Vous êtes sur la page 1sur 32

Reservoir systems of the Pennsylvanian lower Atoka Group (Bend Conglomerate), northern Fort Worth Basin, Texas: High-resolution

facies distribution, structural controls on sedimentation, and production trends


Tucker F. Hentz, William A. Ambrose, and David L. Carr

AUTHORS Tucker F. Hentz  Bureau of Economic Geology, University of Texas at Austin, University Station, Box X, Austin, Texas; tucker.hentz@beg.utexas.edu Tucker F. Hentz is a research associate at the Bureau of Economic Geology, Jackson School of Geosciences at the University of Texas at Austin. His areas of interest include sequencestratigraphic analysis and siliciclastic depositional systems. Hentz received his B.A. and M.S. degrees in geology from Franklin & Marshall College and the University of Kansas, respectively. William A. Ambrose  Bureau of Economic Geology, University of Texas at Austin, University Station, Box X, Austin, Texas; william.ambrose@beg.utexas.edu William A. Ambrose is a research scientist at the Bureau of Economic Geology, Jackson School of Geosciences at the University of Texas at Austin. His areas of interest include unconventional energy minerals, siliciclastic depositional systems, and stratigraphy. Ambrose holds B.S. and M.A. degrees in geosciences from the University of Texas at Austin. David L. Carr  Bureau of Economic Geology, University of Texas at Austin, University Station, Box X, Austin, Texas; david.carr@beg.utexas.edu David L. Carr is a research associate at the Bureau of Economic Geology, Jackson School of Geosciences at the University of Texas at Austin. His areas of interest include siliciclastic sedimentology and stratigraphy and their application to exploration, production, and gas-storage activities. Carr received his M.A. degree in geosciences from the University of Texas at Austin.

ABSTRACT This study defines the depositional systems of mature lower Atoka Group reservoirs, structural influence on their sedimentation, and sand-transport patterns at a higher degree of resolution and over a significantly larger part of the play area than previously conducted. The reservoir systems are characterized by pronounced variations in depositional style, even between stratigraphically adjacent systems. They represent a variety of on-shelf siliciclastic depositional facies, including gravelly braided river, fluvial-dominated delta, and low-sinuosity incised river deposits. Penecontemporaneous, high-angle, basementrooted reverse faults and genetically associated folds of the Mineral WellsNewark East fault system exerted direct control on the orientation of complex fluvial-channel and deltadistributary sand-transport pathways and the geometry of deltaic depocenters. Multiple contemporaneous source areas, including the Ouachita fold belt to the southeast, the Muenster arch to the northeast, and the south flank of the Red River arch, also contributed to the complexity of sandstone trends in the lower Atoka play area.

ACKNOWLEDGEMENTS This study was funded by the Geology Foundation of the University of Texas at Austin and the State of Texas Advanced Resource Recovery (STARR) project. The authors express their appreciation to Eric C. Potter, William L. Fisher, Douglas C. Ratcliff, Jeffrey A. Kane, Luciano L. Correa, Michael R. Hudec, and H. Scott Hamlin for their valued advice and input during the study. Jack D. Beuthin and John F. White, Devon

Copyright 2012. The American Association of Petroleum Geologists. All rights reserved. Manuscript received June 20, 2011; provisional acceptance August 24, 2011; revised manuscript received September 20, 2011; final acceptance October 4, 2011. DOI:10.1306/10041111078

AAPG Bulletin, v. 96, no. 7 (July 2012), pp. 1301 1332

1301

Energy Corporation, helped clarify key structural aspects of the study area. We also thank Bradford E. Prather and Mihaela S. Ryer for their peer reviews and Stephen E. Laubach for his comments. Manuscript editing was by Lana Dieterich. John Ames and Joel Lardon prepared the illustrations under the direction of Cathy Brown, Media Manager. Publication was authorized by the Director, Bureau of Economic Geology. The AAPG Editor thanks the following reviewers for their work on this paper: Bradford E. Prather and Mihaela S. Ryer.

Bubble maps of normalized per-well first-year production and total cumulative production allow qualitative conclusions regarding geologic controls on production distribution. Most wells with optimal gas production occur within two northwest-trending production fairways that coincide with primary sandstone trends of one or more reservoir systems. Highest perwell oil production exists where lower Atoka reservoir facies occur above oil-prone Barnett Shale source rocks (vitrinite reflectance <1.1% Ro) in the western and northwestern parts of the study area. Widespread fault-bounded, karst-produced sag structures that extend vertically from the source rocks through the lower Atoka Group most likely served as hydrocarbonmigration conduits and formed traps for both oil and gas.

This article is dedicated to the memory of John A. Jack Jackson, whose pioneering geologic insight led to the discovery of Boonsville field.

INTRODUCTION The two current primary producing stratigraphic zones in the northern Fort Worth Basin are the Mississippian Barnett Shale and the younger Pennsylvanian (lower Atokan Stage) lower Atoka Group (Bend Conglomerate) (Figure 1). Natural gas production from the Barnett Shale of the Newark East field that began in about 1998 has established it as one of the currently most active and prolific natural gas plays in the United States (Montgomery et al., 2005; Pollastro et al., 2007). In contrast, the shallower siliciclastic lower Atoka play is mature and has been steadily producing primarily natural gas from multiple fields since the early 1950s. Boonsville field, the focus of this article, encompasses approximately 2300 mi2 (6000 km2) in Wise County and parts of adjacent counties (Figure 2) and is by far the largest of the lower Atoka fields in the play, as gauged by total gas and oil production, areal extent, and number of producing wells. As of January 2011, lower Atoka reservoirs in all fields had collectively produced more than 3.2 tcf of natural gas and more than 36.3 million bbl of oil from more than 5700 total wells (IHS Energy, Inc., 2011). Although Boonsville field and the smaller lower Atoka fields are in advanced stages of production, operators continue to explore the interval for infield prospects. However, no published data exist regarding the play-scale distribution of individual reservoir units of the lower Atoka succession to help guide drilling. To address this issue, we conducted a study of the chronostratigraphic framework of the interval in all of Wise County and easternmost Jack County (960 mi2 [2500 km2]) of the northern Fort Worth Basin (Figure 2) to construct highresolution maps of gross sandstone trends and depositional systems. The study area covers the core region of the lower Atoka

1302

Reservoir Systems of the Lower Atoka Group, Northern Fort Worth Basin

Figure 1. Lithostratigraphy of the northern Fort Worth Basin. Terminology is derived from several sources discussed in the article. GR = gamma ray; SP = spontaneous potential; Res = deep resistivity.

play, which coincides with the area of greatest well concentration in Boonsville field.

OBJECTIVES The main objectives of this study are to construct a regional chronostratigraphic framework of the lower Atoka Group, gain a more precise view and

a better understanding of the lower Atoka Group reservoirs sandstone distribution and depositional settings, and investigate geologic controls on regional gas and oil production trends of the unit. Tasks that were used to achieve the objectives included (1) correlating key chronostratigraphic surfaces as a basis for reservoir analysis, (2) constructing gross sandstone maps of selected chronostratigraphic intervals to define general reservoir trends and
Hentz et al. 1303

Figure 2. (A) Location of study area, Boonsville field, and regional structural elements associated with the northern Fort Worth Basin. (B) Location of wells, including cored wells, used in correlation and mapping within the study area. Cross sections AA and BB are shown in Figures 6 and 7, respectively. The Boonsville three-dimensional seismic area of Hardage et al. (1996a, b, c) and McDonnell et al. (2007) are also delineated.

interpret depositional settings using core data to better constrain lithologic and genetic interpretations from well logs, (3) assessing structural control on sedimentation, and (4) constructing bubble maps of per-well hydrocarbon production to characterize regional production patterns.

DATABASE AND METHODS For all correlation and mapping in our study area, we used wireline logs from 705 evenly spaced (1 3 mi [1.64.8 km]) wells (Figure 2). Because many of the study wells do not penetrate the top of the Marble Falls Limestone, the top of the shaly Smithwick facies (Figure 1) was used as the interval base to maximize the number of map control points. Primarily gamma-ray logs were used to ensure consistent and accurate interpretation of sandstone
1304

and shale intervals within the lower Atoka succession. Gross sandstone calculation from most wells was based on the gamma-ray log, with a sandstone cutoff value of 75 API units. During an earlier phase of this study, we used well logs covering the entire northern half of the Fort Worth Basin, with spacings of generally 10 mi or more (16 km) to map regional thickness and structural trends of the lower Atoka Group (Figures 3A, 4A). We described and interpreted slabbed whole cores from two lower Atoka wells in the west-central part of the study: Oxy A-4 Tarrant and EP Operating 3 Tarrant County WB (Figure 2). These represent two of the three publicly available whole cores from this succession in the study area. Maharaj and Wood (2009) described the core from the Threshold 33 Yates well, located approximately 3 mi (4.8 km) south of the Oxy A-4 Tarrant well (Figure 2). Data derived from these cores enabled corroboration of

Reservoir Systems of the Lower Atoka Group, Northern Fort Worth Basin

Hentz et al. Figure 3. (A) Isochore map of the lower Atoka Group, northern Fort Worth Basin. The fault locations outside the study area are modified from Thompson (1982). (B) Isochore map of the postSmithwick lower Atoka section in the study area (Figure 1). The fault locations are derived from Figure 4B. U = upthrown side of fault; D = downthrown side. 1305

1306 Reservoir Systems of the Lower Atoka Group, Northern Fort Worth Basin Figure 4. (A) Structural contour map of the top of the lower Atoka Group, northern Fort Worth Basin. The fault trends of the Mineral WellsNewark East fault system outside the study area are modified from Thompson (1982). (B) Structural contour map of the top of flooding surface 3 in the study area (Figure 5). The uncontoured area in the north is the region of super-Smithwick, lower Atoka Group nondeposition. U = upthrown side of fault; D = downthrown side.

depositional facies interpreted regionally from gross sandstone maps and identification of flooding surfaces (FSs). Two types of gas and oil production bubble maps were constructed for production analysis: first-year gas and oil production and cumulative production. A small area (26 mi2 [67 km2]) of publicly available 3-D seismic coverage in the far westcentral part of the study area (Figure 2) has been studied for structural (Hardage, 1996; Hardage et al., 1996a, c; McDonnell et al., 2007) and sedimentologic (Maharaj and Wood, 2009) attributes. We drew on these studies for insights related to our regional investigation.

GEOLOGIC SETTING Lithostratigraphy The Fort Worth Basin (Figure 2) of north Texas is a late Paleozoic foreland basin that developed between the northeast-trending Ouachita fold belt and Bend arch along the southern North American continental margin (Walper, 1982; Ewing, 1991). Subsidence and sedimentation in front of the Ouachita compressional fold thrust belt led to the accumulation of a maximum of approximately 12,000 ft (3600 m) of primarily upper Paleozoic strata (Ng, 1979). Within the basins Pennsylvanian Series, the Morrowan and Atokan stages extend from the top of the Barnett Shale upward to the base of the Caddo limestone (lower Desmoinesian) of the lowermost Strawn Group (Gardner, 1960; Blanchard et al., 1968; Thompson, 1988) (Figure 1). However, no clear consensus currently exists on the lithostratigraphic nomenclature of this succession. It has been variously termed the Bend Group and Atoka Group (e.g., Sellards et al., 1932; Lahti and Huber, 1982). Moreover, the basins sandstone- and/or conglomerate-bearing postMarble Falls lower Atokan Stagethe focus of this articleis also variously wknown as the Bend Conglomerate, Boonsville conglomerate, and Big Saline (e.g., Lovick et al., 1982; Martin, 1982). Following the stratigraphic convention of Turner (1957), Gardner (1960), and Blanchard

et al. (1968), later used in regional studies by Lahti and Huber (1982) and Thompson (1982), we herein use the term lower Atoka Group to define strata between the top of the Marble Falls Limestone and the top of the uppermost sandstone and/or conglomerate unit in the lower half of the siliciclastic Atoka Group succession (Figure 1). The overlying lower part of the upper Atoka Group has a distinctive well-log signature throughout most of the study area that helps further constrain the upper boundary of the lower Atoka interval stratigraphically. This overlying zone is characterized by a deep resistivity curve that exhibits a gradual upward-increasing deflection, recording an upwardcoarsening shale-dominated interval (long known by operators as the Pregnant shale because of its distinctive log-curve profile) (Figure 1). This shaly zone is capped by the Davis sandstone (Herkommer and Denke, 1982), a widespread locally productive unit in the play area. We term the basal part of the lower Atoka Group in the study area the Smithwick facies after its original usage by Turner (1957) as a subsurface term to differentiate this interval of consistently interbedded calcareous shale, shaly limestone, and minor calcareous sandstone (as determined by well-log analysis) from the underlying Marble Falls Limestone and the overlying coarse-siliciclastic lower Atoka reservoir section (Figure 1). The Smithwick facies is comparable in lithology to the Smithwick Shale where it crops out in the Llano uplift at the southern terminus of the Fort Worth Basin (Sellards et al., 1932; Erlich and Coleman, 2005), but the absence of rigorous correlation into the central and northern parts of the basin has prevented clear resolution of the units subsurface lithostratigraphy. The Smithwick facies extends throughout the study area and displays a generally consistent well-log expression that changes gradationally with regional change in thickness. The lower Atoka Group is dominated by welldeveloped sandstones and conglomerates interpreted to be of fluvial, deltaic, and fan-deltaic origin (Lahti and Huber, 1982; Thompson, 1982; Hentz et al., 2006, 2007; Maharaj and Wood, 2009). Thin discontinuous limestone beds are a minor component of the interval. Reservoirs exist throughout
Hentz et al. 1307

the Atoka Group, although most completed zones occur in the lower half of the succession. The thickness of the lower Atoka Group varies considerably, both regionally and within the study area. In the northern half of the Fort Worth Basin, the succession generally thickens from northwest to southeast, ranging from less than 200 ft (<61 m) to more than 2600 ft (>790 m) in thickness (Figure 3A). Within the study area, the isochore interval that was mapped (Figure 3B) does not include the Smithwick facies at the base of the lower Atoka Group (Figure 1). Nevertheless, the primary trend of southeastward thickening is evident, although with notable local variations. This super-Smithwick section is absent in the northernmost part of the study area, but it is as much as approximately 1450 ft (442 m) thick in the southeasternmost part. Regional and Local Structures The northern Fort Worth Basin is bordered by the Muenster and Red River arches to the north, the Ouachita fold belt to the east, and the Bend arch to the west, all features that developed in association with late Paleozoic Ouachita deformation (Flawn et al., 1961; Walper, 1982) (Figure 2). In this region, the lower Atoka strata generally dip northwestward toward the Muenster archs west border fault zone, which consists of a series of reactivated basement faults characterized by displacement down to the southwest (Flawn et al., 1961) (Figure 4A). Faults and Folds Parallel northeast-trending faults, many of which were first mapped and interpreted as normal faults by Thompson (1982), displace the lower Atoka interval in Palo Pinto, Parker, Wise, Tarrant, and Denton counties (Figure 4A). The longest of the faults, first termed the Mineral Wells fault by Thompson (1982) and later, informally, the Mineral WellsNewark East fault system by Pollastro et al. (2004, 2007), extends from central Palo Pinto County through northwest Parker County into southeast Wise and west Denton counties. Our higher resolution structure mapping, however, reveals that the northeastern part of the Mineral WellsNewark East fault system in Wise County is
1308

an assemblage of parallel, regionally discontinuous, northeast-trending faults (Figure 4B). Moreover, we document an opposite direction of apparent offset on the faults that Thompson (1982, her plate XI) mapped in our study area, which is discussed below. Three distinct northeast-trending faults, two of which are associated with northeast-plunging folds, compose the Mineral WellsNewark East fault system in the southeastern part of the study area (Figure 4B). Steward (2007) documented that geoscientists at Mitchell Energy Corporation (Mitchell Energy), a major early operator in Boonsville field, had identified the fault trends from two-dimensional (2-D) seismic data in the late 1980s and subsequently by three-dimensional (3-D) data and well control in the early 1990s, although these findings were not published. The northeast fault, the longest fault within the study area (23 mi [37 km]), also exhibits the greatest amount of offset between the nearest study wells (maximum of 250 ft [76 m]). Our mapping of subAtoka strata in the study area shows that these faults and folds exist stratigraphically from at least as deep as the Barnett Shale and extend through the Marble Falls Limestone into the lower Atoka Group in nearly the same geographic locations. The same structures are more weakly defined at the top of the Ordovician strata (Figure 1) probably because of the existence of fewer study wells that penetrate this deeper horizon. Structure-contour maps of correlated FSs just above the lower Atoka Group show that the faults and folds terminate in the upper part of the lower Atoka succession (about FS 10 11 [Figure 5]), consistent with the Atokan age of latest regional faulting in the southern and central Fort Worth Basin (Ewing, 1991). Fault termination near the top of the lower Atoka Group is also expressed as a systematic upward decrease in fault offset between the nearest study wells, ranging from approximately 50 ft (15 m) to a maximum of approximately 250 ft (76 m) at the base of the unit to approximately 25 to 110 ft (834 m) in the upper part. Narrow (maximum of 3 mi [4.8 km] in width) northeast-plunging synclines formed on the downthrown sides of the northeastern and

Reservoir Systems of the Lower Atoka Group, Northern Fort Worth Basin

southeastern faults in the study area (Figure 4B). Maximum fault offset between study wells is enhanced by the effects of gradual fold downwarp on the two faults immediate northwestern flanks, although the distribution of contour lines directly across the faults indicates apparent offsets of no more than 100 to 150 ft (3146 m) (Figure 4B). The thickness of lower Atoka strata is consistently greater on the downthrown (northwest) sides of all three faults, particularly along the major northeastern fault (Figure 3B). A northeast-plunging anticline is also associated with the upthrown side of the northeastern fault (Figure 4B) and is the site of distinctly thinner lower Atoka strata adjacent to the fault. Collectively, these factors indicate that faulting and folding were contemporaneous with lower Atoka sedimentation. The regional influence of faulting on sedimentation of the lower Atoka Group was first recognized by Thompson (1982). Because of the inferred genetic association of folds with the faults in the study area, we propose that these structures record predominantly compressional deformation and that they are southeastdipping reverse faults. In the late 1980s and early 1990s, geophysicists at Mitchell Energy, using 2-D and 3-D seismic data, also interpreted the faults to be en echelon up-to-the-southeast, and nearly vertical, but they characterized them as wrench faults that exhibit both normal and reverse characteristics at different positions along their trends (Steward, 2007). Sullivan et al. (2006, 2007) similarly characterized the structures as wrench faults from seismic analysis. However, recent interpretation of a larger, proprietary, 3-D seismic data set by Devon Energy Corporation indicates that the faults in the study area were probably originally normal faults that were reactivated and rotated during later regional compression and are currently expressed as

Figure 5. Type log of lower Atoka Group in the study area. The location of the well is shown in Figure 2. Interval flooding surfaces (FS) 8 to 9 only locally appear as two progradational successions; regionally, the interval typically comprises a single upward-coarsening unit (Figures 6, 7). Thin (<10 ft [<3 m]), discontinuous limestone beds locally occur in FS 4 to 5 and FS 7 to 8 successions. GR = gamma ray; Res = deep resistivity. Hentz et al. 1309

near-vertical reverse faults (J. F. White, 2011, personal communication). Inasmuch as our mapping shows that the faults and folds occur in nearly the same geographic locations at the tops of the Ellenburger Group, Barnett Shale, and Marble Falls Limestone (Figure 1), we concur that the fault planes are nearly vertical (>80) and most likely inherited their configuration from steeply dipping basement faults. Our assertion is supported by Elebiju et al. (2010), who documented parallel northeast-trending faults displacing the Ellenburger Group (Figure 1) and underlying Precambrian crystalline basement from two 3-D seismic volumes in southwest and east-central Wise County, where our mapped faults occur. They concluded that these basement faults propagated into the younger Paleozoic succession of the northern Fort Worth Basin. Sullivan et al. (2006, 2007) further surmised that these faults were reactivated during deposition of the Atoka Group siliciclastics. Reverse faults of the Mineral WellsNewark East fault system in the study area record the final stage of deformation in a northwest-oriented compressional stress regime formed by the northwesterly advance of the Ouachita fold belt during the middle Pennsylvanian sub-epoch (Cheney and Goss, 1952; Flawn et al., 1961; Walper, 1982). Termination of the faults in the upper part of the study succession suggests that major fault-forming stresses in the study area, probably also regional, diminished or ended during the early Atokan Age. Karst-Collapse Sags Small, circular, sag structures (16403940 ft [500 1200 m] in diameter) in the lower Atoka Group that formed by stratal deformation above collapsed paleocaves within the Ellenburger Group (Figure 1) were first documented by Hardage et al. (1996a) in the 26-mi2 (67-km2) Boonsville 3-D seismic area of west-central Wise and east-central Jack counties (Figure 2). Subsequent identification of the same features by Sullivan et al. (2006, 2007) in another larger seismic volume covering most of the southwest quadrant of Wise County indicates that they are widespread in this part of the Fort Worth Basin. Collapse sags within the lower Atoka succession are probably best developed in the western half of
1310

the study area, where the upper Paleozoic strata are in the low to moderate range of their regional thickness and where they directly overlie the karsted Ellenburger Group at the regional sub-Mississippian unconformity (Figures 1, 3) (Pollastro et al., 2007, their figures 6, 16A; Steward, 2007). The structures are defined by concentric faults that extend from the Ellenburger strata vertically upward 2500 to 3500 ft (7621067 m) through the Barnett, Marble Falls, and lower Atoka successions (McDonnell et al., 2007). Fault offset varies within and among the structures, but in one example given by McDonnell et al. (2007, their figure 3), offset ranges from 150 to 250 ft (4676 m) probably the range of maximum offset (R. G. Loucks, 2011, personal communication). Spatially, the structures are 1640 to 4920 ft (5001500 m) apart in the Boonsville 3-D seismic area (McDonnell et al., 2007), although elsewhere they can be more than 1 mi (>1.6 km) apart (Sullivan et al., 2007, their figure 3). The largest structures remained active during deposition of the lower Desmoinesian Strawn Group (Figure 1) and probably affected lower Atoka depositional patterns on a local scale (Hardage et al., 1996a).

CHRONOSTRATIGRAPHY Flooding Surfaces and Reservoir Systems Twelve correlated horizons within, and one just above, the lower Atoka Group correspond to gammaray maximums that cap regionally occurring, upwardfining, retrogradational intervals above thicker, upward-coarsening, progradational, and blocky aggradational successions (Figure 5). They are interpreted to represent regional FSs just above allocyclic transgressive units. Three additional regional FSs associated with the Caddo limestone, a regional, unnamed, progradational interval, and the Davis sandstone (Figure 1) were correlated within the upper Atoka Group to aid in regional correlation consistency. The lower Atoka FSs, designated FS 1 to FS 12, can be correlated throughout the study area (Figures 6, 7), and we interpret each of the 11 individual genetic depositional units bounded by

Reservoir Systems of the Lower Atoka Group, Northern Fort Worth Basin

Figure 6. Northwest-southeast cross section AA parallel to the regional increasing-thickness trend (Figure 3) in the study area. The line of the section is shown in Figure 2. FS = flooding surface; GR = gamma ray; Res = deep resistivity. Hentz et al. 1311

Figure 7. Southwest-northeast cross section BB extending to the western border fault of Muenster arch (Figure 2), where lower Atoka reservoir systems intertongue with inferred alluvial-fan deposits derived from the uplift. The line of the section is shown in Figure 2. FS = flooding surface; GR = gamma ray; Res = deep resistivity.

these surfaces to be isochronous. The units are generally 50 to 90 ft (1527 m) thick but range from a maximum of approximately 40 ft (12 m) in the northern and northwestern margins of the study area to as much as 190 ft (58 m) in the southeast, consistent with the regional lower Atoka south1312

eastward-thickening trend in the basin (Figure 3A). These genetic units contain a single sandstone and/ or conglomerate-bearing reservoir interval that occurs either as multiple closely spaced beds with an upward-coarsening well-log signature (e.g., FS 78) or as a single massive unit with a blocky,

Reservoir Systems of the Lower Atoka Group, Northern Fort Worth Basin

blocky-serrate, or upward-fining (e.g., FS 23) log expression. For the purposes of this article, we herein term these chronostratigraphic units reservoir systems (Figure 5). Unlike the other reservoir systems, FS 11 to 12 is a composite succession containing three FS-bounded, progradational, sandstone-bearing units. Because the system is characterized by this consistent regional log-facies expression and all three units are commonly perforated (and productive), however, we correlated it as a single unit. Similar, although local, lower Atoka chronostratigraphic frameworks based on FSs were established by Hardage et al. (1996b, c) and Maharaj and Wood (2009) in the extreme west-central part of our study area as part of their investigations of the 26-mi2 (67-km2) Boonsville 3-D seismic data set (Hardage, 1996; Hardage et al., 1996a, c) (Figure 2). All 12 of our FSs extend into the area of seismic coverage (Figure 6), and some are represented in the study cores from that area. Many of our surfaces coincide with the surfaces of these studies. Map Units Three lower Atoka reservoir systems were chosen for mapping gross sandstone distribution and are described in chronologic order: reservoir system 1 (FS 23), reservoir system 2 (FS 67), and reservoir system 3 (FS 78) (Figure 5). These intervals are presented primarily because they are represented in the cores available for this study and they are also among the most commonly completed reservoirs of the lower Atoka Group.

setting (Lahti and Huber, 1982; Thompson, 1982). The underlying Morrowan and older shale-rich slope and basinal facies existed farther to the southeast in an elongate foredeep oriented parallel to the Ouachita fold belt (Walper, 1982). However, by the time of Atoka deposition, the rate of sediment input from the structurally rejuvenated fold belt began to exceed the rate of foredeep subsidence and caused progradation of abundant lower Atoka siliciclastics northwestward from the Ouachita highlands into the study area (Kier et al., 1979). Reservoir System 1 (Flooding Surface 23) Sandstone Geometry and Distribution Reservoir system 1 is characterized by a primary northwest-trending, broad, 7- to 8-mi (11- to 13-km)wide belt of sinuous sandstone bodies with more than 20 ft (>6 m) of gross sandstone that collectively form an anastomosing pattern (Figure 8). Tributary depositional axes with 10 to more than 20 ft (3 to >6 m) of gross sandstone merge from the northeast to form the main depositional axis. Gross sandstone values are locally more than 50 ft (>15 m) within the main depositional axis, but values typically decrease abruptly to less than 10 ft (<3 m) over distances of less than 1 mi (<1.6 km). As with all three of the mapped reservoir systems, reservoir system 1 was not deposited over much of the northern part of the study area (Figure 8). The lower Atoka strata gradually lap out against a broad low-relief paleohigh in the northernmost part of Wise County (Figure 3B). Only the lower Atoka Smithwick facies exists in this region, which is inferred to have composed exposed piedmont terrain on the southern flanks of the Red River and Muenster arches (Figure 2). Log Response and Lithology The gamma-ray log response of the FS 2 to 3 interval along the primary depositional axis is typically a single 10- to 60-ft (3- to 18-m)-thick sandstone unit with a blocky or blocky-serrate profile and a sharp flat (erosional) base (Figure 9). Outside the main depositional axis, the single sandstone unit thins and pinches out into a shale-dominated interval, where the log response is serrate or spiky.
Hentz et al. 1313

DEPOSITIONAL SETTING AND DISTRIBUTION OF RESERVOIR SYSTEMS Areal distribution and inferred depositional patterns displayed in gross sandstone maps of the three selected reservoir systems record a variety of lower Atoka sandstone trends and depositional systems. Pronounced variation in the depositional styles of stratigraphically adjacent reservoir systems 2 and 3 is particularly noteworthy. All lower Atoka deposits in the study area accumulated in a shelf

Figure 8. Gross sandstone map of reservoir system 1 (flooding surface [FS] 23) recording anastomozing fluvial-channel pattern of braid-plain system sourced by the Ouachita fold belt to the southeast, Muenster arch to the northeast, and exposed low-relief terrain underlain by Smithwick facies on the southern flanks of Red River and Muenster arches to the north. The location of faults of the Mineral WellsNewark East fault system is derived from Figure 4B. U = upthrown side of fault; D = downthrown side.

The FS 2 to 3 interval in the core from the Oxy A-4 Tarrant well (Figures 2, 8) consists of a lower 16-ft (5-m) section of multiple 1- to 4-ft (0.3- to 1.2-m)-thick beds of planar-tabular-bedded chertclast conglomerate and cross-bedded and planartabular-bedded coarse- and very coarse grained sandstone, with an overall blocky gamma-ray response (Figure 10). The base of this coarse-grained interval is a prominent erosion surface at 5566 ft (1697 m). Although the overall grain-size profile of the conglomeratic and coarse-grained sandstone
1314

section in the FS 2 to 3 interval is blocky, individual beds commonly have erosional bases and upward-fining grain-size profiles. The section also contains a 6-in. (15.2-cm) in-situ coal bed abruptly overlain by another 10-ft (3-m) section of gravelly sandstone and chert-clast conglomerate (Figure 10A). The upper part of the cored section comprises burrowed very fine and finegrained sandstones and shales capped by a 3-ft (0.9-m) coal seam (Figure 10C) with a thin siltstone parting.

Reservoir Systems of the Lower Atoka Group, Northern Fort Worth Basin

Figure 9. Cross section perpendicular to the main-channel axis of the braided-river reservoir system 1 (flooding surface [FS] 23). GR = gamma ray; Res = deep resistivity.

Depositional Setting Thompson (1982) interpreted the lower Atoka interval to be a fluvial-dominated fan-delta system sourced primarily by the Ouachita fold belt, with the Muenster arch (Figure 2) being a secondary source. Lahti and Huber (1982) proposed a fluvialdeltaic origin, but with the Muenster and Red River arches being source areas of the succession. Unlike these two investigations, which interpret the depositional setting on the basis of gross sandstone maps of the entire lower Atoka Group, our study used higher well density supplemented by core data to define separate depositional episodes within the unit. Primarily because of the FS 2 to 3 sandstone trends overall southeast-northwest orientation (Figure 8), we concur with Thompson (1982) that the Ouachita fold belt, approximately 60 mi (97 km) to the southeast, was the primary source area for the lower part of the lower Atoka Group. However, reservoir system 1 does not exhibit characteristics of a delta system, such as downslope

(northwestward) bifurcation of gross sandstone depositional axes (i.e., distributary channels) or upward-coarsening, progradational, grain-size log and core profiles. Following McPherson et al. (1987), we invoke a braid-plain setting for this coarse-sediment fluvial system with a distant highland source. Abrupt change in log facies in the northeasternmost part of the study area along the west border fault of the Muenster arch, though, is inferred to record locally developed alluvial-fan and/or fan-delta facies intertonguing with nearly the entire lower Atoka shelf succession (Figure 7). Similar facies relationships exist adjacent to other contemporary Pennsylvanian uplifts in north Texas and south Oklahoma, such as the fan-delta deposits (granite wash) on the northern flank of the Amarillo uplift (Dutton, 1982). Main depositional axes in the FS 2 to 3 interval, composed of conglomerate and cross-bedded coarse- and very coarse grained sandstone in the Oxy A-4 Tarrant core, exhibit wide, sinuous, and
Hentz et al. 1315

Figure 10. Core description and photographs of braid-plain, bed-load fluvial, and peat-swamp deposits of reservoir system 1 (flooding surface [FS] 23) represented in the Oxy A-4 Tarrant well located in Figures 2 and 8. (A) Description and log response from 5537 to 5569 ft (16881697 m). (B) Chert-clast conglomerate at 5550 ft (1692 m) comprises the coarsest deposits of the braid-plain fluvial system. (C) Coal at 5542 ft (1689 m) records peat-swamp facies. GR = gamma ray; Res = deep resistivity.

anastomosing gross sandstone patterns (Figure 8), characteristic features of braided fluvial systems (Smith and Smith, 1980; Makaske, 2001; Bashforth et al., 2010). Conglomerate and sandstone beds within the FS 2 to 3 interval also commonly have sharp, erosional, gamma-ray curve and core bases (Figures 9, 10A), and represent migratory bar forms within traction-load braided river systems, similar to those described by Lunt and Bridge (2004) and Lunt et al. (2004) in the Sagavanirktok River, Alaska. An erosion-based, upward-fining, conglomeratic, and coarse-grained sandstone interval cored from the Threshold 33 Yates well, which is located approximately 3 mi (4.8 km) south of the Oxy A-4
1316

Tarrant well (Figure 2), was described by Maharaj and Wood (2009, their figure 8) as representing coarse-grained fluvial facies. The section occurs within our FS 4 to 5 interval, bears most of the sedimentologic characteristics of the FS 2 to 3 interval, and represents a depositional setting similar to that of the FS 2 to 3 interval. A thin (6-in. [15.2-cm]) coal bed near the middle of the core section and a 3-ft (0.9-m) coal seam at the top of the section in the Oxy A-4 Tarrant core (Figure 10A, C) record periods of interfluvial peat-swamp development. The abrupt upward transition from conglomerates and very coarse grained sandstones to very fine grained sandstone,

Reservoir Systems of the Lower Atoka Group, Northern Fort Worth Basin

muddy siltstone, and coal in the FS 2 to 3 interval suggests channel abandonment on a short time scale, consistent with avulsion in an anastomosing system (Prez-Arlucea and Smith, 1999; Makaske, 2001). The upper coal seam, overlain by inferred distal delta-front facies, is interpreted to represent the initial stages of a regional FS 3 in this lower part of the lower Atoka succession. Structural Influence on Sedimentation Segments of the Mineral WellsNewark East fault system were active during early Atoka deposition and controlled regional channel orientation of reservoir system 1. Sandstone abundance and the general northeast orientation of tributary-channel axes between parallel fault segments B and C (Figure 8) record funneling of sand from the secondary Muenster arch source area via the downthrown side (syncline) of fault C (Figure 4B). Also, in contrast to the generally consistent regional northwest trend of the main channel axis in the western half of the study area, main channel orientations were markedly disrupted by faults A through C in the east. Faulting appears to have also caused broadening of the overall braided river system in the east. Thickening of sandstones within the prominent syncline on the downthrown side of fault B (Figures 3B, 4B) is notably absent. Either fault B was not active during deposition of reservoir system 1 or the associated syncline was the locus of thick silt and mud accumulation between major fluvial tributaries (Figure 8). Reservoir System 2 (Flooding Surface 67) Sandstone Geometry and Distribution Sandstone-body distribution in the FS 6 to 7 interval records two trends in the study area: (1) a primary high-gross sandstone trend of multiple narrow (1.52.5 mi [2.44.0 km]), northwest-northeastelongate belts in the southern and western parts of the study area and (2) a secondary trend of discontinuous northeast-oriented sandstone bodies extending across the west-central part of the study area (Figure 11). Gross sandstone values are as high as 60 ft (18 m) in the primary area of sand accumulation in the southeast and range from ap-

proximately 10 to 30 ft (39 m) in the northwestand secondary northeast-oriented trends. Log Response and Lithology Reservoir system 2 comprises multiple thinner sandstone bodies, in contrast to reservoir system 1 (Figure 12). In this intervals main depositional trends in the south and east, the sandstone bodies form consistently upward-coarsening successions overlain by thinner upward-fining intervals that collectively range from less than 30 to 90 ft (<9 27 m) in thickness. The FS 6 to 7 section occurs in the deepest part (55105473 ft [16791668 m) of the EP Operating 3 Tarrant County WB core, which is near the fringe of the secondary sandstone trend (Figure 11). It consists of an upwardcoarsening section with a 13-ft (4-m) interval of fossiliferous dark-gray to black shale and siltstone with calcareous nodules (Figure 13B) that grades upward into sparsely burrowed fine-grained sandstone and siltstone with small-scale ripples and lenticular bedding (Figure 13C). The succession is capped by an 8-ft (2.4-m) upward-coarsening section of pebbly and shelly medium- to very coarse grained sandstone with inclined bedding (Figure 13D). Bioclastic material in this upper interval consists dominantly of brachiopod and crinoid fragments. The thickest and coarsest individual sandstone bodies in the upper part of the core interval have sharp bases (Figure 13A), suggesting that they are erosional. Depositional Setting We interpret the FS 6 to 7 interval to represent fluvial-dominated deltaic deposits on the basis of its sandstone-body geometry of northwest-northeastbifurcating narrow depositional axes (distributary channels), upward-coarsening log character, and the vertical superposition of facies that record progradation. The intervals overall facies succession, bed forms, and stacking patterns are similar to those of other fluvial-dominated deltaic deposits (e.g., Van Heerden and Roberts, 1988; Tye and Coleman, 1989; Neill and Allison, 2005). The vertical succession in the FS 6 to 7 cored interval comprises prodelta shales at the base overlain by upwardcoarsening shelly and burrowed distal delta-front
Hentz et al. 1317

Figure 11. Gross sandstone map of reservoir system 2 (flooding surface 67) recording a north-northwestwardprograding fluvialdominated deltaic system. The location of faults of Mineral WellsNewark East fault system is derived from Figure 4B. U = upthrown side of fault; D = downthrown side.

sandstones and channel mouth-bar facies consisting of planar-stratified medium- and coarse-grained sandstone (Figure 13A, D). The gross sandstone geometry of the FS 6 to 7 interval (Figure 11) is comparable to that of highstand, fluvial-dominated, deltaic deposits of the Woodbine Group in the East Texas field (Ambrose et al., 2009). The examples in the East Texas field, however, are smaller in scale than those in the FS 6 to 7 interval, with distributary channel depositional axes less than 2000 ft (<600 m) wide in contrast to those in the FS 6 to 7 interval, which are approxi1318

mately 5000 ft (1524 m) wide in the west-central part of the study area. In addition, the distal part of the deltaic facies tract near the EP Operating 3 Tarrant County WB core (Figures 11, 13) has a minor strike-aligned (northeast-trending) gross sandstone component, suggesting marine reworking of delta-front facies. The sandstone beds at the top of the FS 6 to 7 interval from 5477 to 5484 ft (1670 1672 m), dominated by planar-stratified mediumand coarse-grained sandstone with abundant crinoid fragments (Figure 13A, D), is consistent with highenergy marine reworking of a channel mouth bar.

Reservoir Systems of the Lower Atoka Group, Northern Fort Worth Basin

Figure 12. Cross section perpendicular to the direction of progradation of fluvial-dominated deltaic reservoir system 2 (flooding surface [FS] 67). GR = gamma ray; Res = deep resistivity.

The overall north-northwestern direction of progradation of reservoir system 2 suggests that the Ouachita fold belt to the southeast, not the Muenster arch, was the primary source area for these deposits. Within interaxial areas, the gamma-ray log expression of the FS 6 to 7 interval is commonly weakly upward coarsening to serrate (Figure 12), recording interdistributary facies, some of which are thin limestone beds (Thompson, 1982, her figures 8, 9) expressed as thin high-resistivity units. Structural Influence on Sedimentation Syndepositional faulting along the Mineral Wells Newark East fault system was also a primary con-

trol on the position and orientation of the FS 6 to 7 deltaic system in the study area (Figure 11). The southeastern extent of the delta was restricted by the downthrown block of fault C, and the primary area of deltaic sand accumulation was the syndepositional syncline associated with fault C (Figure 4B). Moreover, faults A and B appear to have collectively exerted regional control over the northwestward extent of the primary deltaic sand depocenter. The syncline associated with fault B (Figure 4B) was the site of considerable distal deltaic sand deposition. North- and northwest-oriented distributary channel sandstone trends in the southernmost part of the study area abruptly diverge
Hentz et al. 1319

Figure 13. Core description and photographs of reservoir system 2 (flooding surface [FS] 6 7) represented in the EP Operating 3 Tarrant County WB well located in Figures 2 and 11. (A) Description and log response from 5510 to 5473 ft (16791668 m). (B) Fossiliferous prodelta shale at 5498 ft (1676 m). (C) Sparsely burrowed siltstone and shale in distal delta-front facies at 5488 ft (1673 m). (D) Granular coarsegrained sandstone in proximal delta-front facies at 5479 ft (1670 m). GR = gamma ray; Res = deep resistivity.

1320

Reservoir Systems of the Lower Atoka Group, Northern Fort Worth Basin

Figure 14. Gross sandstone map of reservoir system 3 (flooding surface 78) recording a southeast-flowing, low-sinuosity, incised river system. The location of faults of the Mineral WellsNewark East fault system is derived from Figure 4B. U = upthrown side of fault; D = downthrown side.

to the northeast between, and parallel to, faults B and C. The longest northwest-oriented distributary channel system breeches the fault segments at the northern and southern terminations of faults A and B, respectively, where offset was minimal. Fault A seems to have exerted less pronounced control over the orientation of delta-front sandstones through which it cuts, although the consistent termination of these sandstone trends just 2 to 4 mi (36 km) northwest of the fault is probably caused by the presence of more local accommodation on the faults downthrown side.

Reservoir System 3 (Flooding Surface 78) Sandstone Geometry and Distribution The FS 7 to 8 interval is dominated by a wide (3 6 mi [4.89.6 km]), mostly southwest-trending belt of 20 to 60 ft (618 m) of high gross sandstone traversing the southern half of the study area, an orientation that is orthogonal to the depositional trends of the two older genetic intervals FS 2 to 3 and FS 6 to 7 (Figures 8, 11, 14). Most of the primary, generally linear, trend of the FS 7 to 8 interval exhibits low sinuosity, in contrast to the
Hentz et al. 1321

Figure 15. Cross section perpendicular to the main-channel axis of the fluvial-estuarineincised channel fill of reservoir system 3 (flooding surface [FS] 78).

anastomosing gross sandstone trend of reservoir system 1 (FS 2 to 3) (Figure 8). In the southeastern part of the study area, however, the primary sandstone trend turns sharply to the south and probably to the east or northeast, parallel to the regional orientation of the fluvial system (Figure 14). This feature is probably a result of syndepositional structural influences, as will be discussed below. South of the primary trend in the map area (010 ft [03 m] of gross sandstone), the thickest gross sandstone of the underlying FS 6 to 7 interval is located (Figures 11, 14); compensational deposition possibly explains this stacking relationship. Log Response and Lithology The dominant gamma-ray log response of the FS 7 to 8 interval along the main depositional axis in the southern part of the study area is overall upward fining, but with a prominent lower zone exhibiting blocky and blocky-serrate responses (Figure 15). The lower sandstone zone, typically the thickest in the intervalcommonly 30 to 45 ft (914 m) thickconsistently exhibits an inferred erosional
1322

base in contact with underlying shale-dominated zones. On the margins of the main depositional axis, particularly in the northern half of the map area, the FS 7 to 8 interval consistently exhibits serrate log patterns of lower gross sandstone values that record thin sandstone beds interbedded with siltstone and shale beds in an overall upwardcoarsening succession (Figure 15). The FS 7 to 8 interval in the EP Operating 3 Tarrant County WB well (Figures 2, 16A) consists of a lower unit of burrowed gray shale at the base that grades upward to an 8-ft (2.4-m) upwardcoarsening section composed of thin sharp-based beds of very fine grained sandstone interbedded with siltstone (Figure 16B). This upward-coarsening section is overlain by a 7-ft (2.1-m) section of black fissile shale and siltstone with lenticular beds of rippled, very fine grained sandstone (Figure 16C) that is in turn truncated at 5445.5 ft (1660.2 m) by a fine- to medium-grained, cross-bedded sandstone that grades upward into fine-grained sandstone with abundant flaser ripples and mud drapes (Figure 16D). The FS 8, not cored in this well, is

Reservoir Systems of the Lower Atoka Group, Northern Fort Worth Basin

Figure 16. Core description and photographs of reservoir system 3 (flooding surface [FS] 78) in the EP Operating 3 Tarrant County WB well located in Figures 2 and 14. (A) Description and log response from 5478.5 to 5425 ft (1669.81653 m). (B) Distal delta-front deposits composed of thin (0.4-in. [1-cm]-scale) beds of very fine grained sandstone with clay clasts and shell fragments interbedded with dark silty mudstone at 5454.5 ft (1662.5 m). (C) Crossbedded fluvial estuarine channel facies at 5445 ft (1660.7 m). (D) Draped lenticular bedding in upper estuarine-fill deposits at 5434.5 ft (1657.5 m). GR = gamma ray; Res = deep resistivity. Hentz et al. 1323

overlain by laminated and mud-draped, very fine grained sandstone at 5432.5 ft (1656.3 m). Depositional Setting We interpret the primary southwest-oriented sandstone trend of reservoir system 3 (FS 78) to represent a generally low sinuosity fluvial system mostly on the basis of its gross sandstone pattern and consistent blocky to blocky-serrate and upward-fining gamma-ray log facies, both of which indicate an aggradational depositional system. The flat inferred erosional base of these facies indicates regional channel incision. These log facies contrast with the upward-coarsening (progradational) patterns of multiple thinner sandstone beds on the margins of the primary high-gross sandstone trend (Figure 15). A pattern of tributary depositional axes merges into the main depositional axis from the north, northeast, and southwest (Figure 14), also supporting a fluvial depositional origin. The river system most likely flowed from the southeast, with the Ouachita fold belt being the main source area. Southerly and southwesterly tributary trends also point toward the exposed terrain to the north and the Muenster arch (30 mi [48 km] to the northeast) as secondary sand sources. On the basis of evidence from core and regional well-log data, we infer the channel-fill depositional system to be an incised-valley system (e.g., Van Wagoner et al., 1990) at least partly composed of estuarine channel-fill facies regionally truncating an older progradational (deltaic) episode of deposition. The primary fluvial system incises the older progradational deposits (Figure 15), and the gross sandstone map (Figure 14) delineates sandstone components of both the channel fill and, to a lesser extent, the progradational facies of the FS 7 to 8 interval. The cored interval in the EP Operating 3 Tarrant County WB well (Figures 14, 16A) is on the periphery of a tributary channel trend where less than 10 ft (<3 m) of gross sandstone is in the northwestern part of the study area. Therefore, only a thin channel-fill interval of the fluvial system is represented, which is composed of a lower section of shaly, prodelta, and upward-coarsening deltafront deposits incised by an upper section of tidally
1324

influenced estuarine channel sandstones (Figure 16). The lower delta-front section is composed of numerous and thin (0.2- to 0.4-in. [0.5- to 1-cm-scale) beds of very fine grained sandstone with minute flattened clay clasts and shell fragments (Figure 16B) that represent hyperpycnal turbidite deposits derived from sediment-laden density flows on the delta front. Recent studies of modern and ancient deltas with a significant fluvial component suggest that density flows, instead of hypopycnal plumes with flocculants, are dominant sedimentary features in the delta front (Olariu and Bhattacharya, 2006; Bhattacharya and MacEachern, 2009). In contrast, the upper sandy part of the FS 7 to 8 interval from 5433 to 5445 ft (1656.41660.1 m) represents tidally modified estuarine channel deposits of a tributary of an inferred incised-valley system. This interpretation is based on (1) its erosional contact with a fine-grained section of burrowed distal-deltaic siltstone and fissile shale and (2) the presence of a fine- to medium-grained, upward-fining sandstone succession with bed forms that range from shale-draped, oversteepened foresets at the base (Figure 16C) to finely laminated and shale-draped, fine-grained sandstone at the top (Figure 16D). Similar facies associations have been described in estuarine channel facies in the Miocene Bear Lake Formation in Alaska (Finzel et al., 2009) and aggradational, tidally influenced, fluvial channel deposits in the lower Holocene Mekong Delta (Tamura et al., 2009). Dalrymple and Choi (2007) described stratification types that are typical in tidally modified fluvial channel deposits, including draped foresets, double-mud drapes, reactivation surfaces, and lenticular bedding, many of which features are present in the FS 7 to 8 core (Figure 16D). Structural Influence on Sedimentation Fault segments of the Mineral WellsNewark East fault system determined the location of the FS 7 to 8 fluvial channel system in the study area. The channels upper reach in the southeastern part of the study area is aligned along the downthrown edge of fault C (Figure 14). Other faults of the Mineral WellsNewark East fault system just outside the map area to the southeast (Figure 4A)

Reservoir Systems of the Lower Atoka Group, Northern Fort Worth Basin

probably also controlled the highly sinuous and incised character of the channel system in this area. The downthrown sides of faults A and Band the syncline associated with fault B (Figure 4B) collectively formed a narrow, northeast-oriented, linear, topographic low that provided the pathway for most of the sands that define the primary elongate channel trend in the central part of the map area (Figure 14). A major tributary is also aligned with part of the downthrown side (syncline) of fault B along the northeastern part of the fault. The primary channel trend shifts abruptly northwestward approximately 3 mi (4.8 km) from the downthrown side of fault B to that of fault A at their local area of termination, recording a syndepositional fault-produced paleotopography.

Atoka reservoir characteristics, production data represented on the two types of bubble maps, and qualitative conclusions regarding the geologic controls on production distribution are herein presented. Reservoir Characteristics Many lower Atoka wells in Boonsville field produce hydrocarbons from two or three reservoir systems. Individual pay zones are typically lenticular in geometry and characterized by a wide variability in areal extent. The trapping mechanism of reservoirs is stratigraphic, reflecting not only the discontinuity of fluvial and deltaic sandstone bodies at the scale of current well spacing (40160 ac) (Hardage et al., 1996a; Carr et al., 1997) but also the highly variable porosity values within individual reservoir bodies caused by heterogeneous cementation (Gardner, 1960; Blanchard et al., 1968; Glover, 1982). Productive zones have porosity values ranging from 3 to 27%, although most reservoirs need to be fractured because of generally low permeability values (<1 md) (Blanchard et al., 1968; Thompson, 1982). The primary source rock of gas and oil in lower Atoka reservoirs is the Barnett Shale (Hill et al., 2007), and the Mineral WellsNewark East fault system probably controlled some migration of the hydrocarbons from the source rock to the lower Atoka section (Jarvie et al., 2005; Pollastro et al., 2007). The fault-bounded, karst-produced sag structures in the lower Atoka succession (Hardage et al., 1996a; Sullivan et al., 2006, 2007; McDonnell et al., 2007) that most likely occur throughout the western part of the study area were also probably hydrocarbon migration pathways. Shale intervals that are regionally cyclically interstratified with the reservoir facies (Figures 57) provide the seals for lower Atoka hydrocarbon traps. Gas Production Lower Atoka wells with the highest first-year gas production (0.75 to >2 bcf) (Figure 17) and highest total cumulative gas production (2 to >6 bcf) (Figure 18) areally coincide and occur mostly
Hentz et al. 1325

LOWER ATOKA PRODUCTION TRENDS Analysis of distribution and trends of hydrocarbon production from the lower Atoka Group in the study area is based mostly on the construction of two types of gas- and oil-production bubble maps. (1) First-year gas and oil production is defined as the cumulative gas or oil production from the first year of each wells history, which is typically a wells best period of performance. Calculating firstyear production enables normalization of production histories of individual wells of varying ages. (2) Cumulative production is the total gas or oil production from each well. All production data are current as of January 2011 (IHS Energy, Inc., 2011). The bubble maps represent all 3911 productive lower Atoka wells within the study area, most of which have produced both gas and liquids. Linking first-year and cumulative production tallies specifically to reservoir systems 1 through 3 and to the other eight reservoir systems (Figure 5) is hampered because multiple zones have commonly been completed in each well, leaving no reliable means of allocating production volumes to specific reservoir systems. Individual completions within each well can be identified, but it is impossible to determine which one (or combination) of them contributed to the per-well production totals represented on bubble maps. Instead, lower

Figure 17. Map of first-year cumulative gas production per well from the lower Atoka Group in the study area. The fault locations are from Figure 4B. U = upthrown side of fault; D = downthrown side.

northwest of the Mineral WellsNewark East fault system within two northwest-trending production fairways 11 and 22 mi (18 and 35 km) long in the central and southwestern parts of the study area, respectively. The fairways are perpendicular to fault orientation and thickness trends (Figure 3B), suggesting little or no regionally consistent association with these features. Although fault offset may have been a local influence on trap formation along parts of the two northern faults (Figures 17, 18), the Mineral WellsNewark East fault system and its associated folds generally do not appear to have exerted regional influence on trap location or well productivity in the study area. Northwest-oriented fractures also may have contributed to reservoir quality close to the faults; fracture zones associated with reverse faults and normal faults reactivated as reverse faults form perpendicular to such struc1326

tures (Hancock, 1985). The faults were probably hydrocarbon migration pathways (Jarvie et al., 2005; Pollastro et al., 2007), although the areal extent of their influence is not documented. Published discussion of lower Atoka reservoirs indicates that local structure is also thought not to have controlled lower Atoka hydrocarbon accumulation at the scale of individual traps (Blanchard et al., 1968; Glover, 1982). However, the effects of the widespread, fault-bounded, karst-produced sag structures on lower Atoka reservoir compartmentalization were surmised by operators by at least the early 1990s (Steward, 2007) and were demonstrated shortly thereafter by Hardage et al. (1996a). The structures most likely also had a direct influence on gas migration from Barnett Shale source rocks to lower Atoka reservoirs. The production fairways are subparallel to structural strike (Figure 4B), also

Reservoir Systems of the Lower Atoka Group, Northern Fort Worth Basin

Figure 18. Map of total cumulative gas production per well from the lower Atoka Group in the study area. The fault locations are from Figure 4B. U = upthrown side of fault; D = downthrown side.

suggesting at least some basic regional structural control on the general distribution of optimal gas production. The clustering of high-quality gas wells within the two fairways probably reflects the concentration of favorable reservoir facies and stratigraphic/ diagenetic traps along parts of the primary sandstone trends of one or more reservoir systems. For example, the northwest-trending braided river channel complex in the western half of reservoir system 1 (Figures 5, 8) almost precisely coincides with the western gas production fairway (Figures 17, 18). However, the eastern gas production fairway lies mostly within a low-sandstone interfluvial area of the same reservoir system. In another instance, the longest northwest-oriented distributary channel system of reservoir system 2 (Figures 5, 11) aligns

with the eastern gas production fairway, although the reservoir systems main sandstone depocenter in the southeast (Figure 11) occurs in an area of few high-quality gas wells. Oil Production In contrast to the clustering of the best lower Atoka gas wells within fairways, the distribution of the best lower Atoka oil wells is more diffuse (Figures 19, 20). However, like the gas wells, the high-volume oil wells are restricted to a region northwest of the Mineral WellsNewark East fault system. These wells occur in a weakly arcuate belt (Figure 20) in the most updip western and northwestern parts of the study area (Figure 4B), where the lower Atoka succession is thinnest (Figure 3B). Like the gas
Hentz et al. 1327

Figure 19. Map of first-year cumulative oil production per well from the lower Atoka Group in the study area. The fault locations are from Figure 4B. The isoreflectance contour (1.1% Ro) derived from Pollastro et al. (2007) marks the approximate areal boundary between oil-prone Barnett shale source rocks (north), where Ro is less than 1.1%, and gas-prone source rocks (south) of the lower Atoka Group. U = upthrown side of fault; D = downthrown side.

wells, the oil wells with the highest first-year oil production (20 to >60 thousand bbl) (Figure 19) and highest total cumulative oil production (50 to >200 thousand bbl) (Figure 20) areally coincide. This area of primary lower Atoka oil production also closely coincides with the region where thermal maturities of the underlying Barnett Shale source rockdetermined by vitrinite reflectance (Ro)define the Barnetts primary area of oil production (Pollastro et al., 2007). Jarvie et al. (2005) determined that the cracking of Barnett oil to gas occurs at about 1.1% Ro and that good correlation exists between 0.6 and 1.1% Ro and areas of historic Barnett oil production. Although the 1.1% isoreflectance contour presented in Pollastro et al.
1328

(2007, their figures 12, 18) is approximate, reasonably good correlation between the areas of oilprone Barnett Shale source rocks (<1.1% Ro) and of most primary lower Atoka oil production within the study area also can be inferred (Figures 19, 20). The presence of oil-prone source rocks stratigraphically below most high-quality oil wells suggests that this relation is at least one of the regional controls on lower Atoka oil distribution. The generally diffuse arrangement of the best oil wells in the western half of the study area may also collectively reflect (1) local discontinuity of stratigraphic traps within thin updip strata of multiple reservoir systems and (2) the spatial distribution in the same area of localized karst-produced sag structures that

Reservoir Systems of the Lower Atoka Group, Northern Fort Worth Basin

Figure 20. Map of total cumulative oil production per well from the lower Atoka Group in the study area. The fault locations are from Figure 4B. The isoreflectance contour (1.1% Ro) derived from Pollastro et al. (2007) marks the approximate areal boundary between oilprone Barnett Shale source rocks (north), where it is less than 1.1% Ro, and gas-prone source rocks (south) of the lower Atoka Group. U = upthrown side of fault; D = downthrown side.

probably formed traps and served as conduits for oil migration from the Barnett Shale source rocks to the lower Atoka reservoirs.

CONCLUSIONS This study defines lower Atoka depositional systems in the mature play area at a higher degree of resolution than has previously been achieved. Selected reservoir systems record varied depositional facies in the lower Atoka succession: braided river, fluvial-dominated delta, and lowsinuosity incised fluvial system. Sediment was derived from source areas primarily to the southeast (Ouachita fold belt) and northeast (Muenster

arch). Piedmont terrain underlain by Smithwick facies on the southern flanks of the Red River and Muenster arches composed a minor source to the north. In contrast to previous studies that typically define a single southeast-dipping normal fault (Mineral Wells fault) within Boonsville field, we delineated a zone of three parallel northeast-trending faults of the Mineral WellsNewark East fault system in the study area. The faults are genetically associated with parallel conterminous folds and are interpreted to be high-angle (>80) southeast-dipping reverse faults rooted in older faults displacing Precambrian crystalline basement. Lithofacies maps of depositional systems demonstrate that faults and associated folds were
Hentz et al. 1329

active during sedimentation and directly controlled sandstone trends and depositional patterns of lower Atoka reservoir systems. Regional isochore patterns of the lower Atoka Group, which show thickening of strata on the downthrown sides of the faults and thinning on the upthrown sides, support this interpretation. Bubble maps of normalized first-year well production performance and total cumulative production indicate that the distribution of bestproducing gas and oil wells within the study area was influenced only minimally by these major structures. Instead, probable principal controls on the areal distribution of hydrocarbon production are fault-bounded, karst-produced sag structures primarily in the western half of the study area that extend through the Barnett Shale source rock into the lower Atoka section, providing conduits for gas and oil migration to localized stratigraphic traps. The areal correlation of most high-quality lower Atoka oil wells directly above oil-prone (<1.1% Ro) Barnett Shale source rocks in the western and northwestern parts of the study area is compelling evidence of this regional influence on oil distribution. Optimal gas-well distribution was probably also directly influenced by the presence of reservoir facies and traps along primary sandstone trends of one or more fluvial and deltaic reservoir systems.

REFERENCES CITED
Ambrose, W. A., T. F. Hentz, F. Bonnaffe, R. G. Loucks, L. F. Brown Jr., and F. P. Wang, 2009, Sequence-stratigraphic controls on complex reservoir architecture of highstand fluvial-dominated deltaic and lowstand valley-fill deposits in the Woodbine Group, East Texas field: Regional and local perspectives: AAPG Bulletin, v. 93, p. 231 269, doi:10.1306/09180808053. Bashforth, A. R., H. J. Falcon-Lang, and M. R. Gibling, 2010, Vegetation heterogeneity on a Late Pennsylvanian braidedriver plain draining the Variscan Mountains, La Magdalena coalfield, northwestern Spain: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 292, p. 367390, doi:10 .1016/j.palaeo.2010.03.037. Bhattacharya, J. P., and J. A. MacEachern, 2009, Hyperpycnal rivers and prodeltaic shelves in the Cretaceous seaway of North America: Journal of Sedimentary Research, v. 79, p. 184209, doi:10.2110/jsr.2009.026.

Blanchard, K. S., O. Denman, and A. S. Knight, 1968, Natural gas in Atokan (Bend) section of northern Fort Worth Basin, in B. W. Beebe and B. F. Curtis, eds., Natural gases of North America: AAPG Memoir 9, p. 14461454. Carr, D. L., R. Y. Elphick, R. A. Johns, and L. S. Foulk, 1997, High-resolution reservoir characterization of mid-continent sandstones using wireline resistivity imaging, Boonsville (Bend Conglomerate) gas field, Fort Worth Basin, Texas: The Log Analyst, v. 38, p. 5470. Cheney, M. G., and L. F. Goss, 1952, Tectonics of central Texas: AAPG Bulletin, v. 36, no. 12, p. 22372265. Dalrymple, R. W., and K. Choi, 2007, Morphologic and facies trends through the fluvial-marine transition in tidedominated depositional systems: A schematic framework for environmental and sequence-stratigraphic interpretation: Earth-Science Reviews, v. 81, p. 135174, doi:10 .1016/j.earscirev.2006.10.002. Dutton, S. P., 1982, Pennsylvanian fan-delta and carbonate deposition, Mobeetie field, Texas Panhandle: AAPG Bulletin, v. 66, no. 4, p. 389407. Elebiju, O. O., G. R. Keller, and K. J. Marfurt, 2010, Investigation of links between Precambrian basement structure and Paleozoic strata in the Fort Worth Basin, Texas, U.S.A., using high-resolution aeromagnetic (HRAM) data and seismic attributes: Geophysics, v. 75, no. 4, p. B157B168, doi:10.1190/1.3435939. Erlich, R. N., and J. L. Coleman Jr., 2005, Drowning of the upper Marble Falls carbonate platform (Pennsylvanian), central Texas: A case of conflicting signals?: Sedimentary Geology, v. 175, p. 479499, doi:10.1016/j.sedgeo .2004.12.017. Ewing, T. E., 1991, The tectonic framework of Texas: Austin, Texas, University of Texas at Austin, Bureau of Economic Geology, 36 p. Finzel, E. S., K. D. Ridgway, R. R. Reifenstuhl, R. B. Blodgett, J. M. White, and P. L. Decker, 2009, Stratigraphic framework and estuarine depositional environments of the Miocene Bear Lake Formation, Bristol Bay Basin, Alaska, onshore equivalents to potential reservoir strata in a frontier gas-rich basin: AAPG Bulletin, v. 93, no. 3, p. 379 405, doi:10.1306/10010808030. Flawn, P. T., A. G. Goldstein Jr., P. B. King, and C. E. Weaver, 1961, The Ouachita System: Austin, Texas, University of Texas at Austin, Bureau of Economic Geology Publication 6120, 401 p. Gardner Jr., R. A., 1960, The Boonsville (Bend conglomerates, gas) field, Wise County, Texas: Abilene Geological Society, Geologic Contributions 1960, p. 718. Glover, G., 1982, A study of the Bend Conglomerate in S. E. Maryetta area, Boonsville field, Jack County, Texas, in C. A. Martin, ed., Petroleum geology of the Fort Worth Basin and Bend arch area: Dallas Geological Society, p. 353364. Hancock, P. L., 1985, Brittle microtectonics: Principals and practice: Journal of Structural Geology, v. 7, p. 437457. Hardage, B. A., 1996, Boonsville 3-D data set: The Leading Edge, v. 15, no. 7, p. 835837, doi:10.1016/0191-8141 (85)90048-3. Hardage, B. A., D. L. Carr, D. E. Lancaster, J. L. Simmons Jr., R. Y., Elphick, V. M. Pendleton, and R. A. Johns, 1996a,

1330

Reservoir Systems of the Lower Atoka Group, Northern Fort Worth Basin

3-D seismic evidence of the effects of carbonate karst collapse on overlying clastic stratigraphy and reservoir compartmentalization: Geophysics, v. 61, p. 13361350, doi:10.1190/1.1444057. Hardage, B. A., D. L. Carr, D. E. Lancaster, J. L. Simmons Jr., D. S. Hamilton, R. Y. Elphick, K. L. Oliver, and R. A. Johns, 1996b, 3-D seismic imaging and seismic attribute analysis of genetic sequences deposited in low-accommodation conditions: Geophysics, v. 61, p. 13511362, doi:10 .1190/1.1444058. Hardage, B. A., J. L. Simmons Jr., D. E. Lancaster, R. Y. Elphick, R. D. Edson, and D. L. Carr, 1996c, Boonsville 3-D seismic data set: Austin, Texas, University of Texas at Austin, Bureau of Economic Geology, 40 p. Hentz, T. F., J. A. Kane, W. A. Ambrose, and E. C. Potter, 2006, Depositional facies, reservoir distribution, and infield potential of the lower Atoka Group (Bend Conglomerate) in Boonsville field, Fort Worth Basin, Texas: New look at an old play (abs.): AAPG Annual Convention Abstracts Volume, v. 15, p. 46. Hentz, T. F., E. C. Potter, and M. A. Adedeji, 2007, Reservoirscale depositional facies, trends, and controls on sandstone distribution of the lower Atoka Group (Bend Conglomerate), Fort Worth Basin, Texas (abs.): AAPG Annual Convention Abstracts Volume, v. 16, p. 62. Herkommer, M. A., and G. W. Denke, 1982, Stratigraphy and hydrocarbons, Parker County, Texas, in C. A. Martin, ed., Petroleum geology of the Fort Worth Basin and Bend arch area: Dallas Geological Society, p. 97127. Hill, R. J., D. M. Jarvie, J. Zumberge, M. Henry, and R. M. Pollastro, 2007, Oil and gas geochemistry and petroleum systems of the Fort Worth Basin: AAPG Bulletin, v. 91, no. 4, p. 445 473, doi:10.1111/j.1365-3091 .2004.00628.x. IHS Energy, Inc., 2011, U.S. production and well history control databases: Englewood, Colorado, CD-ROM. Jarvie, D. M., R. J. Hill, and R. M. Pollastro, 2005, Assessment of the gas potential and yields from shales: The Barnett Shale model, in B. J. Cardott, ed., Unconventional energy resources in the southern mid-continent, 2004 symposium: Oklahoma Geological Survey Circular 110, p. 3750. Kier, R. S., L. F. Brown Jr., and E. F. McBride, 1979, The Mississippian and Pennsylvanian (Carboniferous) systems in the United StatesTexas: U.S. Geological Survey Professional Paper 1110-S, 45 p. Lahti, V. R., and W. F. Huber, 1982, The Atoka Group (Pennsylvanian) of the Boonsville field area, north-central Texas, in C. A. Martin, ed., Petroleum geology of the Fort Worth Basin and Bend arch area: Dallas Geological Society, p. 377399. Lovick, G. P., C. G. Mazzine, and D. A. Kotila, 1982, Atokan clastics: Depositional environments in a foreland basin, in C. A. Martin, ed., Petroleum geology of the Fort Worth Basin and Bend arch area: Dallas Geological Society, p. 193212. Lunt, I. A., and J. S. Bridge, 2004, Evolution and deposits of a gravelly braid bar, Sagavanirktok River, Alaska: Sedimentology, v. 51, p. 415432. Lunt, I. A., J. S. Bridge, and R. S. Tye, 2004, A quantitative,

three-dimensional depositional model of gravelly braided rivers: Sedimentology, v. 51, p. 377414, , doi:10.1111/j .1365-3091.2004.00627.x. Maharaj, V. T., and L. J. Wood, 2009, A quantitative paleogeographic study of the fluvio-deltaic reservoirs in the Atoka interval, Fort Worth Basin, Texas, U.S.A.: Gulf Coast Association of Geological Societies Transactions, v. 59, p. 495509. Makaske, B., 2001, Anastomosing rivers: A review of their classification, origin, and sedimentary products: EarthScience Reviews, v. 53, no. 3, p. 149196, doi:10.1016 /S0012-8252(00)00038-6 . Martin, C. A., (ed.), 1982, Petroleum geology of the Fort Worth Basin and Bend Arch area: Dallas Geological Society, 442 p. McDonnell, A., R. G. Loucks, and T. Dooley, 2007, Quantifying the origin and geometry of circular sag structures in northern Fort Worth Basin, Texas: Paleocave collapse, pull-apart fault systems, or hydrothermal alteration?: AAPG Bulletin, v. 91, p. 1295 1318, doi:10.1306 /05170706086. McPherson, J. G., G. Shanmugam, and R. J. Moiola, 1987, Fan-deltas and braid deltas: Varieties of coarse-grained deltas: Geological Society of America Bulletin, v. 99, p. 331 340, doi:10.1130/0016-7606(1987)99<331 :FABDVO>2.0.CO;2. Montgomery, S. L., D. M. Jarvie, K. A. Bowker, and R. M. Pollastro, 2005, Mississippian Barnett Shale, Fort Worth Basin, north-central Texas: Gas-shale play with multitrillion cubic foot potential: AAPG Bulletin, v. 89, p. 155 175, doi:10.1306/09170404042. Neill, C. A., and M. A. Allison, 2005, Subaqueous deltaic formation on the Atchafalaya shelf, Louisiana: Marine Geology, v. 214, p. 411430, doi:10.1016/j.margeo.2004 .11.002. Ng, D. T. W., 1979, Subsurface study of Atoka (Lower Pennsylvanian) clastic rocks in parts of Jack, Palo Pinto, Parker, and Wise counties, north-central Texas: AAPG Bulletin, v. 63, no. 1, p. 5066. Olariu, C., and J. P. Bhattacharya, 2006, Terminal distributary channels and delta front architecture of river-dominated delta systems: Journal of Sedimentary Research, v. 76, p. 212233, doi:10.2110/jsr.2006.026. Prez-Arlucea, M., and N. D. Smith, 1999, Depositional patterns following the 1870s avulsion of the Saskatchewan River (Cumberland Marshes, Saskatchewan, Canada): Journal of Sedimentary Research, v. 69, no. 1, p. 6273. Pollastro, R. M., R. J. Hill, D. M. Jarvie, and C. Adams, 2004, Geologic and organic geochemical framework of the Barnett-Paleozoic total petroleum system, Bend arch Fort Worth Basin, Texas (abs.): AAPG Annual Meeting Program, v. 13, p. A113, CD-ROM. Pollastro, R. M., D. M. Jarvie, R. J. Hill, and C. W. Adams, 2007, Geologic framework of the Mississippian Barnett Shale, Barnett-Paleozoic total petroleum system, Bend archFort Worth Basin, Texas: AAPG Bulletin, v. 91, no. 4, p. 405436, doi:10.1306/10300606008. Sellards, E. H., W. S. Adkins, and F. B. Plummer, 1932, The geology of Texas, volume I, stratigraphy: University of Texas at Austin Bulletin, v. 3232, p. 99101.

Hentz et al.

1331

Smith, D. G., and N. D. Smith, 1980, Sedimentation in an anastomosed river system: Examples front alluvial valleys near Banff, Alberta: Journal of Sedimentary Petrology, v. 50, p. 157164. Steward, D. B., 2007, The Barnett Shale play, phoenix for the Fort Worth Basin: A history: Fort Worth Geological Society and North Texas Geological Society, 202 p. Sullivan, E. C., K. J. Marfurt, A. Lacazette, and M. Ammerman, 2006, Application of new seismic attributes to collapse chimneys in the Fort Worth Basin: Geophysics, v. 71, no. 4, p. B111B119, doi:1190/1.2216189. Sullivan, E. C., K. J. Marfurt, C. Blumentritt, and M. Ammerman, 2007, Seismic geomorphology of Paleozoic collapse features in the Fort Worth Basin (U.S.A.), in R. J. Davies, H. W. Posamentier, L. J. Wood, and J. A. Cartwright, eds., Seismic geomorphology: Applications to hydrocarbon exploration and production: Geological Society (London) Special Publication 277, p. 187203. Tamura, T., Y. Saito, S. Sieng, B. Ben, M. Kong, I. Sim, S. Choup, and F. Akiba, 2009, Initiation of the Mekong River delta at 8 ka: Evidence from the sedimentary succession on the Cambodian lowland: Quaternary Science Reviews, v. 28, p. 327344, doi:10.1016/j.quascirev.2008.10.010. Thompson, D. M., 1982, Atoka Group (Lower to Middle Pennsylvanian), Fort Worth Basin, Texas: Terrigenous depositional systems, diagenesis, and reservoir distribu-

tion and quality: University of Texas at Austin, Bureau of Economic Geology Report of Investigations 125, 62 p. Thompson, D. M., 1988, Fort Worth Basin, in L. L. Sloss, ed., Sedimentary coverNorth American Craton: U.S.: Geological Society of America, Decade of North American Geology Series, v. D-2, p. 347352. Turner, G. L., 1957, Paleozoic stratigraphy of the Fort Worth Basin: Abilene and Fort Worth Geological Societies Joint Field Trip Guidebook, p. 5777. Tye, R. S., and J. M. Coleman, 1989, Depositional processes and stratigraphy of fluvially dominated lacustrine deltas: Mississippi delta plain: Journal of Sedimentary Petrology, v. 59, p. 973996, doi:10.1306/212F90CA-2B24 -11D7-8648000102C1865D. Van Heerden, I. L., and H. H. Roberts, 1988, Facies development of Atchafalaya delta, Louisiana: A modern bayhead delta: AAPG Bulletin, v. 72, p. 439453. Van Wagoner, J. C., R. M. Mitchum, K. M. Campion, and V. D. Rahmanian, 1990, Siliciclastic sequence stratigraphy in well logs, cores, and outcrops: Concepts for high-resolution correlation of time and facies: AAPG Methods in Exploration Series 7, 55 p. Walper, J. L., 1982, Plate tectonic evolution of the Fort Worth Basin, in C. A. Martin, ed., Petroleum geology of the Fort Worth Basin and Bend arch area: Dallas Geological Society, p. 237251.

1332

Reservoir Systems of the Lower Atoka Group, Northern Fort Worth Basin

Vous aimerez peut-être aussi