Numerous studies of the vertical distribution of sulfur in coal have been done for coals around the world, encompassing various geologic periods and coal rank. Increased sulfur at the top and bottom of coal seams appears to be the rule rather than the exception. Suggate (1995) in a study of subbituminous Miocene Epoch coal in New Zealand observed that sulfur content decreased from roof to floor, with a minor increase near the floor. The percent sulfur at the top of coal varied from > 6% to < 2%, however the sulfur in the middle of the coal seam was always 1.5 to 2.0% less. A study of New Zealand Cretaceous coals deposited under marine conditions (Sherwood et al. 1992) also noted higher sulfur at the top of the seam. A bituminous coal of Cretaceous Period from Alberta, Canada averaged only 0.3% sulfur, and the coal was deposited "well removed from marine clastic influences" (Langenberg et al. 1992). This coal showed sulfur elevated at the base and to some extent at the top. The Tertiary "Big George" coal deposit in the Powder River Basin of Wyoming is a 200-ft (60 m) thick subbituminous coal associated with freshwater depositional environments. This coal has total sulfur typically below 0.3%, however the bottom 0.6 ft (0.2 m) has over 1% sulfur (Kent, 1986). The trend of higher sulfur in the upper and lower portions of a coal seam have also been observed in the Eocene lignite coal deposits in Texas (Arora et al. 1980).

Another example in the literature of high sulfur at top and bottom of coal is a coal in the Cherokee Group (Pennsylvanian) in southeastern Iowa (Biggs and Bruns, 1985). The upper 28% of the coal contained 67% of the pyrite, the middle 47% contained 19% of the pyrite and the lower 25% contained 14% of the pyrite. The coal was bounded top and bottom by pyrite-rich shales. The paleoenvironment is not discussed, but most coals in this part of Iowa have associated marine sediments (Howes, 1990).

Similar trends to those described above have been observed in coal seams of the northern Appalachian basin. Reidenour et al. (1967) found higher sulfur at the top and (sometimes) bottom of Clarion and lower Kittanning coals in Clearfield County that have roof rocks that were deposited in a brackish depositional environment. Appalachian coals interpreted to have been deposited in a freshwater depositional environment also show high sulfur at the top and bottom of the seam. This has been observed for the Pittsburgh (Donaldson, et al., 1979) (Figure 8.55), Redstone (Hawkins, 1984), and Waynesburg (Donaldson et al., 1985b) (Figure 8.56) coals of northern West Virginia and southwestern Pennsylvania. High sulfur has also been observed at the top and bottom of the upper Freeport coal by Cheek and Donaldson (1969) in northern West Virginia.

Figure 8.56 Cross-section showing total sulfur content of the Waynesburg coal in a strip mine in northern West Virginia. Note that the highest sulfur is at the top of each of the two benches of coal. From Donaldson et al. (1979).

The findings for vertical variations in sulfur concentration within a coal seam are consistent with the distribution of sulfur in modern thick peat deposits. Modern peat deposits frequently show high sulfur at the base of the deposit, for example Staub and Esterle (1994) attribute the high sulfur at the base of a peat deposit on Sarawak, East Malaysia to initial conditions being "mangrove and Nipa vegetation" growing on a mudflat under brackish to marine conditions with waters rich in sulfate. This organic rich environment would have near neutral pH and abundant iron, conditions favorable for pyrite formation. Neuzil et al. (1993), likewise found the highest sulfur in the bottom of the peat in the Riau and West Kalimantan provinces of Indonesia. Sulfur values for the basal peat were ~0.3 % (with a high of 0.56 %), whereas the upper peat was less than 0.2 %. The underlying sediments were of marine origin and the authors suggest this may have supplied sulfur to the lowest peat horizons. Esterle and Ferm (1994) looked at thick peat deposits at Sarawak, Malaysia, and Sumatra, Indonesia. Sulfur in the basal one meter of the deposit was 2.4 to 4.5%, and was 0.3% or less above this zone.

It can not be assumed that high sulfur in the upper portion of a coal bed, or high sulfur within a coal bed, are evidence of marine influence. Paleoenvironmental interpretations using sulfur alone, may not be valid. The fact that high sulfur is frequently found at the top and bottom of coal seams, regardless of paleoenvironment, is important from a mining standpoint. The top and bottom of a coal seam are the most likely to be left behind on the mine site as "pit cleanings" because of high sulfur or ash, or as coal that is not recoverable in the mining process. The acid potential from this source must be considered in any evaluation of potential acid-materials problems.

All of the discussions so far have involved geologic processes that occurred hundreds of millions of years ago, during the Pennsylvanian. These include the controls that paleoclimate and paleodepositional environment had on overburden mineralogy. This section will deal with a much more recent geologic process, the near surface physical and chemical weathering of rock which has occurred within the past million or so years. The significance of this influence on the distribution of carbonate and sulfide minerals (pyrite) can be as great as those that occurred in the more distant past. The discussion will look at weathering south of the glacial margin and within the glaciated portion of the Appalachian Plateau. Weathering results in the near surface removal of carbonates and sulfide minerals; carbonates by dissolution and sulfides by oxidation. This zone is usually recognizable by the yellow-red hues (indicative of oxidized iron) of the rocks. Generally, in the unglaciated portions of the Appalachian Plateau, the intensely weathered zone extends to 20 to 60 ft (6 to 20 m) below the surface.

Chemical weathering of bedrock is enhanced by physical factors such as stress-relief fracturing on hill slopes and bedding-plane separations due to unloading. Clark and Ciolkosz (1988) have suggested that periglacial conditions during the Pleistocene contributed to the shattering of near-surface rock. Kirkaldie (1991), in a study of weathering of bedrock on the northern Appalachian Plateau, observed that the maximum thickness of weathered bedrock beneath glacial deposits was similar to the maximum thickness of weathered bedrock in unglaciated regions. The distribution of thickness of the weathered zone beneath the glacial deposits, however, was different, with less weathered rock typically beneath glaciated areas (Figure 8.57). This is attributed to erosion of much of the weathered surface material by glaciers. The findings of Kirkaldie suggest that (1) the bedrock weathering occurred prior to the last glacial period, or that (2) in the non-glaciated areas the rate of weathering has been balanced by the rate of erosion. No studies of weathering rates have been done for the Appalachian Plateau, so the second theory has not been tested. Whatever the method, and whatever the duration of time, shattering of rock increases surface area and thus provides greater area for contact with water and air, and weathering is accelerated. These processes acting together also increase the permeability of the weathered zone. As discussed in Chapter 10 the ground water associated with the weathered zone is dilute, in terms of dissolved solids, because readily soluble products have been removed by chemical weathering.

Figure 8.57 Distribution of the thickness of weathered bedrock in glaciated and unglaciated portions of the Appalachian Plateau in Pennsylvania. The large box encloses the 25^th and 75^th percentiles. The horizontal line within the box is the median, the "whiskers" indicate the range of data. Data are from Kirkaldie (1991).

Chemical weathering is also influenced by lithology. Coarser, more permeable lithologies may allow oxidation to extend to a greater depth. Kirkaldie (1991) measured the depth of the highly weathered zone, noting the type of cover, amount of cover, lithology, and topographic position. In unglaciated terrain, he noted that the maximum thickness of highly weathered rock was 28.9 ft (8.8 m) in sandstone, and only 11 ft (3.3 m) in shale. However, the lithologic difference was not the only variable.

This weathered-rock zone exists throughout the Appalachian Plateau. In spite of this it has been little studied. Smith and colleagues (Grube et al., 1972; Smith et al., 1974; and Singh et al., 1982) investigated the effects of weathering on the Mahoning sandstone in northern West Virginia, and noted a "pyrite-free weathered zone approximating 20 feet (6 m) of depth below the land surface..." (Smith et al., 1974, p. 3). Singh et al. (1982), in addition to noting the pyrite-free zone, also note a loss of "alkaline earth" elements within 20 ft (6 m). This is illustrated for calcium in Figure 8.58. The loss of calcium is best explained by a loss of calcareous minerals (calcite and dolomite). Brady et al. (1988; 1996) in a study of the upper Kittanning and lower Freeport overburden in Fayette County, PA noted a similar loss of calcareous rocks in the near-surface weathered zone to about 7 m depth. Hawkins et al. (1996) noted weathering to depths of 30 to 60 ft (10 to 20 m). The loss of carbonates due to weathering is also discussed in Chapter 10.

Figure 8.58 Total sulfur and total calcium in sandstone overburden, as a function of depth below the surface, for four sites in Preston County (northern), West Virginia. From Singh et al. (1982).

A weathering profile is evident at the top portion of most drill logs shown in this chapter. It is hard to determine precise rules of thumb for the depth of leaching of carbonates and oxidation of pyrite because these minerals can only occur where they were originally present (before weathering). If no pyrite was ever present within a stratigraphic horizon, its absence is not due to weathering, but to the fact that it was never there in the first place. The same is true for calcareous strata. It is evident from the stratigraphic figures that rarely do NPs greater than 30 ppt CaCO₃ or sulfur greater than 0.5% occur within 20 ft (6 m) of the surface. A good example of this effect is the Blue Lick coal in Figure 8.26. Where the coal is under shallow cover it has less than 0.7 percent sulfur, however, in OB-C where the coal occurs under about 60 ft (20 m) of cover the coal averages over 2 percent sulfur. Most of the exceptions are where a horizon equivalent to a limestone, or strongly calcareous zone lies close to the surface. Some drill holes that encounter this zone at less than 20 ft (6 m) may still contain carbonates. The same is true for very high sulfur zones.

Figure 8.59 shows the effects of weathering on the sulfur content of the upper Freeport coal and the roof rock (shale) in an area of Fayette County. There is a clear loss of sulfur at shallow depth (17 to 20 ft; 5 to 6 m) in both the coal and the shale.

An accurate knowledge of the extent (depth) of the weathered zone is important from an overburden sampling standpoint. Overburden sampling should be done to the extent that it adequately represents the weathered zone and unweathered bedrock. This will entail drilling overburden analysis holes at maximum cover to be mined and at lesser cover. Weathering is also discussed in Chapter 9.

An understanding of the effects of weathering on the distribution of pyrite, sulfate salts, and carbonates is important in (1) accurately defining their distribution within unmined overburden, (2) being able to design a mine plan to prevent postmining problems, and (3) accurately predicting postmining water quality. An understanding of the weathering profile is just as important as understanding the lateral and vertical distribution of strata and their pyrite (sulfur) and carbonate content.

Glacial deposition was discussed earlier. Glaciers not only deposit, but can also erode. Erosion can include previously deposited glacial sediment, bedrock, and soil. Weathered bedrock is easily erodible and tends to be thinner in glaciated regions because of removal by glacial erosion. This is illustrated in Figure 8.57 where the median thickness, as well as the interquartile range of weathered bedrock is lower for areas that have been glaciated than for areas that have not been glaciated.

As with bedrock, the weathering of tills and other glacial sediments results in a change in the geochemistry and therefore in their effect on water quality. Weathering produces a vertical mineralogical gradation within tills. Leighton and MacClintock (1930, 1962) described the weathering of tills by dividing them into five horizons (Figure 8.60). Horizon 5 is unweathered till. It is still its original color (usually a shade of gray in northwestern Pennsylvania) and carbonates are still present. In horizon 4, pyrite, and other iron minerals, are oxidized. A till’s characteristic oxidized color, a shade of yellow, brown, or red, is caused by iron oxide (such as hematite) or hydrous iron oxide (such as goethite and limonite) replacements of iron magnesium silicates and iron sulfides (Hallberg, 1978). Oxidation of manganese minerals (Hallberg, 1978) generally produces a darker brown color. Tills high in carbonate also tend to oxidize to a darker brown compared to low carbonate tills (White, 1982). Carbonate minerals are still present in horizon 4.

In horizons 3, 2, and 1, till is oxidized and totally leached of carbonates. Horizon 2 is chemically decomposed till. Silicates have been degraded. It is still recognizable as till. Horizon 1 is the A and upper B horizons of the soil where major leaching, added organic material, and addition of iron oxide and clay have modified the composition and texture. The parent material is no longer recognizable.

Horizons 4 and 5 are those of benefit in the prevention of acid mine drainage. Carbonate minerals are still present in those horizons, although partial leaching may have occurred. Horizon 4 may be partially leached in the coarse sand fraction (Hallberg, 1978) before leaching is apparent in the clay fraction.

Weathering begins immediately upon deglaciation. A later glacial advance over the existing till may erode

part or all of the weathered horizons of the preexisting till. If multiple tills are encountered during overburden removal, a buried soil or weathered horizon, with carbonate minerals leached, may exist between two calcareous zones. Figure 8.37 illustrates this situation. Unit 6 is a weathered, leached zone between calcareous zones 4 or 5 and 7. Units 10 and 11 are again a leached zone between the calcareous units 9 and 12. In Figure 8.38 and Figure 8.61, the NP distribution with depth at the McCoy and Spagnolo mines show low NP zones between high NP zones. These are partially leached weathering zones at the top of older, buried tills.

Likewise, if a till is thin and overlies another weathered till (a likely scenario in the glaciated coal region), leached zones of both tills may be superimposed, resulting in an excessive depth of leaching. Figure 8.62

Depth (feet)

(11.58 Meters) Neutralization Potential (ppt)

Figure 8.61 Distribution of NP for till stratigraphy at the Spagnola mine, Mill Creek Township, Mercer County. Mine site is at location "B" on Figure 8.36.

illustrates this situation with three tills. Over most of the exposure, calcareous till is not encountered until the third till is reached (unit 6).

Each till has a characteristic average depth of leaching (Figure 8.60), providing no erosion of till has occurred since the beginning of weathering, and no other sediment has been deposited on top to halt the weathering process. The carbonate content of the Mapledale Till is so low that weathering horizon 4 cannot be determined in the field, and the depth of leaching is difficult to determine. The Mapledale Till is oxidized to a depth of 10 to 16 feet (3 to 5 m) (White et al., 1969).

Because a number of factors control the depth of leaching (and other weathering characteristics), the same till may be leached to different depths in different locations. All other factors being equal (which they usually aren’t), older tills are leached of carbonates to a greater depth than younger tills. The nature of the parent material affects the rate of weathering. Coarser, more permeable sediments allow greater penetration of weathering agents (Hallberg, 1978). Landscape position can affect runoff vs. infiltration. Also, on steep slopes, erosion may proceed at a rate equal to weathering and a weathering profile may not develop.

Many of the weathering processes described above for western Pennsylvania are relevant to the Anthracite Region because the paleoclimate and present climate for both regions were/are essentially the same. The geologic structure of the Anthracite Region, however, as part of the Ridge and Valley Physiographic Province, is dramatically different than the flat-lying rocks of the Appalachian Plateau. Due to the combined effects of the folded and faulted geologic structure and the resistance of the sandstone lithologic units to erosion, the weathering processes have formed a different topographic expression in the Anthracite Region.

Classic and recent works on weathering in the Ridge and Valley Province have concentrated on topographic development and glacial and periglacial features. In the first issue of National Geographic, Davis (1888) discussed the geologic structure, erosional history and topography of valleys and ridges of central Pennsylvania and the "dissected plateau further west" (p. 14). A later more detailed discussion of the subject is in Davis (1909). Descriptions of Appalachian slope form development are contained in Hack (1960, 1979). Additional information on weathering in the Ridge and Valley is found in Thornburry (1965, 1969), Clark and Ciolkosz (1988), and Sevon (1989). The four anthracite fields are preserved in synclinal basins that are essentially surrounded and "defended" by sandstone ridges. These ridges are more resistant to erosion than the shales and coals of the Pottsville and Llewellyn Formations. The slope forms of the ridges are typically mature (i.e., convexo-concave), but some free faces occur, such as the Harveys Creek water gap in the Northern Field.

The Pleistocene glaciation and climate left several features that can be observed today. These include till deposited in the Northern Anthracite Field (Hollowell, 1971, and Braun, 1997), and periglacial effects which extend south to the other anthracite fields. Periglacial activity has, in most places, lowered ridge tops as much as a few tens of meters (Braun, 1989). As a result, earlier weathering horizons have been removed, leaving a very small depth of weathered rock. Some of this weathering has resulted in the deposition of colluvium on the lower slopes. Colluvium in the Ridge and Valley province ranges in thickness from one meter on the upper ridge slopes, to 30 meters at the base of the slopes and in valleys. Twenty-seven percent of the soils in the Ridge and Valley province are colluvial, versus only 13 percent in the Plateau (Ciolkosz, 1978). Ridge tops have little or no weathered material. The lower one-half to three-fourths of the ridge slopes are colluvial. Residual soils occupy valley floors (Ciolkosz and others, 1979). In most places in Pennsylvania, at least two ages of colluvium are present, which includes reworked older regolith (Braun, 1989; Sevon, 1989). There is rarely any residual weathered bedrock beneath colluvium or glacial sediments in the Ridge and Valley (W. Sevon, personal communication, 1997), and where present, it is only a few feet thick (Hoover and Ciolkosz, 1988).

Because the bedrock, as far as is known, is low in pyrite and the tills lack carbonates, the effects of weathering in the Anthracite Region on mine drainage is probably not as significant as in the bituminous region.

Lithology is controlled by geologic factors such as paleoclimate and paleo-depositional environment. Sandstones are deposited in high energy environments, whereas shales and siltstone are deposited in quieter environments. Donaldson et al. (1985a) report the presence of several types of sandstones throughout the Pennsylvanian: lithic, calcareous, and quartz-rich (Figure 8.48). Quartz-rich sandstone is found predominantly in the lower Pennsylvanian Pottsville Group formations, but also occurs in the Allegheny and lower part of the Conemaugh, decreasing in frequency upward (Skema, personal communication, 1998). Lithic sandstones are found throughout the section. Calcareous sandstone is common from the middle of the Conemaugh through the Dunkard. Sandstone can form in a variety of environments, including alluvial and distributary (deltaic) channel sands, and marine barrier-bar sands. Typically sandstone is composed of quartz with varying amounts of feldspar, mica, and organic debris. Frequently they contain a high concentration of quartz because of this mineral’s resistance to mechanical and chemical weathering. In this section we will examine the relationship between rock type and water quality. Abundant sandstone relative to finer-grained rock types, such as shale and siltstone, has been linked to acid mine drainage in the Allegheny Group. We will also examine what, if any, relationships exist between rock type and water quality in the other coal-bearing stratigraphic Groups of western Pennsylvania.

Many published studies of Allegheny Group mines with abundant sandstone overburden attest to a problem with water quality. Examples of Allegheny Group mine sites with sandstone overburden that produced acidic drainage are: Clarion coal (Dugas et al., 1993; Cravotta, 1991; Site 9 in Brady, et al., 1990; Henke, 1985; J.H. Williams et al., in press); lower Kittanning coal (Guo et al., 1994; Durlin and Schaffstall, 1993); upper Kittanning coal (Brady et al., 1988) and upper Freeport coal (Kania et al., 1989).

Williams et al. (1982) did an analysis of the geologic controls on water quality within the lower Allegheny (Kittanning/Clarion "formations") and upper Allegheny (Freeport "formation") for, primarily, the northern portion of the bituminous coal field. Figure 8.63 is the model for the Kittanning/Clarion, showing six different lithologic (overburden) relationships. Table 8.10. Mean values for select water quality parameters demonstrating differences of water quality with different overburden lithology. "Group" numbers refer to specific lithologic sections shown on Figure 8.63 and Figure 8.64. Data are from Williams et al. (1982).

These models, and accompanying water quality as displayed in the above table, show the important influence of marine and freshwater limestones and calcareous rocks on postmining water quality. They also show the effects of predominantly sandstone overburden, especially for the Freeport sites (Model 2) which typically produce acidic drainage. Comparatively, sites with little or no sandstone typically produce alkaline drainage (Models 1, 3, and 4). Williams et al. (1982) found that the "role of sandstone as an acid-producer cannot be determined from this Kittanning study, due to accompaniment of either acid-producing, high-sulfur shale, or limestone" with the sandstone. Kittanning/Clarion models 2 and 4 are both acid producing. Model 2 represents primarily brackish shales in the absence of calcareous strata. Model 4 represents the sandstone overburden, with acid-producing shale below. Model 3 is similar to situations where the marine Vanport limestone overlies the Clarion coal. Models 5 and 6 include freshwater limestone, such as the Johnstown limestone that occurs below the upper Kittanning coal, thus the alkaline water. The nonmarine portion of the Kittanning model is very similar to the Freeport model.

Figure 8.63 Schematic diagram representing the Kittanning/Clarion model of Williams et al. (1982). From Williams et al. (1982).

Figure 8.64 Schematic diagram representing the Freeport model of Williams et al. (1982). From Williams et al. (1982).

The work of Brady et al. (1988) is consistent with Williams et al.’s (1982) upper Allegheny "Freeport" model. Brady et al. looked at the overburden above the upper Kittanning coal in the Stony Fork watershed in Fayette Co., PA. Mine sites with predominantly channel sandstone overburden produced acidic drainage. The sandstone lacked calcareous minerals or cements. Overburden in areas away from the sandstone channels contained calcareous shales and muddy limestone, and mining in these areas resulted in alkaline drainage. The mines and resulting water quality in the Stony Fork watershed are discussed in detail in Chapter 9.

Typically sandstone is low in sulfur, even when acid producing. Channel sands can contain eroded material, including ripped up mats of peat (present day "coal spars") and lag deposits consisting of fossilized, and often coalified, logs. Individually these coal inclusions can be high in sulfur, but during typical overburden sampling this high sulfur is diluted by the surrounding inert quartz and other minerals that compose the sandstone. Thus these sandstones may contain acid-forming material (coal spars), but the samples have low overall sulfur concentrations. Occasionally there are pyrite-rich sandstones. Some of these are black and high in organic carbon, still others are light in color, but high in sulfur. The light-colored high-sulfur sandstones seem to occur just above coal or organic-rich shale. An example of this was shown in Figure 8.13, hole A-5.

Within the Allegheny and Pottsville Groups, sandstones generally contain low concentrations of calcareous minerals. There are some (rare?) exceptions to this however. Figure 8.13 shows two drill logs with calcareous sandstone in the lower Kittanning overburden, drill logs A8 and OB-1 from the Kauffman site in Clearfield County. The calcareous minerals are present as cement in the channel bottom deposits. Examination of thin sections reveals that minor amounts of siderite occur with calcite in the interpore cement (V. Skema, 1996, personal communication). Some calcite at the Kauffman site is also associated with fractures. Most of the sandstone at this mine, however, is not calcareous. Sandstones in the Conemaugh, Monongahela, and Dunkard Groups frequently contain calcareous minerals (Figures 8.2, 8.26, 8.31 and 8.34). Some sandstones have lost their carbonates through weathering.

diPretoro (1986) found a relationship between postmining net alkalinity and percent sandstone. All but one mine site within his study area (northern WV) containing greater than 63% sandstone produced negative net alkaline (i.e., acid) drainage. Sixty-seven percent of sites with less than 30% sandstone had positive net alkalinity (Figure 8.65). All but one of the sites with high percent sandstone were Allegheny Group coal mines. All but one of the Dunkard Group (Waynesburg coal) mines were alkaline. Many of the Waynesburg sites were less than 50% sandstone (in contrast to Waynesburg mines in Pennsylvania, where sandstone is the major overburden lithology, Figure 8.2 and Figure 8.34). Figure 8.66 shows a comparison of net alkalinity, for upper Freeport mines in West Virginia, with <50% sandstone and those with >50% sandstone. Median net alkalinity for sites with less than 50% sandstone is 105 mg/L, and for sites with more than 50% sandstone it is -172 mg/L.

An examination of 41 mine sites in western Pennsylvania by the Department of Environmental Protection (DEP) and the Office of Surface Mining (OSM) also shows relationships between percent sandstone and water quality (Figure 8.67). This data is somewhat biased in that, for the most part, it represents sites that were predicted to produce alkaline drainage (in theory acid-producing sites would not have been issued because of regulatory requirements), therefore acidic sites are probably underrepresented. Acidic sites are

Figure 8.65 Net alkalinity as a function of percent sandstone for surface mines that encountered the Waynesburg, upper Freeport and lower Kittanning coals. Sites are in Preston County (northern), West Virginia. Most sites with greater than 63% sandstone are acidic, and most sites with less than 20% sandstone are alkaline. Data are from diPretoro (1986).

present however, and make up 36% of the mines studied. Despite this bias toward "good" mines, we feel several observations can be made. Twelve of 13 (93%) mines with less than 20% sandstone are alkaline. Five of 8 mines with greater than 63% sandstone have net alkaline water, which is in contrast with diPretoro’s findings where nearly all sites with >63% sandstone produced acidic water. A reason for this difference is that four of the alkaline, high-percent sandstone sites in the DEP/OSM study are in the Monongahela and Dunkard Groups. As Figure 8.2 shows, calcareous sandstones occur in this stratigraphic interval, but are generally lacking in the Allegheny Group. The DEP/OSM data underrepresents high percent sandstone Allegheny Group sites, because experience has shown that they frequently produce acidic drainage, thus the permits were not issued.

The data from the two studies (diPretoro, 1986; and the DEP/OSM study) have not been combined into one data set because they employed different methods of "weighting" to determine percent sandstone. Separation of the studies allows comparison of data from two adjacent, but separate areas. Both studies show that when there is a low percentage of sandstone the mine drainage is generally alkaline (Figure 8.65 and Figure 8.67). Analysis of the DEP/OSM data suggests this is not related to higher NPs in the finer-grained rocks, nor to lower sulfur in these rocks. Some of the low-percent sandstone overburden, in fact, has higher sulfur than the high percent sandstone overburden (Figure 8.68).

Figure 8.67 Net alkalinity as a function of percent sandstone for 41 surface mines in western Pennsylvania. Mines in the Conemaugh, Monongahela and Dunkard Groups are all alkaline, regardless of percent sandstone. Most sites with less than 20% sandstone are alkaline.

Figure 8.68 Maximum potential acidity as a function of percent sandstone. There is essentially no relationship between the two variables, with the possible exception of higher MPA at low percent sandstone. Data are from 41 surface mines in western Pennsylvania.

Also there is no relationship between net neutralization potential and percent sandstone (Figure 8.69). The finer-grained rocks would contain more clay minerals

than sandstone and these clay minerals may contribute to acid neutralization through silicate weathering, ion exchange, and adsorption of H⁺ on negatively charged clay surfaces, thus better quality water.

There are probably several reasons Allegheny Group overburden with a high sandstone content produces acidic drainage. As mentioned, sandstones in the Allegheny group typically do not contain calcareous cements or minerals. This is probably related to paleoenvironmental influences that were not conducive to calcareous mineral deposition. Sandstones, which are composed largely of quartz, lack any appreciable ion exchange capacity and any ability to neutralize acid that would be generated by this process, and thus lack acid-buffering capability. Massive sandstones can have small discrete zones with high pyrite content (such as coal stringers) that may be important in acid mine drainage (AMD) generation. The importance of these zones may be missed in a sulfur analysis of the rock due to dilution of the zone by the surrounding "inert" sandstone. diPretoro (1986) suggested that low pH water generated by sandstones is favorable for AMD catalyzing bacteria, and sandstone, which during mining breaks into large blocks, allows greater permeability to air and water. The greater permeability to air by sandstone overburden is verified by Guo and Cravotta (1996). A site with blocky sandstone had a minimum O₂ in spoil gas of 18 volume percent (vol %). A second site with shale overburden had a minimum O₂ in spoil gas of less than 2 vol %. The minimum values for both sites were measured at ~10 m below the surface. Another way sandstone can affect overburden is where noncalcareous channel sandstones cut out and replace calcareous strata (Brady et al. 1988). As discussed above, the Monongahela and Dunkard Group sandstones are frequently calcareous.

Although there are certain rules of thumb regarding the relationship between sandstone and mine drainage quality, site-specific information is necessary to accurately predict water quality from a particular mine site. For example, sandstone above the lower Kittanning in Clearfield County is normally not calcareous, but some can be on rare occasions; and the sandstone above the Waynesburg coal which though situated in a dominantly calcareous part of the section is not necessarily calcareous. Sandstone and its relationship to mine drainage quality is stratigraphically specific and may be regionally specific. It would be interesting to compare the findings discussed above for the northern Appalachians with other coal-producing regions.

Figure 8.69 Neutralization potential as a function of percent sandstone. Data from 41 surface mines in western Pennsylvania.

Frequently the highest sulfur strata are coals, boney coals, and other organic-rich rocks. Typically these organic-rich rocks are immediately above, below, or within a coal seam (e.g., a parting). Figure 8.49 shows sulfur content of roof and floor rocks compared to sulfur in the adjacent coal. Low-sulfur coals, such as those in the Powder River Basin and the southern Appalachians, have low-sulfur roof and floor. High-sulfur coals, such as occur in Pennsylvania and the Illinois Basin, may or may not have high-sulfur roof and floor. The sulfur in the associated rocks in some instances is higher than the sulfur in the coal.

Maksimovic and Mowrey (1993) examined the characteristics of the floor and roof materials from over 450 underground and surface coal mines across the US. The majority of immediate roof rocks were fine-grained sediments, shale, limestone, and fireclay. Sandstone only made up a minor fraction of the immediate roof rock. Sandstone was more significant a component in the main roof, making up 50% of deep mine roofs, but only 29% of surface mines. This difference may possibly be because deep mines are preferentially developed where the roof is sandstone because of more stable roof conditions, whereas surface mines may avoid areas with massive sandstone bodies because of increased mining costs. In all cases, the immediate floor is largely composed of fine-grained sediments. The fine-grained rocks associated with the coal, whether roof rock, floor rock, or a parting, are often transitional between coal and the rock types above and below the coal. This transitional nature resulted in rocks that are high in organic matter, and probably formed in a higher pH environment than existed in the coal swamp. These conditions, as previously discussed, can be conducive to pyrite formation, since organic matter is a necessary ingredient for sulfate reduction.

The proximal relationships between high-sulfur strata, and coal or rider coal seams can be seen in virtually all the drill logs depicted in this chapter that show sulfur data. An exception to the high sulfur occurring immediately above the coal can be seen in Figure 8.13, where the highest sulfur interval occurs 3 ft (1 m) or more above the low sulfur overburden which overlies the coal. The high-sulfur zone (1.15% in A-8, 0.92% in OB-1, and 4.4% in OB-4) is associated with the brackish fossil Lingula. This is consistent with the work of Guber (1972) who found that shales and siltstones overlying the lower Kittanning coal containing the phosphatic brachiopod Lingula had higher sulfur than rocks not having Lingula (Table 8.11). Morrison (1988) also found on average that brackish shales had higher sulfur (2.40%) than marine (0.95%) or freshwater (0.15%) shales. Our experience confirms that brackish shales are typically high in sulfur.

Table 8.11 Contingency table of observed frequencies of Lingula- and non-Lingula-bearing rocks by three sulfur classes. The rocks are from the interval above the lower Kittanning coal, Clearfield County. Data are from Guber (1972).

High-sulfur strata are also associated with marine shales such as the lower Kittanning and Brush Creek overburdens shown in Figures 8.11, 8.12, 8.23 and 8.24. Many of these same shales are calcareous. The top and bottom of some of these shale facies may be brackish. The thickness of high-sulfur zones in marine and brackish environments is often greater than that in the conchostracan-bearing overburden. It is important to accurately account for this high-sulfur material when designing plans to prevent acid drainage problems, and when predicting postmining water quality.

Four geologic processes account for overburden mineralogy. Two of these, paleoclimate and paleodepositional environment, date back hundreds of millions of years to the Pennsylvanian Period. The other two, surface weathering and glaciation, are more recent, occurring within the past few million years. Paleoclimate can account for the mineralogy of some terriginous rocks, but probably has little, if anything, to do with the paleodepositional controls that influenced pyrite and carbonate content of marine/brackish sediments. Weathering can reduce the original carbonate and sulfur content of rocks by dissolution and oxidation processes respectively. Glaciation can influence overburden by erosion of weathered bedrock and by the deposition of calcareous tills. The mineralogy of the rocks, particularly the presence of carbonates and pyrite, ultimately influences postmining water quality.

Earlier discussions argued that the calcareous carbonates are more important in controlling water quality from surface mines than is pyrite. The presence of only 1 to 3% carbonate minerals can influence whether acidic or alkaline drainage is produced (Brady et al., 1994; and Perry and Brady, 1995). The amount of sulfur present is not directly related to acid production except in the absence of calcareous strata.

Despite early recognition of the importance of calcareous strata (Leitch et al., 1932 and Crichton, 1923), most subsequent research has concentrated on studying the reactivity of various forms and features of pyrite. This seemed logical, because pyrite is required for the formation of acid mine drainage. The number of studies on pyrite reactivity is in the hundreds if one includes kinetic (leaching) tests. A few representative examples on relative pyrite activity are Caruccio’s studies of pyrite morphology (1970) and trace element content (1972), Hammack et al.’s (1988) studies of pyrite origins (genesis); and Scheetz’s (1992) investigations of pyrite molecular structure. Chapter 7 lists hundreds of references for various studies on "kinetic" tests. Few comparable studies have been done on the role of carbonates on mine drainage neutralization and inhibition, despite the clear role they play.

The complexity of carbonate mineralogy, with extensive solid solution substitution, and the role of this variable on resultant mine drainage chemistry is poorly

understood. It is known that the NP test (see Chapters 6 and 11) does not adequately distinguish forms of carbonates (Skousen et al., 1997b). The iron carbonate "siderite" may be temporarily alkalinity producing, but with oxidation and precipitation of the iron component, it is ultimately not alkalinity producing (Cravotta et al., 1990). Possible exceptions are where significant solid solution substitution for the iron, by Mg or Ca occur, or where siderite is mixed with a calcareous carbonate. A revised NP procedure reduces the falsely high readings for NP when siderite is present (Skousen et al., 1997b). The work by Skousen et al. is a follow-up of studies done by Morrison et al. (1990a; 1990b; 1994). The field studies by Brady et al. (1994) and Perry and Brady (1995), as well as references contained therein, suggest that not only do calcareous minerals neutralize mine drainage, but they also inhibit pyrite oxidation. Laboratory studies are consistent with these findings (e.g., Williams et al., 1982). The role of carbonate mineralogy on mine water chemistry is clearly a subject deserving more study.

Others factors, in addition to mineralogy, such as mining methods, mine site hydrology, and controls on spoil pore gas composition can also significantly influence postmining water quality. The role of these variables (and others) is discussed in Chapters 1, 3, 12, 13, 14, 15, 16, and 17.

Several studies have examined water quality as it relates to stratigraphic position of coal seams. Most surface mines in Pennsylvania have extracted coals in the Allegheny and Monongahela Groups. Table 8.12 shows comparisons of pH for upper Pennsylvanian (Conemaugh and Monongahela Groups, plus the Permian? Dunkard Group) coal mines, and the upper Allegheny and lower Allegheny Group. Two of these studies looked at water quality from deep mines in Pennsylvania four studies looked at water quality from surface mines in Pennsylvania and West Virginia, and one study included water quality from both surface and deep mines. The last line of Table 8.12 is the mine drainage quality data from Table 8.2 which have been adapted and summarized for comparison with the other studies. Samples of streams and water supply wells have been deleted from our summary.

Deep mines, as discussed in Chapter 9, can produce poorer quality water than surface mines on the same coal seam, if they are "free draining." The deep mines included in these studies are "pre-Act" and mostly up-dip mines. The study by diPretoro (1986) had the lowest percentage of alkaline sites from the upper Allegheny. This is probably because a high percentage of these sites had predominantly sandstone overburden. The DEP/OSM study and this study have the highest percentage of alkaline sites for the Lower Allegheny mines. This is probably because these studies include a bias toward sites that would produce acceptable water quality, bad permits being denied (and thus no postmining water quality). As a generality, the upper Pennsylvanian and upper Allegheny mines in Pennsylvania produce more alkaline drainage than lower Allegheny mines.

As discussed above, the water quality reflects the geologic processes that have been at work. The upper Pennsylvanian/Permian? strata has freshwater limestones and calcareous shales. The upper Allegheny, with the exception of northern West Virginia, generally has calcareous shales and freshwater limestones. The lower Allegheny is frequently marine or brackish, and often lacks calcareous strata. The Emrich and Thompson (1968) study also tried to look at regional trends in water quality by coal seam. They found that lower Kittanning deep mines in Armstrong and Jefferson Counties frequently produced alkaline drainage. We have found that surface mines in this same area also generally produce alkaline drainage (area of Figure 8.12), and this is discussed in Chapter 9.

A side by side comparison of Figure 8.2 and Table 8.2 (both in pocket at rear of book) shows the relationships between stratigraphic intervals and water quality of the bituminous coal region of Pennsylvania. Table 8.2 can also be compared to total sulfur and NP data of drill holes shown in Figure 8.4 through Figure 8.34, for a more detailed evaluation of the relationships between stratigraphic characteristics of specific surface mine sites and the postmining water quality of those sites. Alkalinity greater than 100 mg/L and acidity greater than 100 mg/L are shown in bold, thus highlighting water samples and stratigraphic intervals with high alkalinity and acidity. The remaining paragraphs of this section describe the relationships from the Waynesburg Formation of the Dunkard Group down to the Pottsville Group, as the data is presented in Table 8.2.

Dunkard Group - The only coal commonly mined in the Dunkard Group is the Waynesburg coal. The postmining water quality from surface mining of the Waynesburg Formation can be highly variable (Table 8.2; in pocket at back of book). The water samples of the Waynesburg interval (Table 8.2) are from D.S. Jones and P. Cestoni (1991, DEP hydrogeologists, personal communication). Samples are from a 2 x 2.5 mile (12.8 km²) area in Greene County. Most samples were collected at the Susan Ann site, and represent water before, during and after surface mining activities. Surface mining began in 1980 and was completed in 1991. The first two samples, collected prior to mining, have alkalinity of 379 mg/L and 340 mg/L, and are from low flow seeps (<1 gpm; <4 L/min). These emanate from a small abandoned deep mine (DM-2) on the Waynesburg coal. The next two samples (Table 8.2) have acidity of 1500 mg/L and 306 mg/L. These are from low flow seeps adjacent to the sealed entry of DM-2 and were collected during mining. The fifth sample, from the same location, has alkalinity of 92 mg/L, iron of 18 mg/L and manganese of 29.4 mg/L. This sample was collected after backfilling and revegetation in 1991. The next sample in Table 8.2 is from a small underground mine (DM-1). It has an alkalinity of 142.0 mg/L and significantly lower metals concentrations then the preceding sample. The flow was ~5 gpm (~19 L/min.) The two samples labeled "F" in Table 8.2 have high alkalinity, relatively high metals concentrations and emanate from spoil. The remaining Dunkard Group samples are from tributary streams. All have alkalinity greater than 250 mg/L.

The relationship between lithology and water quality of the Waynesburg Formation is evident from a comparison of Table 8.2 to Figure 8.34. Data in Figure 8.34 are from the Susan Ann mine and a nearby site called Fox. There are significant thicknesses of calcareous sandstones, siltstones and shales with NP values greater than 50 ppt. These account for the high alkalinity concentrations in postmining water quality. There are also significant amounts of strata with high sulfur (>0.5%). These include the Cassville shale above the Waynesburg coal (e.g., 1.39% S in Fox # 3) and the high-sulfur sandstone in hole PS-3. These strata account for the high acidity and metals concentrations in some mine discharges.

Monongahela Group - The water quality of the Monongahela Group is shown in Table 8.2. The lithologic and geochemical characteristics of the Group are depicted in Figures 8.2, 8.26, 8.31 and 8.32. Freshwater limestones including the Benwood, Fishpot, and Redstone, and other calcareous strata are abundant in the Monongahela. For example, Figure 8.32 shows drill holes with thick limestones and other high NP strata associated with the Sewickley and Redstone coals. As would be expected from the abundance of carbonate strata, the water quality associated with all but the lowest part of the Pittsburgh Formation is highly alkaline with relatively low metals concentration (Table 8.2). In fact, the sample from Upper Tyrone Township, Fayette County of a spring near the cropline of the Redstone coal has an alkalinity of 626 mg/L, the highest of any samples found in DEP permit files. While the metals concentrations of this sample are relatively low, the sulfate concentration is 1440 mg/L, making it the dominant anion, rather than bicarbonate.

The abundance of carbonate strata in the overburden of the Redstone coal, along with the high alkalinity in natural background water, and the ubiquity of alkaline mine drainage associated with this seam in southwestern Pennsylvania has resulted in overburden analysis rarely being needed to make permit decisions. Thus acid-base accounting data for the Redstone are rare.

Farther east, in Somerset County, the geologic structure and stratigraphic correlations become more complex, and overburden analysis data from the Monongahela Group is more common. For example, the Blue Lick Coal is routinely mined in Somerset County. This is a split or rider of the Pittsburgh seam which is not mineable in counties to the west (Shaulis, 1993). The generally high alkalinity of the Blue Lick 3 and Krause sites (Table 8.2) are representative of the postmining water quality of the Morantown to Sewickley coal overburden in this area. Drill holes OB-A, OB-C, and OB-5 in Figure 8.26 are examples of the overburden for this interval in Somerset County.

Some active and abandoned, surface and underground mining activities on the Pittsburgh coal have produced serious acid mine drainage problems (see water samples from Allegheny and Westmoreland Counties in Table 8.2). This scenario occurs in areas where the Pittsburgh coal is mined alone (see OB-1 and OB-2, Figure 8.31). There are a few calcareous strata in these drill holes, but the high-sulfur (> 1%) shales and sandstones overlying the coal apparently overpower the calcareous strata, thus the production of AMD. This type of AMD scenario also occurs where calcareous Redstone and Sewickley strata are present, but mining does not encounter them. For instance, where deep mines only disturb high-sulfur strata adjacent to the coal, or where a surface mine disturbs only low cover strata that have had the carbonates removed by weathering. A classic example of where a remining operation was able to improve water quality by mining a greater amount of cover and thus encountering calcareous strata, and improve mine drainage quality, is the Solar site discussed in Chapter 9 (and in Skousen et al 1997a). The discharge from a large Pittsburgh deep mine improved from a pH of 2.5 prior to remining to a pH of 7.5 following remining.

Conemaugh Group - The postmining water quality of surface mines within the Conemaugh Group is shown in Table 8.2. The lithology and geochemistry of the overburden are depicted in Figure 8.24 and Figure 8.25. The earlier discussion of Conemaugh stratigraphy documented that this thick sequence of rocks is generally "barren" of mineable coals. Few coals are mined in the Casselman Formation. The only part of the bituminous region where these coals reach mineable thicknesses is in extreme eastern Somerset County. The coals that have been mined are the Morantown (which is actually a split of the Pittsburgh coal and thus in the Monongahela Group), the Wellersburg coal (Figure 8.2 and Figure 8.25), the Barton, and Federal Hill coals. The last two are rarely mineable. Postmining water quality for the Wellersburg overburden is shown on Table 8.2. The alkaline nature of the water is consistent with the calcareous nature of the overburden depicted in Figure 8.25.

The Glenshaw Formation contains several mineable coals and numerous marine and freshwater limestones (Figure 8.2). The principal coals mined are the upper and lower Bakerstown and the Brush Creek, and the water quality associated with surface mines for these coals in Somerset and Fayette Counties are shown in Table 8.2. Drill holes DH-3 and DH-12 in Figure 8.25 are from Southampton Township, Somerset County and water from the same mine is shown in Table 8.2. Three samples of lower Bakerstown pit water and two samples of upper Bakerstown pit water are included in Table 8.2. All have alkalinity greater than acidity (the most alkaline of these is 118 mg/L from a lower Bakerstown pit pool). A postmining discharge from the Stateline site has alkalinity as high as 210 mg/L and a pH of 8.1. This demonstrates that the thick calcareous strata shown above the lower Bakerstown in DH-12 and DH-13 (Figure 8.25) are capable of producing alkaline postmining water quality, despite the presence of some high sulfur coal and shale (e.g. 9.18% in DH-12). Drill holes 5 and 11F in Figure 8.25 are from nearby Brothersvalley Township. The four water samples from the Brothersvalley site are of preexisting discharges with moderate acidity and considerable manganese and aluminum concentrations. These samples were collected prior to the remining operations on the site. These water samples show that acid discharges can occur on some mines where the overburden quality is generally good (Drill Holes 5 and 11F). It is not known whether the water quality following remining improved.

The last six water samples from Conemaugh strata in Table 8.2 are from two sites in Wharton Township, Fayette County where the Brush Creek coal has been surface mined without postmining water quality problems. These samples are from the same site as Drill Holes DH-19 and OB1 in Figure 8.23.

Allegheny Group - The postmining water quality characteristics of surface mine sites in Allegheny Group strata are represented by approximately 100 samples in the bottom half of Table 8.2. The coals of the Allegheny Group are the most extensively mined seams of western Pennsylvania and consequently more is known about the relationships of the stratigraphy, lithology, mineralogy, and postmining water chemistry of this stratigraphic interval, from published geologic reports and unpublished data in DEP permit files, than any other coal-bearing stratigraphic interval in Pennsylvania.

The areal extent of mining of the upper Freeport and lower Kittanning coals in western Pennsylvania is large. The magnitude of permit file data for this area can be estimated from DEP data bases: (1) mylar map records, (2) a computer system which tracks all permitting, inspection, compliance, and enforcement actions for all mines active since 1985, and (3) an overburden analysis database that contains rock chemistry analyses for permits issued since 1993, plus some historical records. The computer tracking system includes 1594 mines on the upper Freeport coal (from 1985 to November 1997), for a total of 339,018 permitted acres. The lower Kittanning coal was permitted for 1142 mines covering 305,187 acres (M. Klimkos, 1997, DEP, personal communication). The overburden analysis database contains 291 drill logs with chemical data for the upper Freeport overburden and 336 drill logs for lower Kittanning overburden.

The following description of upper and lower Allegheny Group water quality will be brief and broadbrush. Table 8.2 provides details for water samples from specific mine sites. The postmining water quality characteristics of some Allegheny Group surface mines has also been described in detail in Chapters 9, 10, and 18 of this book.

Upper Allegheny - There is a relationship between the stratigraphy and postmining water quality of the upper Allegheny Group (stratigraphy shown in Figures 8.2, 8.17 through 21, and water quality shown in third panel from the bottom of Table 8.2). The three freshwater limestones of the upper Allegheny (Johnstown limestone, lower and upper Freeport limestones) and associated coal, calcareous shales, siltstones, sandstones, and underclays are shown in these drill holes with relatively high NP values. Most of the water samples for this stratigraphic interval in Table 8.2 have alkalinity concentrations greater than 100 mg/L and several samples are greater than 300 mg/L. For example, the Laurel Hill # 1 and # 2 sites in Jackson Township, Cambria County are the same sites as the four drill holes shown in Figures 8.17 and 8.18, and the toe of spoil seep with a pH of 8.1 has an alkalinity of 484 mg/L. The Shero site in West Carroll Township, Cambria County, has postmining water quality with an alkalinity of ~375 mg/L. This site is in close proximity to the Fisanick site shown in Figure 8.19. The Fruithill, Witherow, Beaver # 1, and Fink/Mays surface mine sites from Clearfield County also exhibit relatively high postmining alkalinity for seeps, spoil discharges, and abandoned deep mines that are hydrologically connected to the surface mines, and in headwater tributary samples. The Nashville site in Indiana County and the Schmunk and Rennie sites from Fayette County in Table 8.2 have similar alkalinity and low metal concentrations in postmining seeps, springs, and tributary samples for the upper Allegheny Group strata. However, the Chanin, Morrison, and Stuart sites in Fayette County, particularly in the upper Kittanning coal, exhibit dramatic variations from high alkalinity to high acidity postmining discharge (Table 8.2) corresponding to facies changes where the typically calcareous overburden strata are replaced by thick channel sandstones lacking carbonate minerals. The Morrison and Stuart sites are in the same vicinity of Wharton Township, Fayette County. The dramatic difference is overburden quality is depicted in Figure 8.20 where drill holes BM-OB2F and BM-OB5F from the Morrison site possesses high NP overburden strata, while the drill holes from the Stuart site are generally lacking high NP strata, especially drill holes GBS-DH2, which is entirely sandstone overburden with total sulfur as high as 2.6%.

Lower Allegheny - The relationships between the stratigraphy and postmining water quality of the lower Allegheny Group can be determined by comparing the second panel from the bottom of Table 8.2 with Figures 8.2, 8.7, 8.12, 8.13, 8.15 and 8.39. The lower Allegheny lacks the freshwater limestones and associated calcareous shale which are characteristic of the upper Allegheny; therefore the postmining water quality of the lower Allegheny is often more acidic as displayed in Table 8.2. However, the marine Vanport limestone in the lower Allegheny Group is the thickest and most commercially significant limestone unit in the Pennsylvanian rocks of the state, and in the counties where it is of appreciable thickness (e.g. Armstrong, Butler, Clarion, Lawrence and Mercer, see Figure 8.8), high alkalinity concentrations occur in mine drainage, springs, wells and streams as shown in Table 8.2 (e.g., Graff and Wilson sites).

The water quality associated with the surface mining of the MK coal reflects the transition in stratigraphy from the lower to the upper Allegheny Group (see analyses in Table 8.2). The overburden strata between the MK and UK coals typically includes the Washingtonville shale overlying the MK coal which is stratigraphically the highest marine or truly brackish bed in the formation, and the Johnstown limestone underlying the UK coal, which is stratigraphically the lowest freshwater limestone in the formation. In a multiple seam mining operation having a bench for each coal, the MK pit water may have considerable alkalinity because the Johnston limestone is the dominant influence on water quality, overwhelming the acid from the acidic shales and siltstones (Figures 8.13 and 8.15).

Lower Kittanning overburden contains no true limestones. The characteristics of the overburden chemistry and postmining water quality of this stratigraphic interval are quite variable as shown in Figure 8.12 through Figure 8.15 and Table 8.2. More is known about paleoenvironmental and mine drainage quality variations for this interval than for any other coal in Pennsylvania. This is due to the large amounts of data in DEP permit files and numerous geologic studies (e.g., Williams, 1960; Williams and Keith, 1963;, Guber, 1972; Hornberger et al., 1981, 1985; Williams et al., 1982; and Morrison et al., 1990a). In some areas where the lower Kittanning overburden was deposited in a marine environment as shown in Figure 8.12, the mine drainage can exhibit appreciable alkalinity. For example, the Snyder site in Table 8.2 is the same site as drill holes OB-1 and OB-2 in Figure 8.12, and the pit water had an alkalinity of 114 mg/L, despite the presence of sulfur content greater than 1% in several shale strata. (The Snyder site is also featured in Chapter 18 of this book). Additional mine drainage data from a marine environment is contained in Table 8.2 for the Shannon site in Clarion County. The lower Kittanning overburden from brackish environments is generally characterized by black shales overlying the coal. These shales can have very high sulfur content, for example 8.34% in drill hole 560A3 in Figure 8.13. Similar holes were drilled on the Kauffman site in Figure 8.13 (e.g. drill holes OB-1 and OB-4). Comparable acidic postmining water is shown in Table 8.2 for the adjacent Betty site. Water data from the Kauffman site was not used because site reclamation is incorporating imported alkaline material, thus postmining water from it is not necessarily representative of the overburden. Additional postmining discharges with high acidity, metals and sulfate concentrations from brackish environments of the lower Kittanning are shown for the Albert #1 and Kelce sites in Clearfield County. Geologic description and water data for brackish sites in the Luthersburg area of Clearfield County are presented in Chapter 9.

Lower Kittanning overburden from a "freshwater" paleoenvironment (probably marginally brackish) is depicted in drill holes 61-B and 61-C in Figure 8.15. Postmining seeps from the Swiscambria site have moderate acidity and metals concentrations. These LK "freshwater" environment sites (also drill holes OB-1 and OB-2 in Figure 8.15), often have shales with lower sulfur contents than the marine and brackish shales, but these sites still tend to produce water with acidity greater than alkalinity. Finally, some lower Kittanning sites may have overburden strata principally comprised of thick channel sandstones. These often produce drainage with high acidity and metals (e.g., Lawrence site from Fayette County; Table 8.2).

Postmining water quality variations of surface mines on the Clarion coals are probably the most complex of any coal mined in Pennsylvania because three of the four geologic factors that account for overburden mineralogy (i.e., paleoenvironment/stratigraphy, glaciation, and surface weathering) can be operative and significant within the same county (e.g., Clarion County). The result is good and poor quality mine drainage in the same vicinity. Stratigraphic correlations and nomenclature of these lower Allegheny coals can be confusing as discussed earlier.

Often the total sulfur content of the Clarion coals and the overlying shales are high, as shown in drill holes DH23-6 (8.33%) and DH18-1 (5.92%) in Figure 8.7 and drill hole DH-4 (5.16%) in Figure 8.39. Other Clarion coal overburden analyses in DEP files (not included in these figures) commonly have total sulfur greater than 8%. Discharges with >1,000 mg/L acidity related to this type of Clarion overburden strata are common. Examples shown in Table 8.2 are the Old 40 and Zackerl sites in Clarion County, the Orcut site in Jefferson County, and the Philipsburg site in Centre County. Where the Vanport limestone is present, the mine drainage quality can be dramatically different as discussed earlier and shown in Figure 8.10.

The influence of unweathered and weathered glacial tills upon overburden mineralogy and mine drainage quality of Clarion coal surface mines is shown in Figures 8.39 and 8.60 and Tables 8.2 and 8.3. Drill hole OB-2 in Figure 8.39 from the Cousins site in Lawrence County is an example of an unweathered till zone (i.e., zone 5 in Figure 8.60) with NP values > 500 ppt CaCO₃. These high values are due to clasts of the Vanport limestone. The pit water sample from the Cousins site on Table 8.2 has an alkalinity concentration of 130 mg/L, while spring samples from the site (Table 8.2) have higher alkalinity concentrations. Drill hole OB-3 - SPAG (Figure 8.39) is from the Spagnolo site shown in Table 8.3. This drill hole encounters till with several zones of NP greater than 30 ppt. This is indicative of zone 4 (oxidized, calcareous till) in Figure 8.60 and the water quality at the Spagnola site has a pH greater than 6.5 and an alkalinity of approximately 30 mg/L (Table 8.3). Most of the glacial till shown in DH-4 and DH-28 from the Pike Township site in Figure 8.39 appears to be weathered and non-calcareous (see Zones 2 and 3), but the thick Vanport limestone below the till would promote alkaline water at this site without influence from the till.

Pottsville Group - The Pottsville Group coals are generally thin and discontinuous. While generalized columnar sections of the Group show the Sharon, Quakertown and Mercer coals, the only Pottsville Group coal that is mined to any reasonable extent in the bituminous region is the Mercer coal zone. A few surface mines in Somerset County are mining a coal labeled the Quakertown but the correlations are tenuous or doubtful. Recent surface mining of the Mercer coals is largely restricted to an area of northwestern Pennsylvania including portions of Clarion, McKean, Lawrence and Mercer Counties. Since there has been so little mining activity on Pottsville Group coals, very little useful data exists. Consequently the description of the relationships between Pottsville Group overburden chemistry and water quality characteristics has been reduced to a comparison of the drill logs for two proposed mine sites that were not mined, (shown in Figure 8.6), to some background water samples from that area, plus some recent and abandoned mine drainage samples of Pottsville Group mines shown in the bottom panel of Table 8.2.

The five samples of old mine discharges for the Mercer coal site in Norwich Township, McKean County are from the same site as drill holes DH 177-1 and DH 177-9 shown in Figure 8.6. The highest alkalinity for these samples is 7 mg/L, and the highest acidity is 73 mg/L; some of these samples have nearly equal acidity and alkalinity values. The two water samples in Table 8.2 from the Mine #201 site, are from the same site as drill holes DH 201-1, DH 202-2 and DH 201-3 in Figure 8.6. These samples are background (premining) springs off the Mercer coal outcrop and exhibit similar alkalinity, acidity, and sulfate concentrations to the Mine #177 background samples described above. The four pit water samples of the Matthews site on the Mercer Coal in Mahoning Township, Lawrence County, have significant alkalinity concentrations (e.g. 206 mg/L) and relatively low metals concentrations. However, the high alkalinity is chiefly attributed to carbonate minerals in the till overburden.

The mildly acidic to highly alkaline water quality available from the recent Mercer coal sites may tend to deceive some readers because it does not represent the full range of mine drainage quality emanating from the Mercer coal zone in Pennsylvania. Therefore, the abandoned mine discharges from the Horseshoe site in Cambria County were included in Table 8.2 to represent the significant acidity and metals concentrations that are possible from Mercer coal mine discharges. The Louden (MP-29 and MP-22) and Page (MP-65) discharges are from large abandoned underground mine complexes on the Mercer coal near Altoona, PA. The Louden discharge has acidity and iron as high as 1835 mg/L and 194 mg/L respectively. The Quakertown coal pit water sample from an active surface mine in Somerset County, at the bottom of Table 8.2, is the only water sample found in DEP files for this interval.

Following is a summary of postmining water quality variations in Pennsylvania attributable to geologic factors. Regional- and local-scale variations in postmining water quality for the bituminous and anthracite mining regions are evident in data from this study and previous studies.

A bimodal distribution of coal mine drainage pH has been observed within both the bituminous and anthracite regions (Figure 8.70a and Figure 1.2, Chapter 1, for bituminous coal mine drainage and Figure 8.70b for anthracite mine drainage). Brady et al. (1997) state: "Although pyrite and carbonate minerals only comprise a few percent (or less) of the rock associated with coal, these acid-forming and acid-neutralizing minerals, respectively, are highly reactive and are mainly responsible for the bimodal distribution…. Depending on the relative abundance of carbonates and pyrite, and the relative weathering rates, the pH will be driven toward one mode or the other." Chapter 1 discusses the pH distribution for the bituminous region.

Water Quality in the Bituminous Coal Region -The large-scale variations in water quality within the Bituminous Coal Region are principally related to the outcrop patterns of the Pottsville, Allegheny, Conemaugh, Monongahela and Dunkard Groups, and to paleoenvironmental changes of specific stratigraphic intervals at a regional scale. The large-scale water quality variations are chiefly attributable to the presence or absence of calcareous rocks. A good example of delineating mine drainage quality by outcrop area is the Monongahela Group. Surface mining most Monongahela Group seams results in alkaline drainage (an exception is acid problems from some Pittsburgh deep mines). The alkaline drainage occurs because of the abundance of thick freshwater limestones that are encountered during mining. Good examples of regional-scale paleoenvironmental variations occur within the lower Allegheny Group. Some examples are where: (1) the marine Vanport limestone is thick and overlays the Clarion coals, as opposed to areas where the Vanport limestone is absent; and (2) marine shales overlying the lower Kittanning coal to produce alkaline mine drainage, as opposed to lower Kittanning shales that are of a brackish or "freshwater" environment resulting in acidic drainage.

By comparing Table 8.2 to Figure 8.2 (both in pocket at back of book), relationships between the geology and water quality of the bituminous region can be discerned. The highest alkalinities in Table 8.2 are associated with: (1) the thick freshwater limestone sequences (e.g., the Redstone limestone) of the Pittsburgh Formation in the southwestern corner of Pennsylvania, (2) the thick marine Vanport limestone of the lower Allegheny Group in the central to northwestern area of the bituminous region, and (3) the freshwater limestones of the upper Allegheny Group (i.e., Johnstown, lower and upper Freeport Limestones) in areas where these limestones and associated calcareous shales are abundant, such as portions of Clearfield, Cambria, Indiana, Westmoreland, and Fayette Counties. In addition, significant alkalinity concentrations in Table 8.2 occur in: (1) the Conemaugh Group where the marine Brush Creek and Woods Run limestones are present, and (2) the lower Allegheny Group mines in northwestern Pennsylvania where the overburden includes significant thickness of calcareous glacial till.

The highest acidity concentrations in Table 8.2 are associated with: (1) the Clarion and lower Kittanning coals, particularly where black brackish shales and thick marine shales predominate in the overburden, such as Clarion, Clearfield, and Centre Counties, and (2) in the Pittsburgh coal and Waynesburg coals of the southwestern portion of the bituminous region, in areas where the carbonate strata are absent or lacking appreciable thickness. Significant acidity concentrations are also shown in Table 8.2 where local facies changes occur, such as in the upper Allegheny Group in areas of Fayette County, where the channel sandstones occur instead of freshwater limestone and calcareous shales.

The medians and ranges of pH, alkalinity, acidity and sulfate concentrations of the major stratigraphic intervals of the bituminous region are shown in Table 8.13. This table has been constructed using data in Table 8.2. Most stream samples were not included from Table 8.2, unless the water reflected drainage from the mine site (e.g., headwater streams during baseflow conditions). Well samples were also deleted, unless they were monitoring wells within mine spoil. Where multiple samples of the same mine discharge point occur in Table 8.2, a median was computed for inclusion in Table 8.13. Unrepresentative premining samples were also deleted. The number of samples (N) shown in Table 8.13, is the number of mine discharge locations used to compute the medians and ranges shown.

Comparison of median pH values in Table 8.13 shows a lower pH for the lower Allegheny Group than for the stratigraphically higher strata. The median pH values for the upper Allegheny, Conemaugh, Monongahela, and Dunkard Groups are within the upper mode of the bimodal frequency distribution for pH shown in Figure 8.70a. These pH values are reflective of the importance of carbonate strata in controlling mine drainage quality. The wide range of pH values and the range of alkalinity and acidity concentrations for the intervals document that some strata within each major interval have the potential to produce alkaline and acidic drainage. The relatively higher median sulfate concentration for the lower Allegheny Group, indicates that this strata has a greater potential for acid production. That is consistent with the extreme acidity and sulfate concentrations in lower Allegheny Group samples in Table 8.2. Many of these samples are also marked by high concentrations of iron, manganese, aluminum, and sulfate.

Anthracite Region Water Quality - Regional variations in mine drainage quality of the Anthracite Region are shown in Table 8.14. The relationships between the postmining water quality and specific stratigraphic intervals of the Anthracite Region are much less well known than those of the Bituminous Region for at least two reasons: (1) the complexity of the geologic structure has impeded the acquisition of stratigraphic data from routine exploration drilling and made correlations of units and associated mine drainage difficult; and (2) a large portion of the mining hydrology of the four anthracite fields is controlled by large-volume, mine pool discharges. The mine drainage from gangways developed in multiple coal beds is commingled in rock tunnels (that crosscut the geologic structure and strata) which interconnect the mine workings. Thus discharges are often a composite water representing multiple coal seams throughout a large mine complex. Despite this, some significant regional variations in mine drainage quality are evident for the anthracite fields (Figure 8.71). These are probably related to mineralogic differences between the fields.

Some Northern Anthracite Field mine waters have significant alkalinity (e.g., Plains Borehole, Table 8.14). This may be attributable to the presence of marine and freshwater limestones and other calcareous rocks in the Northern Field. A few postmining dis-charges of the Northern Field have low pH and high acidity (Loomis Bank discharge), although high acidity discharges are relatively rare in the anthracite fields. Many large volume discharges of the Northern Field have circumneutral pH with nearly equal concentrations of acidity and alkalinity (e.g., Jermyn Outfall,Table 8.14). Although cirumneutral, some of these discharges have relatively high concentrations of iron, manganese or aluminum, and because of large flows they have high pollution loads.

Figure 8.71 Boxplots showing differences in (A) pH and (B) sulfate for discharges from the four anthracite fields in eastern Pennsylvania. The box is bounded by the 25^th and 75^th percentiles and the median is represented by a horizontal line within the box. "Whiskers" and "*" (outliers) define the range. Data are from Growitz et al. (1985).

There are 14 major discharges in the Eastern Middle Field. Mine drainage from four of these are shown in Table 8.14. There is no significant alkalinity in any of the discharges. As far as is known, there are no limestones or other calcareous strata in this region. The highest alkalinity is 13.8 mg/L and the highest acidity concentration is 194 mg/L. No severe AMD (pH < 3.0, acidity > 1000 mg/L) is known in the Eastern Middle Field. The Eastern Middle Field appears to lack both calcareous rocks and high-sulfur rocks. The Jeddo Tunnel discharge, in the Eastern Middle Field, is the largest mine discharge in the state (Table 8.14), and generally has an acidity concentration >100 mg/L and a flow >50,000 gpm (>190 m³min^-1). Though the acidity concentration is not "high," because of the flow the acid load is large.

The water quality of the postmining discharges of the Western Middle and Southern Anthracite Fields is somewhat more mysterious than that of the Northern and Eastern Middle Fields. Some discharges have significant alkalinity, but no carbonate stratigraphic units have been reported in these fields. The Packer V discharge in the Western Middle Field has alkalinity of 160 mg/L and iron of 20.9 mg/L in Table 8.14. The Richards discharge near Mt. Carmel has a pH of 3.7 and an acidity of 70 mg/L. Other discharges of the Western Middle Field exhibit circumneutral pH, with iron concentrations of 10.6 to 30.5 mg/L in Table 8.14. Because some of these discharges drain large interconnected underground mines spanning square miles, various anthropogenic sources may also contribute to water quality. However the North Franklin and the Doutyville tunnel discharges are located in a mostly rural area.

Several mine discharges of the Southern Anthracite Field have significant alkalinity concentrations, including the Wadesville, Eagle Hill, and Kaska discharges (Table 8.14)). In fact the water pumped from the Wadesville shaft has an alkalinity of 414 mg/L. This is one of the most alkaline mine waters found in Pennsylvania. It is almost certain that a detailed study of stratigraphy in this area would reveal calcareous strata or calcareous secondary mineralization. Several Southern Field discharges have significant acidity concentrations (Bell, Newkirk, Porter Tunnel and Markson discharge, Table 8.14). A study by C.R. Wood (1996) shows that many abandoned underground mine discharges in the anthracite fields have improved in water quality between 1975 and 1991.

A final factor that may affect the relationships between postmining water quality and stratigraphy in the Anthracite Region is the stratification of mine pool water. The mine pools consist of water accumulated in void spaces within abandoned underground mines, and deep pools or lakes in abandoned surface mines that are hydrologically connected to abandoned underground mines. These mine pools typically are chemically stratified into "top water" and "bottom water." The stratification of anthracite mine pools is discussed in Barnes (1964), Erickson et al. (1982) and Ladwig et al. (1984). Additional discussions on the areal extent, volume of impounded water, and interconnections (breeches) between mine pools are contained in a series of studies by Ash (e.g., 1949, 1954).

The top water discharges are typically of circumneutral pH, although some samples in Table 8.14 may have elevated iron, manganese, or aluminum. Top water is believed to reflect shallow groundwater systems, with relatively short residence times, where most of the flow is confined to the upper part of the mine pool. The bottom water typically has higher concentrations of acidity, metals, and sulfate than the top water of the same mine pool. Bottom waters are indicative of longer residence times, less circulation (and less oxygen). For example, the Markson and Good Spring No. 1 mine pool discharge samples shown in Table 8.14 are from adjacent collieries within the Donaldson Syncline in the Southern Anthracite Field. The mine maps of these two collieries indicate that the coal seams mined, mining engineering factors, and geologic conditions of the collieries are essentially the same; yet the Good Spring No. 1 discharge has a pH of 6.2 (and sulfates of 112 mg/L) and the Markson discharge has a pH of 3.4 (and sulfates of 491 mg/L) (Table 8.14). The Good Spring No. 1 discharge is top water and the Markson discharge is bottom water with a distinct hydrogen sulfide aroma. The samples of the Markson and Good Spring No. 1 mine pool discharges were collected on the same date in relatively low flow conditions and are within a few mg/L of the average sulfate values from five years of monthly samples.

Figure 8.71a depicts variations in the pH of mine discharges for the four anthracite fields. The Eastern Middle Field has the lowest median pH and the least variability in pH, consistent with an absence of carbonate strata. Figure 8.71b shows that the Eastern Middle Field discharges also have the lowest sulfate concentrations and the least variability in concentration. The other fields show a wider range in pH and sulfate, although the Southern Field typically has lower sulfate than the Northern and Western Middle Fields.

For purposes of this discussion, local-scale variations in geology and mine drainage quality are those which occur within a single mine site, or between adjacent mine sites, or within a relatively small geographic area. These local-scale variations are much more prevalent and explainable in the Bituminous region, although distinct differences in mine drainage quality have been reported from adjacent active underground mines in the Anthracite region. There are at least four types of local-scale variations in geology that account for differences in mine drainage quality: (1) abrupt facies changes where freshwater calcareous strata (i.e., from a lacustrine or upper delta plain paleoenvironment) are replaced/cut out by channel sandstones, (2) more subtle facies/depositional environment changes occur where the thickness and purity of carbonate strata or the presence and thickness of high sulfur zones, within the coal seam or overlying shales, can very within a mine site, (3), weathering of bedrock which results in loss of carbonate and sulfur minerals, and (4) variations in the thickness and depth of weathering of glacial tills which result in loss of carbonate minerals.

An example of the effects of abrupt facies changes is shown in Figure 8.20 and reflected in the water quality from the neighboring Morrison and Stuart surface mines (Table 8.2). The mines extracted the upper Kittanning coal in Wharton Township, Fayette County. The calcareous overburden strata of the Morrison site is reflected in the alkaline nature of the seep (pH 7.0, alkalinity 308 mg/L and low metals concentrations). The Stuart site overburden was principally channel sandstones. The postmining seepage from this site is acidic (pH 2.8, acidity 1290 mg/L and high metals concentrations).

The more subtle local variations in facies, carbonate mineral, and sulfur content of overburden strata cannot be readily seen in Table 8.2. However the ranges and medians of acidity, alkalinity, and pH values from Table 8.2, summarized in Table 8.13, demonstrate that low pH and circumneutral pH, and alkaline and highly acidic values occur in all of the major stratigraphic groups in the bituminous region. The high acidity concentrations in generally alkaline stratigraphic sequences may appear anomalous, but they probably represent local variations where significant pyrite is present and carbonate minerals are lacking.

Geologic variability in portions of Lawrence, Venango, and Mercer Counties illustrate some of the reasons for local variations in mine drainage quality. The water quality differences result from differences in thickness and weathering depth of tills, and abundance of carbonate and sulfide minerals in till and bedrock. Mine drainage quality in Table 8.2 for the Cousins site on the Clarion coal in Hickory Township, Lawrence County show the influence of calcareous till (alkalinity of 130 mg/L in pit water), as does the Mathews site on the Mercer coal in Mahoning Township, Lawrence County (pit water alkalinity as high as 206 mg/L). This is consistent with the alkaline nature of till on the Cousins site (with NP > 500 ppt). Drill hole OB3-SPAG in Figure 8.39 from the Spagnola site in Millcreek Township, Mercer County contains a few zones with NP > 50 ppt. This calcareous overburden resulted in alkaline drainage. By contrast, the Oddfellow site in Pine Township, Mercer County, with a lack of calcareous rock, produced acidic drainage. Drill hole DH-28 from the Oddfellow mine shows thick till (and essentially no bedrock) overlying the Clarion coal, with only one zone with NP greater than 50 ppt; while DH-4 exhibits equally thick till with no zones of NP greater than 30 ppt, but the till is underlain by ~15 ft (5 m) of Vanport limestone with NP values from 920 to 956 ppt. A surface mine on the Clarion/Brookville coal with no evidence of thick till or Vanport limestone in the overburden, is the old Riddle site from Mineral Township, Venango County. An unnamed tributary to Sulphur Run receiving water from the Riddle site has a pH of 3.0 and an acidity of 240 mg/L (Table 8.2).

Worst case examples of acidity, metals, and sulfate concentrations that can be produced from Clarion coal overburden, without the benefits of Vanport limestone or calcareous tills on site, are the Old 40 and Zacherl sites in Table 8.2, located a short distance to the southeast in Clarion County. These two sites have the highest acidity concentrations of the any stratigraphic interval shown in Table 8.2. Thus, the importance of carbonates in controlling mine drainage quality is evident in both local and regional water quality for the bituminous and anthracite fields of Pennsylvania.

This Appendix proposes a revision of Williams (1960) "Environmental Faunal Classification" by reexamining the paleoecology of conchostracans, a fossil used by Williams (1960) to define freshwater depositional environments, and by adding an additional faunal group.

Williams (1960) pioneered the use of fossil fauna as a paleosalinity indicator in the Pennsylvanian of Pennsylvania. Table 8.15 is a list of faunal facies. Faunal Groups 1 through 4 are from Williams. Faunal Group 5 has been added by the authors. Faunal Group 4 was interpreted as "Fresh-water" by Williams. A reexamination of what is known about the most abundant fossil Order in Faunal Group 4, conchostracans, suggests that this group lived in a marginally brackish, not freshwater, environment. This is important from an overburden chemistry standpoint. As discussed in the body of the text, ocean water is high in sulfate and freshwater contains very small amounts of sulfate. Sulfate, through reduction, supplies the sulfur that forms pyrite. Where sulfate is low it can limit the amount of pyrite that is formed. Thus salinity can control the amount of sulfur available for the formation of pyrite.

Faunal Groups 1 through 3 retain the same interpretations that Williams gave them. Faunal Group 1, which is often associated with limestone, was probably deposited in deeper water than Faunal Group 2. Crinoids, which are stenohaline (i.e., not tolerant of brackish conditions or variable salinity) have been added to Faunal Group 1.

The marginally brackish environment may have been "paralic." Paralic is defined as "pertaining to intertongued marine and continental deposits laid down on the landward side of a coast or in shallow water subject to marine invasion, and to the environments (such as lagoonal or littoral) of the marine borders" (Bates and Jackson, 1987). Skema (1997, personal communication) has suggested the term "marginally-brackish" because it infers salinity (of interest to overburden chemistry), whereas paralic infers geographic position. The evidence for conchostracans occupying a marginally-brackish environment during the Pennsylvanian, in addition to the comments by Tasch, include the following:

1. Progressions from one paleoenvironment to another, such as those found in the rocks above the lower and middle Kittanning coals, are completely transitional. The transition is from truly marine conditions such as Williams’ Faunal Group 1, which has a wide variety of marine fauna, to a more restricted fauna in Faunal Group 2, to the brackish fauna of Group 3, to the conchostracan-bearing Group 4. This transition can be lateral or vertical. The paleoenvironmental maps (Figure 8.11 and Figure 8.16) show this lateral transition. Skema (1997, personal communication) has seen conchostracan-bearing strata grade vertically into Lingula-bearing strata, and back to conchostracan-bearing strata.
2. The faunal groups can interfinger laterally. An excellent illustration of this is the cross-sections in Plate 2 of Glass et al. (1977). The cross-sections show conchostracans and Lingula occurring at the same stratigraphic horizon above the middle Kittanning coal only a few miles (km) apart.
3. Many of the rocks and coals associated with conchostracan fossils are high in sulfur (>0.5%). The sulfur content of the upper Kittanning, lower and upper Freeport coals and overlying strata are higher than typical for freshwater coals and sediments. This is consistent with some marine influence.

Table 8.15 Faunal groups showing typical fossils for specific depositional environments. Salinity increases downward in the table. Also shown are the carbonates that are typically associated with the rocks from these faunal groups.

With the inclusion of marginally brackish environments, some marine influence is present in Pennsylvanian Period rocks in Pennsylvania from the Pottsville Group through at least the Skelly horizon of the Glenshaw Formation. The importance of this reinterpretation of Faunal Group 4 is that the higher than freshwater salinity that influenced the faunal facies would also influence the amount of sulfate available for pyrite formation. The marine influence present for Faunal Group 4 was less than that of brackish or truly marine sediments. This is reflected in the generally thinner sequences of high-sulfur strata associated with conchostracan-bearing rocks. Marginally brackish facies would be exposed to marine influence for less time, from transgressive to regressive portions of a cycle, than more fully marine sequences, and the shallower depths of water allowed for less sediment deposition. Thus the thinner zone of high sulfur rock.

This chapter has benefited from the input of many individuals. Vic Skema and Jim Shaulis, of the Pennsylvania Geologic Survey, have patiently answered numerous questions about the stratigraphy of western Pennsylvania and have generously provided unpublished data. Bob Smith, also with the PaGS, has kindly provided unpublished mineralogic data on limestones of Pennsylvania and X-ray diffraction data on rocks. Particular debt is owed to many DEP District Mining Operations hydrogeologists for contributions of data. We thank them all and in particular Joe Tarantino, Scott Jones, Tim Kania, and Dave Bisko. Steve Ebersole, Bob Weir, Michael Klimkos and Ray Roy helped with the figures. Stacy Thorne improved the manuscript in innumerable ways.

An embryonic form of this chapter began as a session for the Office of Surface Mining’s Acid-Forming Materials course back in 1989. This course allowed Keith Brady to be exposed to mine water chemistry and coal geology outside of Pennsylvania. He would like to thank Ann Walker, course coordinator, for the opportunity to broaden his horizons.

We thank our reviewers Dr. Arthur Rose and Joel Morrison, Penn State University, and Viktoras Skema, Pennsylvania Geologic Survey. Your comments improved the accuracy and readability of the text. As always, any shortcomings are those of the authors.

Several figures appearing in this chapter were taken or modified from figures appearing in the International Journal of Coal Geology, Vol. 5 (1985). These include Figure 8.45, 8.46 and 8.47 and Table 8.8 which are from Cecil et al. (1985). They originally appeared on pages 210, 205, 212, and 210 respectively of that journal. Figure 8.48 appeared in Donaldson et al. (1985) on page 189 of the same journal. Climate curves used in the construction of Figure 8.44 are from Cecil et al. (1985) on page 224 and Donaldson et al. (1985) on page 189 in Vol. 5 of the same journal, and the curve from Philips and Peppers (1984) appeared on page 230 of Vol. 3 of the International Journal of Coal Geology. The above cited figures from the International Journal of Coal Geology are used with the kind permission of Elsevier Science - NL, Sara Burgerhartstraat 25, 1055 KV Amsterdam, The Netherlands.

Figure 8.8 is used with the permission of R. Bergenback; Figure 8.9, Figure 8.63, Figure 8.64 are used with permission of E.G. Williams; and Figure 8.22 with permission of S. Weedman. Figure 8.53 originally appeared in Special Publication No. 7, Internat. Assoc. of Sedimentologists on page 370. It is reproduced with the permission of Blackwell Science Ltd., Oxford, UK. Figure 8.51 is used with permission of H. Gluskoter. Figures 8.27, 8.28, 8.29, 8.30, 8.33, 8.55 and 8.56 are from the Carboniferous Coal Guidebooks, A. Donaldson et al., eds. (1979). These figures are used with permission of the West Virginia University Department of Geology and Geography. Figure 8.40, 8.41 and 8.58 are from the book "Acid Sulfate Weathering" and are used with permission of the Soil Science Society of America, Madison, Wisconsin. Figures 8.3 and 8.54 are used with permission of the Geological Society of America, Boulder, Colorado.