale. The individual
objectives were: 1) to identify the source of chemically degraded percolate,
2) to identify and contrast the basic chemical properties which control chemical
concentration and character of percolate, 3) to compare caisson results with
lab-based leachings, 4) to select a chemical parameter as an indicator of other
chemical parameters, 5) to determine the potential of trace metal contamination
of groundwater and 6) to identify and compare processes that potentially
dominate acid production and loss at the spoil particle scale.
Analysis, Sampling and Storage Methods
The analytic methods used were as follows,
For each spoil layer, total S was determined by a high temperature combustion,
while standard tests (ASTM, 1971) were used to determine sulfate, pyritic and
organic-S content of the spoil matrix.
93
-------
On spoil extracts and percolate, the chemical analysis included electrical
conductivity (EC), total acidity, pH, SO,, Ca, Mg, Mn, Al, ferrous Fe, total
soluble Fe, Cd, Cu, Zn, Pb, Hg.
Total acidity to pH 7.3 and 8.2 were determined by NaOH titration and a
glass-calomel electrode. The sample was first pretreated with 30% H«0? and
boiled (EPA, 1974). Electrical conductivity (EC) was determined at ~25C by a
Yellow Springs instrument Model 31 conductivity meter equipped with a Beckman
VS01 conductivity cell of 0.1 cell constant.
All metals, except ferrous iron were analyzed with an instrumentation
Laboratory (IL) 351 Atomic Absorption (AA) Spectrophotometer using a background
correction where possible. Al, Ca, Mg, Mn, total soluble Fe, Cd, Cu, and Zn
were analyzed by flame AA (Emmel, 1977). Hg, Pb, and Cr were analyzed with and
IL 555 carbon furnace AA using the method of known addition (Emmel, 1976).
Because the background was too high for some lead samples when using standard
methods, a slower atomization rate was employed. Sulfate concentrations were
determined by AA measured loss of added barium due to BaSO, precipitation (Borden
and McCormick, 1970). Ferrous and total soluble Fe were determined spectrophoto-
metrically as the orthophenanthroline complex on a Coleman Model 14 Universal
Spectrophotometer (APHA, 1965).
The spoil percolate was obtained from caisson lysimeters whereas spoil ex-
tracts were basically the result of lab-based spoil leaching procedure.
The sampling methods were as follows.
The spoil material analyzed was subsampled from the spoil used Co reconstruct
the caisson profile.
Extract sampling was done in two closely related ways. For Table 3.12 a
10-kg subsample of each spoil layer was placed in a 20 cm diameter PVC cylinder
94
-------
and leached with 1000 ml of water and the leachate was filtered. For trace metals
in Table 3rl5' (total soluble Fe, Mh, Ca, Zn, Cd, Cr, Pb, Hg), approximately 80 g
of each spoil layer was placed in a glass bottle and reciprocally shaken with
distilled water (80 ml) no less than 8 hr/week over 4 weeks. The sample was
centrifuged and the clear supernatant was analyzed.
The spoil percolate in the caisson was sampled from lysimeters. Three 50 x
2
610 mm lysimeters with a catch area of 0.0305 m , added during the winter of
1977-78 to augment the original system described in the previous section, were
placed at three levels in each caisson (positions H, M. L). The lysimeters were
operated and percolate sampled continually through all runs. When the lysimeters
outflow rate exceeded one sample every 10 minutes, the sampling frequency was
reduced to one sample/10 minutes. Sampling ceased upon cessation of lysimeter
outflow or extensive flooding of the lysimeter due to a rising water table. In
subsequent tables and figures, samples from the water table maintained in the
60 cm thick bottom sand layer are "well analysis" whereas those labeled "water
supply" refer to the added rain water. Well and water supply samples were taken
immediately prior to the initiation of each run.
Sample storage was usually not a problem except for caisson percolate which
could generate large numbers of samples. Other samples were refrigerated at 3 C
and analyzed within a few days. Caisson percolate samples were collected in
polyethylene bottles, centrifuged and part of the supernatant was analyzed on
site for pH, ferrous Fe, redox potential, total acidities at pH's 7.3 and 8.2.
The remainder of the sample, was immediately placed in air tight plastic bag,
refrigerated (3 C) for transport, frozen within 48 hours and usually analyzed
within one week after collection. Samples selected for trace metal analysis
were treated similarly except for storage up to 12 weeks before analysis.
95
-------
There was no significant effect of storage on the total acidity and total soluble
iron concentrations as determined experimentally.
Chemical Properties of the Spoil
The chemical properties of the spoil are determined in two ways. One way is to
measure the pyrite, sulfate and organic sulfur content of the 21 spoil layers and one
soil layer. The sulfate is a measure of the potential salt content which could be
lost to percolate waters and the pyrite is a measure of the potential acid production
Another way is to measure the solutes extracted in a water-spoil equilibration or
leaching, usually at a low water to spoil ratio. In this case, primarily acidic
salts (iron, manganese and aluminum sulfates), neutral salts (calcium, magnesium, or
sodium sulfates), or sulfuric acid analyzed in the extract represent maximum concen-
trations due to a lack of dilution, extensive contact or long equilibration times.
The spoil and water extract properties are presented in Tables 3.11-3.13.
The pyrite, sulfate and organic sulfur content range between two extremes (Table
3.11.), the lowest set by the soil (layer 0) and the highest by a black shale
(layer 21). Sulfate or pyrite-S concentration of the other layers, appear to be in
about the range of 0.05 to 0.15% and would be expected to exert a somewhat similar
impact on these concentrations in percolate.
Extracts of each spoil layer contain substantial total acidity, total salts, and
sulfates. Spoil layer 21 potentially dominates the system with by far the highest
concentrations of the chemical parameters (Table 3.12) . Extracts of spoil layer 8
contains a surprisingly high total acidity content compared to the sulfate content of
the spoil layer. In fact, there is substantially more variability in the total
acidity or total SO, extracted among spoil layers 1-20, then would be expected on the
basis of the pyrite S or sulfate S concentration taken from Table 3.11.
96
-------
TABLE 3.11. SPOIL PYRITE ANALYSIS
Caisson Layer
1 0
1
2
3
4
5
6
7
8
9
10
2 11
12
13
14
15
16
17
18
19
20
21
Total
soluble
Fe
mrr 1 0
mg/ x,
7.4
26.3
32.7
38.7
46.8
33.8
43.1
54.0
24.4
30.4
43.0
49.2
45.1
49.1
36.6
42.8
28.2
22.2
29.2
20.7
22.8
677.5
Pyritic-S
.012
.076
.095
.112
.136
.098
.125
.157
.071
.088
.125
.143
.131
.143
.106
.124
.082
.064
.085
.060
.066
1.967
Sulfate-S
.012
.065
.082
.079
.089
.066
.066
.076
.094
.070
.053
.054
.062
.053
.066
.049
.059
.056
.049
.077
.055
.738
Organic-S
.012
.047
.033
.070
.065
.063
.052
.056
.036
.048
.053
.051
.050
.051
.043
.043
.041
.032
.040
.111
.033
1.095
Total
.045 soil
.188 spoil
.210
.261
.290
.227
.243
.289
.201
.206
.231
.248 spoil
.243
.247
.215
.216
.182
.152
.174
.248
.154
3.800 black shale
-------
TABLE 3.12. VALUES OF pH, ELECTRICAL CONDUCTIVITY (EC), TOTAL ACIDITY
CA, MG, AND S04 FOR TWENTY-ONE LAYERS OF RECONSTRUCTED SPOIL PROFILE
No
0
1
2
3
4
5
6
7
8
9
10
Sand
11
12
13
14
15
16
17
18
19
20
21
Sand
Depth
m
0
0.51
0.74
0.91
1.12
1.32
1.55
1.83
2.08
2.29
2.52
2.74
0
0.28
0.48
0.69
0.91
1.17
1.44
1.68
1.89
2.15
2.45
2.82
PH
3.91
3.35
3.29
3.10
3.17
3.37
3.18
3.05
2.35
3.17
3.35
2.97
2.98
3.21
2.56
3.42
3.14
3.19
3.21
3.36
3.37
2.79
EC
ymhos/cm
410
275
740
800
995
660
620
445
3400
998
870
905
820
900
2000
280
1280
660
845
2050
1100
6000
Total
acidity
2.89
1.03
5.09
5.12
6.62
3.10
3.45
2.90
45.44
8.04
5.86
6.68
6.22
5.92
15.72
1.00
8.72
3.78
5.38
14.24
6.86
67.08
Ca
Caisson
1.71
0.71
1.96
1.73
3.08
1.63
1.81
0.66
5.25
3.11
2.78
Caisson
3.08
2.65
3.98
4.44
0.84
6.02
2.55
3.29
10.30
4.90
23.33
Mg
1
0.53
0.34
2.96
0.02
4.07
1.98
1.95
0.57
6.98
3.40
3.23
2
3.40
3.15
3.60
4.16
0.34
4.41
2.08
3.26
9.67
3.80
24.58
S°4
0.88
2.63
8.87
8.77
11.87
6.57
6.79
'4.50
54.85
11.89
9.69
11.09
10.19
10.89
20.88
2.80
14.90
7.70
9.70
33.38
12.80
134.81
98
-------
TABLE 3,13. TRACE METAL CONCENTRATION OF SPOIL EXTRACT*
Layer
Blank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
§
PH
4.02
3.67
3.40
3.46
2.86
3.49
2.80
3.69
2.90
3.77
3.08
3.41
3.41
3.57
3.28
3.58
4.03
3.41
3.77
3.40
2.65
-
Cd
<.004
<.009
<.004
.012
.012
.029
.022
.010
<.004
.026
<.009
.015
.016
.015
.010
.010
.005
<.004
.006
.019
.011
.049
.010
Cr
<.001
<.001
<.001
.002
<.002t
.013
.001
<.001
<.001
.006
<.001
.005
.004
.001
<.001
.001
<.001
.003
<.001
<.001
<.001
.240$
0.05
Cu
<.02
<.02
<.02
.08
.30
.84
.48
.10
.08
.64
.02
.45
.46
.28
.10
.20
.12
.04
.10
.10
.19
.74
1.0
Fe
<.2
<.2
<.2
<.2
.4
4.5
.4
.4
<.2
1.0
<.2
1.6
1.2
.3
<.2
.3
<.2
<.2
.3
<.2
<.2
3700.
0.3
Hg
<.02
<.02
<.02
<.02
<.02
.06
<.02
<.02
<.02
<.02
<.02
<.02
<.02
<.02
<.02
<.02
<.02
<.02
<.02
<.02
<.02
.05
0.002
Mn
<.02
35.4
28.3
64.6
39.4
161.
83.8
59.6
27.3
128.
32.3
85.8
79.8
56.6
40.4
61.6
40.4
29.3
40.4
69.7
63.6
319.
0.05
Pb
<.001
.003
.003
.012
.009
.020
.015
.009
.010
.018
.003
.027
.013
.011
.003
.011
.008
.007
.006
.007
.012
.010
0.05
Zn
<.02
.65
.75
1.95
2.75
5.13
4.00
2.70
1.22
5.00
1.15
4.30
3.80
3.50
1.65
3.15
2.05
1.65
2.10
2.75
2.85
26.0
5.0
- = not determined.
* = concentration either mg/& extractant or mg/g spoil; mass ratio of extractant to spoil equals 1:1.
t = insufficient sample.
T = flame analysis; rest determined by carbon furnace atomic adsorption.
§ = EPA drinking water standards.
-------
Trace metal concentrations, Cd, Cr, Cu, total soluble Fe, Hg, Mn, Pb and Zn, in
spoil were analyzed (Table 3.13). The EPA Interim Primary Drinking Water Standards
(EPA, 1975) for Cd, Cr, Hg and Pb and the EPA Proposed Secondary Drinking Water
Standards (EPA, 1977) for Cu, Fe, Mn and Zn are included for comparison. The trace
element contents of 21 spoil extracts relative to the drinking water standards were
equaled or exceeded in 14 layers for Cd, 1 for Cr, none for Cu, 11 for total soluble
Fe, 2 for Hg, all for Mn, none for Pb and 3 for Zn. Because the analytic limits for
mercury analyses were 10 times higher than the standard, we suspect, but cannot prove
that the mercury standard was exceeded more than twice. Again, spoil layer 21
extract provided the highest concentrations.
Chemical Properties of the Spoil Percolate
The mean values (x) and standard deviation (Sd) of pH, total acidity (pH 8.2),
ferrous Fe, Mn, Al, Ca, Mg and SO, concentrations in percolate from lysimeters (L-l,
2, 4, 6, 8, 9), water supply and well are presented for all runs (Table 3.14).
Table 3.14 shows the water quality to generally deteriorate with depth in both
caissons, dramatically so at lysimeter L-9. The potentially greatest acid producing
spoil layers from the previous section are: primarily spoil layer 21 located
immediately above lysimeter L-9 in caisson 2, and to a lesser degree spoil layer 8,
which drains into lysimeter L-8 and L-9 in caisson 1. There appear to be two
processes affecting water quality degradation. First, the percolate quality gradu-
ally deteriorates from L-l to L-6 in both caissons. Because fluctuation of potential
acid products concentration in any spoil layer overlying L-l to L-6 is not extreme,
the acid product concentration in the percolate steadily increases with spoil depth.
Thus, the source of the chemically degraded percolate is general and no contributary
zone can be clearly delineated. Second, the percolate quality generally deteriorates
dramatically from L-6 to L-9 in both caissons. Thus, if the fluctuation of acid
100
-------
TABLE 3.14. CHEMICAL C11AKACTER1STICS OF SPOIL DRAINAGE AT SELECTED DEPTHS WITHIN CAISSONS ] AND 2*
Sanipl tng
depth
pll
sdt
Total
X
acidity
Sd
roeq/l
Caisson it
Water
supply
l.-l
L-2
L-4
1.-6
L-fl
L-9«
Hell"*
Caisson 2ft
Water
supply
L-l
I.-2
L-4
I.-6
I.-8
L-98
Wei If
IS, It the a
6.0 0.0
3.8 1.0
5.7 0.5
4.2 0.0
4.1 §
3.4 0.1
3.6 0.0
6.8 0.4
6.3 0.4
3.8 0.2
3.8 O.L
3.5 0.1
3.6 0.2
2.6 0.1
2.7 0.1
2.5 0.1
0.2
0.2
0.2
0.5
1.4
5.0
12.1
14.5
0.9
3.1
3.2
1.9
6.3
22.
82.
157.
§
0.9
0.2
0.0
§
1.0
0.2
4.3
0.1
1.9
3.3
0.4
4.2
IB.
50.
33
Ferroua Fe
X
0
0
0
-
1
2
1
100
1
2
3
4
2
207
278
2600
Sd
.8 9
.1 0.1
.0 0.0
-
.7 9
.2 9
.9 0.9
68
.7 0.8
.4 1.2
.1 4.1
.0 3.8
.6 1.9
9
191.
1580
Mn
X
.2
3.5
0.4
-
16.
99.
330
0.2
14.7
18.
17.5
43.
117
258
456
Sd
9
5?0
0.1
-
§
6.
11.
0.1
15.0
23.
7.
13.
3
134.
14.
ta
X
0.0
O.I
0.4
-
_
22.
84.
0.3
0.0
12.2
1.5
8.1
32.
68.
169.
266
Ca
Sd
9
0.
0.
-
_
9
14.
9
0.
10.
0.
2.
17
55.
79
53
X
mg/t
2.4
1 11.0
3 10.6
-
-
88.
270
490
0 1.8
0 80
7 48
1 61
200
277
392
400
Sd
9
5.8
3.4
-
-
9
24.
5
0.0
57.
70
17
74
49
126
140
Mg
_
X
0.1
1.2
2.0
-
-
70
210
250
0.1
41
32
31
105
135
308
511
Sd
9
1.0
1.7
-
-
9
28
9
0.0
29.
42
25.
25
116
°86
SO
„
X
13.5
6.8
5.3
-
-
740
2300
3800
29.
550
410
310
1270
3300
7000
10500
4
''d
9
44.
15.
"
—
1120
410
9
*
410
580
270
460
2130
3250
1800
t Included 7/26/77, 9/U-L6//7 run a.
§Qt\e observation.
'Data available for flooding condition only.
**Uatec table contlnuouaLy maintained In a and layer (below spoil).
fflncludea 8/18/77, 10/5-7/77, 8/23/78 runs.
-------
products concentration in selected strata is extreme, i.e., total acidity for layer
8, caisson 1 or layer 21, caisson 2, these spoil layers have the most impact and
control percolate quality. Under these conditions, the contributing zones are
specifically and clearly delineated.
The effect of different incubation periods and water applications associated
with each run undoubtedly affected, but did not override the acid products con-
tributing pattern for each caisson. Possibly, sufficiently extreme incubation
conditions and water applications were not employed to stress the chemical system
and, thus break down the pattern described here. However, the experimentally
employed conditions adequately represent most of the extreme water application
and incubation conditions that would be experienced in the field (see hydrology
section).
A similar analysis of caisson percolate shows the same pattern of trace metal
contamination (Table 3.15). The level of contamination was generally similar to
that determined for the corresponding spoil extracts in Table 3.13. The results
compare well for Cd, Fe, Hg, Mn at all lysimeters; Cr at lysimeter L-9, caisson 1
and lysimeters L-6, L-8, caisson 2; Pb at lysimeter L-9, caisson 1 and lysimeters
L-8, L-9, caisson 2; and Zn at lysimeter L-9, caisson 2. Contamination in the
remaining combinations compared to spoil extract are considerably higher,
particularly for Cu and Zn. Analyses of caisson wells which accumulate all
percolate during the runs, corresponded with lysimeter data for caisson 2 but not
for caisson 1. The reduced solubility and chemical precipitation caused by the
high pH (7.0) of caisson 1 well waters probably accounted for the reduced trace
metal contamination. The cause of the high pH is likely due to acid neutraliza-
tion by carbonate minerals. Note the high Ca and Mg concentrations relative to
the low total acidity (Table 3.14).
102
-------
Sample
TABLE 3.15. TRACE METAL CONTAMINATION OF CAISSON PERCOLATE AND
WELL WATERS TAKEN FROM CAISSON I AND 2 FOR RUN 1
Time
PH
Cd
Cr
Cu
Fe
Hg
Mn
Pb
Zn
mg/£
Caisson 1
o
L/J
Water
supply*
L-9*
Well*
Caisson 2
Water
supply*
L-6
L-8
L-gJ
Well*
0820
1300
0800
0853
1000
0948
0936
0800
6.1
3.6
7.0
6.0
3.6
2.5
2.4
2.7
<.004
.021
<.004
<.004
.019
.022
.061
.055
<.001
.020
.013
<.001
.032
.150
.400
.180
<.02
.25
<.02
<.02
.14
.41
5.1
2.40
<0.02
5.4
6.0§
<0.02
2.7
564.
1490.
1083.
<.02
<.02
<.02
<.02
<.02
<.02
<.02
<.02
<0.02
51.
336.
<0.02
28.
118.
419.
459
t
t
.028
t
.600
.012
.012
.050
t
8.40
0.19
t
6.00
9.40
17.3
11.2
*Sampled before start of leaching cycle. Water level in well maintained at approximately a con-
stant level within sand layer between runs.
tNo reading, insufficient sample.
TRising water table intersects lysimeter (1st sample).
§Questionable answer due to analytic interferences from other solutes and extensive precipitate
formation. Ferrous iron concentration of well determined onsite at time of sampling was 51
mg/A.
-------
Chemical Interrelationships of Spoil and Spoil Percolate
The spoil chemical properties set a potential which controls percolate quality
subject to the percolation rate or volume. The spoil chemical properties can be
dominant under many conditions and, thus provide a means of classifying and
estimating its impact on percolate quality.
»
The average total acidity of percolates taken from different depths within each
caisson (Table 3.14) were compared with total acidity of spoil extracts (Table 3.12)
from corresponding spoil layers (Figures 3.18 and 3.19). The total pyrite and total
sulfate content from Table 3.11 of each layer are included for comparison. Data
analysis attempted to determine if discrepancies between the two independent
measures of total acidity existed and if under certain circumstances, basic spoil
Total Pyrltlc and Sullate Sulfur, 9/100 g
0.01 0.1
i.o
T
Caisson I
O Mean t itandard deviation ol total oc'dtt
- 1.0
— Z.C
-I I.I
•j4.c r
T
-U.o a
i
JT.O
A lotal «uilal«. S d«« lop axil label!
& lotal pynt«, S (v«* lop axil labvtl
-f <*>»•• ilandard deviation
Slanoard n...ol,on co«l.«x*l oil «al.
1.0
10-C
1.0 TO.O
Total Acidity ipH82) meg/1
Figure 3.18. Relation of total acidity from lysimeter
drainage with spoil and spoil extract
analyses for corresponding spoil layers.
104
-------
a
D
a
Total Pyrltlc and Sulfate Sullur. g/ioog
0.01 0.05 0.1 1 0
1 1 r
O Maan t standard deviation of total icldlty
• Total acidity of (Iran ailract (Rogowakl, 1977)
A Total lulfat*. S (<•• top axil labal)
& Total pyrile. S daa top axis labaj)
|r»«« standard daviatton
-•fl*-- Standard deviation continual off scale
Caisson 2
1.0
2.0
3.0
4.0 r
5.0 i
a
«
6.0 Q
7.0
8.0
9.0
10.0
10.0 100.0
Total Acidity . meq/l
Figure 3.19. Relation of total acidity from lysimeter
drainage with spoil and spoil extract
analyses for corresponding spoil layers.
properties such as total pyrite content could be used to approximate the observed
percolate quality.
The relationships of total acidity in percolate and spoil extracts, although
initially separated at caisson 1 spoil surface, converged with depth (Figure 3.18).
An analogous comparison for caisson 2 shows both relationships to be generally
coincident (Figure 3.19).
The greatest divergence between these relationships in both caissons was where
the total acidities were lowest, particularly less than 3.0 meq/l on caisson 1.
Consistent with previous observations in Table 3.14, the total acidity in the
caisson 1 percolate increased progressively with depth to the 2.3 m level (L-8),
at which point the spoil extract parameter became a good measure. There appeared
105
-------
to be no correspondence between the total acidity and either total pyrite or total
SO^ content other than a decrease in all parameter values observed generally within
1 meter of the spoil surface. In caisson 2, the total acidity of percolate, total
acidity of the spoil layer extract, and the total pyrite and SO contents clearly
indicate the individual dominant influence of spoil layer 21, but not the other
layers (Figure 3.19).
It should be noted that while pyrite content, as given by percent S in Table
3.11, of layer 21 is one order of magnitude greater than in overlying layers,
pyrite content of layer 8 is not. However, SO,, a pyrite oxidation product, is
considerably more concentrated in layer 9 relative to overlying layers. Possibly
the similarity in magnitude of acid production between layers 8 and 21 results
from an originally greater pyrite availability in layer 8 due to greater exposed
surface area or a greater amount of framboidal pyrite (Caruccio, 1973).
Thus, the total acidity determined for each spoil layer extract provides a good
estimate of the total acidity of percolate, if the potential acid products contribu-
tion of the spoil layer is reasonably high (>5 meq/1). At low concentrations,
hydrologic and spatial variability diffuse the contributions of many spoil layers.
Measures of total sulfate or total pyrite appear to be useful predictions of the
total acidity of percolate when great differences in sulfate content or acid
producing exist, respectively. Otherwise, these two parameters appear insensitive
to changes in total acidity of percolate.
The pattern, volumes and rates of flow would be expected to affect the inter-
relationship between spoil layer chemistry and percolate, especially under extreme
conditions. Under conditions of high water application rates (0.143 mm/sec -
caisson 1; 0.055 mm/sec - caisson 2), the percolate outflow quality decreased at a
rapid rate to a stable level in most cases (Figure 3.20). The total acidity of
sequentially sampled percolate from the H and L lysimeters followed a common
106
-------
- '"r Co...en 1
f I
I I I I I I
b.
H — High lystmatar
M— Medium lysimatar
L — Low lysimatar
• -Well
Ol LyBi««t«r ruMniafl •! hffh
5) rat*, ••M*pl*rf p«rl«4lc»My.
Appr««iM«l«tr WOO ml
All milllm ..i-p
I I I I I I
• 140 300 420 soa 700
1400 • t4« 2*0 4X0 S4O 700 1400
Accumulated Lylimclar Outflow Volu
1100
mis
Figure 3.20. Total acidity as related to flow duration on caissons 1
(7/18/78) and 2 (8/23/78).
decreasing curvilinear relationship. At all lysineters where sufficient samples were
taken, the total acidity upon extended leaching stabilized according to lysimeter
location. Expressed in meq/1, these were approximately: caisson 1, H-1.0, M-
5.0, L - 7.0; caisson 2, H - 3.5, M - Unknown, L - 12.5 to 15.5. Although consider-
ably higher than the stabilized total acidities, the initial total acidity values for
the H and L lysimeters also relate to lysimeter location. The amount and the rate at
which rainwater is applied (Rogowski, 1978) does have substantial impact on the total
acidity of strip mine spoil percolate during the initial part of the run. However,
the total acidity of the percolate quickly stabilizes relative to its position in the
spoil or the presence of a dominating spoil layer. The common curve shape for
different depths suggests that the same processes control total acidity concentration
in percolate.
107
-------
One reason for the readily discernible and, in some cases, not greatly reduced
effect of spoil chemical properties on percolate quality even under high percola-
tion rates, may be due to partial and changing leaching patterns of the overlying
coarse spoil throughout the event. The results indicate that flow occurred in
caisson 1 on the average through only 3% of the overlying volume, while flow
occurred in caisson 2 through 30% of the overlying volume at the H level (near
the surface), 2% at the M level and 6% at the L level. These values were calcu-
lated in the following way. The lysimeter depth, the time between the start of
rainfall application and the first lysimeter outflow, rate of lysimeter outflow
and rainwater application relative to lysimeter outflow were used to estimate the
initial percolation rate and final outflow rates. The average porosity of spoil
material in caisson 1 was 0.384 and in caisson 2 was 0.447 by volume. Thus, rain-
water application rates corrected for porosity would be .372 (.143/.384) and 0.123
mm/sec for caissons 1 and 2, respectively. Outflow rate/application rate ratio
for H and L lysimeters, caisson 1, becomes then .04 and .02. Similarly, for H, M
and L lysimeters in caisson 2, the ratio becomes .30, .02 and .06, respectively.
For these lysimeters to produce outflow, saturated flow must have occurred over
some portion of the porosity immediately above the empty lysimeter cavities.
These computations support a hypothesis (Rogowski, 1978) that flows in coarse
spoil materials occur only through a portion of available pore space. A word of
caution, under unsaturated flow conditions some divergence of flow in the vicinity
of lysimeters cavities is likely.
Another reason high percolation rates did not mask the expected percolate
quality based on spoil chemical properties, is that acid product concentrations
at or near the spoil particle surface are in excess of flow transport capacities.
Under the most extreme leaching conditions simulating a rapidly moving saturated
flow system, spoil layer 21 was leached in the laboratory (Figure 3.21). The
108
-------
40
CT
0)
>,
•H
T3
20
O
H
120 240
Leachate, mm
350
Figure 3.21. Total acidity as related to quantity of leachate
under saturated conditions for spoil 21.
flow rate was 12.5 ml/min for an 85 g sample, with a pore volume of 25 mis. Incubat-
ing the column under an oxygen-free (N_) atmosphere for two weeks and repeating the
leaching step essentially regenerated an approximately equivalent amount of acid
products, implying that diffusion of acid products from the particle interior to the
surface is an important process.
Processes Controlling Acid Product and Acid Losses from the Spoil Particle
If a single spoil particle from the worst spoil layer can be considered to
represent the behavior of that layer, we would probably observe the following.
Immediately following particle formation due to the mining process, the spoil
particle would probably have pyrite and pyritic reaction products distributed
with some degree of ordered uniformity throughout the particle. Initially and
rapidly, acid products at and close beneath the particle surface would be lost
109
-------
to percolating or surrounding waters due to the short diffusion distance. If
extensive ongoing pyrite oxidation was the primary source of acid products, that
lost to drainage waters would likely be controlled by the kinetics of the pyrite
oxidation reaction (we assume percolation is not limiting if the spoil particle
is in an unsaturated zone). After the pyrite or indigenous acid products at or
near the particle surface are removed, the accumulation rate of acid products at
the particle surface would likely become progressively more diffusion controlled,
i.e., controlled by CL diffusion in and/or acid products diffusion out. We
would expect acid products lost from the particle, or the spoil to be initially
high and then reduce over time to some base level. This behavior has been
observed for stripmine drainage (Vimmerstedt and Struthers, 1968). Thus, the acid
products contribution to percolating or surrounding waters would initially be
controlled by reaction kinetics but would gradually shift to diffusion control.
The specific objectives of this section are to: 1) determine at the spoil
particle scale if the diffusion processes potentially control acid product movement
to the particle surface or 0- diffusion to the reaction sites; and 2) compare some
of the published acid production rates for spoil based on pyrite content to the
observed rate of pyrite oxidation.
The basic approach was to alternatively leach and incubate spoil-containing
columns in the laboratory under oxygen free (N«) or atmospheric (20% 0-)
conditions. The rate of acid production and acid product regeneration during
these incubations, and the leaching curve characteristics during the leaching
step, provided the data which was either used directly for comparing the
importance of acid production versus acid products leaching or indirectly as
input into a diffusion equation.
110
-------
Laboratory Setup and Procedure
The physical setup consisted of 13 columns, each being continuously incubated
at room temperature (~25C) for periods of 5.5 - 3305 hours. Either water saturated
air (20% 0~) or N~ gas, was passed through these incubation columns. The long-term
gas-flow rate through these series connected columns following the preliminary high
rate purging step for the N- incubation, was 5 ml/min for N~ and a 40-80 ml/min
for air. Both rates were measured at the exhaust of the last column in the
series. In the N~ incubation system, a 10 cm water-pressure equivalent head was
maintained using a manostat bottle. This, in addition to the 5 ml N_/min flow
rate through the system prevented air entry. Because the total void volume of 7
columns (most ever run in a series) was less than 0.5 liters, the total air
volume was replaced at least once every 12.5 minutes at the air flow rate. Each
column was incubated from one to seven times with each incubation being followed
by a leaching cycle. In some cases, the treatments (air or NL) were switched
between incubations (Table 3.16).
The columns are polyethylene cylinders of 3.2 cm internal dia x 30 cm length,
'stopped with neoprene stoppers and end packed with glass wool. These columns were
series connected by tygon tubing with gas inflow at the bottom and exhaust at the
top.
The columns were filled with approximately 75 gms of sieved, washed spoil 21.
Spoil 21 characteristics and source areas were described earlier in this section
and the hydrology section. Spoil 21 was taken from previously undisturbed shale
strata immediately overlying the coal and contains 2% pyritic sulfur, 0.7% sulfate
sulfur and 1.7% organic sulfur. Before packing the columns, the spoil has been
sieved to remove the fractions less than 2.0 and greater than 8.0 mm. The wash
procedure was used to remove most of the surface associated salts and acid
products before N. and air incubations were initiated. The wash step consisted of
111
-------
Column
no.
1
2
3
i,
5
6
J
8
9
10
11
12
13
Sequence 1
hrs/gas
43/N
137/N
1482/N
46/N
1484/N
67/N
21/A
139/A
25/A
W63/A
47/A
45/A
70/A
Sequence 2
hrs/gas
268/N
25/N
2523/N
115/N
3305/N
123/N
1100/A
790/A
115 /A
-
122/A
245/A
1QQ/A
Sequence 3
hrs/gas
310/N
19/N
-
20/N
-
17/N
1969/N
1968/N
22/A
-
17/A
310/A
16/A
Sequence 4
hrs/gas
16/N
6/N
-
4/N
-
409/A
-
-
6/A
-
408/N
15/A
-
Sequence 5 Sequence 6 Sequence 7
hrs/gas hrs/gas hrs/gas
-
745/A
-
763/N
-
6/A 15/A 7/A
-
-
954/N
-
6/N 16/N 7/A
-
- - -
*Time * hours incubation; N « 1001 N. atmosphere; A - ambient atmosphere (-202 0 ).
sequentially leaching (every two minutes) one retention volume (~25 mis) with dis-
tilled water until the leachate was less than 125 ymhos/cm electrical conductivity
(EC). This was the starting point for the first incubation. The starting point
for subsequent incubations was at the end of the extraction cycle described as
follows.
The column was bottom filled with distilled water to the top of the spoil,
immediately distilled water was added to the top of the column, then 25 mis were
drained from the column. This removed the water contained in the column void
space below the spoil which had not contacted the spoil material. The draining
of the first 25 mis was the zero or starting time. Following this, one retention
volume (25 mis) was drained from the column every two minutes and chemically
analyzed. The time to drain the 25 mis was approximately 5 seconds. The reten-
tion volume, i.e., void space within the spoil, ranged from 22-25 mis/column.
The two minute sampling frequency was carried out until the EC was less than or
equal to 125 iamhos/cm. Once collected, the 25 ml leachate sample was immediately
analyzed for EC and pH. In excess of 50 composited samples, each composite taken
from different columns at different stages of elution, were analyzed for Ca, Mg,
112
-------
Mn Al, total soluble Fe and SO,. Two or three individual samples, taken in series
from the same column and of approximately the same value, were composited to
provide sufficient sample for analysis and a more stable result. These data were
used to develop: 1) an EC - total salts relationship (sum of the Ca, Mg, Al, Mn,
total soluble Fe and SO, concentrations) with equation and appropriate statistics
(Figure 3.22, 3.23) and 2) an EC - SO, relationship with the equation and
appropriate statistics (Figure 3.24, 3.25). Figures 3.23 and 3.25 provide an
expanded scale to observe the low values in these relationships. The SO, or total
salts concentrations predicted by either equation below 850 EC was sufficiently
close so that only the equations in Figures 3.22 and 3.24 were used.
15000
12500 —
10000
3 f 750°
5000
2500
Statistics for log transformation of the
relationship TS-a(EC)b
r2 - ,983"
a - 0.024t
b • 1.49t
texceeds 0,011 probability level
95Z approximate confidence
bounds about regression line
2000
4000
EC (umhos/on)
6000
8000
Figure 3.22. Relationship of total salts (TS)
with electrical conductivity (EC)
for selected column leachates.
113
-------
220
440
EC (umhol/cm)
660
B80
Figure 3.23. Relationship of total salts (TS) with
electrical conductivity (EC) for
selected column leachates (expanded).
12000
10000
8000
f 6000
4000
2000
Statistics for log transformation of the
relationship S04-a(EC)'°
r - .984
a - .027^
b - 1.44T
1 exceeds 0.01Z probabil-
ity level
95Z approximate confidence
bounds about regression
line
2000 4000
EC (umhos/cm)
6000
8000
Figure 3.24. Relationship of SO^ with electrical
conductivity (EC) for selected
column leachates.
114
-------
600
500 _
400 _
f 300 —
200 -
100 _
250
500
EC (umhos/ca)
750
1000
Figure 3.25. Relationship of SO, with electrical
conductivity (EC) for selected
column leachates (expanded).
The water-filled porosity for spoil layer 21 was determined on spoil samples
that were stored prior to use under a water saturated condition in excess of one
year. These were drained, rinsed with distilled water, rolled in absorbent towels
until the reflected sheen of free surface water from the particle surface
disappeared, then subjected to a weighing-drying-weighing sequence. Drying was at
105C until a stable weight loss was achieved, followed by drying at 125C as a
check. The water-filled porosity was calculated by dividing water-lost volume
(105C) by dry spoil volume where dry spoil equals spoil mass divided by 2.63, the
measured particle density. The water-filled porosities of replicates were 11.5%
and 11.9%, providing an average 11.7% (v/v).
Calculation Method
The basic equation used to delineate the acid salt source and acid producing
zones in a water saturated spoil particle is a previously published diffusion
115
-------
equation (Crank, 1964). The assumed geometry is a cylinder in which diffusion proceeds
in the x and y along the cleavage between planes but not z direction perpendicular to
the planes (Figure 3.26). The spoil particle is considered isotropic in the x, y
directions and the system at steady state (dQ /dt = 0).
2IlDt(C -C,)h
Qt = In (r'/r ) (3'1}
b a
QC = total mass (mg) of salt lost per particle and is measured during
extended leaching (where dQ /dt = 0) following column incubation
(Figure 3.27). The measured salts leached from the column times
.257 gins/total column mass provide the salt leached from the
individual particle. The individual particle mass of 0.257 gm
is based on a cylinder of 0.5 cm dia and height (sieved fraction
of 2-8 mm) and a measured spoil particle density of 2.63 gms/cc.
The total-salts loss rate on a per particle basis for two columns
containing spoil 21 averaged 1.41 x 10~ mg/sec (one result was
1.62 x 10 mg/sec for a 2500 ml solution ranging in 94-76 EC
delivered in 3000 sees from 222 gms spoil and the other was
1.20 x 10~ for a 200 ml solution ranging from 91-72 EC
delivered in 840 sees from 75 gms spoil.
t = time interval (equals one where Q is expressed on a per sec
basis).
h = cylinder height (0.5 cm).
-7 2
D = is 6 x 10 cm /sec and is a composite diffusion coefficient (D)
for total salts based on the published D values for the dominant
SO^ salts; FeSO^, MgSO , CaSO and MnSO,. This composite D was
taken as 5 x 10 times 0.12 (water-filled porosity). Published
116
-------
_V*,CC.MxJ-bcci>)
Figure 3.26. Assumed geometry and labelling
of individual spoil particle.
1040
880
720 -
560 -
O
.c
E
3
Ul
400 -
240 -
Leachate Volume, ml
100 150 200
1 1 1 1 1 T
250
8 12
Time, min
dQ,/dt
I I L
16
20
Figure 3.27. Example of column leaching sequence
following N» incubation.
117
-------
D's in free water are as follows: 3.9 x 10 for FeSO. at 15C
4
(Bruins, 1929); 3.5 x 10~6 for MnSC>4 at 15C (Bruins, 1929);
5.1 x 10~6 for MgS04 at 20C (Bruins, 1929); 7.1 x 10~6 for MgS04
at 25C (Chemical Rubber Company, 1969). Using ion mobilities
,1/2
(X) at 25C, infinite dilution, and the equation X/X, = (M./M-)"
where M = molecular weight, D for CaSO, was calculated to be
times that for MgSO, (Daniels and Alberty, 1962).
r = total radius of cylinder (0.25 cm).
r = radius of noncontributing or unweathered portion (unknown).
3.
C = concentration at cylinder surface (r,) when dQ /dt = 0. This C
D . D t
value appears diffusion, not dilution controlled upon extended
leaching (Figure 3.27) and equals 85 EC or 16 mg/1 total salts.
C = concentration at break in the curve obtained by plotting C
SL O
versus time of N? incubation (Figure 3.28 and 3.29). Break
defined by extrapolating both legs of Figure 3.29 to their
intersection (1000 EC or 710 mg/1 total salts). The C (starting
C for each leaching sequence representing different N? incubation
times) was determined by 3 point linear extrapolation (Figure
3.27). The concentrations of the first 3 sequential 25 ml
leachates were extrapolated back to the y axis where leaching
time equals 0 or the leachate volume approaches 0, i.e. , no
2
dilution. The coefficient of determination (r ) for this 3 point
fitting ranged from 0.95-1.00 for 49 of 51 leaching sequences
which included both atmospheric and N. incubations. For two
2 2
sequences, r =0.90 and 0.62. The r for 65% of the sequences
equaled or exceeded .990. The raw data was plotted and found to
provide very good fits in nearly all cases.
118
-------
10,000 h
E
o
E
d
0,000
6,000
2,000
0
0
500
_L
.20% 02
A100% N2
1000 1500 2000
Time, hours
_L
2500 3000
Figure 3.28. Relationship between 0- availability, incubation time and the acid
product concentration in leachate (extrapolated to zero dilution).
-------
3000
2500 -
2000 -
S
u
CO
o
jr
3
u
w
1500
1000 -
500 _
100 200
Time, hours
300
Figure 3.29. Expanded scale relationship between 0
availability, incubation time and the
acid product concentration in leachate
(extrapolated to zero dilution).
120
-------
The greatest depths or lower boundaries of acid producing zones of salt zones
within the spoil particle were calculated by rearranging equation (3.2) into:
r/r = e3'14 Dt(Ca - Cb)/Qt (3.2)
D 3.
Upon substituting,
»
/ 1.55 x 105 D sec/cm2 (3.3)
rK'r* = e
b a
The basic equation and terms used to calculate the depth of 0~ penetration in a
water saturated spoil particle in equation (3.2) with the following differences.
Q = 2.5 x 10 mg 0- consumed per particle per second at steady state.
Basic data obtained from Figure 3.30 was 0.45 mg S0,/hr
(0.50-0.05). See final paragraph, this section, for details on how
Figure 3.30 was constructed. The 0~ consumed per SO, generated was
0.583 gins 0,,/gm SO, , a conversion factor derived directly from the
-2 +2 +
assumed stoichiometry FeS2(s) + 3.5 0 + HO -»• 2SO + Fe + 2H .
/* -i
D = 2.2 x 10~ cm /sec. This is the diffusion coefficient of 0 in
-5 2
free water at 25C (1.8 x 10 cm /sec; OSU, 1971) corrected for
the water-filled porosity (.12).
Cfe = 0 mg/1.
C = 10 mg/1 at 25C (OSU, 1971).
cl
The total sulfate produced for each incubation used to construct Figure 3.30
was calculated by integrating a fitted equation describing the relationship of
SO. with time or volume (curve in Figure 3.27 is a good example) between the lower
limit of t = 2.0 minutes (first leaching volume of 25 nils) and the upper time
limit where EC = 150 umhos/cm (arbitrarily chosen). SO, was calculated from EC
for each two minute intervals on the basis of the equation in Figure 3.24. The
121
-------
Ii .<2.24>
| I3.6>» I11-")
. 20%Oz
o
1.0
O) 0.8
J
.1 0-6
0.4
'3 0.2
<
0.0
10.5
I
I
|
1
--JQ.05
1.0
10
100 1000
Time, hours
10,000
Figure 3.30. Relationship of acid production rate
for different incubation periods
under air and N9 atmospheres.
The integration was performed by summing trapezoids. The coefficient of determination
2
(r ) for the goodness of fit of the equations ranged from 0.86-0.99 with only 2 of 51
falling below 0.90 and with over one-half equal to or exceeding 0.980. All relation-
ships were significant at the 1% probability level with most significant at the 0.01%
probability level.
Experimental Results
The experimental results show air incubation to produce substantially more acid
products or salts than does oxygen-free (N-) incubation once past the first 50 hours
(Figures 3.28 and 3.29). This applies to the "starting" or undiluted concentration
of the first extract (C ) as determined by 3 point extrapolation. Before the first
50 hours or below 1000 ymhos/cm EC the two curves appear practically indistinguishable.
After substantial incubation time, the curve representing the ISL incubated system
would be expected to flatten out according to mass action principle in which the
122
-------
buildup reaction products at the spoil surface would depress and finally stop net
movement of similar salts to that surface. This expected behavior did not happen
here. Instead the EC increased slowly and linearily from the 100-3305 hour
incubation period. We have no explanation and can only speculate that a slow
rate of pyrite oxidation continued due to a nongaseous oxygen source or other
•
oxidant. Apparently, this EC increase is kinetically rather than equilibrium
controlled because the straight line fit implies that the relationship is
independent of EC (Figure 3.28). However, this curve changes from curvilinear
to linear in the range of 900-1200 EC which likely represents the equilibrium
stabilized EC if the other acid-producing kinetically controlled system was not
operating.
The total acid produced over each incubation period is expressed as a rate of
SO, mass generated per hour (Figure 3.30). SO, was chosen because it is the
product of pyrite oxidation, and can be used to measure pyrite oxidation if other
sources of SO, are not substantial. This also assumes that SO. is not lost by
4 4
precipitation within the spoil particle which is reasonable, considering the
usually low CaSO, concentrations in spoil extracts upon extended leaching. It is
apparent that there is a great deal of variability in the rate of SO, generated
up to about 200 hours incubation time (Figure 3.30). For the long-term air
incubation, SO, production from oxidation is 0.45 mm/hr for a 75 gm spoil. The
acid production under the air system is 10 times that of the O^-free system. The
steady production of SO, in the 07-free system corresponds to the slow but steady
increase in EC observed in Figures 3.28 and 3.29.
We propose the hypothesis that acid production rate is likely controlling
when the pyrite is exposed (the diffusion pathway length is short). As
weathering progresses assuming no major physical breakdown of the particle, the
diffusion pathway length increases until diffusion of either acid products out
123
-------
or 02 in is limiting. Using essentially the data of Tables 3.17, 3.18 and other
parts of this study, we calculated the depth limits of the acid products source
or to which 0- penetrated. This assumes a primarily water-saturated particle so
that solute diffusion can occur, and 0~ diffusion is controlled through the water
rather than gaseous phase.
Diffusion of Acid Products to the Particle Surface or'P., into
the Particle as the Controlling Process
From equation (3.3) we calculated r /r as described earlier using data from
0 3.
the oxygen-free (N ) incubation part of the study. The ratio of r, /r is the
<• b a
ratio of the total particle radius to the unweathered radius (Figure 3.26). Thus,
at 1.0, there is no weathered portion because the radius of the cylindrical
particle and the unweathered portion are the same. Large values of r /r signify
D 3.
that much of the particle is weathered (or depleted) of either the more readily
reactive pyrite or acid products. The unweathered and weathered volume was calcu-
lated for different diffusion coefficients (D) corrected for porosity (Table 3.17).
Obviously, the depth of weathering or contributing volume is very small for those
chemicals with small diffusion coefficients and is larger or total for those with
large diffusion coefficients. The acid products contributing volume and weathering
depth were calculated to be 0.022 cm and 17%, respectively. This was based on a
corrected diffusion coefficient of 6 x 10~ . Thus, 83% of the average spoil
particle was unweathered or unleached and probably contained substantial reactive
pyrite and acid product concentrations.
In a similar approach, the r /r , Ar and % contributing volume were calculated
D Si
based on the penetration of 0_ into a water-saturated cylindrical spoil particle.
The depth of CL penetration and contributing volume was calculated to be 0.06 cm
and 42%, respectively (Table 3.17). This was based on a corrected diffusion
coefficient of 2.2 x 10 . Based on these calculations, 58% of the average spoil
124
-------
TABLE 3.17. THE WEATHERED VOLUME (CONTRIBUTING) AND DEPTH (ir) AS
RELATED TO DIFFERENT DIFFUSION COEFFICIENTS*
r It Dt
cm /sec
2320 5 x 10"5
4.7 1 x 10~5
2.17 5 x 10"6
1.32 2.2 x 10"6
1.17 1 x 10"6
1.097 6 x 10"7
1.053 5 x 10"7
1.015 1 x 10"7
1.008 5 x 10~8
1.002 1 x 10~8
*Used equation [3|.
tDiffusion coefficient corrected
systems at 25C times 0.12).
1(0.25 cm - r ).
r
a
cm
.0001
.053
.115
.190
.214
.228
.231
.241
.248
.2496
for porosity
, Contributing
Art volume!
cm %
".25 "100
.197 95
.135 . 79
.060
-------
particle remained experimentally in an oxygen free system and, thus probably retained
substantial reactive pyrite.
Where diffusion potentially limits the acid products contribution of subsurface
spoil to groundwater, an obvious question is raised. Because the depth by diffusion
is limiting and diffusion only occurs through surface area of the particle which to
some degree can be manipulated, could the contributing volume of spoil be substan-
tially decreased by increasing the size of the average spoil particle? Using
-7 2
D = 6 x 10 cm /sec, the contributing volume related to average particle size was
determined (Table 3.18). Table 3.18 represents the acid product source zones rather
than the depth of 0 penetration. The contributing volume was greatly affected by
particle size. Excluding extremes such as particle sizes less than .05 cm and
greater than 5 cm radius, the absolute and relative (%) contributing volume change
greatly. This suggests that relatively small changes in particle size can greatly
alter the rate of acid production or acid products loss per unit volume spoil bank.
Rate of Pyrite Oxidation as the Controlling Process—
From the air incubated part of the study, the long term pyrite oxidation express-
ed as SO, production was found to be .45 mg/hr for a 75 gram spoil sample. Based on a
3.8% pyrite content, this converts too .16 mg SO, generated/gm pyrite/hr. This
compares to: 0.13 mg SO,/gram pyrite/hr reported by Braley (1954) for aerated
moistened -8 + 40 mesh sulfur ball; 0.11 mg SO /gram pyrite/hr reported by Clark (1966)
for submerged -40 + 50 mesh pulverized pyrite at 20C (calculated from Clark's equation
2/3
dSO /dt = 0.023 [D.O.] where dissolved oxygen equals 10 mg/1); 0.06 mg SO,/gram
pyrite/hr reported by OSU (1970) for submerged sulfur ball at 25C (averaged 35 yg 0
consumed/gram pyrite/hr for 8-10 mg/1 dissolved oxygen converted to SO. basis).
Although we cannot directly compare these to our results because of the different
experiment conditions, pyrite sourves and physical experimental setup, our results
126
-------
although high are generally in agreement. Moreover, Braley (1954) reported SO, pro-
duction in submerged pyrite to be well below that obtained from aerated moistened
pyrite which could substantially increase the rate of SO, production reported by
Clark (1966) and OSU (1970) if subjected to our rather than their experimental
conditions. On the basis of these results, acid production from this fresh spoil
»
subjected to air incubation appears to be controlled by the reaction rate rather
than 0- diffusion into the spoil particle.
Interpretation of Results
The spoil material used in this experiment contains substantial concentrations
of both pyrite (3.8%) and acid products (3.1%). If deep placed, especially where
submerged or frequently flushed, the initial acid product losses to groundwater can
be very large, even in the absence of 0_. In fact, the loss could be hydrologically
controlled by percolation or groundwater flow rates rather than chemically controlled.
Subsequently, as leaching progresses, diffusion quickly becomes potentially limiting
because of the very low diffusion coefficient of these acid products.
Pyrite oxidation in the spoil particle appears potentially controlled by either
the pyrite oxidation or 0_ diffusion rate. For new, unweathered spoil, the
experimentally determined pyrite oxidation rate compared reasonably well with
published values and the 02 diffusion rates never appeared to be controlling. Thus,
where 0, is in excess supply and the particle remains water saturated, the initial,
and the subsequent rate of acid production from pyrite will likely be controlled by
the pyrite oxidation rate. As weathering removes the shallow pyrites, the 0-
diffusion or acid product diffusion rate should become controlling.
In cases where diffusion is limiting, the basic particle size is important in
controlling che rate of acid product and acid loss to groundwaters since this
reduces the contributing volumes relative to the total spoil mass.
127
-------
Chemical Parameter Interrelationships
The basic thrust of this unit is to investigate interrelationships between
chemical parameters that are useful for simplifying data collection. Often for a
selected site or within limits, chemical parameters are intercorrelated so that
one can be used as a surrogate measure for several.
Analyzing caisson percolate, total acidity at pH 8.2 was found to be linearly
correlated with total acidity at pH 7.3, SO,, total soluble Fe, Mg, Mn and Al but
not with Ca, ferrous Fe or pH (Table 3.19). The relationship between total acidity
determined at pH 7.3 and 8.2 is very precise with a slope near 1.0 and intercept
near zero (Figure 3.31), indicating that both procedures measure the same thing in
this particular system. The relationship with SO, is not' exactly linear, but for
our purposes can be considered as such (Figure 3.32). However, there are several
limitations. At concentrations less than 10 meq/1, the basic relationships either
change in slope and/or become more variable for SO,, Mg, Mn, Al and total soluble
Fe (Figures 3.32, 3.33 and 3.34). For Ca, the decline in concentration (Figure 3.33)
with increasing total acidity, which probably results from CaSO, precipitation,
precludes the use of a linear relationship. Comparison of sample pH with the .log of
total acidity assumes that a single disassociation constant controls the hydrogen ion
concentration in solution which is not likely true over a substantial pH range.
Within the aforementioned limits, the total acidity can be selected as a measure of
other chemical parameters. Thus, total acidity data subsequently presented can also
be interpreted in terms of SO., Mg, Al and total soluble Fe concentrations.
Data in Figure 3.35 shows a rough correlation between trace element concentra-
tion and pH, supported by the studies of Massey (1972) who found the log of Zn, Cr
and Ni concentration to be negatively correlated with pH in aqueous spoil extracts.
To establish a simple method for estimating the potential of trace metal contamina-
tion in chis spoil system, selected trace metal concentrations from spoil layer
128
-------
Total Acldliy (pH ).3>. meq/l
0
n>
P>
M
c
f-f
fD
Ul
0
i-h
rt
O
rt
PI
pi
0
H-
d-
M*
rt
" >
M 2 £
P> £
rt 5
H-
O "
0 ? -
w „ S
H- "
T3
1 =
n> i> °
rt !;
fD
(D
0 •
o
rt
O
S
y. i« i » « i i
.*•••
9 _
*
,
•. ~"
— _
*
i _
H,
•
i i i.i i i .
-------
300
270 -
240
210
180
150
- 120
o
en
90
60
30
0*
30
60 9O 120 150
Total Acidity (pH 8.2), meq/l
180 210
Figure 3.32. Relationship between total acidity and
sulfate concentration.
130
-------
Ca
A
Mg
r
c
U 30
o
c
o
u
Mn .
. • V
•*»'. Ff' ' \ 1
. ••*•
1 1 1
fe
.1
..v
60 90 120 ISO ISO
Total Acidity (pM a.J), meq/i
Figure 3.33. Relationship of calcium, magnesium,
manganese and aluminum with total
acidity.
(T
4)
0 30 -
0 30 30 90 120 150 ISO
Total Acidity (pH 3 j). meq/l
Figure 3.34. Relationship between
total soluble iron
and total acidity.
131
-------
u.o
0.8
_
"5> « ,
e 0-4
c
rsi
OT 0.2
0
0.0
-0.2
1.
-0.1
-0.3
~ -0.5
0)
5 -0.7
3
<-> -0.3
01
0 1.1
-i.a
-1.5
I • 1 1 1 1 i 1
-
A
-A ~
* * * *
A A A
_ A A-
A A
— —
A
1 1 1 1 1 1 Ik
8 2.9 3.0 3.2 3.4 3.8 3.9 4.0
1 1 1 1 1 1 1
A
— ± —
A *
_ A -
A
~ * A ~
* * A
A A -
_ —
A
— —
1 1 1 1 A Al
'2.8 2.9 3.0 3.2 3.4 3.8 3.9 4.0
2.2
— 20
o>
E
1 "
09
0
~" 1.6
1 " 1 1 1 1 1 1
_ A —
- -
* A A
* -
_ —
A A A\
A
_ A _
At ^
O 1 1 1 1 1*1 1
2.8 2.9 3.0 3.2 3.4 3.8 3.9 4.0
pH
Figure 3.35. Relationship of log Zn, Cu and
Mn concentrations with pH.
extracts (Table 3.13) were compared with the corresponding pH and SO. concentration.
SO, concentration was chosen for several reasons. First, it correlated well with
total acidity over the complete concentration range and the SO, loss due to CaSO,
precipitation was not substantial. Second, specific conductance which could be
used is not a valid measure of total salts concentration over a range where the
chemical composition and ionic species change substantially. Third, pyritic
132
-------
minerals which may also serve as a source of trace metals are the predominant source
of SO. in this system. Data from spoil layer 21 was omitted because it was a single
point of extreme value.
In all comparisons, the point scatter was considerable; not at all similar to
the high precision achieved by Massey and Barnhisel (1972). The plots of Zn, Cu
and Mn vs SO, concentration (Figure 3.36) showed generally less scatter than did
the corresponding log Zn, Cu and Mn vs pH plots (Figure 3.35). The reader may note
that the apparently similar scatter is numerically much larger on the log-log scale.
The sulfate concentrations can provide reasonable estimates of the Zn, Cu and Mn
concentrations provided they are not extrapolated much beyond the concentration
range upon which they were tested.
133
-------
90
4.0
_
0>
£ i-o
e
N
2.0
10
g
0
o.a
o
6 0.4
3"
(J
-
_
-
~
-
°
140
120
£ 100
£ so
s
40
40
20
_
—
-
_
-
-
0
1 I I 1
A -!
A
A
A
A
A
AA* A
*VA
AA
A*
—
A
! 1 1 1
SOO 1000 1SOO 200O 23
I I I I j
A
A * A"
A
***
4k I A I I I
iOO 1000 1300 2000 23
1 \ \ \ '
A
-
-
. A
* A
. A
^Ai
A ^ A
1 ! 1 1
500 1000 1SOO 2000 25
SO4> mg/l
Figure 3.36. Relationship of Zn, Cu and
Mn concentrations with
sulfate concentrations
134
-------
REFERENCES
1. Allison, L. E. Organic Carbon. In: Methods of Soil Analysis Part 2, C. A.
Black, ed. American Society of Agronomy, Madison, Wisconsin. 1965.
2. American Public Health Association, Inc. Standard Methods for the Examination
of Water and Wastewater. Twelfth Edition, 1965.
3. ASTM. Standard Methods of Coal Analysis. Am. Soc. for Testing Materials,
Philadelphia, Pennsylvania, 1971.
4. Borden, F. Yates, and Larry H. McCormick. An Indirect Method for the Measurement
of Sulfate by Barium Absorption Spectrophotometry. Soil Sci. Soc. Am. Proc.,
34(4): 705-706, 1970.
5. Braley, S. A. Summary Report of Commonwealth of Pennsylvania, Department of
Health, Industrial Fellowships No. 1-7, August 20, 1946-September 30, 1953.
Mellon Institute of Industrial Research, Pittsburgh, Pennsylvania, 1954.
279 pp.
6. Bruins, H. R. Coefficients of Diffusion in Liquids. In: International
Critical Tables of Numerical Data, Physics, Chemistry and Technology,
E. W. Washburn, ed., 1929. 5:63-77.
7. Caruccio, F. T. Characterization of Strip-mine Drainage by Pyrite Grain Size
and Chemical Quality of Existing Groundwater. In: Ecology and Reclamation
of Devastated Land, Volume I, R. J. Hutnik and C. Davis, eds. Gordon and
Breach Publishers, New York, 1973. pp. 193-226.
8. Chemical Rubber Company. Handbook of Chemistry and Physics. 50th Edition,
R. C. Weast, ed. The Chemical Rubber Company, Cleveland, Ohio, 1969.
9. Ciolkosz, E. J., R. L. Cunningham, G. W. Petersen, and R. C. Cronce. Character-
istics Interpretations and Uses of Pennsylvania Minesoils. Progress Report 381.
The Pennsylvania State University, Agricultural Experiment Station, University
Park, Pennsylvania, 1983. 88 pp.
135
-------
10. Clark, C. S. Oxidation of Coal Mine Pyrite. J. Sanit. Eng. Div., Am. Soc.
Civ. Eng., 92:127-145, 1966.
11. Crank, J. The Mathematics of Diffusion. Oxford University Press, Amen House,
London E.G. 4, 1964. 347 pp.
12. Daniels, F., and R. A. Alberty. Physical Chemistry. John Wiley and Sons, Inc.,
New York, 1962. 744 pp.
13. Dixon, R. M., and A. E. Peterson. Water Infiltration Control: A Channel System
Concept. Soil Sci. Soc. Am. Proc., 35(6) :968-973, 1971.
14. ElBoushi, I. M. Geologic Interpretations of Recharge through Coarse Gravel and
Broken Rock. Unpublished Ph.D. Dissertation, Stanford University, Stanford,
California, 1966.
15. Emmel, Robert H., John J. Sotera, and Ronald L. Stux. Atomic Absorption
Methods Manual, Volume 1. Standard Conditions for Flame Operation.
Instrumentation Laboratory, Inc. (publisher), Wilmington, Massachusetts,
1977. 80 pp.
16. Emmel, Robert H. , Martha Fogg Bancroft, Stanley B. Smith, Jr., John J. Sotera,
and Timothy L. Corum. Atomic Absorption Methods Manual, Volume 2. ' Flameless
Operations. Instrumentation Laboratory, Inc. (publisher), Wilmington,
Massachusetts, 1976.
17. Gill, M. A. Analysis of One-dimensional Non-Darcy Vertical Infiltration.
J. Hydrol., 31:1-11, 1976.
18. Green, R. E., and J. C. Corey. Calculation of Hydraulic Conductivity: A
Further Evaluation of Some Predictive Methods. Soil Sci. Soc. Am. Proc.,
35(1):3-8, 1971.
19. Hill, D. E., and J.-Y. Parlange. Wetting Front Instability in Layered Soils.
Soil Sci. Soc. Am. Proc., 36(5):697-702, 1972.
136
-------
20. Jordan, C. F. A Simple Tension Free Lysimeter. Soil Sci. , 105(2)-.81-86, 1966.
21. Keys, W. S., and L. M. MacCary. Application of Borehole Geophysics to
Water-Resources Investigations. Techniques of Water-Resource Investi-
gations of the U.S. Geological Survey Book 2, Chapter El. U.S. Government
Printing Office, Washington, DC 20402, 1971.
22. Leps, T. M. Flow through Rock Fill. In: Embankment-Dam Engineering,
Casagnande Volume, R. C. Hirschfeld and S. L. Paulas, eds. John
Wiley and Sons, Inc., New York, 1973.
23. Marshall, T. J. A Relation between Permeability and Size Distribution of
Pores. J. Soil Sci., (9):l-8, 1958.
24. Massey, H. F. pH and Soluble Copper, Nickel and Zinc in Eastern Kentucky Coal
Mine Spoil Materials. Soil Sci., 114(3):217-221, 1972.
25. Massey, H. F., and R. I. Barnhisel. Copper, Nickel and Zinc Released from
Acid Coal Mine Spoil Materials of Eastern Kentucky. Soil Sci., 113(3):
207-212, 1972.
26. Mein, R. Modeling of the Infiltration Component of the Watershed. Unpublished
Ph.D. Thesis, University of Minnesota, Minneapolis, Minnesota, 1971.
27. Millington, R. J., and J. P. Quirk. Transport in Porous Media. Int. Congr.
Soil Sci., Trans. 7th, Madison, Wisconsin, 1960. 1.3:97-106.
28. Ohio State University Research Foundation. Acid Mine Drainage Formation and
Abatement. Water Pollut. Control Res. Ser. DAST-42-14010 FPR-04/71, U.S.
Environmental Protection Agency, Washington, DC, 1971. 83 pp.
29. Ohio State University Research Foundation. Sulfide to Sulfate Reaction
Mechanism. Water Pollut. Control Res. Ser. 14010 FPS-02/70, Federal
Water Quality Administration, U.S. Department of Interior, Washington,
DC, 1970.
137
-------
30. Pedersen, T. A., A. S. Rogowski, and R. Pennock, Jr. Physical and Hydraulic
Characteristics of Some Minesoils. Soil Sci. Soc. Am. J., 44:321-328, 1980.
31. Philip, J. R. Stability Analysis of Infiltration. Soil Sci. Soc. Am. Proc.,
39(6):1042-1049, 1975.
32. Quisenberry, V. L., and R. E. Phillips. Percolation of Surface Applied Water
in the Field. Soil Sci. Soc. Am. J., 40(4):484-489, 1976.
33. Raats, P. A. C. Unstable Wetting Fronts in Uniform and Nonuniform Soils.
Soil Sci. Soc. Am. Proc., 37(5):681-685, 1973.
34. Raney, W. A. Field Measurement of Oxygen Diffusion through Soil. Soil Sci.
Soc. Am. Proc., (14):61-65, 1949.
35. Ritchie, J. T., D. E. Kissel, and Earl Burnett. Water Movement in Undisturbed
Swelling Dry Soil. Soil Sci. Soc. Am. Proc., 36(6):874-897, 1972.
36. Rogowski, A. S. Hydrologic Parameter Distribution on a Mine Spoil. Symposium
on Watershed Management 1980, Boise Idaho, July 21-23, American Society of
Civil Engineers, New York, 1980. pp. 764-780.
37. Rogowski, A. S. Water Regime in Strip Mine Spoil. In: Surface Mining and
Fish/Wildlife Needs in the Eastern United States, D. E. Samuel, J. R.
Stauffer, and W. T. Mason, Jr., eds. Proceedings of a Symposium, FWS/OBS-
78/81, sponsored by West Virginia University and the Fish and Wildlife
Service, USDI, 1978. pp. 137-145.
38. Rogowski, A. S. Estimation of the Soil Moisture Characteristic and
Hydraulic Conductivity: Comparison of Models. Soil Sci., 114(6):
423-429, 1972.
39. Soil Survey Staff. Soil Taxonomy. Agriculture Handbook 436, U.S. Department
of Agriculture, Soil Conservation Service. U.S. Government Printing Office,
Washington, DC, 1974. p. 469.
138
-------
40. Smith, R. E., and D. A. Woolhiser. Mathematical Simulation of Infiltrating
Watersheds. Hydrology Paper 47, Colorado State University, Fort Collins,
Colorado, 1971.
41. U.S. Environmental Protection Agency. National Secondary Drinking Water
Regulations. Federal Register, 42(62):17143-17146, 1977.
42. U.S. Environmental Protection Agency. Water Programs, National Interim
Primary Drinking Water Regulations, Part IV. Federal Register,
40(248) :59570, 1975.
43. U.S. Environmental Protection Agency. Methods for Chemical Analysis of
Water and Wastes. EPA-625/6-74-003a, 1974.
44. Vimmerstedt, J. P., and P. H. Struthers. Influence of Time and Precipitation
on Chemical Compositioi\ of Spoil Drainage. In: Proceedings, Second
Symposium on Coal Mine Drainage Research, Pittsburgh, Pennsylvania. Coal
Industry Advisory Commission to Ohio River Valley Sanitation Commission,
1968. pp. 152-163.
139
-------
-------
SECTION 4
FIELD STUDIES
PHYSICAL AND HYDROLOGIC SETTING
In Section 2 we have concentrated on the physical and chemical description of
experimental site. In this section we will emphasize the overall response of the
site to climatic influences as observed during the study period. To evaluate site
response properly we need first to consider the actual physical setting and
properties of a site as a whole. Figure 4.1 shows the cross section and aerial
view of the site, delineating areas that have been deep mined, stripmined and left
undisturbed, as well as indicating the position of wells, seep and stream relative
to each other. Chronologically, Kittanning B coal was mined first from drift mines
p-Strippud.
B-Oeop Mined . f
^
80 120
Meters
160
Figure 4.1.
Schematic diagram of the
experimental site.
140
-------
extending back into the hill (Figure 4.la). Subsequently, the northern portion of the
area was stripmined. Little coal was recovered because of previous deep mining
activity. This area was reclaimed in 1967 before new Pennsylvania legislation
requiring topsoiling went into effect and now supports a stand of beech, pine and
aspen. The area is bordered on its northernmost edge by a narrow strip of undisturbed
land, and on the south by a buried high-wall. The southern portion of the site was
stripmined for Kittanning C coal only. The deep mined layer was left undisturbed and
the whole area was reclaimed under new regulations, topsoiled and seeded to grass. As
part of fanning operations the proprietor spread manure over the eastern half of the
site which now supports a heavy grass and legume stand (Figure 4.2).
We have measured infiltration on the site using small and large ring infiltrometers
(Rogowski, 1980). Figure 4.3 shows relative distribution of infiltration over the area
with a peak in the east central portion. In general we observed that infiltration and
hydraulic conductivity were quite variable both in time and in space. We have also
found (Pedersen et al., 1980) that when sufficient moisture was present pan evaporation
adequately represented reclaimed site evapotranspiration. At other times the evapo-
transpiration could be simulated using Ritchie's (1972) ET model. Figure 4.4a and b
shows typical soil water variation over the area and depth, while Figure 4.4c indicates
an overall shape of the water table sloping towards northeast. Figures 4.5a and b
show a characteristic distribution pattern of solar radiation and precipitation over
the site respectively, while Figure 4.6 illustrates site response to a spring storm.
During that storm we observed essentially no runoff until a high intensity burst
of rain 18-23 hours after rain first begun. Vegetated area responded to this burst
faster than the nonvegetated one, but runoff rate on the nonvegetated part was 100 times
larger. Simultaneously with the occurrence of runoff, water table, 72 m below the
ground surface, began to rise and continued rising for more than five days following
the occurrence of this 29 mm storm.
141
-------
i
i!
!*
l!
BSD
EDO
ISO
100'
50
SO 100
ISO
Figure 4.2.
Vegetative cover
density (T/ha) at
the experimental
site.
142
-------
eso
eoo
1SO -
1 oo f
iso eoo
Figure 4.3. Relative distribution of infil-
tration at 60 minutes for a 65
mm summer storm (200 = 43 mm,
150 = 38 mm, 100 = 31 mm,
50 = 23 mm).
143
-------
soTOOISOeoo
so TooISOeoo
Figure 4.4. Wacer content at the surface (a) and at 0.9 m (b) and water
table evaluations during a monitoring period in the summer,
site scales are in meters.
144
-------
Runoff, mm/hour
H-
OP
n>
*~
CO
H-
n
fB
(t)
cn
T3
O
0
CO
fT
O
cn
T3
H-
TO
to
rr
O
a
o
M
a
*j
0
u
*.
7
2 S S b» Rain(al1 Intensity, mm/hour
0 -* M U *•» >4«W*WW-*0
1 1 lo.
t>d
n v»g*tit*d nunoir Ar«i^
Ev«nt of 4/14/80
_ 29mm. 21 hours
Ar«. - 0.2 ha
1 1 1 1 1 1 1 'l I i
1 1 1 1 i i i
j
7-
o
c
a I i
3t
m
v : ^=^ ^\
Non A. PJ
H- l-(
(D p) CU
^ rt n
TD H- rr
ft> O (D
fl 0 f«
H- H-
S ^ co
(D PJ rt
0 -— - H-
rt n
P) P)
M 0 D-
CL H-
CO CO
H- O rt
rt i-h t-<
O H-
• 13 f
^ C
(D rt
O H-
H- O
T3 3
H-
rt T3
P) ft>
rt rt
H- rt
O (D
o
Ml
(U CO
rt O
cr-
Watertabla Elevation Chang
mm
Accumulated Rain, mm
-------
In our modeling effort we observed (Rogowski and Weinrich, 1981) that on top-
soiled profiles (such as spoil C in Figure 4.1) we could adequately model infiltra-
tation and redistribution of applied water with numerical techniques using moisture
characteristics corrected for coarse fragment content and hydraulic conductivities
derived either from seepage flux measurements or experimentally determined values.
On nontopsoiled profiles (such as spoil B in Figure 4.1) numerical techniques
appeared inadequate. A better fit to experimental data was obtained by subtracting
from the flux the amount of water retained by fine particles and the amount
retained on the surface of coarse fragments and then considering the flux to be the
function of gravity alone. However, often even this approach was not wholly
satisfactory. At times when spoil was reasonably dry (water content ^ 0.1 by vol)
water apparently moved only through the larger channels in the coarse spoil
materials without appreciably wetting the rest of the profile.
CHEMICAL SETTING
Occurrence of acid mine drainage from deep mines and strip mines, and resulting
degradation of water quality in streams of a mined area, is caused primarily by the
oxidation of iron sulfide mineral, pyrite. Two mechanisms for pyrite oxidation
have been demonstrated. The first mechanism involves oxygen in the spoil atmosphere
reacting directly with pyrite and water,
+ 3.5 0 + H0 -»• Fe2+ + 2SO~ + 2H+ (4.1)
the second mechanism uses ferric iron as the electron acceptor,
14Fe3+ + FeS2 + 8H20 -»• 15Fe2+ + 2SO^~ + 16H+ (4.2)
For this reaction to produce significant levels of acid, ferric iron must be present.
Although at the normally low pH's of strip mine materials (pH<4), ferrous (2+)
-------
rather then ferric (3+) iron is the dominant ion, certain bacteria use the ferrous-
ferric oxidation reaction as an energy source. Because ferric iron is needed as
input to equation 4.2, the rate limiting step for equation 4.2 may be,
7 bacteria -
14Fe H- 14H + 3.5 C>2 > 14FeJ + 7H20 (4.3)
which is often considered to be bacterially catalyzed within strip mine spoils.
Equations 4.2 and 4.3 can reduce algebraically to 6.1; thus, regardless of
whether the reaction is bacterially catalyzed or not the final products of pyrite
oxidation remain the same. Acid mine water emptying into a stream will cause a
decrease in stream biota, but does not directly result in any visual degradation.
It is the iron in the acid mine drainage that creates significant visual impact by
precipitating in the streams to color them red or to form yellow-orange stains
called "yellow-boy" on stream bottoms and sides.
We measured oxygen profiles at the experimental reclaimed stripmined site.
Figures 4.7a and b illustrate a typical 0- profile in August and October
respectively approximately 700- and 800 days from the start of experiment (Jaynes
et al. , 1982). During this time of the year water levels in wells are dropping
and increased penetration of oxygen into the spoil is generally observed. With
the approach of winter the surface spoil layer cools and plant respiration slows
or ceases. Thus, 0_ uptake decreases in the surface and 0^ can penetrate deeper
into the spoil. The increased 0_ concentrations in the deeper parts of the spoil
although slight, can apparently induce an increase in pyrite oxidation.
The results presented in Figures 4.1-4.6 illustrate a typical study site
response to the climatic influences. They form the background against which we
will now discuss the overall impact of mining and reclamation on quality of the
groundwater at the experimental site in space and in time.
147
-------
Figure 4.7. Typical oxygen profiles in August (a), and
October (b) at the experimental site.
WATER QUALITY
Site Response in Time
Figure 4.8a, b, c and d shows the overall response of mined and reclaimed experi-
mental site during the 1000-day study period. The bars in Figure 4.8a give the
monthly rainfall amounts and the points indicate weekly water levels. The bars are
positioned so as to reflect the maximum precipitation during a given month. For
example, if a greater than 25 mm rain occurred on the 20th, the bar representing that
month's precipitation would be positioned 2/3 of a way through the interval. If all
148
-------
JULIAN DAYS
244 71 271 t!4 314 !4«i 34=1 244
JUL IAN DAYS
271 114 314
S14 I 1 1 1 1 1
512-1 WELLS 51. 52 AND 53
3 31QJ_ j^ ^ t ^ ^ ^
2 30*4%/" >P| " ^-'| '^=~> "
j 30d_j ' 4QO {
- i 570
- 304_| 600
2 5G2J
?nn j
i 1 1 i 1 !
514, i 1 1 1
1 14 J 1 1 1 1 1 1
_ 312-j WELLS 54. 56. 53 AND 51 [
_ 30S Jv- -,' '^ _ V-'^ ''X-^ .->,„'" "
- 5D4-i [
_ 3Q2_| "ifcjW*<***ll'lSiitf*'**>ll-i«M_ M,a>f*< ^~
?nn ,
I i I I I r
.. ^14 1 1 1 1 !
WELLS 101 AND 103
L 312J
WELLS 104. 105 AND lOd
U
SC"1^
DAYS =3QM START
_ :D4_
—
— , i
1 ' , i , . • !
i ' ! ' ; ! I i j !
i i i 1 1 i 1 1 i i i 1 i i
|
1 ! 1
|
1 1
j|
-150
I
• UIQQ
[
j -50 C
l : j
1 I II i J „
200
3AYS FSOM START
Figure 4-8a. Overall response of the experimental sice during
the study period: water levels.
149
-------
JULIAN DAYS
2*4 71 271.. 114 314
mntv i - 1 - 1 - 1
34*
JULIAN DAYS
271 IH 314 141 541
1QDOJ
.••.••--.•"•"."..' ..'•;.•'• "•'£*'. ••'
I WE.i-S 54. 56. 55 ANO 51
*.--f. ------- 1
"- ID-
51. 52 AND 53
_j , , 1 1 , —
m i
3 -. • ... . -,--.
_j . .- • ' _>.,-.- ;;_/;• . . .
y .-'''*•
3
J
i
3
-i WELLS 104. 105 ANO 106
1000,
_
_ 100_
-
_
1D_
~
_
_
III!)
1 1 1 i
SEEPC. 3 ANO STREAMS)
•"V.". .
1 .. ' ' . • • . '
( t «
t '—
'
.
1 1 1 1 1 1 f-
0 200 400 600 500 1000 1200 0 200 400 600 500 1000 1200
DAYS FROM START DAYS FROM START
Figure 4.8b. Overall response of Che experimental site during
the study period: Iron (Fe) concentration.
150
-------
JULIAH OAYS JULIAN DAYS
244 7S 271- 114 314 141 54°, 244 71 271 114 314 HI 541
DC_ i i ; i i •. COQJ : , i i i _
I
WELLS ?4. ?6. ?5 AND
1DDIU
IDCL"
WELLS F1. ?2 AND ?3
I-
-| WELLS 1D4. ID? AND
a
_*-
t i_
SEEPC. ) AND STREAMS
200
r i I i I r i i r i
600 500 1000 1200 0 200 400 600 300 1DOO 1200
OAYS FROM START
OAYS FROM START
Figure 4.8c. Overall response of the experimental site during
the study period: Aluminum (Al) concentration.
151
-------
JUL I Af. ".'•': JUL iAN DAYS
2+4 71 271. II* 314 141 341 244 71 271 114 314 141 341
| i ; | , , IQDCj i i I I ; ,_
iac_.
WELLS F4. 96. 9Z AND F1
ID.
lOOEi
ID
•y
WELLS f '.. =2 AND ?3
• ' ' . ' •*•••«•.•. .'«^«.V.~""""" .
WEuLS 1D4. 105 AND 106
i i i i
SEEP<. J AND STREAM<03
-\ -,'-r-v -nj.^ -,-,..-. * i
J _^;-/.., ; ,;TiV ^jf^'-^j-i ,
_ IDJ .•'*•• -* V
0 200 400 iSDO 500 1000 1200 0 200 400 iSOD 500 100Q 1200
DAYS FROM START DAYS FSOM STAST
Figure 4.8d. Overall response of the experimental site during
study period: Sulfate (SO.) concentration.
152
-------
rains were less than 25 mm the bar would be positioned in the middle of an interval.
The points that indicate weekly water levels are grouped into four groups of wells.
Wells 51, 52, 53 are located in a reclaimed area, stripmined for lower Kittanning
B-coal. Wells 104, 105 and 106 are located in the area deep mined for B-coal. For
wells 101 and 103, located on the fringes of deep mined area, the information is
scanty, highly variable, and not included in subsequent parts of Figure 4.8. Even
though water level response in wells 101 and 103 appears reasonably smooth the
picture is misleading. Water level would drop during the dry period below the well
bottom and, would not be recorded. Thus, the results from these wells represent the
response during high water periods only. Wells 54 and 56 (dots), and 58 (stars *),
in the upper right hand corner of Figure 4.8a, are on the fringes of stripmined
area and well 59 (•) samples water table overlying undisturbed at that point,
Clarion A-seam.
In subsequent parts (b, c and d) of Figure 4.8 we show on the same scale dis-
tribution of Fe, Al, and SO, in the well water and in the seep and stream draining
the B-coal area. In all drawings the horizontal dashed lines represent the mean
distribution for a given group of wells. If more than one distribution is present
two dashed lines appear. In what follows we compare the response observed in
stripmined wells 51, 52, 53, 54 and 56 with that of deep mined wells 104, 105 and
106, we contrast their behavior with that of an undisturbed area (well 59), and
examine probable effects of acid mine drainage on seep and stream.
Water levels in all wells, except well 59 (•) are approximately the same, and
times of occurrence of minima and maxima generally coincide. In Table 4.1 days
from start of experiment, appropriate dates, and times of major storms and times
of occurrence of these well minima and maxima are listed. The data show that the
maxima in wells both on strip and on deep mined spoil follow occurrence of major
storms, while minima appear to result from 60-day or longer dry spells (rains <25
153
-------
TABLE 4.1. TIME OF OCCURRENCE OF MAJOR STORMS AND ASSOCIATED
WELL ELEVATION MINIMA AND MAXIMA
Day
5
100
ZOO
300
400
500
600
700
300
900
1000
Month
Sept
Dec
Mat
June
Occ
Jan
April
Aug
Nov
Feb
May
Calendar
090678
121073
032079
062879
100679
011480
042380
080180
110980
021781
052881
Julian
249
344
79
179
279
14
114
214
314
48
148
Dace
Rain ^ 25 mm
286,343
24,57,69
144
204,249,271,278
330,359
105
133,199
299
61,81
Well Min
(Julian)
323
205
65
293 (large)
33
90 (small)
Well Max
66
284
112
335 (snail)
60
147
mm). The response appears to be more dependent on the presence or absence of rain than
season of the year. In that sense the spoil-deep mine system appears to respond faster
to weather extremes than does the natural system. The fast response of mined area is
not altogether surprising since the spoil can hold only about one-fourth as much water
between 10 and 1500 kPa as the undisturbed materials it replaces. In contrast, the
response of the undisturbed well 59 (solid •) is considerably damped, while that of
well 58 (stars *) which is on the-very fringe of stripmined area and quite shallow,
occurs immediately after the major storms possibly in a form of interflow from mined
area just above it.
In Figure 4.8b we observe behavior of Fe at a point. Generally the highest con-
centration of Fe both in wells and in the seep and stream occurs during the time when
water level is at the minimum. The concentration of Fe appears highest in the deep
mined area (104, 105, 106), followed by that of stripmined area (51, 52, 53). Con-
centration of iron (Fe) in the fringe wells (54, 56, and 58) is less than or equal to
that in seep and stream, while the concentration in the undisturbed well 59 (•) is
the lowest. The most significant aspect of Fe behavior is the large variability
observed in the response. Concentration in the stream tends to be lower than in the
154
-------
seep, most likely because of dilution by less contaminated waters derived from drainage
above the seep. The lowest values of Fe concentration in the stream at times (i.e., °^
day 200) related to dilution by precipitation, at other times (i.e. , ^ day 500) they
occur during low water level and reflect low effluent concentrations.
Figure 4.8c shows the Al response of the site. Al levels in the undisturbed
well 59 (•) are quite low on the order of 1 or 2 mg/l. The highest concentrations
occur in parts of the deep mined area (well 105—upper dashed line in the lower left
hand corner of Figure 4.8c) and in the stripmined area (51, 52, 53), Little or no
lowering in Al concentration on the fringes of mined area (54, 56, 58) is evident
except perhaps at well 58 where thick soil overburden is present. The seep and
stream effluent appear to have Al concentrations similar to those found in well 106
(lower dashed line in the left hand corner of Figure 4.8c). The overall response,
compared to Fe shows less variability, where less well defined maxima and minima
follow the water level maxima and minima.
Finally in Figure 4.8d gives sulfate (SO,) concentrations for the same three groups
of wells, the seep and the stream. The sulfate concentrations show the least scatter
and smallest variability compared to Fe and Al. The highest average concentrations are
observed in the deep mined area (wells 104, 105, 106) concentrations are identical in
the stripmined area and on its fringes (wells 51, 52, 53, 54 and 56) and concentrations
decrease markedly in well 58 (*) and particularly in well 59 (•) on undisturbed
material. Interestingly, the seep and stream effluent concentrations of SO,, while
lower than the deep- and stripmined water table concentrations are substantially
higher than concentrations in the undisturbed area (59). Sulfate concentrations appear
more nearly constant with little or no tendency to follow water level minima and
maxima.
In summary therefore, Figure 4.8 shows that the water levels in the wells at the
mined and reclaimed site responded primarily to rainfall events ^ 25 mm. Iron (Fe)
155
-------
concentrations varied most and were higher at the deep mined than at the stripmined
site; aluminum (Al) concentrations varied less and were highest in parts of deep
mined and stripmined areas; sulfate (SO.) concentrations varied least and were
highest in deep mined area.
Controlling Processes
The variation of chemical concentration of well water with time can result from
chemical or hydrologic changes in local groundwater or from throughflow or ground-
water of different composition originating elsewhere. For local groundwater
hydrologic dilution or concentration would be primarily through the addition of
percolate. Where the quality of added water is better than that of the groundwater,
the chemical concentration decreases due to dilution, but the load stays roughly
constant or may even increase. Where chemicals are being added or removed from
solution, the concentrations will be correlated with load but not with total volume.
Because the water levels and percolate contributions are variable, it is likely that
the process determining the water quality in the well may switch back and forth,
depending on climate, season and even event. The gain or loss of chemicals to
groundwater from diffusion, resolution or chemical precipitation proportionately
increase or decrease both concentration and load independent of changes in groundwater
volume. Separating these effects from increases due to the addition of highly con-
taminated percolate depends mostly on the rate of the process. In strip mined systems,
rates of increase due to percolate inflow are usually rapid and short term, being
basically a response to a hydrologic event. The chemical processes such as those
controlled by diffusion or pyrite oxidation kinetics are much slower and steadier.
Diffusion of salts contained in spoil materials can cause a steady chemical contribu-
tion to a groundwater, thus increasing both concentration and load, simultaneously.
Similarly, in cases where solubility, redox or other chemical additions or
156
-------
subtraction processes largely independent of volume or flow rate are operating, the
load and concentration plots versus time should provide similar relationships. One
difficulty with this interpretation is that apparent accretion or drainage of
groundwater at a point can seemingly increase or decrease the load over time before
the system attains an equilibrium. Table 4.2 summarizes the controlling processes
by looking at the concentrations, water volume, or percolate flux to the well and
load relationships, and assuming that water through flow is not controlling unless
stated otherwise.
TABLE 4.2. CONTROLLING PROCESSES AFFECTING GROUNDWATER QUALITY ON STRIPMINES
System
Concentration (C)
4
Water Level (H)
or Flux (Q)
A Load Correlates with
oC AQ
(Al) Increasing
(82) Decreasing
(A2) Increasing
(C3) Stable
(A3) Increasing
(Bl) decreasing
(B3) Decreasing
(C2) Stable
Stable Yes
Increasing No
Increasing Yes
Decreasing No
Decreasing \f
Stable
Decreasing
Increasing
Yes
No Acid produced without water
addition, e.g., pyrite
oxidation or salt diffusion.
No Dilution effect, direct
inflow of good quality
percolate.
Yes Substantial percolate inflow
with acid products to the
groundwater.
Yes Groundwater draining out of
system.
U Acid products added from
spoil matrix while waters
draining away.
No Acid product loss, e.g.,
chemical precipitation.
Groundwater through flow.
Yes Groundwater through flow
and/or variations of Al
or A2.
— If 6 Load correlates with AC, then (Al) is the type of system present, excluding drainage loss. If it
correlates with &Q, drainage loss must be dominant.
In general, we can measure concentration experimentally by taking a sample of
the well water at a given time and infer the load by assuming similar concentration
throughout the water column of unit area and of height equal to that of the water
table at a well point. To do this properly we must know accurately the lower well
boundary at each point and be able to estimate accurately the absolute height of the
157
-------
water table above its lower boundary. Where water table is sloping, or where con-
centrations vary with depth, considerable errors can be made by assuming the wrong
depth. Fortunately, coal seams in western Pennsylvania are approximately level over
short distances such as our study site, and spoil materials are sufficiently porous
so that we can apply uniform concentration and constant depth lower boundary
assumptions. Under these conditions Load = Concentration x Depth, where depth is
expressed on a volume basis.
For the stripmined wells (51, 52, 53) and the deep mined site (wells 104, 105,
106) iron, aluminum, and sulfate loads appear to follow the pattern of concentra-
tions shown in Figure 4.8b, c and d respectively. When concentrations increase the
loads increase as they decrease 'the loads decrease. The system appears quite
dynamic, signifying ongoing leaching, weathering and oxidation of pyrite.
In Figure 4.9 we look closer at the 500 to 700 day segment of SO, concentrations
and water table elevations in Figure 4.8 for three of the wells. We took well 106
to represent the deep mined area, well 51 to represent stripmined area and well 59 to
represent undisturbed area. In an attempt to find the controlling mechanism of acid
generation, we compared changes in the water table elevations (solid line) with
changes in concentration of SO, (dashed line) using criteria set out in Table 4.2.
At each of the main time segments shown in Figure 4.8 we listed a possible operating
system (Al through C3). The letters above the line designate a more likely system,
those below a less likely one. The analysis suggested that on a deep mined portion
of the site we have acid products added to the gro-undwater as it drains away followed
by an influx of acid percolate during accretion. On a stripmined site we can assume
either a stable concentration of ISO,] suggesting groundwater throughflow and/or
drainage, or, more likely, an acid percolate influx during accretion accompanied by a
groundwater throughflow during decline. The most likely situation on the undisturbed
158
-------
o—o—-
0 -
SOQ S20 540 560 500 800 620 510 580 $00 700
DAYS
Figure 4.9. Water table elevations (solid line) and SO/
concentrations (dashed line) during 500-700
day time segment, points represent experi-
mental values of SO, concentration.
site is a groundwater throughflow and drainage with a hint of acid products arriving
via percolate or from the adjacent matrix.
Site Response in Space—
Mean values—Figure 4.10 shows the distribution of mean values associated with
some of the chemical constituents of groundwater at the study site. The orientation
of the site in relation to North and to monitoring wells is shown in the upper left
hand corner map in Figure 4.10. In general, all the maps show that lowest values of
pH and highest for acidity and chemical concentrations center around wells 105 and
106 in the deep mined area. The steepest gradient appears to exist for pH and Fe,
the least steep for acidity at pH 8.2, SO, and Mn; the gradient for Ca, Mg, and Al
159
-------
Figure 4.10. Average changes in pH (units), and iron (Fe), manganese
(Mn), magnesium (Mg), aluminum (Al),
in mg/1, outside scale is in meters.
(Mn), magnesium (Mg), aluminum (Al), and sulfate (SO,)
is intermediate between these extremes. Although largest variance values (not shown)
(standard deviations ± 90 meq/1) were observed for Fe in the deep mined area, values
almost as large for Mg centered around the highest infiltration zone in the northeast
part of the area shown in Figure 4.3. Several of the chemical variables such as Mn
and Al show an increase in variance towards plot borders, for others (SO, , pH.)
variance increases from North to South in the deep mined area.
Another way of assessing spatial behavior is to examine chemical concentrations
in wells on specific dates during low and high groundwater levels. Table 4.3 shows
160
-------
TABLE 4.3. ACIDITY , pH, AND CONCENTRATION OF SELECTED CHEMICAL CONSTITUENTS IN WELL WATER
ON A MINE AND RECLAIMED SITE MEASURED DURING HIGH AND LOW WATER LEVELS'^
t
Well Ho.
51
52
53
54
55
56
5?
104
105
106
38
59
101
103
10?
Stie«» (2)
Seep (1)
Seep (3)
Seep (4)
1
10.08
10.19
.9.8?
10.02
10.01
10.06
10,07
11.16
10.05
10.44
7.64
2.10
10.03
10.08
11.96
W.t«r
2
9.51
9.70
8.99
8.9?
9.46
9.41
-
10.12
9. 28
10.16
-
2.01
9.72
9.29
12.67
level
3
10.20
10.23
10.02
10.47
10. 19
10.19
10.18
10.57
10. IB
10.48
7.57
2.21
10.16
10.24
13.12
Acidity
<"*,//)
4
B.18
8.41
6.96
-
-
-
-
-
B.31
8.76
-
1.19
-
-
-
1
2.96
2.81
2.97
2.84
2.66
2.97
-
2.81
2.20
2.55
3.64
5.94
-
7.1?
7.34
2.72
2.93
-
-
2
2.99
2.95
2.92
3.10
-
1.03
-
-
-
2.95
-
6.23
6.56
-
-
2.79
2.91'
2.98
2. 11
3
2.92
2.97
1.46
3.08
2.70
3.12
4.14
2.65
J.41
3.12
-
5.80
-
7.88
6.82
2.71
2.7?
2.75
2.63
4
1.05
2.92
3.40
-
-
-
-
-
3.19
2.97
-
5.13
-
-
-
2.60
2.78
2,85
2. 78
.1
6.
49.
22.
16.
16.
10.
-
16.
127.
19.
4.
0.
-
-
-
15.
7.
_
-
4
3
8
5
3
9
9
9
8
0
5
4
B
2
12.1
16.1
18.9
11. J
-
15.3
-
-
-
19.0
-
0.2
2.6
-
-
12.7
3.96
14.2
19.0
1
8.12
.16.9
12.7
12.2
-
9.4
2.8
15.4
90.9
15.6
-
0.4
-
-,
2.1
14.9
11.1
10.?
18.6
4
10.8
41.7
12.1
_
_
-
-
-
17.8
27.5
-
0.8
-
-
-
16.1
12.9
16.9
21.5
i
. 11.
117.
130.
41.
14.
22.
-
108.
860.
250.
0.
0.
-
-
-
80.
42.
_
-
7
5
6
5
4
1
5
0
0
2
2
4
a
Fe
(P-B/O
2
100.0
168.0
169.0
2B.5
-
18.1
.
_
_
117.0
-
0.2
0.2
-
-
141.0
19.9
175.0
231.0
3
57.
450.
141.
12.
56.
7.
-
115.
1255.
255.
_
0.
_
-
_
116.
)05.
161.
207.
5
0
0
2
1
B
0
0
0
7
0
0
0
0
4
405
415
455
-
.
,
_
.
34 i
515
_
4
_
_
_
161
112
221
271
.0
.0
.0
.0
.0
.5
.0
.0
.0
.0
1
25.0
105.0
60.2
44.4
27.5
41.1
_
29.6
105.1
5.1
11.8
1.1
..
0.4
,
17.8
7.1
„
-
Hn
2
21.0
79.8
57.1
47.4
„
51.6
„
_
_
4.6
_
0.4
1.0
_
_
11.2
1.4
10.2
6.1
1
30.6
55.6
41.8
38.2
11.8
14.7
12.2
27.5
(1.2
4.6
_
0.4
_
_
1.?
11.6
11.5
11.3
6,1
i.
27.0
111.1
12.6
f
_
„
_,
62.7
7.6
„
2.5
_
«
10.7
B.O
10.2
8.2
Uoll Ho.
51
52
51
54
55
56
57
104
105
106
56
59
101
103
10?
Strom (2)
Seep (1)
Seep (1)
Seep (4)
C. Kg
<«g/') («8/O
1
104.
151.
162.
182.
125.
152.
-
114.
287.
112.
66.
44.
-
56.
48.
102.
64.
-
... :
3
7
B
9
6
B
6
4
7
3
2
3
2
5
1
2
127.0
110.0
137.0
185.0
-
223.0
-
-
-
170.0
-
46.2
10.0
-
5.1
110.0
52.1
90.4
48.2
1
116.0
217.0
92.5
175,0
66.1
121.0
117.0
64.4
259.0
169.0
-
36.1
-
60.3
14.1
92.5
87. 4
110.0
66.3
4
173.0
255.0
207.0
-
-
-
-
-
350.0
249.0
-
68. 4
-
-
-
113.0
90.4
108.0
94.5
1
83.4
-
150.0
156.0
119.0
145.0
J
60.1
148.0
168.0
40.0
28.5
-
11.1
21.5
52.2
11.6
-
~~
2
87.0
297.0
126.0
149.0
-
204.0
-
-
-
150.0
-
11.6
3.0
-
3.0
47.4
16.1
45.6
41.1
1
93.3
206.0
101.0
130.0
58.4
131.0
118.0
49.0
264,0
158.0
-
22.1
-
5.1
7.2
49.8
41.4
50.6
31.2
4
109.0
164.0
128.0
.
-
-
-
-
275.0
119.0
-
52.2
_
_
-
49.0
40.2
49.0
45.0
1
24.0
217.5
104.0
95.0
68.0
19.0
-
59.0
164.0
35.0
20.1
_
_
_
1.2
57.5
25.1
_
-
Al
2
48,0
141.0
70,5
71.0
_
102.0
_
_
-
44.0
_
0.6
_
_
_
42.1
10.1
41.0
57.0
1
36.0
148.0
42.5
10. 1
44.0
60.0
12.0
34.0
141.0
42.0
„
_
„
—
_
38,0
37.2
51.0
44.0
-4
120.0
203.0
104.0
.
*
_
_
_
157.0
16.0
1.0
_
_
46.5
41.0
48.0
38. 5
1
19.0
102.0
47.0
40.5
31.0
13.0
29.0
68.0
44.0
9.4
3.0
_
_
26.0
11.0
_
-
SO
2
23.0
77.0
17.0
11.0
_
44.5
_
41.5
80.0
17.0
l.H
1.0
0.4,
21. i
6.S
25.0
10.0
>0
1
25.5
74.0
11.0
15.5
25.0
11.0
22.0
28.5
127.0
42.0
4.6
0.5
27.5
21.5
27.5
27.5
4
53.2
73.2
54,7
_
.
..
_
62.5
48,6
S.I
28.5
24.0
27.5
31.0
A
Acidity at pll 8.2 In
t
Water level given aa l.utght above (or be tou-negat ive) > 300 u benchmark.
'ij.t.r level. 8lven lor, 1 - 400 (10/8/79), 2 - 570 (1/24/80). 3 - 600 (4/21/80), 4 - 775 (10/13/80 d.y. «roo nt.rt.
-------
water table elevations, pH, acidity and chemical concentrations in all the wells for
four dates. Dates 1 and 3 correspond to relatively high water levels; dates 2 and 4
correspond to low water levels (Figure 4.8a). For the purposes of analysis the
results were grouped into those for spoil wells (51, 52, 53, 54, 55 and 56), deep
mine wells (104, 105, 106), undisturbed material wells (57, 59, 101 and 103), the
seep, and the stream. When missing values were encountered it meant that insufficient
water was present to obtain a. sample.
Average values for selected concentrations of SO,, Fe, and Al at high and low
water levels are shown in Table 4.4. The results are not conclusive. While con-
centrations of SO., Fe and Al in undisturbed wells are generally low, higher values
on the average of acidity, SO,, Fe and Al occur during low water levels on spoil
wells. In contrast on deep mine wells the average values of acidity and average
concentrations of Fe, Al and SO. are either lower or about the same at lower water
4
levels. Based on the results of this study we have concluded that both the strip-
mined and deep mined areas respond similarly to major storm events (^ 25 mm).
Although on the average little or no difference is apparent between results at high
and low water levels in the wells, iron concentration appears considerably higher
during high water level in the deep mined wells, while acidity, SO,, Fe and Al are
all higher at low water levels in spoil wells.
CONCLUSIONS
Distribution of groundwater elevations in time at the experimental site showed
that highest levels on both the stripmined and deep mined areas generally follow
the occurrence of major storms, while lowest levels appeared to result from 60-day
or longer spells with rains < 25 mm. Iron (Fe) data showed most scatter. Highest
concentrations of Fe on stripmined wells and in the stream occurred when water level
was at the minimum. Aluminum (Al) and SO, were less variable and showed at times
162
-------
TABLE l+.i. AVERAGE WATER LEVELS, ACIDITY AND CONCENTRATIONS OF SO,, FE AND AL AT
HIGH AND LOW WATER LEVELS FOP. GROUPED WELLS
Well Bottom Water Level Acidity
High Low High
Lou
(m) (meq/1)
Spoil
Deepmined
Undisturbed
Seep
Stream
MEAN
7.7 10.1 8.8 18
8.1 10.5 9.3 48
7.0 8.8 7.0 2
18
15
9.8 8.4 20
25
28
1
15
14
14
S04
High
(mg/1)
41
56
4
23
27
30
Lou
50
58
4
24
25
32
Fe
Hign
(mg/1)
107
477
<1
172
98
171
Low
247
326
2
167
152
178
Al
High
(mg/1)
74
86
11
58
- 39
54
Lou
107
79
1
44
43
55
some tendency to follow water level maxima and minima. We have found that despite much
scatter and variability in the data, SO, concentration was most highly correlated with
other variables and could be used to estimate each of the other chemical parameters for
selected sampling sites. Closer scrutiny of 500 to 700 day SO, data suggested that on
the deep mined site the controlling mechanism of acid generation was addition of acid
products during groundwater drainage followed by an influx of acid percolate during
groundwater accretion. On stripmined portion of the site the most likely mechanism of
acid product gain and loss were acid percolate influx during groundwater accretion
accompanied by the groundwater throughflow during decline.
163
-------
REFERENCES
1. Jaynes, D. B,, A. S. Rogowski, H. B. Pionke, and E. L. Jacoby, Jr. Atmosphere
and Temperature Changes within a Reclaimed Coal Strip Mine. Soil Sci., 136(3):
164-176, 1983.
2, Pedersen, T. A., A. S. Rogowski, and R. Pennock, Jr. Physical Characteristics
of Some Minesoils. Soil Sci. Soc. Am. J. , 44:321-328, 1980.
3. Ritchie, J. R. Model for Predicting Evaporation from a Row Crop with
Incomplete Cover. Water Resour. Res., 8(5):1204-1213, 1972.
4. Rogowski, A. S. Hydrologic Parameter Distribution on Mine Spoil. Watershed
Management Symposium, ASCE Irrigation and Drainage Division, July 21-23,
Boise, Idaho, 1980. 'pp. 764-780.
5. Rogowski, A. S., and B. E. Weinrich. Modeling Water Flux on Strip-Mined
Land. Trans. ASAE, 24(4)-.935-940, 1981.
164
-------
-------
SECTION 5
RECOMMENDATIONS
INTRODUCTION
Potentially there are several chemical and hydrologic problems associated with
placement of acid spoil materials. The rationale for a deep placement well below
the soil surface, and preferably below a water table, is to prevent or minimize
oxidation of pyrite to sulfuric acid and associated salts by reducing the supply of
oxygen. If, however, substantial sulfuric acid or associated salts are already
contained within the spoil because of present or previous mining, handling and
reclamation operations (or if large supplied of indigenous salts exist, placement
below a water table) may actually increase the rate of acid and salt leaching.
Specific placement of acid- and salt-containing spoil should be aimed at preventing
contact with percolating water or rising water tables. We recommend placement
based on chemical and physical spoil properties that may affect water percolation
and 02 diffusion rates in the profile. Both the deeper placement of acid spoil and
coarser particle size can substantially reduce the amount of acid drainage.
Placement above the water cable with emphasis on percolate control may be better
for high sulfate spoils, while placement below the non-fluctuating water table may
be better for pyritic spoils.
Acid generation from pyrite oxidation is common in many mining wastes world
wide. Although such materials may occur anywhere in the overburden profile, they
are frequently composed of shales overlying the coal layer.
Present federal regulations in the United States require that acid- and toxic-
forming materials be covered during the backfilling process with a minimum of 1.2 m
165
-------
of the best available non-toxic and non-combustible material. Other requirements
are that the acid- and toxic-forming materials (1) be placed so as to minimize
contamination of onsite or offsite groundwater systems, and (2) be placed so as not
to pose a. threat of water pollution.
In the State of Pennsylvania, the acid- or toxic-forming spoil must be placed
either below the water table if the water table does not fluctuate substantially, or
on a clay layer above the zone of highest anticipated water level and covered with a
soil or soil-like material. The type of placement is usually written into the mining
permit, depending on the projected position and behavior of the post-mining water
table.
The purpose behind specific placement of the acid- and toxic-forming spoil is to
keep it out of the plant root and surface runoff zones and simultaneously to reduce
its potential impact on the groundwater. Techniques for reducing this impact on
groundwater generally rely on atmospheric or hydrologic isolation. Isolation from
the atmosphere (atmospheric isolation) attempts to reduce the 0, resupply rate, thus
slowing down pyrite oxidation that ultimately results in acid mine drainage. One
method of atmospheric isolation is placement of acid spoil several meters below the
ground surface but still in an unsaturated zone, i.e. increasing the pore length for
0? diffusion. The diffusion coefficient of 0- through the atmosphere at 25°C is
2 -i
20.6 mm s (Ohio State University, 1971). The effectiveness of increasing the pore
length depends on the 0,, resupply rate relative to the pyrite oxidation rate.
Atmospheric isolation is better achieved by submerging highly pyritic materials below
the water table. This type of placement is effective because the diffusion coeffi-
cient of 0~ in water is about 4 orders of magnitude less than in air. In contrast,
isolation from flow (hydrologic isolation) emphasizes control over transport of acid
products rather than the rate of acid production. Methods such as clay enveloping
have been used to isolate extremely acid spoil. At times, a clay envelope is
166
-------
deliberately constructed in the unsaturated zone above the water table to minimize
leaching even further.
Once the spoil is reclaimed, the system undergoes rapid change. Within a short
time after backfilling and reclamation, the spoil may subside causing an increase in
density (Rogowski and Jacoby, 1979), plant growth begins and the spoil undergoes
external and internal erosion, weathering and changes in outflow. It is usually
during this initial short-term period that the loss rates of acid products
(Vimmerstedt and Struthers, 1968) and sediment are the highest. After this period
(usually less than a year), flow parameters stabilize and the system seems to
gradually approach a steady state (or as much of a steady state as can exist in a
system subject to climatic variability). To minimize the potential impacts, both' the
short- and long-term effects should be considered.
Our objective was to evaluate how spoil placement may potentially affect the
quality and quantity of drainage and groundwater. To do so, we will consider the
effects of spoil placement and spoil particle size on oxygen and acid product
movement through a prototype spoil system consisting of two 3-m high caissons
filled with field sampled spoil. In addition, we will draw on our previously
published work so that acid leaching can be more completely discussed.
WHAT WAS DONE
The profiles described in this section serve as illustrative examples, they
were sampled near Kylertown, Pa., on a reclaimed spoil site derived from the
Pennsylvania System (Allegheny Series) materials. The prototype profiles were
instrumented for monitoring water content and sampling of leachate (Rogowski and
Jacoby, 1979; Pionke et al., 1980a). One profile was topsoiled to simulate field
conditions, and in second profile a 0.4-m layer of acid shale obtained from just
above the coal layer was placed 2 m below the spoil surface. The coarse nature
167
-------
of the spoil materials, and the properties of both the topsoil and spoil are discussed
in detail in Rogowski (1977) and Pedersen et al. (1978, 1980).
The profiles were leached with rainwater, simulating conditions in the field.
Between leachings, the contained materials were exposed to ambient atmospheric and
temperature conditions for different periods of time. Water content was monitored
using a neutron probe and leachate samples were analyzed for pH, total acidity, total
salts, SO , Ca, Mg, Mn and Fe (Pionke et al., 1980a).
Chemical leachate data used in this paper are from published data by Rogowski
(1977) and Pionke et al. (1980a,b). Leachates from different layers were analyzed
by Rogowski (1977) and Pionke et al. (1980a) for pH, soluble salts and sulfate
content. Results represent several runs at different water application rates
interspersed with variable-length incubation periods. The work of Pionke et al.
(1980b) consisted of a laboratory study on sulfate salt diffusion and leaching
mechanisms at the spoil particle scale. Data from Lovell et al. (1978) provide the
sulfate and pyrite contents of the overburden layers sampled from active highwalls.
Sulfate and pyrite sulfur were analyzed by standard methods (ASTM, 1971) Sulfate S
was extracted in HC1 followed by pyrite S extraction in HNO_. Sulfur analysis was
done gravimetrically by precipitation as BaSO,.
SUBSIDENCE OF MATERIALS AND CHANGES IN DENSITY AND POROSITY
We found that bulk spoil density increased following the first water application
to the newly reconstructed spoil profiles. The topsoiled profile subsided 6 and the
nontopsoiled profile 10% on average. The density change decreased with depth, being
largest near the spoil surface. Some topsoil moved by internal erosion into the
underlying spoil profile (Rogowski and Jacoby, 1979); this affected the hydrologic
behavior of the topsoiled spoil. Infiltrating water did not move as a front but
rather in cracks. A partially impeding soil layer formed at the interface of soil
168
-------
and spoil led to unsaturated conditions in the underlying spoil during subsequent
water applications. When the percolating water reached the bottom of the profiles,
the water table began to rise flooding the spoil. Thus under topsoiled conditions,
the properties of the surface layer and the rising water table determine the water
regime in the lower profile (Rogowski, 1978).
_3
Changes in porosity, assuming constant particle density of 2.65 g cm , are
related to changes in bulk density (Baver et al., 1972, p. 185).
Pore space (%) = (l - (bulk density/particle density)] X 100 (5.1)
Thus, changes in bulk density following water application calculate out to as much as
a one-fourth reduction in the spoil pore space. This in turn would lead to less pore
space being available for oxygen diffusion and water flow.
Compaction of the surface topsoiled layer, especially when coupled with success-
ful revegetation, may limit the amount of water transferred to deeper-placed
materials. If deeper-placed materials are coarse, a denser, less coarse layer
immediately above a coarse acid layer will further inhibit water percolation and
oxygen diffusion. Internal transfer of fines, known as piping which affects the
profile hydrologic behavior, can be controlled by judicious selection of the spoil
particle size most suitable for each layer. In reclamation work, it may be desirable
to use the standard filter design criteria (Cedergren, 1967, p. 175)
D (of spoil) D,_ (of spoil)
< 4~5 < (5-2)
DQ, (of soil) D.,, (of soil)
OJ J.J
Restated in words, this means that: a 15% size (D, ,-) of spoil should be less than
4-5 times the 85% size (Doc) of the topsoil, and also the 15% size (D1C) of spoil
OJ ±J
should be more than 4-5 times the 15% size of the topsoil. The soil and spoil
materials used in the caissons were highly susceptible to internal erosion. In the
169
-------
field, piping and surface erosion of fines leads to large amounts of coarse frag-
ments left on the surface; a common sight at many mines.
WATER MOVEMENT IN THE PROFILE
On the topsoiled spoil we readily ponded the water, applying 500 mm in 2.3 h.
On nontopsoiled spoil water did not pond even when applied at rates in excess of
1000 mm h .We therefore applied 536 mm of water intermittently with garden
sprinklers in 51.6 h. The amounts of water applied represented about half of the
average annual precipitation at the field site.
Figure 5.1 shows initial water content in the topsoil spoil profile during
wetting, 0.5 h after water application started, and water content profiles at 24,
72, 144, 480 and 648 h. Water movement into the spoil during infiltration
appeared to be controlled by the fine/coarse interface between the topsoil and
•5 —
<5
I
«t
03
£.
I
o
!.f
2.0
? «,
Mil
y
/
rl
, i\i I
,
.10
.20 .30
r Conient (m / m )
Figure 5.1. Water profiles in the topsoiled
spoil prior to and following
ponding.
170
-------
spoil layers. Subsequently, during the redistribution and evaporation phases,
little change occurred in the profile water content below 1 m depth. Most of the
changes resulting from evaporation occurred initially at the surface and in the
layer immediately below. It appears, therefore, that topsoiled spoils, such as
these studied by us, may have lower infiltration rates, may store more water in the
finer soil materials in the immediate plant root zone, and may greatly influence
both the underlying spoil water content and the rate of seepage of water to the
water table. The behavior is not unlike that of a much-studied phenomenon in
natural soils where finer-texture layers overlie coarser ones (Miller, 1973).
In contrast, infiltration on nontopsoiled spoil can proceed very rapidly. Water
percolated directly to the water table, which began to rise very shortly after the
water application started (0.6 h). Figure 5.2b shows the response of the water table
to water application and the times, indicated by arrows, corresponded to similarly
identified water content profiles in Figure 5.2a. The general shape of these profiles,
particularly in the surface layers, suggests rapid water percolation, probably by
gravity alone (rather than as a wetting front) and through the larger pores and cracks
between the coarse fragments. The vertical lines with hash marks in Figure 5.2 (a and
b) describe the layers in which the field profile was sampled and reconstructed in the
caissons. The first layer from the bottom consisted of sand and the second of coarse
acid shale fragments.
Comparison of the "Flooded" curve at 51.6 h with the "Redistributed" curve at
144 h in Figure 5.2a illustrates an interesting feature. It appears that during
flooding, percolating water did not completely fill all the voids possibly trapping
some air. Subsequently, during the redistribution phase, water from large voids in
the acid shale drained into the underlying layer. However, the finer-textured layer
above the coarse acid shale could not release enough water to fill the large voids
in the shale. This suggests that such coarser lenses situated within a finer matrix
171
-------
0.10 0.20 0.30
Water Comem (Oy vol
Figure 5.2. Water profiles (a) in nontopsoiled spoil following
wetting, and (b) corresponding elevations of water
table and amounts of water applied.
may at times become dewatered during the water redistribution phase, particularly if
they connect through larger cracks and pores with the outside air at the spoil surface.
We may speculate that if the coarse fragments consist of acid-producing materials, such
dewatering may induce preferential mass-air movement into such spoil pockets enhancing
oxidation. However, the amount of potential acid drainage generated this way would
also depend, as will be seen later, on the relative size of the coarse fragments.
High permeability of the spoil can lead to a rapid build-up of the water table
and subsequent rapid discharge of acid mine drainage. To moderate or prevent such
172
-------
rapid build-up and discharge, we can employ selective placement of less permeable
materials in specific portions of the profile to produce less permeable layers and
reduce both the percolation and the discharge rates. However, a reduction in dis-
charge can be offset by a build-up of acid products which, when eventually flushed
out, may be quite concentrated. At times, coal is not mined completely to the edge
of an outcrop. A certain width of it may be left, forming a natural internal
barrier. Such an internal dam, when coupled with a less permeable layer, may
submerge the acid layer and reduce the potential of pyrite oxidation. Current
regulations in the United States, requiring topsoiling, revegetation and a stipula-
tion that coal should not be stripped to the edge of deposit, reflect these types of
practices. It appears that on materials similar to those studied, acid mine
drainage potential will lessen through judicious placement of spoils, attention to
particle size distributions, topsoiling, revegetation, and some control of flow
rates.
CHEMICAL EFFECTS OF ATMOSPHERIC AND HYDROLOGIC ISOLATION
Placement of acid spoil below the water table can often reduce both the initial
and the long-term acid loss. When submerged, the diffusion rate of 0- is extremely
— 3 2 —1
low (D = 1.8 x 10 mm s in water) (Ohio State University, 1971). The low dif-
fusion coefficient exerts a dominant control on the pyrite oxidation rate by limiting
the 0,, input to reactions 5.3 and 5.4. This is the single most accepted reason given
for placing acid spoil in a saturated zone.
FeS2(s) + 3.5 0£ + H20 = 2S02~ + Fe2+ + 2H+ (5.3)
Fe2+ + 0.25 0 + H+ = Fe3+ + 0.5 H20 (5.4)
173
-------
Submergence is a sound approach, assuming that almost all of the sulfur in the
acid spoil is pyritic or organic. However, this is not the case when the spoil
contains substantial amounts of acid products and salts resulting from atmospheric
exposure, either during previous or present mining activities, or contains salts
indigenous to the formerly-intact strata. Note the high concentrations of SO, in
» H
unweathered and slightly weathered unmined coal overburdens in Table 5 .1. The
TABLE 3-1- THE SULFATE, SULFIDE AND ORGANIC SULFUR CONTENT FOR UNMINED COAL OVERBURDEN
Sulfur content (*)
Geogracmic location
Somerset Gountv, Pa..
near Somerset
Clarion County, Pa.,
near West Freedom
Clearfield County, Pa.,
near Westover
Clearfield County, Pa.,
near Kylertown
Ll thology
B-1X.
B-8X,
B-11X,
B-16X,
A-3X,
C-1X,
06,
08,
B-1X,
Lover
Middle
Middle
Upper
Upper
Upper
Upper
Upper
Lover
Kittanning
Klttanning
Kit tanning
Klttanning
Clarion
Freeport
Freeport
Freepo rt
Kir. tanning
Sulf
0.
1.
2.
0.
5.
0.
0.
1.
0.
:ate
01
,37
.13
18
.39
.03
.05
.02
.74
SuJi
1.
3,
8.
3
13
0.
1
1
1
fide
.64
.20
.08
.10
.17
.24
.85
.18
.98
Organic
0.17
5.12
1.68
1.22
7,04
0.06
2.42
0.03
1.10
Weathering*
S
S
u
s
V
s
s
s
N
The number-hyphen-letter combinations, e.g. S-l, refer to lithologic position in the overburden strata.
X refers to snsde strata immediately overlying coal.
'Percent sulfur by weight.
From visible physical evidence: U * unweachered; S » slightly weathered; N •> not classified.
The first 8 sooils are from Lovell et al. (1978), the last one is from Pionke et al. (1980b).
dissolved SO resulting from pyrite oxidation and subsequent secondary relations will
generally be in the form of FeSO , GaSO,, MgSO,, rU2(SO ) and MnSO , with the CaS04
being dominant, especially in the presence of limestone. All strata designated X in
Table 5.1 represent strata immediately overlying coal. Based on the 3 western
Pennsylvania sites studied by Lovell et al. (1978), these contain the highest pyritic
and sulfate concentrations found in any strata except for the coal itself. In the
2+
absence of limestone, Pionke et al. (1980b) have observed the Fe , Ca, Mg and Mn
concentrations to be dominant and approximately equal in the leachates from the most
acidic spoil layers. Assuming a 1:1 Fe to Ca mix of sulfates in layer C-8 of
174
-------
_3
Table 5.1 (1.02% SO -S, 1750 kg m density, 0.30 m thickness), 1 ha of C-8 material
would contain 127 metric tons of FeSO, and 114 metric tons of CaSO,. In the extreme
4 4
case of layer A-3X of Table 5.1 this would amount to 673 metric tons of FeSO and
602 metric tons of CaSO,. Within solubility limits, placing sizable quantities of
this type of material below the water table would introduce large quantities of both
acidic (FeSO,, Al-tSO ) , MnSO,) and non-acidic (CaSO , MgSO ) salts directly into
the groundwater system. This would also be true for spoils containing relatively low
sulfate concentrations. For example, a 0.1% SO, concentration would yield about 24
metric tons of combined FeSO, and CaSO, per ha. The placement of SO,-containing
spoil below a water table, even in a slowly-moving groundwater system, may cause
initially rapid leaching losses of SO, followed by a decrease according to a
leaching-type curve (Figure 5.3a). In such a system, sulfate leaching losses,
initially controlled by the hydrologic transport, become progressively more control-
led by diffusion as the spoil-particle exterior is leached (Pionke et al., 1980b).
Consequently, the initial SO, losses would be major and the long-term losses prob-
ably minor. Thus, the decision whether to place the most acid spoil layer below the
water table may depend on the spoils' sulfate content and the type of control
(short-term or long-term) desired.
Placement of the acid spoil below a soil surface but above the water table
potentially exposes this spoil to a more rapid 0? resupply rate and greater acid
production, according to the reactions given in equations 5.3 and 5.4. However,
it reduces the direct hydrologic exposure because of unsaturated conditions,
intermittent flows, and only partial contact of flowing water with affected zones.
This may reduce the expected high initial sulfate losses, but can increase the
long-term losses due to increased atmospheric exposure and the reduced rate of
leaching efficiency.
175
-------
j
120 740
L«*ch«te. mm
Percolate mm
Figure 5.3.
Total acidity as a function of (a)
leachate, and (b) percolate quantity
at 1.2 m (•), 2 m (A), and 2.5 m (o)
below the spoil surface.
The 02 resupply rate in spoil results primarily from CL diffusion. The usual
exception is a thin mixing zone caused by temperature changes and wind profile near
the surface. Within the 0? diffusion zone, the oxygen resupply rate will depend on
the diffusion path length and the porosity of overlying materials. Assuming a
constant diffusion coefficient, D, in the diffusion equation, F = D 3C/3x where F is
the transfer rate of diffusing substance per unit area, and 3C/8x is the concentra-
tion gradient normal to the surface, we can estimate both the effects of increasing
the path length (depth of placement) or decreasing porosity (spoil settlement). The
6-10% volume reduction described earlier approximately translates into a 25%
decrease in pore volume and cross-sectional pore area. The F values, or potential
0 resupply rates, would thus decrease accordingly. Changing the path length has
potentially an even greater effect. Spoil placement at a depth of 9 rather than 3 m
176
-------
would decrease the 3C/3x by increasing the 3x value. Assuming uniform spoil bank
porosity and that only one spoil layer consumes 0?, the greater depth of placement
would potentially decrease the CL resupply rate to approximately one-third the
original rate. However, this rate could still exceed the 0 consumption rate
according to equations 5.3 and .4, and thus the deeper placement would not reduce
acid production. Deeper placement, if effective, would probably provide a long-term
rather than a short-term control. Rapid spoil settling, changes in density and
reduction in pore volume would probably cause the greatest short-term changes.
Moreover, the reclaimed spoil would initially contain a reservoir of approximately
20% CL in the pore space air. Assuming all 0 consumed within 6 months by reaction
5.3 to form a 47% CaSO,-53% FeSO, (1:1) mix, 28 metric tons ha~ of the CaSO.-FeSO.
44 44
mixtures would be produced by a 9-m thick spoil profile, with a 50% initial pore
volume.
Perhaps the most effective control on acid loss to groundwater from spoil
placed above the water table may be the water flux rather than oxygen flux control.
This assumes a certain type of chemical-hydrologic relationship. If we plot con-
centration of acid products in leachate versus leachate or percolate volume, we can
get 2 basically different relationships, which are exemplified in Figure 5.3
(a and b).
The leaching-type curve (Figure 5.3a) can characterize a moving groundwater
system. This curve was generated by measuring the acid product content in
leachate from a spoil-filled column subjected to rapid, saturated throughflow
(Pionke et al., 1980b). Decreasing the leachate volume within the linear part of
the curve, e.g., from about 90 to 30 mm, would provide an approximately constant
acid product load (concentration X water flow) delivered to groundwater. Thus, the
acid product load lost to the groundwater system would be controlled by the resupply
177
-------
rate of acid products from spoil to the surroundings, which appears to be diffusion
controlled.
Figure 5.3b is the observed relationship between the amount of percolation under
unsaturated flow conditions and total acidity of that percolate taken from lysimeters
placed at different depths in Caisson 2 (Pionke et al., 1980a). In contrast to
Figure 5.3a, reducing the percolate in Figure 5.3b would greatly reduce the acid
product load delivered to the water table. For example, reducing the percolate to
one-half (12-6 mm) substantially reduces the acid product load delivered to each
lysimeter, and reduces it to about 1/2 the original load in 2 of the 3 lysimeters.
This relationship characterizes the field situation where the load would be control-
led by the percolation or flow rate.
The reason for the apparently stable percolate quality of Figure 5.3b is not
known. We speculate that as percolation proceeds, the pathways of water flow
through the spoil change, thus continuously leaching a percentage of fresh spoil.
More likely, acid salt concentrations at or near the spoil particle surface are in
excess of the flow transport capacities. Percolate control in the unsaturated
system might therefore be particularly useful for treating spoils that contain high
acid product concentrations. Its impact would probably be both initial and
long-term.
INTRAPARTICLE DIFFUSION PROCESSES AFFECTING ACID PRODUCTION AND LOSS
Up to this point, we have considered either deep placement of the acid spoil en
masse to reduce pyrite oxidation, or changes in percolation to control initial and
long-term impacts on groundwater quality. Turning now to the individual spoil
particle scale, where the particle is partially weathered, diffusion processes can
potentially control acid production in situ or acid product loss to percolating or
surrounding waters. Pionke et al. (1980b), using spoil B-1X (at the bottom of
Table 5.1) and the published diffusion coefficients for the common SO, salts,
178
-------
calculated the depth of the SO.-salt-contributing zone within a spoil particle. This
approach differs from that of Caruccio (1975) or Geidel (1979) in that we postulated
a cylindrical model of spoil particle and applied to it diffusion theory to explain
observed results. In contrast, the work of Caruccio (1975) and Geidel (1979) is
based on experimental data from selected crushed (to pass 2-mm sieve) spoil
materials, without an attempt at a theoretical development.
Assuming a cylindrical particle (Figure 5.4), conceptually, a boundary (r )
3.
exists within this particle that separates the weathered outer zone from the
unweathered interior zone. Inside this boundary, the SO, concentration with depth
is relatively uniform. Outside this boundary, SO, and 0- gradients exist that
control the transport rate of SO, to the particle surface, or of 0~ to the pyrite-
containing interior. The weathered outer zone, of depth (r,-r ), was calculated
D 3.
using a diffusion equation and experimentally-collected input data. Converting
the depth (r,-r ) to volume for different-sized cylindrical particles, the
Q 3
effective SO,-contributing volumes in Figure 5.5 are readily calculated. Figure
5.5 shows that the effective SO,-contributing volume to total spoil particle volume
(%) decreases rapidly with increasing particle size. The considerably faster
Weathered Outer ? r'°' °?*
t , . , . , tnooeontnbutmg)
^one iConlribuliftQl
Most direct
diffusion p*
Figure 5.4. Assumed structure of
cylindrical spoil
particle.
179
-------
diffusion of CL into the particle would shift the curve to the left but still provide
the same basic relationship.
0.11 L
1.0 05 0-23
P0rt«cl« Rodtu* mm
Figure 5.5. Contributing spoil volume as a function
of cylindrical particle radius.
The values in Figure 5.5 apply strictly to spoil B-1X of Table 5.1 under sub-
merged conditions. However, similar relationships would be expected for other
spoils provided that diffusion processes were controlling, i.e. the mass movement-
rate of water through the spoil particle was not significant. The impact of
increasing the particle size, and thus decreasing the percentage of contributing
volume, is basically a long-term one because the outer portion of the particle must
become partially weathered and leached. The initial impact would probably be minor,
because both the indigenous acid products and pyrite in the freshly fractured spoil
would be in abundant supply at or near the particle surface.
The effectiveness of diffusion control was not studied for the unsaturated
condition. In this case, the total pore space within the particle would be likely
to contain both water and air. Potentially, the 0? would diffuse much faster into
the particle because of the far greater diffusion coefficient of 0~ in the air.
180
-------
In contrast, the diffusion coefficients of SO, salts would be much less due to the
lower water content. However, other conditions such as mass movement of water into
the particle upon wetting, or out of the particle upon drying, with concomitant air
entry and the continuity of air or water filled pores, could greatly alter the
relationships between the spoil particle size and percent effective contributing
volume.
We conclude that a number of spoil properties need to be considered before
deciding how to handle acid spoils. Generally, these properties relate to, and
often control, percolation and 0? diffusion rates into the spoil bank. These
include porosity, water retention, and particle size distribution of the topsoil,
the layer below the topsoil, and the most acid spoil layer. Depth below the
surface and differences in particle size distributions can substantially reduce
the amount of percolate or slow the amount of CL delivered. Analysis of both SO,
and pyrite concentrations in overburden can identify the most acid strata for
subsequent isolation and can indicate whether placement should be above or below
the water table." Placement above the water table, with emphasis on percolate
control, may be a better method for highly SO,-contaminated spoil, whereas place-
ment below a nonfluctuating water table may be best for highly pyritic spoils low
in SO,. In addition, the rate of acid production and loss from spoil can
potentially be managed by controlling the spoil particle size.
181
-------
-------
REFERENCES
1. ASTM. Standard Methods of Coal Analysis. American Society for Testing
Materials, Philadelphia, Pa., 1971.
2. Baver, L. D., W. H. Gardner, and W. R. Gardner. Soil Physics. Fourth
Edition, John Wiley and Sons, Inc., New York, 1972. 498 pp.
3. Caruccio, F. T. Estimating the Acid Potential of Coal Mine Refuse. In:
The Ecology of Resource Degradation and Renewal, M. J. Chadwick and G. T.
Goodman, eds. Blackwell Scientific, London, 1975. pp. 197-205.
4. Cedergren, H. T. Seepage, Drainage, and Flow Nets. John Wiley and Sons,
Inc., New York, 1967. 489 pp.
5. Geidel, G. Alkaline and Acid Production Potentials of Overburden Materials:
The Rate of Release. Reclam. Rev., 2(3/4):101-107, 1979.
6. Lovell, H. L., R. R. Parizek, D. Forsberg, M. Martin, D. Richardson, and
J. Thompson. Environmental Control Survey of Selected Pennsylvania Strip
Mining Sites. Final Report, Argonne National Laboratory, 9700 S. Cass Ave.,
Argonne, 111. 60439, Contract No. 31-109-38-3497, 1978. 468 pp.
7. Miller, D. E. Water Retention and Flow in Layered Soil Profiles. In:
Field Soil Water Regime. Spec. Publ. No. 5, Soil Science Society of
America, Madison, Wis., 1973. pp. 107-117.
8. Ohio State University Research Foundation. Acid Mine Drainage Formation and
Abatement. Water Pollution Control Research Service, DAST-42-14010
FPR-04/71, U.S. Environmental Protection Agency, Washington, D.C., 1971. 83 pp.
9. Pedersen, T. A., A. S. Rogowski, and R. Pennock, Jr. Comparison of
Morphological and Chemical Characteristics of Some Soils and Minesoils.
Reclam. Rev., 1:143-156, 1978.
10. Pedersen, T. A., A. S. Rogowski, and R. Pennock, Jr. Physical Characteristics
of Some Minesoils. Soil Sci. Soc. Am. J., 44:321-328, 1980.
182
-------
11. Pionke, H. B., A. S. Rogowski, and C. A. Montgomery. Percolate Quality of
Strip Mine Spoil. Trans. ASAE, 23:621-628, 1980a.
12. Pionke, H. B., A. S. Rogowski, and R. J. DeAngelis. Controlling the Rate of
Acid Loss from Strip Mine Spoil. J. Environ. Qual, 9:694-699, 1980b.
13. Rogowski, A. S. Acid Generation within a Spoil Profile: Preliminary
Experimental Results. In: Seventh Symposium Coal Mine Drainage Research, •
NCA/BCR, The Coal Building, 1130 Seventeenth St., N.W., Washington, B.C.
20036, 1977. pp. 25-40.
14. Rogowski, A. S. Water Regime in Strip Mine Spoil. In: Surface Mining and
Fish/Wildlife Needs in the Eastern U.S., D. E. Samuel, J. R. Stauffer,
C. H. Hocutt, and W. T. Mason, Jr., eds. Proceedings of a Symposium,
December 3-6, Morgantown, West Va., 1978. pp. 137-145.
15. Rogowski, A. S., and E. L. Jacoby, Jr. Monitoring Water Movement through
Strip Mine Spoil Profiles. Trans. ASAE, 22:104-109, 114, 1979.
16. Vimmerstedt, J. P., and P. H. Struthers. Influence of Time and Precipitation
on Chemical Composition of Spoil Drainage. In: Proceedings, Second
Symposium Coal Mine Drainage Research, May 14-15, Pittsburgh, Pa., Bituminous
Coal Research, Monroeville, Pa. 15146, 1968. pp. 152-163.
183
-------
TECHNICAL REPORT DATA
(Please read liizlrucnons on tnt reiersc be lore completing/
1. REPORT MO.
FPA-60n/7-84-044
|3. RECIPIENT'S ACCESSIOJ»NO.
A. TITLE AND SUBTITLE
Hydrology and Water Quality on Stripmined Lands
5 REPORT DATE
March 1984
6. PERFORMING ORGANIZATION COO£
7. AUTHOR(S)
A. S. Rogowski and H. B. Pionke
8. PERFORMING ORGANIZATION REPORT NO.
8
9 PERFORMING ORGANIZATION NAME AND ADDRESS
Northeast Watershed Research Center
USDA-ARS, 110 Research Building A
University Park, Pennsylvania 16802
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
EPA-IAG-D5-E763
12. SPONSORING AGENCY NAME AN")
Office of Environmental Processes and Effects Research
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC 20460
13 TYPE OF REPORT AND PERIOD COVERED
FINAL
14. SPONSORING AGENCY CODE
EPA/600/16
is. SUPPLEMENTARY NOTES
r r u. e ivi c IN i **n n > i>i w i c-a
This project is part of the EPA-planned and coordinated Federal Interagency
Energy/Environment R&D Program^
16. ABSTRACT
Studies were conducted to evaluate physical properties of spoils resulting from
surface-coal mining and reclamation operations. Physical properties were determined
at 10 sites randomly located within a 4-ha experimental area. Microlysimeter data
indicated that evapotranspiration (ET) on minesoil could be approximated by class-A
pan evaporation or by model results.
Two large caissons were also used to measure water movement in reconstituted spoil
profile. Settlement was considerable and changes in temperature and bulk density were
good indicators of water movement within the spoil and topsoiled spoil profiles. Flow
through larger channels did contribute significant amounts of water to the water table.
Total acidity provided a reasonable estimate of other major chemical parameters
contained in spoil drainage.
The concentrations of Cd, Cr, total soluble Fe, Hg, Mn and Zn in spoil extracts
exceeded EPA water quality drinking standards from one (Cr) to all (Mn) spoil layers.
The standards for Pb and Cu were not exceeded.
Specific placement of acid- and salt-containing spoil should be aimed at
preventing contact with percolating water or rising water tables.
(Circle One or More)
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
f; —•.-S-p —r«i .--s:n»»fn
------- |