INTRODUCTION

Depositional models provide a predictive framework for assessing resources, siting wells, and planning completion and field-management strategies that is vital for successful recovery of coalbed methane. The objective of this paper is to present a depositional model for Oak Grove field of central Alabama (Figure 1), which is among the largest and most productive coalbed methane fields in the world. Oak Grove field is in the eastern part of the Black Warrior basin, a late Paleozoic foreland basin that formed adjacent to the Appalachian-Ouachita orogen (Thomas, 1985, 1988). The field contains diverse tectonic structures, including normal faults and Appalachian folds and thrust faults (Pashin, 1991). This paper tests the relationship of depositional patterns to those structures and propo es ways in which sedimentologic and tectonic models can be incorporated into exploration and production management programs for coalbed gas.

The Oak Grove mine is in the east-central part of Oak Grove field and was established in 1974. There, the Blue Creek bed of the Mary Lee coal group (Figure 2) is mined at an approximate depth of 350 m (1150 ft). In 1977, 23 coalbed methane wells were drilled as part of a U. S. Bureau of Mines pilot program to reduce methane-related mine hazards and to enhance mine productivity. The program demonstrated the feasibility of coalbed methane as an economic resource and thus gave birth to the modern coalbed methane industry. Oak Grove field was formally established in 1980 in the area around the mine, but fewer than 50 wells were on line from 1981 to 1985, and less than 1 bcf of methane was produced annually. Shortly thereafter, permitting increased as the coalbed methane industry blossomed and the field was expanded to its present size in 1988. As of December 31, 1992, 640 wells had produced 62 bcf of coalbed methane, and a total of 881 wells had been permitted.

Oak Grove field contains a wealth of subsurface data that can be used to test existing depositional models and to formulate new models of coal-bearing strata. Early stratigraphic models stressed the cyclic nature of coal-bearing strata, which accumulated in regional marine-nonmarine depositional packages, or cyclothems, that formed in response to eustasy and tectonism (Udden, 1912; Weller, 1930, 1956; Wanless and Shepard, 1936). During the 1960s, it became apparent that emphasizing the cyclic nature of coal-bearing strata was inadequate for characterizing coal resources and solving the many geologic problems associated with

End_Page 960------------------------------

coal mining. Sedimentologic models of modern environments became widely available at this time, and construction of facies models based on depositional systems resulted in sophisticated, process-oriented approaches to characterizing coal-bearing rocks (Williams and Ferm, 1964; Ferm, 1970). These approaches have been applied successfully to a variety of mine-related problems, such as characterizing roof stability and understanding controls on the distribution and quality of mineable coal (e.g., Horne et al., 1978). Indeed, coal is part of a dynamic sedimentologic system that represents interaction between biologic and physical processes. For example, lateral offset of thick coal and sandstone bodies and associated bed splits (Figure 3) is considered to reflect interplay of the fluvial- eltaic system and the peat-forming biological community by avulsion of channels into former swamp areas and establishment of swamps where sandy streams once flowed (Ferm and Cavaroc, 1968).

Tectonism can also be a critical factor determining the stratigraphic and sedimentologic characteristics of coal-bearing strata. Control of coal thickness and geometry by synsedimentary movement of folds and normal faults (Figure 3) has long been recognized in the Appalachian region (Williams and Bragonier, 1974; Horne et al., 1978) as well as in European coal basins (Ciuk, 1968; Broadhurst and Simpson, 1983). In the Black Warrior basin, coal-exploration cores have been instrumental in identifying fault control of coal-body geometry (Weisenfluh and Ferm, 1984). According to this structural model, the thickest coal beds are present in upthrown fault blocks, whereas the coal is thin and split in the adjacent downthrown blocks (Weisenfluh and Ferm, 1984; Ferm and Weisenfluh, 1989).

Most studies of the distribution, thickness, and geometry of coal in the Appalachian region have relied strictly on outcrop and core data from actively mined intervals. For this reason, those studies have generally been restricted to only one or two economic coal zones. Many Appalachian coal basins contain abundant data from the petroleum industry, particularly

Fig. 1. Index map of study area showing location of Oak Grove field and other coalbed methane fields, Black Warrior basin, Alabama.

End_Page 961------------------------------

geophysical well logs, that can be used to construct depositional models for deep coal resources. The advent of coalbed methane as a viable energy resource, moreover, has made developing such models timely and practical. Using the closely spaced geophysical well logs in Oak Grove field allows evaluation of several coal zones in the same area. This approach provides a means of determining the variability of depositional style that can accompany synsedimentary tectonism along a given set of geologic structures.

Fig. 2. Sample well log of the Black Creek-Cobb interval showing coal groups, depositional cycles, and named coal beds used in sedimentologic analysis. Permit 6033-C, Amoco Production Company, USX Amoco-898 (18,7) 63-11 #34, Sec. 36, T18S, R7W, Jefferson County, Alabama.

End_Page 962------------------------------

METHODS

The Black Creek-Cobb interval of the Pottsville Formation (Figure 2) is the major production target for coalbed methane in the Black Warrior basin (McFall et al., 1986; Pashin, 1991) and is the focus of this study. More than 700 gamma-ray and density logs are available in Oak Grove field and provide a database for determining the geologic controls on the distribution and thickness of coal; 4 to 16 logs are available per square mile. Three major rock types, coal, sandstone, and mudstone, were identified in the well logs on the basis of distinctive geophysical signatures and comparison with core.

Standard gamma-ray density logs are available for the Black Creek-Cobb interval throughout much of the Black Warrior basin in Alabama, but these logs are inadequate to determine coal thickness because of large source-sender spacing and fast logging speed (Thomas and Womack, 1983; Sestak, 1984). However, high-resolution (20 in. = 100 ft) density logs are available for more than 500 wells in Oak Grove field and record bed thickness with a minimum resolution of 0.09 to 0.12 m (0.3-0.4 ft). Comparison with cores revealed that beds thinner than 0.01 m (0.03 ft) can be detected with limited reliability at the lower limit of log resolution.

Cross sections based on an interpretation of well logs were made to examine stratigraphic relationships within the Black Creek-Cobb interval, and a structural contour map of the top of the Mary Lee coal bed was made to determine structural and tectonic relationships in Oak Grove field. Numerous maps showing the thickness and distribution of coal and associated siliciclastic rocks also were made, and Pashin (1991) published a complete map set. Maps of maximum coal thickness depict the thickest coal bed in a given stratigraphic interval regardless of position within that interval. Isopach and net-sandstone isolith maps of selected siliciclastic intervals were made to establish the detrital architecture of the coal groups. These siliciclastic intervals are named after the coal beds they eparate (e.g., Jagger-Blue Creek siliciclastic interval).

GEOLOGIC SETTING

Stratigraphy

Economic coal resources in the Black Warrior basin are in the Lower Pennsylvanian Pottsville Formation and are in a series of stratigraphic clusters called coal groups (McCalley, 1900) (Figure 2). Coal groups have formed the basis of most stratigraphic subdivisions of the Pottsville (Culbertson, 1964; Metzger, 1965). However, Pashin (1991) recognized that the Pottsville can be subdivided further into numerous marine-terrestrial cycles of regional extent. The cycles have as much as 100 m (350 ft) of marine mudstone at the base that typically coarsens upward into sandstone. At the top of each cycle is the interbedded mudstone, sandstone, underclay, and coal that makes up a coal group. Pottsville cycles are mainly of New River age (Eble et al., 1991) and have an average duration of 0.2 t 0.5 m.y. (Pashin, in press).

During the Pennsylvanian, peat accumulated on the coastal plain bordering the Appalachian-Ouachita orogen (Ferm et al., 1967; Horsey, 1981; Thomas, 1988). Coal beds are thickest and most numerous in the eastern part of the basin where all coal mines and degasification fields are developed. By contrast, marine rocks predominate farther west where conventional hydrocarbons are produced (Pashin, 1991). Concentration of economic coal resources in the eastern part of the basin has been interpreted to reflect proximity to an orogenic sediment source, which helped maintain fluvial-deltaic platforms where peat could accumulate. Rapid subsidence adjacent to the Appalachian orogen further aided accumulation of thick peat. Farther west, where coal resources are limited, peat could accumulate onl at maximum aggression when most or all of the basin

Fig. 3. Schematic diagram showing fluvial and structural control of coal-body geometry. Fluvial models suggest that offset of thick coal and sandstone is related to avulsion of sand-laden streams into swamp areas and establishment of swamps where those streams once were. By contrast, structural models suggest that coal and sandstone bodies are stacked along faults, with sandstone and split coal predominating in the hanging-wall block.

End_Page 963------------------------------

was emergent (Pashin, in press).

In Oak Grove field, the Black Creek-Cobb interval comprises four coal groups and seven depositional cycles (Figure 2). Each coal group has a distinctive pattern of coal beds in density logs and is thus readily identified. For example, the Black Creek coal group typically has a thinning-upward density signature, whereas the Pratt coal group characteristically has a thickening-upward signature. The Mary Lee coal group, by contrast, contains four thick coal beds that extend across most of the field, whereas the Cobb group contains one thick, widespread bed that is in places accompanied by one or two thin beds. Bed names are most easily applied in the Mary Lee, Gillespy, and Curry cycles, where individual coal beds are widespread. In the other cycles, however, coal-body geometry is too co plex for bed names to be of use.

Structural Geology and Tectonics

Oak Grove field is in the easternmost part of the Black Warrior foreland basin along the leading edge of the Appalachian thrust belt. Three prominent Appalachian folds, the Blue Creek anticline, the Coalburg syncline, and the Sequatchie anticline, are in Oak Grove field and strike with an approximate azimuth of 40° (Figure 4). The Blue Creek anticline is near the southeast edge of the Black Warrior basin and extends along the southeast margin of Oak Grove field. Coal beds that are degassed deeper than 450 m (1500 ft) crop out along the northwest limb of the anticline (Pashin and Sarnecki, 1990). Dip of the anticlinal limb is generally 15 to 40°, and the limb is in places broken by thrust faults with more than 100 m (300 ft) of displacement (Blair, 1929; Miller, 1934).

Intensive drilling in Oak Grove field helps delineate structural relationships at the southwestern termini of the Coalburg syncline and the Sequatchie anticline, the farthest foreland structures of the Appalachian thrust belt in Alabama (Figure 4). The Coalburg syncline is immediately northwest of the Blue Creek anticline and shares common limbs with the Blue Creek and Sequatchie anticlines. Though essentially flat-bottomed, the syncline is strongly asymmetrical, and the axial trace lies only 1.6 to 4.8 km (1-3 mi) northwest of the Blue Creek anticline. The axial trace of the syncline is disjunct and is broken

Fig. 4. Structural contour map of the top of the Mary Lee coal bed showing the southwest terminus of the Sequatchie-Coalburg thrust sheet in the eastern part of the field and normal faults formed by flexural extension in the western part. Line A is the Sequatchie anticline; line B is the Blue Creek anticline, and line C is the Coalburg syncline.

End_Page 964------------------------------

by numerous horsts and grabens along the southeast margin of the field. In south-central Oak Grove field, the synclinal axis ends abruptly at a normal fault. The Oak Grove mine is developed in a southeast-dipping homocline between the axes of the Coalburg syncline and the Sequatchie anticline and northeast of a series of en echelon normal faults.

In north-central Oak Grove field, the Sequatchie anticline is a simple fold that plunges southwest and has more than 120 m (400 ft) of structural relief (Figure 4). Approximately 8 km (5 mi) northeast of Oak Grove field, the anticline verges northwest, and approximately 48 km (30 mi) farther northeast, a thrust fault emerges in the forelimb of the structure. Like the Coalburg syncline, the Sequatchie anticline terminates at normal faults. The anticlinal axis curves toward the south near its terminus, and the bounding faults are shorter and less consistent in orientation than in south-central Oak Grove.

Two distinctive sets of normal faults are present in Oak Grove field (Figure 4). In the western part of the field, subparallel faults define a horst-and-graben system. Fault displacement is as great as 60 m (200 ft), and some faults extend for more than 30 km (20 mi). In the central part of the field, the faults defining the southwest ends of the Coalburg syncline and the Sequatchie anticline are shorter and have an en echelon, right-stepping plan. Strike of the faults is 320° to 340°, and dip is generally greater than 60°; throw is locally 25 m (80 ft) and generally decreases toward the north-central part of the field. In addition, trace length generally decreases from 6.5 km (4 mi) in the south to 1.6 km (1 mi) in the north, and fault strike becomes less consistent to ard the north. In the Oak Grove mine, normal faults with displacement less than 3 m (10 ft) contain horizontal slickensides (McDaniel, 1986).

The large-scale folds (Figure 4) apparently are detached structures that formed in the Pennsylvanian and Permian during the Alleghanian orogeny (Rodgers, 1950; Thomas, 1985). The Blue Creek anticline is thought to overlie the frontal ramp of an upper-level decollement in Mississippian or Pennsylvanian strata that may be a splay of larger thrust faults to the southeast of Oak Grove field, whereas the Sequatchie anticline is thought to have formed at a frontal ramp of a basal decollement in Cambrian shale below the Coalburg syncline (Thomas, 1985). Most frontal ramps in the Appalachians of Alabama are thought to be localized above normal basement faults formed during Iapetan rifting (Thomas, 1991). The open nature of the Sequatchie anticline in Oak Grove field suggests that the structur is a detachment fold. Northeast of the field, where the anticline verges sharply northwest, the structure does not fit well into standard models of thrust geometry and may incorporate a minor backthrust (Cherry, 1990).

Most normal faults in the Black Warrior basin, like those in western Oak Grove field, have been attributed to flexural extension (Bradley and Kidd, 1991) related to deformational loading in the Ouachita orogen (Hines, 1988). Recent work in Deerlick Creek coalbed methane field suggests that, in the eastern part of the basin, those faults are thin-skinned structures that extend downward to a basal detachment in the lower part of the Pottsville Formation (Wang et al., 1993). Termination of Appalachian folds at en echelon normal faults in central Oak Grove field (Figure 4) suggests that some of the faults compose a sinistral, transtensional tear-fault system marking the southwest limit of the basal decollement below the Coalburg syncline. Indeed, horizontal slickensides on normal faults i the Oak Grove mine support a transtensional origin related to Appalachian thrusting. Moreover, the southwestward transition from a thrust-cored fold with complex internal geometry to a simple, southward-arcing detachment fold establishes decreasing forward translation near the terminus of the Sequatchie-Coalburg thrust sheet.

CONTROLS OF COAL THICKNESS AND GEOMETRY

Lower Black Creek Cycle

The lower Black Creek cycle contains a thinning-upward sequence of five to eight coal beds (Figure 2). Nearly all beds thicker than 0.3 m (1 ft) are in the lower half of the coal-bearing part of the cycle, and the thickest bed is present at or near the bottom of the coal group. Net coal thickness in the lower Black Creek cycle typically ranges from 1.2 to 3.1 m (4-10 ft) (Figure 5), and maximum bed thickness is locally greater than 1.5 m (5 ft). As a rule, net coal thickness increases with maximum bed thickness.

The lower Black Creek contains two major sets of coal beds (Figure 6). Bed geometry in the lower set is variable, whereas beds in the upper set are continuous and contain splits with less stratigraphic relief. Although the lower Black Creek typically contains these two sets of coal beds, they do not form regionally significant marker units. Rather, stratigraphic variability of the lower Black Creek simply decreases upward.

Beds in the lower set merge into one of the thickest coal beds in Oak Grove field (Figure 6). Stacked sandstone bodies predominate between coal beds at the north end of the cross section, and the associated siliciclastic intervals fine and thin southward as the coal beds merge. In the southern part of the cross section, however, another series of stacked sandstone bodies is present above the thickest coal.

The sandstone bodies between coal beds at the north end of the cross section (Figure 6) are interpreted

End_Page 965------------------------------

Fig. 5. Net-coal isolith map of the lower Black Creek cycle.

Fig. 6. Stratigraphic cross section AA^prime showing facies relationships in the upper and lower Black Creek cycles. Exceptionally thick coal in the lower Black Creek cycle is typically formed by joining of coal beds near the base of the coal group. In the upper Black Creek cycle, coal and sandstone thickness have a strong inverse relationship. See Figure 7 for location of cross section.

End_Page 966------------------------------

to be stacked fluvial deposits, and the fining and thinning of the associated siliciclastics as the coal beds merge are interpreted as the products of flood-basin deposition. Stacking of the sandstone bodies and separation of the bodies by coal suggests multiple episodes of abandonment and reestablishment of the fluvial system.

Features in the lower Black Creek cycle fit well into the classic compactional model of Ferm and Cavaroc (1968) in which differential compaction of peat and siliciclastic sediment provides avulsion sites for channel axes, thereby juxtaposing thick sandstone and coal. Where the coal beds merge (Figure 6), a large volume of compactible peat had accumulated. Compaction of that peat apparently promoted establishment of a new fluvial trend that is represented by the three stacked sandstone bodies in the southern part of the cross section.

Stacking of sandstone bodies between coal beds in the northern part of the cross section (Figure 6) suggests that the peat was domed above river-bank level, thus causing avulsion and establishment of fluvial axes beyond the line of cross section. Only after the peat dome collapsed or the fluvial deposits built above the level of the dome could the fluvial system avulse toward the south, thereby forming stacked sandstone bodies above thick, rapidly compacting peat. Upward decrease in stratigraphic variability within the lower Black Creek cycle suggests that topography became progressively subdued and that development of avulsion sites by peat compaction was no longer a dominant process near the close of lower Black Creek deposition.

Upper Black Creek Cycle

The upper Black Creek cycle contains one to six (generally 2 or 3) thin coal beds that cap a coarsening-upward sequence that is approximately 30 m (100 ft) thick (Figures 2, 6). Individual coal beds in the upper Black Creek are typically thinner than 0.3 m (1 ft). Net coal thickness ranges from less than 0.2 m (0.5 ft) in the west-central part of the field to more than 1.2 m (4 ft) in the easternmost part (Figure 7).

The sandstone isolith map of the upper Black Creek cycle depicts a major east-west-trending belt of thick sandstone (Figure 8). The 30-ft contour on the net sandstone isolith map outlines an east-west-trending, sinuous sandstone body that is locally 3 km (2 mi) wide and more than 15 m (50 ft) thick; the main sandstone axis narrows and joins another sinuous axis in the westernmost part of the field. Localized lobate and apron-like sandstone bodies are present in west-central Oak Grove field along the flanks of the main axial trend. Sandstone and coal thickness have a strong inverse relationship in western Oak Grove field (Figures 6, 7, 8). As many as

Fig. 7. Net-coal isolith of the upper Black Creek cycle.

End_Page 967------------------------------

Fig. 8. Sandstone isolith map of the upper Black Creek cycle.

Fig. 9. Stratigraphic cross section BB^prime showing facies relationships in the Mary Lee through Cobb cycles relative to the fault bounding the northeast margin of the en echelon fault system in central Oak Grove field. Note that in the Mary Lee cycle, the Jagger coal terminates against the fault, whereas in the Pratt cycle, few thick coal beds are in the hanging-wall block, whereas numerous thin coal beds are in the footwall block. In the Cobb cycle, coal beds are more numerous in the hanging-wall block than the footwall block. See Figure 10 for location of cross section.

End_Page 968------------------------------

six beds are present in the south, whereas only two beds are present in the north where correlative sandstone is thickest.

The axial sandstone (Figure 8) is interpreted to represent a sinuous fluvial channel, and the flanking lobate and apron-like bodies are interpreted to represent crevasse-splay complexes. Joining of the two sinuous sandstone bodies in westernmost Oak Grove may define part of a tributary system. The upper Black Creek fluvial system is similar in scale to that of the Tombigbee and Alabama rivers of southwest Alabama, which have meander belts that are approximately 3 km (2 mi) wide north of where they join to form the Mobile River.

The inverse relationship between sandstone thickness and the thickness and number of coal beds (Figures 6, 7, 8) indicates strong fluvial control of coal distribution. Increase in coal thickness away from the fluvial axis suggests that the fluvial system controlled where peat could accumulate. The uniform increase in the number of coal beds away from the fluvial axis plus the lateral continuity of the siliciclastic intervals separating the coal beds suggests that peat accumulated only in parts of the flood-basin system that were isolated from clastic influx and that the flood basins were extensive.

Mary Lee Cycle

The Mary Lee cycle contains four coal beds; named in ascending order, the coal beds are Jagger, Blue Creek, Mary Lee, and New Castle (Figure 2). All four beds are widespread, but thickness patterns differ markedly among the beds and the associated siliciclastic intervals. The Jagger coal bed ranges in thickness from 0 to 1 m (0-3 ft) and is generally 0.5 to 0.6 m (1.5-2 ft) thick in Oak Grove field. The bed terminates or merges with the Blue Creek bed near the margin of the en echelon fault system in central Oak Grove field (Figure 9). The Jagger bed is absent or not identifiable in much of the footwall block, which contains the Oak Grove mine, and is also not distinguishable in most of western Oak Grove field. However, the bed is present east of the mine and along the axial trace of he Sequatchie anticline north of the mine. The Jagger-Blue Creek siliciclastic interval, where present, is typically 6 to 9 m (20-30 ft) thick (Figure 10); however, the interval pinches out near the fault and where the Jagger and Blue Creek beds join. The interval is absent in the footwall block in eastern Oak Grove field but thickens gradually in the eastern part of the map area. The interval is approximately 6 m (20 ft) thick along

Fig. 10. Isopach map of Jagger-Blue Creek siliciclastic interval, Mary Lee coal group.

End_Page 969------------------------------

the axial trace of the Sequatchie anticline in north-central Oak Grove field.

Unlike the Jagger bed, the Blue Creek bed, which is the primary mining and degasification target in Oak Grove field, extends throughout the map area (Figure 11). The Blue Creek bed is generally thinner than 1.2 m (4 ft) and is thicker than 1.8 m (6 ft) in the south-central part of the field. In central Oak Grove field, coal thicker than 1.8 m (6 ft) has a dendritic geometry in the footwall block. In the Oak Grove mine, coal thicker than 1.8 m (6 ft) is present in a curving channel system, which facilitated exceptional mine yield (McDaniel, 1986). West of the mine, the trunk of the channel system is locally deeper than 18 m (60 ft) and truncates the Jagger bed at the southern channel margin. Near the northern margin, however, the Blue Creek bed is at the level of the Jagger bed (Figure 12).

The Mary Lee coal bed (Figure 4) is between 1.2 and 2.5 m (4-8 ft) above the Blue Creek bed throughout Oak Grove field (Figures 9, 12). The Mary Lee bed is typically 0.5 to 0.8 (1.5-2.5 ft) thick in eastern Oak Grove field and is only 0.3 m (1 ft) thick in the western part of the field. Thickness trends in the Mary Lee bed do not parallel the fault, but the thickest Mary Lee coal coincides with Blue Creek channel-fill coal. Within the trunk channel, thickness of the Mary Lee bed has a maximum value of 4.5 ft.

The Mary Lee-New Castle siliciclastic interval is generally 9 to 18 m (30-60 ft) thick (Figure 13). Isopach contours do not parallel the fault trend, and like the Mary Lee bed, the interval is thickest in the area of Blue Creek channel-fill coal (Figures 12, 13). The Mary Lee-New Castle interval is thinner than 3 m (10 ft) in northernmost Oak Grove field east of the tear-fault system and along the axial trace of the Sequatchie anticline.

The New Castle coal splits profusely, but a major bed is traceable through most of the field (Figures 9, 14). The New Castle is thinner than 0.6 m (2 ft) east of the tear-fault system and is thin or absent in the area containing most of the channel-fill coal bodies (Figures 11, 12, 14). In core, the seat earth below the New Castle bed is not rooted near where it pinches out. The bed is continuous in western Oak Grove field and is generally thicker than 1 m (3 ft) in the horst-and-graben system. On the basis of comparative thickness, the New Castle is a more significant coalbed methane target than the Blue Creek in much of this area.

Sedimentologic evidence indicates a strong interplay of fluvial, biological, and tectonic processes during deposition of the Mary Lee cycle. Development of the channel network delineated by thick Blue Creek coal in the uplifted area northeast of the en echelon fault system (Figure 11) suggests that the network

Fig. 11. Isopach map of Blue Creek coal bed, Mary Lee coal group.

End_Page 970------------------------------

Fig. 12. Stratigraphic cross section CC^prime showing facies relationships in the Mary Lee cycle where the Blue Creek bed is exceptionally thick in the uplifted part of the Sequatchie-Coalburg thrust sheet. The Blue Creek bed is interpreted to fill a valley that incised the uplifted part of the thrust sheet, truncating and locally forming a terrace on the Jagger coal. Differential compaction related to development of a peat-filled valley system apparently had a subsequent sedimentation in the Mary Lee cycle. See Figure 11 for location of cross section.

Fig. 13. Isopach map of Mary Lee-New Castle siliciclastic interval, Mary Lee coal group.

End_Page 971------------------------------

was an incised tributary system. Sharp pinch-out of the Jagger-Blue Creek siliciclastic interval (Figure 10) indicates that the tributaries occupied a steep-walled valley. Some of the tributary channels skirt the axial trace of the Sequatchie anticline.

Fibrous, rooted peat, the precursor to most coal beds in the Black Warrior basin, is typically resistant to erosion. Thus, presence of the Blue Creek bed at the level of the Jagger near the northern margin of the main channel (Figure 12) suggests that terraces formed on peat beds during valley incision. Apparent joining of the Jagger and Blue Creek beds throughout much of Oak Grove field implies that the Jagger bed provided channel floors in a large area. Termination of the Jagger coal and the Jagger-Blue Creek siliciclastic interval near the northeast margin of the en echelon system (Figures 9, 10) indicates that the coal and associated siliciclastic rocks were eroded or never deposited on the upthrown block and were preserved in the downthrown block. Presence of the Jagger coal and he Jagger-Blue Creek siliciclastic interval east of the mine (Figure 10) suggests that the area subsided more rapidly than the area immediately east of the major faults, thus defining a relative uplift in the mine area. This uplift evidently contributed to incision of the Blue Creek paleovalley and apparently was related to elevation of the area around the Sequatchie anticline.

Thick channel-fill coal, like that in the Blue Creek paleovalley system, can be used to estimate peat-to-coal compaction ratio in ancient sequences (Cobb et al., 1981). Assuming the channel was filled completely with peat, the difference in coal thickness in the channel from that outside the channel may represent the amount of peat that filled the channel. Therefore, the ratio of channel depth to the difference in coal thickness yields the peat-to-coal compaction ratio. In cross section CC^prime, the Blue Creek coal in the channel is as thick as 2.8 m (9 ft), whereas adjacent to the channel it is approximately 0.9 m (3 ft) thick (Figure 12). From dividing the channel depth (18 m; 60 ft) by the difference in coal thickness (1.8 m; 6 ft), the compaction ratio is 10:1. The same value was derived by Cobb et al. (1981) for a Pennsylvanian channel-fill coal in Kentucky.

Valley incision apparently had a strong impact on sedimentation for the remainder of Mary Lee cycle deposition. The Mary Lee bed represents a blanket of peat that traversed the map area, but thickening of the Mary Lee bed and the overlying Mary Lee-New Castle siliciclastic interval in the area of the incised tributary system (Figures 12, 13) indicates compactional control of sedimentation. Significant variation of thickness in the Blue Creek and Mary Lee beds

Fig. 14. Isopach map of New Castle coal bed, Mary Lee coal group.

End_Page 972------------------------------

within the tributary trend, however, indicates that compaction did not proceed uniformly and that parts of the paleovalley may have been underfull or even overfull with Blue Creek peat. Hence, the compaction ratio derived in the preceding paragraph uses the thickest coal in the channel system and thus is a maximum value.

Contours on the isopach map of the Mary Lee-New Castle siliciclastic interval do not follow the fault bounding the en echelon system (Figure 13), suggesting that fault-related topography was minimal, perhaps because sedimentation outstripped differential subsidence. Splitting and local thickening of the New Castle coal in the westernmost tributaries (Figures 12, 14) indicates continued compactional control of sedimentation, but absence of the coal in the eastern part of the tributary trend (Figure 14) and thickening of the coal in western Oak Grove field establishes that controls on peat accumulation had changed. One interpretation is that the New Castle pinches out because the compactional subsidence rate in the tributary system exceeded the sedimentation rate, resulting in a lake in tead of a swamp. Coal lacking rooted seat earth near where the New Castle bed pinches out may represent log accumulations or peat flotants at the lake margin.

Gillespy and Curry Cycles

The Gillespy and Curry beds (Figure 2) are among the most reliable stratigraphic markers in the Pottsville Formation because they are widespread, closely spaced, and scarcely split (Pashin, 1991, in press) (Figure 9). Even where coal is absent, the stratigraphic position of the beds is readily recognized in well logs by the presence of coarsening-upward sequences below the stratigraphic position of the coal. In eastern Oak Grove field, however, the beds split more abundantly and are thicker than they are elsewhere in the Black Warrior basin (Figure 15).

Throughout much of the field, the Gillespy and Curry beds are very thin, are at the lowest limit of log resolution, and have a combined thickness of 0.3 m (1 ft) or less. In the eastern part of the field, however, where the beds thicken and split (Figure 15), net coal thickness locally exceeds 1.2 m (4 ft), and the beds are viable completion targets. In central Oak Grove field, the Curry cycle thickens by more than 6 m (20 ft) across the boundary of the en echelon fault system (Figure 9). Net sandstone thickness is greater than 9 m (30 ft) in an east-west belt that is approximately 5 km (3 mi) wide, and sandstone is

Fig. 15. Stratigraphic cross section DD^prime showing facies relationships in the Gillespy, Curry, and Pratt cycles in eastern Oak Grove field. In this area, coal is thick in most cycles, and beds split toward the southeast. These relationships are interpreted to represent down-warping of the ancestral Coalburg syncline during Black Creek-Cobb deposition. See Figure 17 for location.

End_Page 973------------------------------

absent in much of north-central Oak Grove field (Figure 16). The sandstone belt traverses Oak Grove field and is developed south of the relative uplift containing the Sequatchie anticline and the Oak Grove mine. In the western half of the field, lobate and elongate sandstone bodies outlined by the 10-ft contour extend north from the sandstone belt. One such body extends along the western margin of the tear-fault system, and in the westernmost lobate body, the 20-ft contour follows the boundary fault of the western Oak Grove horst-and-graben system.

The sandstone belt (Figure 16) is interpreted to represent the axis of a fluvial system, and the flanking lobate and elongate sandstone bodies are interpreted to represent the associated crevasse-splay systems. Because the sandstone belt is approximately 5 km (3 mi) wide, the fluvial system may have been the largest in the map area during Black Creek-Cobb deposition. Diversion of the fluvial system around the uplift containing the Sequatchie anticline and the Oak Grove mine indicates that synsedimentary tectonism had given rise to significant topographic relief.

Thickening of the Curry cycle across the boundary of the en echelon fault system (Figure 9) further suggests that structural growth occurred, thus differentiating the uplifted Coalburg-Sequatchie thrust sheet from the main part of the Black Warrior basin. Moreover, restriction of thick sandstone to the hanging wall of the en echelon tear-fault system (Figure 16) indicates that topographic relief sufficient to control the position of the fluvial system had developed. Similar structural control is apparent immediately west of the thrust sheet, where a lobate sandstone body extends northward adjacent to the tear faults. Structural control is also apparent in western Oak Grove, where a lobate crevasse-splay deposit is bound partly by faults.

Pratt and Cobb Cycles

The Pratt cycle contains four to eight coal beds throughout most of Oak Grove field; only the Black Creek cycle contains more coal beds (Figure 2). Coal beds in the Pratt cycle split profusely and are discontinuous, and coal-body geometry changes near the northeast margin of the en echelon tear-fault system (Figures 9, 15). Coal beds are thickest and least numerous in the hanging wall, a configuration that contrasts sharply with earlier reports of structurally controlled coal-body geometry in the Black Warrior basin (Weisenfluh and Ferm, 1984; Ferm and Weisenfluh, 1989).

Net coal thickness in the Pratt cycle ranges from less than 1.5 m (5 ft) in the northern part of Oak Grove field to as much as 6.8 m (22 ft) in the eastern part. Similarly, maximum coal thickness ranges from less than 0.5 m (1.5 ft) in the northern part of the

Fig. 16. Sandstone isolith map of the Gillespy-Curry siliciclastic interval.

End_Page 974------------------------------

field to approximately 2.8 m (9 ft) in the eastern part (Figure 17). Net coal thickness increases from 2.2 to 2.8 m (7-9 ft) across the boundary of the en echelon fault system, and maximum bed thickness increases sharply from 2 to 5 ft in the downthrown block (Figures 9, 17). Maximum coal thickness exceeds 1.2 m (4 ft) in a rectangular area southwest of the bounding fault and decreases as the beds diverge away from the fault (Figure 9).

The Cobb cycle generally contains three or fewer coal beds in Oak Grove field (Figure 2), and only one of those beds is typically thicker than 0.1 m (0.4 ft). Net coal thickness ranges from less than 0.5 m (1.5 ft) to more than 1.5 m (5 ft) (Figure 18). More coal beds are present in the hanging wall than in the footwall, but splitting is not precisely coincident with the fault (Figure 9). Even so, net coal thickness doubles from less than 0.5 m (1.5 ft) in the upthrown block to more than 0.9 m (3 ft) in the downthrown block (Figure 18).

In the Pratt and Cobb cycles, increase in coal thickness in the hanging-wall block (Figures 9, 17, 18) suggests that differential subsidence favored peat accumulation, but the fluvial response to subsidence differed greatly in each cycle. In the Cobb cycle, splitting of beds in the downthrown block indicates that differential subsidence favored clastic influx; subsidence was probably enhanced by ongoing compaction of thick Pratt peat in the downthrown fault block. Topography may have been especially subdued during Cobb deposition, thereby facilitating distal flood-basin sedimentation on the footwall block, because bed splitting does not coincide precisely with the fault trace. In the Pratt cycle, however, merging of coal beds suggests that the en echelon fault system was sheltered fro clastic influx.

Splitting of coal beds in the Gillespy, Curry, and Pratt cycles in easternmost Oak Grove field (Figure 15) indicates that increased subsidence rate acted in concert with alternating episodes of clastic influx and peat accumulation. The western margins of the thick coal bodies in this area do not coincide with known faults, so subsidence may have been a response to folding associated with an ancestral Coalburg syncline. In addition to splitting and thickening of coal in the Gillespy, Curry and Pratt cycles, presence of thick coal in the Black Creek coal group (Figures 5, 7) and southeastward thickening of the Jagger-Blue Creek siliciclastic interval of the Mary Lee cycle (Figure 10) indicate that downwarping was operative in this area throughout Black Creek-Cobb deposition.

DISCUSSION

Charles Butts (1926) was first to suggest that Appalachian folding was contemporaneous with

Fig. 17. Maximum-coal thickness map of the Pratt cycle.

End_Page 975------------------------------

Paleozoic sedimentation in Alabama. Subsequent investigators confirmed that synsedimentary tectonism operated in this area, with reactivation of normal, rift-related basement faults from the Cambrian to the Mississippian (Thomas, 1968, 1986; Thomas and Neathery, 1982) and blind thrusting during the Early Pennsylvanian (Pashin and Carroll, 1993). Detachment folding has long been postulated as a control on Pennsylvanian sedimentation in the eastern part of the Appalachian basin (Williams and Bragonier, 1974) and has even been interpreted to have caused stream diversions that permitted widespread peat bodies to accumulate in the distal Appalachian foreland (Belt and Lyons, 1989; Wise et al., 1991).

Widespread structural control of sedimentation in Oak Grove field suggests that an ancestral Sequatchie-Coalburg thrust sheet was present and active during deposition of the Black Creek-Cobb interval of the Pottsville Formation. Three domains with thick coal can be related to those structures and reflect contemporaneous biologic, fluvial, and tectonic processes (Figure 19). The eastern domain contains thick coal associated with folding of part of the Coalburg syncline, whereas the southern domain contains thick coal bodies bound by the en echelon tear-fault system. In the northern domain along the axial trace of the Sequatchie anticline, coal is generally thin, except for the Blue Creek bed of the Mary Lee cycle, which is interpreted as the fill of a valley incised into the uplifted p rt of the thrust sheet.

The southwest terminus of the ancestral Sequatchie-Coalburg thrust sheet differed from the modern structure in several ways. For example, the ancestral Sequatchie anticline may have been much broader than the modern anticline, perhaps reflecting an early stage of detachment folding, because the area of thin coal in north-central Oak Grove field extends along the tear-fault system well northwest of the modern axial trace (Figures 10, 11, 17). Relief of the anticline apparently was substantially less than that which exists today, but the depth of the Blue Creek paleovalley indicates that part of the thrust sheet was elevated more than 18 m (60 ft) above the area west of the tear-fault system. Although the Coalburg syncline now extends through to the en echelon fault system, the eastern omain of thick, downwarp-related coal is separated from the fault-bound coal, suggesting that the ancestral syncline was separated from the tear faults by a southward extension of the Sequatchie anticline. This extension is especially well shown in the Pratt coal isolith map (Figure 17) and may signify minor compression and development of an incipient lateral ramp prior to development of the present-day transtensional fault system.

Unconsolidated sediment has limited capability to transmit layer-parallel stress. Thus, Pottsville sediment

Fig. 18. Net-coal isolith map of the Cobb cycle.

End_Page 976------------------------------

probably deformed passively above structures in competent Cambrian-Ordovician carbonate rocks that were active at depth. Indeed, short fault length and modest displacement suggests that the modern tear-fault system is a shallow expression of a more coherent fault system at depth. Although the general trend of the tear-fault system is visible in some subsurface maps (Figures 14, 17), questions remain whether faults broke surface during Black Creek-Cobb deposition or whether soft sediment was simply draped over the southwest margin of the Sequatchie-Coalburg thrust sheet.

The northeastern margin of what is now the en echelon system exerted the strongest control on depositional patterns of any single structure in the study area and is thus the best candidate for a synsedimentary surface fault (Figure 19). Well logs from the Jagger coal to the top of the Cobb cycle demonstrate 16 m (52 ft) of growth and an increase of net coal thickness of 1.9 m (6.2 ft) across the fault. This estimate of fault growth is a minimum value because compaction and topographic relief were not considered. Modern displacement of the fault is approximately 18 m (60 ft), suggesting that most growth took place during Black Creek-Cobb deposition. The other en echelon faults in south-central Oak Grove field apparently had little or no influence on sedimentation, and thus appear young r than the study interval. One possibility is that these faults formed during continued transtension in the hanging-wall block of the synsedimentary fault.

Extensional faults in western Oak Grove field apparently had little effect on depositional patterns. However, one fault with displacement greater than 60 m (200 ft) may have influenced the thickness of the New Castle coal (Figure 14) and the thickness of sandstone in the Curry cycle (Figure 16). Hence, most extensional faults in western Oak Grove field appear to postdate Black Creek-Cobb desposition, but some of these faults may have been in an early stage of development.

Synsedimentary activity of the Sequatchie anticline provides new insight into the nature and timing of Appalachian thrusting in Alabama. Development of the Sequatchie-Coalburg thrust sheet during the Early Pennsylvanian establishes that thrusting in Alabama reached its farthest cratonward extent in concert with the Lackawanna phase of the Alleghanian orogeny as defined by Geiser and Engelder (1983), and not the better known Main phase,

Fig. 19. Tectonically controlled domains of coal distribution in Coak Grove field, Black Warrior basin, Alabama. The eastern domain contains thick coal associated with downwarping of the ancestral Coalburg syncline, whereas the southern domain contains thick coal bodies bound by the en echelon tear-fault system. In the northern domain along the axial trace of the Sequatchie anticline, coal is generally thin except for the Blue Creek bed of the Mary Lee cycle, which apparently fills a valley that eroded the uplifted part of the thrust sheet.

End_Page 977------------------------------

which took place from the Late Pennsylvanian to the Early Permian. Dating of thrusts in other regions establishes that these structures can be active intermittently over tens of millions of years, with individual episodes of shortening generally lasting between 2 and 10 m.y. (Wiltschko and Dorr, 1983; Burbank and Raynolds, 1988; Jordan et al., 1993). The seven depositional cycles of the Black Creek-Cobb interval span approximately 1.4 to 3.5 m.y. Therefore, the Black Creek-Cobb interval may record only part of one tectonic episode in the Sequatchie-Coalburg thrust sheet.

The history of Appalachian thrusting in Alabama was complex because of the diverse assemblage of folds, forward thrusts, and backthrusts, as well as the presence of major cross-strike structural discontinuities. The thrusting sequence in the southernmost Appalachians is unclear, but (Thomas et al., 1989) suggested that some out-of-sequence (break-back) structures are present in Alabama. Stratigraphic evidence from the internal part of the thrust belt suggests that some structures inboard of the Sequatchie anticline were active during Pottsville deposition (Pashin and Carroll, 1993) Hence, stratigraphic and sedimentologic evidence support simultaneous movement of multiple structures in the Appalachian orogen during the Early Pennsylvanian.

Estimates of missing section based on vitrinite reflectance suggest that as much as 3 km (1.9 mi) of sediment was eroded from the Oak Grove area and that nearly 4 km (2.5 mi) of sediment was eroded from the internal part of the thrust belt (Telle et al., 1987; Levine and Telle, 1991). Hence, the structures exposed at the surface today were once buried deeply below the Alleghanian coastal plain. A similar situation exists in the Mediterranean, where synsedimentary thrusts that were active during the Oligocene and Miocene are now buried below the locus of sedimentation in the Apennine foreland basin (Ricci Lucci, 1986). The distal Apennine thrusts now have little or no surface expression, as may have been the case in the Appalachian foreland as the Alleghanian clastic wedge approached m ximum thickness. Though this structural analogy may be plausible, a critical difference is that the distal Apennine thrusts formed in a marginal basin dominated by turbidites, whereas the Appalachian thrusts of Alabama formed in a continental basin containing mainly shallow-marine and terrestrial strata.

The striking variety of depositional styles in Oak Grove field indicates that sedimentologic and tectonic models must be applied in concert and flexibly to characterize coal distribution adequately. Classic fluvial models, like that of Ferm and Cavaroc (1968), can be applied readily to coal bodies in the lower and upper Black Creek cycles in western Oak Grove field. The structural models of Weisenfluh and Ferm (1984) and Ferm and Weisenfluh (1989) apply in a general way, but synsedimentary movement along the northeast margin of the en echelon fault system gave rise to a different style of coal geometry and thickness distribution in each depositional cycle (Figure 9).

In keeping with earlier models of structurally influenced coal-bearing strata, more coal beds are present in the Mary Lee and Cobb cycles in the hanging-wall block than in the footwall block; and in the Curry cycle, sandstone is generally restricted to the hanging-wall block adjacent to the fault (Figures 9, 16). Contrary to these models, however, the thickest coal in the Pratt and Cobb cycles is in the hanging-wall block (Figure 19). The Pratt cycle further contains fewer coal beds in the hanging-wall block (Figure 9). Only in the Blue Creek bed of the Mary Lee cycle, where thick coal is preserved in an abandoned tributary system, is coal thickest in the uplifted footwall block. In Oak Grove field, coal is generally thickest in structural lows associated with the Sequatchie-Coalburg hrust sheet, suggesting that differential subsidence promoted peat accumulation. In the Mary Lee cycle, however, channel incision provided sufficient topographic relief for thick peat to accumulate on the uplifted part of the thrust sheet.

CONCLUSIONS

The Pottsville Formation in Oak Grove field presents an opportunity to examine the sedimentologic response to tectonism in the Appalachian foreland and to determine the role that depositional and tectonic models should play in exploring and managing coalbed methane resources. Comparing depositional and structural features in Oak Grove field indicates an early phase of detachment folding in the Sequatchie-Coalburg thrust sheet during deposition of the Black Creek-Cobb interval. Thick coal predominates in areas of enhanced subsidence associated with the ancestral Coalburg syncline and the en echelon tear-fault system at the southwest terminus of the thrust sheet. Although tectonism was effective throughout Black Creek-Cobb deposition, the sedimentologic expression of that tectonism diff red markedly among cycles. This difference is especially apparent in the Mary Lee cycle, where thick peat filled a valley incised into the uplifted part of the thrust sheet.

This diversity of depositional style in Oak Grove field reflects the manifold variables inherent in biologic, fluvial, and tectonic processes. Existing depositional models offer many solutions that can be used to decipher the effects of synsedimentary tectonism, but a high degree of flexibility is required to make the most of the opportunities offered by a specific structural and stratigraphic setting. Until recently, knowledge of synsedimentary tectonism in coal-bearing sequences has been derived almost exclusively

End_Page 978------------------------------

from outcrop transects across normal faults and from drill data restricted to mineable coal zones. New data generated by coalbed methane drilling has lessened these restrictions by providing closely spaced control that is continuous among several coal zones. Hence, as data are compiled from other coalbed methane fields, a more robust understanding of the three-dimensional relationship between tectonism and sedimentation will emerge.

Developing depositional and tectonic models on the basis of subsurface data is critical for coalbed methane exploration and development, because these models provide a means for characterizing, conceptualizing, and predicting the distribution of deep coal resources. However, the role of depositional and tectonic models in producing coalbed methane is not straight forward. In the Black Warrior and San Juan basins, for example, exceptional production coincides more strongly with favorable hydrologic, structural, and paleogeothermal settings than with specific depositional settings (Pashin, 1991; Kaiser, 1993). In other words, thick coal by no means guarantees economic gas production. Even so, understanding the geometry and extent of production targets is vital to the economic success of a coalbed methane field, because where geometry and extent are known, well sites, completion zones, and the technology to be applied to those zones can be selected wisely. Additionally, by combining sedimentologic information with structural and hydrologic information, models of coal thickness and geometry can be a valuable tool for modeling subsurface flow paths and for simulating the behavior of coalbed methane reservoirs.