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
-------