Geology of the Haynesville Gas Shale in East Texas and West Louisiana

Edited by Ursula Hammes and Julia Gale


This AAPG Memoir on the Haynesville Shale of East Texas, USA, provides an overview of the main geologic, stratigraphic, sedimentologic, geomechanical, micro-seismic and engineering characteristics of the Haynesville mudrocks. Experts on mudrocks from academia, industry, and government contribute with a variety of topics that describe the Haynesville Shale from basin- to nanaoscale, reflecting the dimensions affecting shale-gas assessment and demonstrating the variety of techniques applicable to shale-gas evaluation. The included papers serve not only as examples of shale-gas analyses of the East Texas shale basin employing different techniques, but also are examples of approaches to evaluating shale basins worldwide.

  1. Page 1

    We present a wide-ranging AAPG Memoir on the geology, geochemistry, petrophysics, and engineering of the Upper Jurassic Haynesville Shale. Eleven papers were solicited from experts in government, industry, and academe to contribute a variety of topics that describe the Haynesville Shale from basin- to nanoscale, reflecting the dimensions affecting shale-gas assessment and demonstrating the variety of techniques applicable to shale-gas evaluation. This memoir is therefore not only an analysis of the East Texas Shale Basin, but also an example of approaches to evaluating shale basins worldwide. We provide a short overview of the Haynesville Formation and East Texas Basin, followed by a synopsis of papers included in the memoir.

  2. Page 5

    Continuous hydrocarbon accumulations in shale reservoirs appear to be characterized by common paleotectonic and paleogeographic histories and are limited to specific intervals of geologic time. In addition, most North American self-sourced shale correlates with geologic time periods of calcitic seas and greenhouse conditions and with evolutionary turnover of marine metazoans. More knowledge about the relations among these controls on deposition is needed, but conceptual modeling suggests that integrating tectonic histories, paleogeographic reconstructions, and eustatic curves may be a useful means by which to better understand shale plays already in development stages and potentially identify new organic-carbon-rich shale targets suitable for continuous resource development. Upwelling and anoxic waters are commonly cited to explain the accumulation and preservation, respectively, of marine organic carbon. In addition, and perhaps alternatively, the broad correlation of self-sourced shale with macroevolutionary trends in land plants and marine metazoans suggests that reduced consumption of organic matter by benthos during periods of high terrestrial and marine organic productivity was responsible. Fundamental to any of the processes that acted during deposition, however, was active tectonism. Basin type can often distinguish self-sourced shale plays from other types of hydrocarbon source rocks. The deposition of North American self-sourced shale was associated with the assembly and subsequent fragmentation of Pangea. Flooded foreland basins along collisional margins were the predominant depositional settings during the Paleozoic, whereas deposition in semirestricted basins was responsible along the rifted passive margin of the U.S. Gulf Coast during the Mesozoic. Tectonism during deposition of self-sourced shale, such as the Upper Jurassic Haynesville Formation, confined (re)cycling of organic materials to relatively closed systems, which promoted uncommonly thick accumulations of organic matter.

  3. Page 25

    Recent discoveries in the Late Jurassic Haynesville and Bossier shales have dramatically increased unconventional gas exploration activity in the mature petroleum provinces of eastern Texas and northern Louisiana. The Haynesville and Lower Bossier shales comprise the uppermost units of a transgressive systems tract of a second-order supersequence, which spans the interval from the top of the Werner Anhydrite/Louann Salt equivalent to the upper Cotton Valley clastics. Depositional variations within the shales are a function of higher-order sequences that resulted from eustatic sea-level fluctuations, paleobasin physiography, and the interplay of local subsidence and sediment input rates. The antecedent topography shaped by underlying carbonates of the Gilmer (Haynesville) Lime (Forgotson and Forgotson, 1976) and subsequent sediment budgets strongly influenced (1) facies development and stacking patterns that vary along the northern rim of the young Gulf of Mexico (GOM) Basin during Haynesville and Bossier time, and (2) the depositional processes, total organic carbon richness, and preservation of the self-sourcing Haynesville and Bossier Shale units. The Haynesville Shale depositional system is an example of a competing carbonate and clastic system that contains contemporaneous retrogradational and progradational facies. In the western part of the system, which is carbonate-dominated and fairly restricted from siliciclastic input, the time-equivalent Gilmer (Haynesville) Lime consists of backstepping carbonate facies. In contrast, to the east, strong progradational stacking patterns, comprised of mainly siliciclastic facies assemblages, dominate in northern Louisiana and western Mississippi because of increased sediment supply from the ancestral Mississippi River, which outpaced subsidence and eustasy. Hence, major bounding stratigraphic events such as higher-order maximum flooding surfaces and condensed sections critical for shale gas exploration appear to change facies laterally whereas the second-order maximum flooding surface, or the turn from retrograding to prograding stacking patterns, appears diachronous along depositional strike. During Bossier time, the youngest carbonates were drowned and siliciclastics became increasingly dominant, expanding westward from northern Louisiana into eastern Texas and ultimately across most of the northern GOM shelf as the Cotton Valley Sandstones and its distal shale equivalents. Depending on the paleophysiography of the depositional shelf setting, some areas of the Haynesville-Bossier system were restricted and relatively sediment starved. These correspond with areas of total organic carbon enrichment and, in turn, lower shale gas exploration risk.

  4. Page 47

    The subsurface Upper Jurassic Haynesville and Bossier Formations comprise three facies associations along the eastern slope of the Gilmer Platform. The lower Haynesville facies association consists of three facies produced by mass-wasting processes: (1) calcirudite/calcarenite, (2) mud-clast calcarenite, and (3) laminated calcisiltite intercalated with laminated calcareous mudrock and bioturbated calcareous mudrock. These facies were deposited by (1) hyperconcentrated density flows/transitional concentrated density flows, (2) hydrated turbidity flows, and (3) distal settling from turbidity flows, respectively. These mass-wasting deposits are the deeper water equivalents of the shallower water Haynesville Lime. The sedimentary dynamics of the mass-wasting processes produced TOC (total organic content)-rich accumulations downslope in the deeper parts of the basin.

    The upper Haynesville facies association also consists of three facies: (1) TOC-rich laminated calcareous mudrock, (2) bioturbated calcareous mudrock, and (3) bioturbated mud-clast calcisiltite. These facies were derived from marine snow deposited and reworked as sediment drifts by bottom currents above and below the oxycline. The Bossier Formation facies association contains (1) massive argillaceous mudrock, (2) bioturbated argillaceous mudrock, and (3) argillaceous claystone. These facies are interpreted as prodelta deposits intercalated with sediment deposited by settling from flood plumes. TOC is relatively high despite sedimentary dilution from deltaic input, indicating high primary productivity of organic matter at the time of deposition. TOC-rich accumulations comparable to the Haynesville Shale are observed in the Bossier Formation on Sabine Island and may exist wherever detrital sediment input has been reduced or diverted by currents. The lower Haynesville was deposited as an upwards-deepening succession during a second-order transgression that started after deposition of the Smackover Formation. Because the upper Haynesville was deposited as a sediment drift with an internally complex sedimentary geometry, no internal cyclicity is apparent, and the position of the second-order maximum flooding surface cannot be established. Deposition of the Bossier marks a significant turnaround when deltaic sediments prograded from the north and buried the mass-wasting and sediment-drift deposits. The distal setting of the facies, evidence of deposition below storm-wave base, the pelagic source of the sediment, and the sedimentary processes involved make application of sequence stratigraphic concepts to the deposits problematic.

  5. Page 69

    A late Oxfordian to early Berriasian nannofossil biostratigraphic framework has been developed from research on D.S.D.P Site 534 and composite outcrop sections in southeastern France and central Portugal. It is built on published nannofossil biostratigraphic studies of the same deepsea site, as well as those from outcrop sections in both southeastern France and northern Italy. These results were first applied to the subsurface in 16 wells in the East Texas Basin (Leon and Robertson Counties) during the 1990s and then in six wells in the Haynesville Basin in northeastern Texas in 2009. The relative stasis in nannofossil evolution during the middle Oxfordian to Kimmeridgian and subsequent major Tithonian radiation necessitated the unconventional use of fossil appearances (i.e., bases) during our examination of mostly ditchcutting samples from these onshore wells. Nannofossil biostratigraphic resolution was further challenged by sparse nannofossil recoveries, which were exacerbated in the older carbonates. The reliability of the individual nannofossil events are assessed for both the research and subsurface sections. Two new Kimmeridgian species, Bucanthus lusitanicus and Calcivascularis cassidyi, are described herein. One new combination is also introduced.

  6. Page 103

    Exploration of the Haynesville and Bossier shale play in east Texas and western Louisiana provides an opportunity for many onshore-focused, U.S. geoscientists to reacquaint themselves with biostratigraphy. Using both vintage and new biostratigraphic data, this study showcases the integration of multiple biostratigraphic disciplines for (1) identification of unique correlation surfaces within the basin; (2) increased resolution and confidence in the age assignments for surfaces defined by logs and seismic; and (3) the recognition of distinct biofacies, which aid in the prediction of preferred intervals and geographic areas for hydrocarbon accumulation within shales. This study incorporates interpretations from five wells using nannofossils, foraminifera, ammonites, and radiolarians. The strengths and weaknesses of each fossil discipline are discussed in the context of a multidisciplinary stratigraphic evaluation. Rock materials for the study were collected from both cuttings and core from stratigraphic intervals interpreted as Haynesville Shale, Bossier Shale, Taylor Sands, and Cotton Valley Group siliciclastics. Biostratigraphic interpretations, which range from lower Kimmeridgian to lower Berriasian, establish more accurate and reliable timelines than have previously been published from a Jurassic basin in North America. Complementing the age interpretations, four key biofacies are recognized and mapped. Some of the biofacies are interpreted to be associated with significant dysoxic to anoxic bottom conditions and potentially areas of corresponding elevated surface productivity, whereas others likely represent changes in sediment supply into the basin from nearby terrigenous clastic and/or detrital carbonate sources.

  7. Page 137

    The Haynesville Shale, an Upper Jurassic (Kimmeridgian) age calcareous and locally organic-rich mudrock, is one of many prominent shale gas plays in North America. As shale plays increase in importance, the ability to define basin-wide, robust, and stable stratigraphic frameworks using data derived from well-bores becomes increasingly critical. Here, the technique of chemostratigraphy is used to define a stratigraphic framework that extends through ten wells ranging from eastern Texas to northwestern Louisiana. Stratigraphic variations in inorganic geochemistry allow clear differentiation of Haynesville Shale from the underlying Smackover Formation, the Gilmer Lime, and the overlying Bossier Formation. More importantly, however, interpretation of the results allows two chemostratigraphic packages and four geochemically distinct units to be defined and correlated within the Haynesville Shale. The lithostratigraphic units are geochemically differentiated using variations in SiO2, Al2O3, MgO, Zr, and Nb, whereas the units within the Haynesville Shale are defined using changes in CaO, Al2O3, MgO, Fe2O3, Rb/K2O and Th/U values, and V enrichments. By integrating the geochemistry with x-ray diffraction and total organic carbon (TOC) results, it becomes apparent that the driving forces behind the changing geochemistry within the Haynesville Shale are the amounts of anoxia in the lower portion of the Haynesville Shale and of CaO input in the upper portion. Cyclical fluctuations in the relative abundances of Zr and Nb are interpreted to represent transgressive—regressive cycles—and provide enhanced correlation within the Haynesville Shale. By combining stratigraphic changes in Zr/Nb values with V enrichments, it is shown that the most severe period of anoxia is associated with the transgressive portion of the oldest cycle. Importantly, this suggests that this stratigraphic horizon is where maximum TOC can be expected. Lateral changes in geochemistry within the Haynesville Shale demonstrate that terrigenous input was highest in the northwest sector of the basin, primarily in East Texas, and anoxia was greatest in the east of the basin, primarily in Louisiana.

  8. Page 155

    The Haynesville Shale in northwest Louisiana and east Texas is a geologically unique gas play in which many petrophysical, engineering, and mechanical properties are close to optimal. With high geopressure gradients ranging from 0.8 to >0.95 psi/ft and reservoir pressures ranging from 8000 to 17,000 psi, it is one of the most prolific shale-gas plays in North America. Through the use of horizontal wells and multiple-stage fracturing, gas production reached >7 Bcf/d in August 2011, and the play has surpassed the Barnett Shale in north Texas as one of the highest gas-producing plays in the United States. The objectives of this study are to investigate the effects of petrophysical, geochemical, geologic, mechanical, and engineering properties, as well as completion practices, on Haynesville Shale production. Core data show that connate water saturations range from 15 to 40% in the Haynesville. Low connate water saturation is attributed to water expulsion by oil and gas during hydrocarbon generation from organic matter within the shale. Nevertheless, slow fluid escape and gas generation at high temperatures resulted in an abnormally high reservoir pressure and pressure gradient, even in this relatively high porosity rock. The effects of the high geopressure gradient have been to increase reservoir pore pressure, to preserve porosity and permeability, and to enhance free gas content and the brittle nature of the gas shales. The average porosity of the Haynesville Shale is high, ~11%, and the free gas content is enhanced by high porosity and gas density. Because of the high formation pressures, effective stresses of the Haynesville are low, and laboratory compression tests show that the rocks are highly brittle at these low effective stresses. Production from the Haynesville is a complex function of geopressure gradient, effective stress, reservoir quality, and completion practices. A wide range of completion parameters, such as length of horizontal well, choke size, number of stages, and proppant volume, have been tested to find optimal production strategies. Large choke sizes, which increase initial potential, can have a detrimental effect on long-term production and smaller choke sizes lower the decline rates and increase long-term well productions. Initial potential and production are higher in the east and south regions with higher pressure, carbonate/silica content and total organic carbon than the northwest region in Texas with lower total organic carbon but higher clay content.

  9. Page 179

    Microstructural studies using a dual beam FIB-SEM instrument reveal dimensions of pores within shales, which are consistent with macroscopic-averaged dimensions resolved by nuclear-magnetic resonance and mercury-injection capillary pressure. These dimensions are on the order of nanometers to hundreds of nanometers. We compare observations on a limited number of samples from the Haynesville to observations on the Woodford, Barnett, and Marcellus Shales. The FIB-SEM imaging uniquely resolves where the pores lie, that is, mainly within kerogen in the Woodford and Barnett and between clay platelets for the Haynesville samples. Measurement of velocities as a function of pressure and calculated anisotropies display a pressure dependence that reflects difference in the microstructure studied. The Woodford samples show a weak velocity dependence on pressure whereas the Haynesville samples show a very strong dependence. Coupled with the validity of the effective pressure law, the pressure dependence of anisotropy may prove useful in monitoring pressure depletion and compartmentalization in the Haynesville Shale.

  10. Page 189

    Modeling the elastic properties of the Haynesville Shale using rock-physics techniques is part of the characterization of this shale that could be used to improve predictions of economic drilling locations. The goal of this modeling is to relate the reservoir properties of interest (e.g., porosity, pore shape, and composition) to the elastic properties. Although this is the same goal as in using rock physics for conventional reservoirs, the approach used here differs. Within the Haynesville Shale, the physical rock properties that most significantly affect the elastic properties appear to be the composition and pore shape. Accordingly, the rock-physics modeling requires an effective-medium theory, notably the self-consistent model, to accommodate these properties. Composition was estimated through a combination of well log and x-ray diffraction (XRD) data. Pore shapes were estimated using estimated stress conditions and numerical studies. The best modeling results explain trends in velocity measurements corresponding to joint variations of composition and pore shape. Accordingly, this rockphysics model could be used in conjunction with seismic data interpretation to identify locations with low velocity and potentially higher organic content and zones with faster velocity more suitable for fracturing.

  11. Page 205

    Petrophysical data collection in the Haynesville Shale in northwest Louisiana consists largely of wireline logging in vertical pilot wells with occasional coring. As in most gas shales, field development is primarily based on these data. Additionally, petrophysical and geologic data from vertical wells are assumed to be laterally consistent along the length of horizontal production wells. Typically these horizontal wells are logged solely with gamma-ray tools that are used for geosteering. However, both lateral and vertical variations in the geologic properties of shale have been demonstrated by core, laboratory, and wireline studies. Published research also suggests that lithofacies stacking pattern recognition is possible in these fine-grained sediments. An investigation has been undertaken in horizontal shale gas wells using a suite of gamma ray, resistivity (EWR), and azimuthal litho-density (ALD) logging-while-drilling (LWD) tools. One objective of this study was to determine if variation of lithofacies in shales could be detected using LWD tools. A proprietary petrophysical lithology model was developed from detailed core lithology descriptions of the Haynesville Shale. This model was applied to the LWD data collected along the lateral sections of four wells drilled in the Haynesville Shale. Vertical stacking patterns of lithofacies were observed in each of the Haynesville wells with this method. Lateral lithofacies variation within wells cannot be confirmed because of spatial and tool resolution issues. However, variation of lithofacies in a mappable stratigraphic interval between wells was observed. Study results appear to support previous research in shale geology and stratigraphy that suggest the recognition of lithologic and stratigraphic patterns can be applied to fine-grained sediments. This work demonstrates that the mapping of stratigraphic variability in shales may be possible by using LWD data in horizontal wells. In addition, lithofacies data generated in this investigation bring into question some of the assumptions generally applied in shale gas field development. The most significant of which is perhaps the idea that beds maintain consistent lithology along the typical lateral well path. In the last of the four wells presented in this report, gamma ray only geosteering was supplemented with input from other LWD sensors. This additional data appear to have improved lateral well targeting. The use of LWD tools beyond gamma ray in shale-gas formations could permit development of more realistic three-dimensional geologic models of shale reservoirs. In addition, these data could facilitate better well lateral placement, improved well completions, formation evaluation, and potentially increase well production and project economics.

  12. Page 219

Purchase Chapters

Recommended Reading