STRATIGRAPHY OF THE BARNETT SHALE (MISSISSIPPIAN) IN THE NORTHERN FORT WORTH BASIN

© Copyright 1982, 2023 James D. Henry

For more than a century, Mississippian-aged rocks have been known and studied in Texas (Gabb, 1862, Shumard, 1863). Plummer and Moore (1922, p. 23) gave the Barnett Shale its name from an outcrop in San Saba County; and Sellards (1933, p. 91) first put the immediatel underlying Chappel Limestone into the literature, based on an exposure near the Central Texas town of San Saba. These two Mississippian units have subsequently been studied and described by a number of different workers (Plummer and Scott, 1937, p. 15; Cheney, 1940; Plummer, 1950; etc.). Unfortunately these efforts have exclusively been concerned with surface outcrops in the Llano area of Central Texas; the stratigraphy of Mississippian rocks in North-Central Texas, especially in the deeper parts of the Fort Worth Basin, has remained poorly known. However, within the last few decades, significant improvements in sophisticated borehole logging techniques and the ever growing availability of deep well control have made possible a much better understanding of the role of Mississippian strata in the geologic history of the Fort Worth Basin.

The present study is concerned with the former of these two formations and their stratigraphic relationships in the northern part of the Fort Worth Basin, primarily in Jack, Wise, and Montague Counties. A majority of the available electrical and porosity logs in this area have been utilized, and the author has had the opportunity to examine Mississippian drill cuttings from numerous wells over the past forty years. It is hoped that this current effort will encourage further detailed studies of this stratigraphic interval which is so modestly represented in the literature.

REGIONAL SETTING

The northern part of the Fort Worth Basin is situated at the confluence of several major structural features whose effects are well represented in the Late Paleozoic stratigraphy of Jack, Montague, Clay, Wise, and Cooke Counties. The most prominent of these features is the Muenster Arch, which separates the Fort Worth Basin from the adjacent Marietta Basin to the northeast. This so-called "arch" is actually a couple of northwest-southeast trending, asymmetrical fault blocks that extend at least from Jefferson County, Oklahoma, to Denton County, Texas. The entire pre-Pennsylvanian Paleozoic section, including a probable minimum of about 1000 feet (300 meters) of Barnett Shale, has been eroded from this arch in eastern Montague and western Cooke Counties so that Precambrian granite, schist, diorite, and diabase (Flawn, 1956, pp. 178-180) are overlain by Late Pennsylvanian (Desmoinesian and/or Missourian) clastics.

Vertical displacement on the order of about 3 miles (5 kilometers) along the southwestern flank of the Saint Jo element of the Muenster Arch is accomplished by two parallel high-angle reverse faults The more northeasterly fault juxtaposes the-Precambrian core of the arch against Ordovician rocks (mostly Ellenburger) beneath the Pennsylvanian unconformity.

The trace of the second major fault, which lies approximately 2 miles (3 kilometers) to the southwest of the first, passes very nearly under the town site of Saint Jo and continues in a southeasterly direction into Cooke County. It is this second fault that causes Ordovician rocks to subcrop opposite the in-place Barnett and Forestburg Formations and consequently marks the northeastern limit of preserved Barnett beds.

About 5 miles (8 kilometers) northwest of Saint Jo, the trace of these faults segues into the eastern end of the Red River Uplift. Tracing the uplift westward from there, the normal-faulted southern edge of that structural trend passes near the town site of Nocona. From there we can follow this structural trend west-southwest until it passes into Clay and Wichita Counties and ceases to be a major feature.  The trace of this fault zone forms the northernmost limit of Barnett beds in the deep Fort Worth Basin. Approaching this fault from the south, progressively older rocks are found to subcrop beneath the pre-Pennsylvanian unconformity. Those areas where the Barnett has been at least partially eroded in Jack and Montague Counties have not been contoured on the regional isopach map on the Barnett Shale presented immediately below.

Figure: After Henry, 1982

On the northwestern flank of the basin, the Electra element of the Red River Uplift is the limiting structural feature. White (1948, p. 30) reports that the Barnett Shale thins to extinction, apparently due to erosion or faulting, along the south flank of the uplift in southern Clay, Archer, and Baylor Counties.  Along the western edge of the basin, the Barnett thins rapidly against the eastern flank of the Bend Arch. The north-south axis of this broad, gentle structure lies today in Archer, Young, and Stephens Counties.  The extreme eastern margin of the northern Fort Worth Basin is obscured by the overthrust rocks of the Ouachita fold belt in Dallas, Collin, and Grayson Counties. This orogeny was initiated in Atokan (Early Pennsylvanian) time and is therefore subsequent to the deposition of the Barnett Shale. In contrast to the Late Pennsylvanian and Permian rocks, which dip generally to the northwest, the regional dip on the Barnett in this study area is about one-half degree to the northeast.

PALEOZOIC HISTORY

The Barnett Shale represents an integral part of the stratigraphy of the Fort Worth Basin, and as such it can best be explained within the context of Paleozoic tectonism in the southern mid-continent.  Shatski (1946) was the first to point out that the elongate structural trend which extends from the Texas Panhandle through Southern Oklahoma—and which borders the Fort Worth Basin on the northeast—is in fact the remnants of an Early Paleozoic aulacogen. (Hoffman, et al., 1974) suggested that the Southern Oklahoma Aulacogen is the abandoned third arm of a triple rift junction that developed in Early Cambrian time over a plume, or hot spot, in the mantle. The other two arms of the triple junction suffered continued rifting and became sites where sea floor spreading led to complete separation of the continental plate The aulacogen, actually a 60-mile (100-km) wide graben system, ceased rifting and began to subside in Middle to Late Cambrian time. The southwestern-most fault boundary of the aulacogen was probably very close to, and parallel with, the present day Muenster Arch. Exaggerated thicknesses of Cambrian extrusives and Ordovician sedimentary rocks were deposited in the deep part of the aulacogen, while much thinner sequences of Ordovician strata were deposited along its northern and southern margins, the latter of which includes most of the present day Fort Worth Basin. Subsidence slowed during Siluro-Devonian time, as evidenced by decreased sedimentary fill in the deep aulacogen. A period of erosion—at most, Late Devonian to Late Mississippian—removed any Siluro-Devonian rocks that may have been deposited in the area that is now known as the Fort Worth Basin. The resulting unconformity on the southern flank of the aulacogen increases in stratigraphic magnitude updip to the southwest (Figure 3).

Figure: After Henry, 1982

The Early MississippianChappel Limestone  was the first formation to be deposited and preserved in the Fort Worth Basin when the aulacogen and adjacent shelf area were again inundated, and the Late Mississippian Barnett Shale was the second. An isopach map of the Barnett (Figure 2 above) indicates that the Cambro-Ordovician Southern Oklahoma Aulacogen was still the primary tectonic factor controlling sedimentation in this area as recently as the latter part of the Mississippian Period. This means that the Fort Worth Basin per se has existed as a discrete geologic entity only since Early Pennsylvanian time at most.

The cause of the renewed terrigenous deposition during the Mississippian Period is not well understood. Hoffman, et al. (1974) insist that dormant aulacogens constitute zones of vertical structural weakness that are easily reactivated by subsequent tectonic events. Thus it seems possible that the Mississippian inundation of the aulacogen—which ultimately resulted in the deposition of the Barnett on a shallow, adjacent shelf area—was caused by subsidence along these rejuvenated zones of weakness.

The low-lying craton to the north and west was the source area for the shale beds of the Barnett. During the deposition of this formation, conditions favored the establishment and growth of at least several dozen discrete organic communities which resulted in the Late Mississippian reef complexes that are commonly referred to as "Mississippi(an)" (which see "The Mississippian Reefs" elsewhere on this website).

Following the deposition of the final Mississippian beds, most of North America, including the southern mid-continent, was subjected to a period of erosion that left a widespread unconformity of uncertain magnitude between the Mississippian Period (Chesterian Epoch) and the Early Pennsylvanian Period (Morrowan and/or Atokan Epoch). In Clay County the upper beds of the Barnett Shale were truncated by this erosional event, and in Montague County the Barnett has been entirely removed from the crest of the Nocona Anticline (Figures 2 and 3). However, in most of the study area this unconformity occupies a higher position in the stratigraphic section, and the Barnett is conformably overlain by the shales and limestones of Chesterian age and  in a few places by thin remnants of the Morrow Shale. (See "The Marble Falls, the Morrow Shale, and the Forestburg" page elsewhere on this website.)

As noted above, the region that is today the northern Fort Worth Basin was actually a shelf area adjacent to the inundated Southern Oklahoma Aulacogen during the latter part of the Mississippian Period. Organic-rich muds and occasional limestones accumulated slowly in a low energy, shallow marine environment. Distinctive bedding patterns can often be traced for several tens of miles throughout this area. Considered regionally, the thickness of the Barnett is primarily a function of its proximity to the edge of the Southern Oklahoma Aulacogen. As can be seen on an isopach map of the total Barnett Formation (Figure 2), the Barnett varies from about 100 feet (30 meters) in southwestern Jack County to 950 feet (290 meters) near the Montague-Cooke County line in the Tex Harvey Oil No. 1 Fields (B. B. B. & C. Survey, A-91). Doubtless an even greater thickness of Barnett was eroded from the crest of the Muenster Arch. The convergence of the 700-, 800-, and 900-foot contours near the Nocona-Saint Jo Fault marks the northeastern limit of the Barnett shelf as it tails off into the basinal aulacogen.

Locally the Barnett thins abruptly over the crests and flanks of the underlying reefs, most of which lie to the west of the 200-foot (61-meter) contour on the Barnett isopach map (Figure 2). The Barnett is nowhere absent over the crest of a reef, although in the Graham Brothers No. 1 Simpson well in southwestern Jack County (C. Stephenson Survey, A-523) it has been reduced to only 30 feet (9 meters).

Throughout the Fort Worth Basin the Barnett is observed in drill cuttings to be a black, organic-rich shale, superficially indistinguishable in color and texture from the younger shales of the Pennsylvanian. However, when moist cuttings are rubbed briskly between the hands, the petroliferous Barnett will release a greasy, bituminous substance that does not occur in the other shales. Macrofossils are rare, being largely restricted to occasional pelmatozoan columnals. This writer was unable, in a single effort, to recover any identifiable microfossils from Barnett drill cuttings in central Jack County.

Where good wireline logs are available, the Barnett cannot be mistaken for any other unit. Resistivity values are unusually high, exceeding those of the Pennsylvanian shales by a factor of at least ten. The compensated density log is also very distinctive; it will be found to have a mean bulk density of 2.50 grams/cubic centimeter throughout the northern Fort Worth Basin. This compares with an average of about 2.58 to 2.62 gm/cc for the overlying Pennsylvanian shales.

Moreover, the Barnett shale beds are characterized by anomalously strong gamma ray counts; readings of 150 to 400 API units are common, thus making it about twice as radioactive as any other shale in the Fort Worth Basin.  Spectral gamma ray logs that distinguish amongst the various elemental sources of the radiation have shown that unusually high concentrations of uranium are responsible for the anomalous readings in the Barnett The other two major sources of subsurface gamma rays, thorium and potassium, are only slightly elevated in the Barnett while unranium content is dramatically increased. Research done since the turn of the millennium indicates that this phenomenon is very likely attributable to—of all things—bacteria and/or fungi.

A present-day genus of bacteria, Geobacter, has been certified by the US Environmental Protection Agency for use in remediating contamination by uranium. Radioactivity is not usually a concern—the common isotope U238 is not radioactive at all—but uranium is a water-soluble, heavy metal element that is chemically toxic to most higher forms of life. Geobacter does not feed on uranium, but it takes it out of its immediate environment and permanently binds it into capsules so as to facilitate some biologically unique detail of its primitive metabolism. Along with the U238 is a minute percentage (0.72%) of the naturally occurring radioactive isotope U235. The bacteria/fungi can not distinguish between the two isotopes, but where the flourish, the end result is a higher than normal concentration of both isotopes left behind when the local bio-environment changes and the microbes die out or move on elsewhere. Some very distant predecessor of Geobacter presumably did very much the same thing, and that almost certainly accounts for the Barnett’s uncommonly hot gamma ray response. Geobacter can exist in several natural habitats, but it prefers anaerobic environments where organic matter is abundant. Surely it is more than a coincidence that the Barnett Shale was deposited on a shallow anaerobic seabed that was uncommonly rich in organic matter. The uranium-fixing bacteria/fungi were in their own special microbial heaven, and they reproduced in astronomical numbers. Where the sedimentary transition to the more normally radioactive overlying Forestburg Formation occurs on wireline logs, the gamma ray response is often seen to be tapering off upwards, as if the microbial population was faltering until it finally yielded to a new and terminally uninhabitable sedimentary regime. But though it temporarily disappeared from most of the basin, those microbes (and/or their descendants) did not become totally extinct; they only moved elsewhere to a more accommodating environment while waiting for a chance to reclaim their former habitat. The much less radioactive sediments that were deposited in their absence can be seen on cross sections A-A' and B-B' below, and the interval is isopached as the "Foreigh Shale" in Figure 5. In much of the basin the radioactive facies returned in the lower Forestburg, typically only a few dozen feet above the top of the Barnett, but the microbes' encore was brief; they only lived long enough to occupy about 10 to 30 feet of vertical section. This relatively brief interval is known to many geologists as the “False Barnett.” The lithology of the False Barnett is very similar but not quite identical to that of the real Barnett shortly below it, therefore suggesting that the high gamma ray count is microbially tied to the Barnett’s earlier anaerobic paleo-environment that was so hospitable to the uranium-fixing bacteria/fungi for so long. When the brief False Barnett depositional facies finally ended, so did the story of these particular bacteria/fungi in the Fort Worth Basin. This phenomenon is common to only a few other bituminous black shales, such as the Woodford of the Anadarko Basin. 

Although the foregoing description is consistently true of the shale facies of the Barnett throughout the study area, well logs reflect subtle changes in lithology that allow the formation to be informally subdivided into laterally persistent, easily recognizable members. In Jack County there are five such members (see cross section A-A' below).

After Henry, 1982

(1) The "basal shale" member is distinctive because of its heavy contamination with fine calcareous material generated by the contemporaneous Barnett reefs. It ranges in thickness from a minimum of about 20 feet (6 meters) up to about 60 feet (18 meters) in the neighborhood of the reef flanks, with which it is laterally correlative. It is nowhere present over the crest of a reef. The top of this member is picked on electric logs by a break in the resistivity curve, and the bulk density log will show an anomalously heavy reading of 2.55 to 2.60 gm/cc.

(2) Immediately overlying the basal shale member is the "major shale" member of the Barnett. With a mean thickness of more than l00 feet (30 meters), it usually represents slightly more than half of the entire Barnett section. This unit is the one most typical of the previously described Barnett lithology, and it is notable for its vertical homogeneity. It thins sharply over the reefs because of nondeposition, and occasionally it is completely absent over the larger ones. The top of this member is picked at a sharp constriction on the electrical log.

(3) The superjacent "minor shale" member is easily recognized by the distinctive trilobate undulatory pattern that it makes on resistivity logs. Lithologically it is virtually identical to the underlying major shale member. This unit varies in thickness from about 40 feet (12 meters) in the far western part of Jack County to about 80 feet (24 meters) in the northeastern corner. Like the major shale member, it thins over reef crests, but there are no known reefs that penetrate up into the minor shale. This would indicate that reef growth had completely ceased by the time this member was deposited and that differential compaction of the surrounding Barnett muds had begun, thus causing the buried reefs to continue having some local influence on the topography of the shallow sea floor.

(4) In the southern half of Jack County the Barnett depositional environment continued without interruption, resulting in the "upper shale" member (my preferred name). This unit, which is previously mentioned above as the "False Barnett," rarely exceeds 30 feet (9 meters) in thickness and thins to 10 feet (3 meters) in the southern part of the county. It is distinguished from the underlying Barnett beds by its somewhat lower resistivity, its slightly lower mean bulk density of about 2.46 gm/cc, and by a significantly reduced content of bituminous matter in the drill cuttings. In most of the study area, the upper shale member is overlain: 1) unconformably by calcareous sandstone and/or sandy limestone beds of the Atokan-age Marble Falls Formation, or 2) conformably by additional beds of Chesterian-age limestones and shales. In fresh drill cuttings these latter rocks appear extremely similar to the underlying Barnett and often make it difficult for the wellsite geologist to identify the exact top of the Barnett while drilling. Once logs have been run, there is no question about the location of the contact between the two units, but the cuttings alone can be, and often are, easily confused.

(5) Throughout most of the northern half of Jack County the upper shale member is separated from the minor shale by a mildly bituminous shale wedge whose lithology differs sharply from that of the usual Barnett facies. This "foreign shale" member exhibits the low resistivity and the relatively high bulk density (2.60 gm/cc) that are not characteristic of the lower members of the Barnett in this area. It is significant that the upper shale member is only slightly thinned above the foreign shale, suggesting that the latter was deposited rather quickly, as compared to the over- and underlying Barnett beds which accumulated relatively slowly. As can be seen from the isopach map in Figure 5 below, the provenance for the foreign shale unit possibly lay to the north at just about the time the Red River Uplift  was just beginning to rise.

Figure 5: After Henry, 1982

The calcareous Foreign Shale interruption can be seen below on the south-north cross section B-B'.

CROSS SECTION B-B': Note Barnett reef at left and calcareous Foreign Shale at mid-right.
After Henry, 1982

The lower three members of the Barnett can be traced from Jack County into southwestern Montague County with little difficulty.

In southeastern Montague County, near what may have been the Mississippian hinge line of the aulacogen/depositional basin, the Barnett undergoes a dramatic change in lithology. A calcareous shale interval develops in the upper part of the formation and apparently becomes limier toward the east-northeast. The above mentioned Tex-Harvey Oil No 1 Fields well was drilled about 2 miles (3 kilometers) southwest of the Muenster Arch boundary fault and penetrated 950 feet (290 meters) of Barnett—the thickest section known at the time in the Fort Worth Basin. The calcareous zone was encountered about 70 feet (21 meters) below the top of the typical bituminous shale facies, and it was mistaken by the operator for the Viola Limestone. The gamma ray-neutron log shows the calcareous unit grading slowly downward into typical Barnett lithology. This is the most basinward deep well yet drilled in the county, and all of the previously discussed formation members are unrecognizable on the logs.
 
In Wise County the three lower members of the Barnett are recognizable on logs in the northwestern part of the county, and in that area, as in Montague County, the top of the minor shale member constitutes the top of the formation. Elsewhere in Wise County these vertical subdivisions are unrecognizable. The regional thickness of the Barnett seems to fluctuate more erratically here than it does in either Jack or Montague Counties (refer to cross section B-B' above), but the formation does not appear to thicken or thin from the top—the changes in thickness generally affect all of the Barnett beds proportionally. The calcareous shale beds in the upper Barnett, which characterize the outer edge of the Barnett shelf in Montague County, are also present in a similar manner in eastern Wise County.
 
Throughout Jack County the Barnett unconformably overlies Early Ordovician limestones and dolomites of the Ellenburger Group. The pre-Barnett subcrop pattern in Clay, Cooke, Wise, Tarrant, and Montague Counties (see Figure 3: Subcrop Pattern Beneath Sub-Barnett Unconformity map above) also includes two carbonate units above the Ellenburger: a dense limestone of relatively low resistivity that may possibly be Simpson equivalent (Middle Ordovician), and, above that, another limestone unit of similar lithology but higher resistivity that is generally considered to be the Viola (Middle Ordovician). These two formations and the upper beds of the Ellenburger Group dip to the northeast at approximately one and a half degrees. The unconformity at the base of the Barnett is an angular one, but the angle is negligibly small—probably less than one degree. Although this hiatus may well include the entire Middle Paleozoic section, the actual period of erosion and/or nondeposition was probably not nearly so long. To the south, in the Central Mineral Region, Barnes, Cloud, and Warren (1946) have reported rocks of Early to Middle Devonian age, and Barnes, et al., (1966) and Miller (1972) have described Silurian sediments. In the Ardmore/Marietta Basin to the northeast and the Anadarko Basin to the northwest, the Siluro-Devonian is represented by the Hunton Group and the Woodford Shale. Though it remains unproven, it is not unreasonable to suppose that rocks of similar age once covered most of North-Central Texas, If that is in fact true, then the pre-Barnett unconformity is the result of, at most, Late Devonian and/or Early to Middle Mississippian uplift and erosion.
 
The lower contact of the Barnett is frequently ambiguous, because of anomalously high gamma ray readings in the upper 30 to 100 feet (9 to 30 meters) of the underlying Ellenburger. This situation is the result of porous beds in the Ellenburger being invaded by dissolved radioactive compounds (mostly uranium with some thorium) derived from the Barnett during post-depositional compaction. This phenomenon is largely confined to dolomite beds, and the intensity of gamma radiation is largely a function of porosity and water saturation.
 
The nature of the Barnett's upper contact in the Fort Worth Basin has been a source of disagreement in the past. Over major structural elements such as the Llano uplift and the Red River uplift (and some lesser structures such as the Nocona Anticline), the Barnett has obviously been eroded, and in some localities it is completely absent.
 
The erosional surface that cuts the upper beds of the Barnett in the northern Fort Worth Basin is a part of a broad regional unconformity that earlier workers have postulated to exist at several different stratigraphic levels in North Central Texas. This erosional surface has been observed to occur first at one depth in the pre-Atokan section, then at another depth somewhere else. (See The Marble Falls, the Morrow Shale, and the Forestburg web page elsewhere on this site.)

In the southeastern quarter of Montague County and the southwestern corner of Cooke County, the Barnett is overlain by a carbonate formation designated the Forestburg Limestone by this author in 1982. This formation is occasionally oil productive (e.g., the eponymous Forestburg and Bingham Creek Fields), and the Barnett Shale beds are undoubtedly the hydrocarbon source rock. The contact of the Forestburg with the Barnett is perfectly conformable; therefore, it is Chesterian in age, and it is overlain unconformably by beds of the Early Pennsylvanian Morrowan or Atokan stages.

The Late Chesterian post-Barnett (i.e., Forestburg) beds have taken the brunt of this erosional hiatus so that the upper contact of the Barnett Shale in most parts of the Fort Worth Basin is conformable. The upper shale member can be traced throughout most of Jack County without any indication of abrupt changes in thickness and thins very gradually toward the west in a predictable manner as do all of the lower members of the typical Barnett Shale facies.

#     #     #