Kevin M. Bohacs

From Oil-Prone Source Rock to Gas-Producing Shale Reservoir - Geologic and Petrophysical Characterization of Unconventional Shale Gas Reservoirs

Many currently producing shale-gas reservoirs are overmature oil-prone source rocks. Through burial and heating these reservoirs evolve from organic-matter-rich mud deposited in marine, lacustrine, or swamp environments. Key characterization parameters are: total organic carbon (TOC), maturity level (vitrinite reflectance), mineralogy, thickness, and organic matter type. Hydrogento-carbon (HI) and oxygen-to-carbon (OI) ratios are used to classify organic matter that ranges from oil-prone algal and herbaceous to gas-prone woody/coaly material. Although organic-matter-rich intervals can be hundreds of meters thick, vertical variability in TOC is high ( 50% of the total porosity, and these pores may be hydrocarbon wet, at least during most of the thermal maturation process. A full understanding of the relation of porosity and gas content will result in development of optimized processes for hydrocarbon recovery in shale-gas reservoirs.

Stratigraphic classification of ancient lakes: Balancing tectonic and climatic controls

Lakes and lake deposits present two fundamental paradoxes: (1) Modern lakes are vastly complicated, but the rock records of lakes are relatively simple; extensive observations reveal three distinct facies associations of common and widespread occurrence. These are referred to here as fluvial-lacustrine, fluctuating profundal, and evaporative facies associations. (2) Most explanations of modern and ancient lakes attribute their nature to climate, but neither modern lake parameters (lake size, depth, and salinity) nor the character of ancient-lake strata (thickness, extent, lithology) correlate with measured or inferred climatic humidity. We propose that it is the relative balance of rates of potential accommodation (mostly tectonic) with sediment + water fill (mostly a function of climate) that controls lake occurrence, distribution, and character. Lake basins may be termed overfilled, balanced fill, or underfilled, depending on the balance between these rates. We conclude that climate and tectonics exert coequal influence on lake deposits at both mesoscales (1 m to hundreds of meters) and macroscales (hundreds to thousands of meters).

Sequence Stratigraphic Distribution of Coaly Rocks: Fundamental Controls and Paralic Examples

Significant volumes of terrigenous organic matter can be preserved to form coals only when and where the overall increase in accommodation approximately equals the production rate of peat. Accommodation is a function of subsidence and base level. For mires, base level is very specifically the groundwater table. In paralic settings, the groundwater table is strongly controlled by sea level and the precipitation/evaporation ratio. Peat accumulates over a range of rates, but always with a definite maximum rate set by original organic productivity and space available below depositional base level (groundwater table). Below a threshold accommodation rate (nonzero), no continuous peats accumulate, due to falling or low groundwater table, sedimentary bypass, and extensive erosion by fluvial channels. This is typical of upper highstand, lowstand fan, and basal lowstand-wedge systems tracts. Higher accommodation rates provide relatively stable conditions with rising groundwater tables. Mires initiate and ©Copyright 1997. The American Association of Petroleum Geologists. All rights reserved.1Manuscript received September 7, 1995; revised manuscript received November 25, 1996; final acceptance May 5, 1997. 2Exxon Production Research Company, 3120 Buffalo Speedway, Houston, Texas 77096. 3Exxon Production Research Company, 3120 Buffalo Speedway, Houston, Texas 77096. Present address: Conoco, Inc., P.O. Box 2197, Houston, Texas 77252. We benefited from the input and assistance of many people. Of special assistance were the teams involved in collaborative studies with Esso Australia and Esso Malaysia: P. Moore, M. Sloan, J. Emmett, B. Burns, A. Partridge, S. Creaney, Hanif Hussein, R. Hill, R. Lovell, and M. Feeley. We also thank the Rock Springs team: R. Beauboeuf, P. McLaughlin, W. Devlin, A. Carroll, Y. Y. Chen, G. Grabowski, Jr., K. Miskell-Gerhardt, M. Farley, R. Webster, and J. Schwalbach. Group members D. Curry and J. Yeakel were always helpful. We also enjoyed and profited from many discussions of these concepts with our colleagues outside Exxon: C. Diessel, R. Boyd, K. Shanley, B. Zaitlin, P. McCabe, M. Hendricks, A. Cohen, and M. Kirschbaum. We thank the reviewers of company reports, whose careful comments on several generations of this work improved it: S. Creaney, M. Feeley, J. Van Wagoner, F. Wehr, J. Yeakel, and A. Young. F. Weber and J. Zullig provided extensive management support. K. Linke translated our sketches into the fine figures herein. We value all the help.

Lake-type controls on petroleum source rock potential in nonmarine basins

Based on numerous empirical observations of lacustrine basin strata, we propose a three-fold classification of lacustrine facies associations that accounts for the most important features of lacustrine petroleum source rocks and provides a predictive framework for exploration in nonmarine basins where lacustrine facies are incompletely delineated. (1) The fluvial-lacustrine facies association is characterized by freshwater lacustrine mudstones interbedded with fluvial-deltaic deposits, commonly including coal. Shoreline progradation dominates basin fill, resulting in the stacking of indistinctly expressed cycles up to 10 m thick. In map view, the deposits may be regionally widespread but laterally discontinuous and contain strong facies contrasts. Transported terrestrial organic matter contributes to mixed type I-III kerogens that generate waxy oil (type I kerogen is hydrogen rich and oil prone; type III kerogen is hydrogen poor and mainly gas prone). The Luman Tongue of the Green River Formation (Wyoming) and the Honyanchi Formation (Junggar basin, China) provide examples of this facies association, which is also present in the Songliao basin of northeastern China, the Central Sumatra basin, and the Cretaceous Doba/Doseo basins in west-central Africa. (2) The fluctuating profundal facies association represents a combination of progradational and aggradational basin fill and includes some of the worlds richest source rocks. Deposits are regionally extensive in map view, having relatively homogenous source facies containing oil-prone, type I kerogen. Examples include the Laney Member of the Green River Formation (Wyoming), the Lucaogou Formation (Junggar basin, China), the Bucomazi Formation (offshore west Africa), and the Lagoa Feia Formation (Campos basin, Brazil). (3) The evaporative facies association represents dominantly aggradational fill related to desiccation cycles in saline to hypersaline lakes and may include evaporite and eolianite deposits. Sublittoral organic-rich mudstone facies are relatively thin but may be (Begin page 1034) quite rich and widespread. The highest organic enrichment coincides with the deepest lake stages. Low input of land plant organic matter results in minimal lateral contrasts in organic content. In some cases a distinctive type I-S (sulfur-rich) kerogen may generate oil at thermal maturities as low as 0.45% vitrinite reflectance equivalent. Examples include the Wilkins Peak Member of the Green River Formation (Wyoming), the Jingjingzigou Formation (Junggar basin, China), the Jianghan and Qaidam basins (China), and the Blanca Lila Formation (Argentina).

Wave-enhanced sediment-gravity flows and mud dispersal across continental shelves: Reappraising sediment transport processes operating in ancient mudstone successions

Recent studies of marine shelf sediment dispersal show that wave-enhanced sediment-gravity flows are widespread phenomena and can transport large volumes of fluid mud rapidly across low-gradient shelves. Flow evolution is controlled by sediment supply, seabed gradient, and spatial distribution of wave energy at the seabed. Using existing flow models, we predict that such flows in mud-dominated sediments will develop a three-part microstratigraphy produced by changing flow conditions, beginning with wave-induced turbulent resuspension, then development of a wave-enhanced sediment-gravity flow, prior to lutocline collapse and suspension settling. Petrographic examination of modern flow deposits collected from the Eel Shelf reveals that resultant beds possess a microstratigraphy consistent with our hypothesis: a silt-rich basal subunit with curved ripple laminae, abruptly overlain by a subunit composed of continuous intercalated silt/clay laminae, and an upper clay-rich drape. Analyses of beds from ancient mud-rich outer-shelf and basinal successions (Cleveland Ironstone, Jurassic, UK, and Mowry Shale, Cretaceous, United States) show that they too contain beds with this three-part organization, suggesting that such flows were active in these ancient settings too. Recognition of these microstructures in these ancient mud-dominated successions demonstrates that sediment in these settings was commonly reworked and transported advectively downslope by high-energy events, contrasting with previous interpretations of these units that deposition was dominated by quiescent suspension settling. Identification of these recognition criteria now allows the products of this newly recognized sediment dispersal mechanism to be identified in other shale-dominated successions.

Lessons from large lake systems-Thresholds, nonlinearity, and strange attractors

Lake systems are the largest integrated depositional complexes in the continental realm: modern lakes have areas up to 374,000 km2, and ancient lake strata extend up to 300,000 km2 in the Cretaceous systems of the south Atlantic and eastern China and the Permian system of western China. The largest lakes do not appear to form a significantly different population in many of their attributes. Their area, maximum depth, and volume closely follow power-law distributions with fractional exponents (–1.20, –1.67, –2.37 respectively), with minimal breaks between the largest lakes and the majority of lakes. Controls on lake size and stratigraphic extent are not straightforward and intuitively obvious. For example, there is little relation of modern lake area, depth, and volume, with origin, climatic conditions, mixis, or water chemistry. Indeed, two-thirds of the largest-area lakes occur in relatively dry climates (precipitation-evaporation ratio [P/E] <1.6). In ancient lake strata, deposits of largest areal extent and thickness tended to form mostly under relatively shallow-water, evaporitic conditions in both convergent and divergent tectonic settings. Geometric and dynamical thresholds appear to govern lake systems as complex, sensitive, nonlinear dynamical systems. Phanerozic examples worldwide indicate that the existence, character, and stacking patterns of lake strata are a function of the interaction of rates of supply of sediment + water and potential accommodation change. Lake-system behavior reflects interactions of four main state variables: sediment supply, water supply, sill height, and basin-floor depth. The stratal record ultimately records five main modes of behavior indicating that nonmarine basin dynamical systems are governed by two fundamental bifurcations and five strange attractors in the sediment + water supply – potential accommodation phase plane: fluvial, overfilled lake, balanced filled lake, underfilled lake, and aeolian/playa. Thus, extremely large lakes are highly dependent on intricate convolutions of climatic and tectonic influences and occur in a variety of settings and climates.

Parasequence types in shelfal mudstone strata—Quantitative observations of lithofacies and stacking patterns, and conceptual link to modern depositional regimes

Mudstone strata have a vast variety of physical, biogenic, and chemical attributes at the lamina to bed scale (approximately millimeters to decimeters thick). Our observations of more than 7 km of Paleozoic to Cenozoic mudstones revealed ordered patterns in this variety, i.e., recurrent associations of lithofacies, bedding style, sedimentary structures, and stratal architecture at bedset to parasequence scales (approximately decimeters to meters thick). We quantified characteristics of each association and their stacking patterns in vertical succession and linked them to sets of depositional processes. Most shelf mudstone strata appear to have accumulated in one of three facies association successions (FASs) that can be related to depositional regimes through characteristic modes of sediment transport and accumulation, as well as variations in benthic-energy and oxygen levels. We interpret these three FASs as records of mud accumulation on different portions of continental shelves that were dominated by storm waves, river floods, or tidal currents. Each FAS records a distinct parasequence type. This approach can help fully integrate insights from oceanographic studies into more robust interpretations of the rock record and rock-property maps.

How to make a 350-m-thick lowstand systems tract in 17,000 years: The Late Pleistocene Po River (Italy) lowstand wedge

The 350-m-thick succession of the Po River lowstand wedge (Italy) associated with the Last Glacial Maximum (deposited over ∼17 k.y) contains stratal architecture at a physical scale commonly attributed to much longer time scales, with complex, systematically varying internal clinothem characteristics. This study investigated clinothem stacking patterns and controls through the integration of seismic reflection data with sediment attributes, micropaleontology, regional climate, eustacy, and high-resolution age control possible only in Quaternary sequences. Three clinothem types are differentiated based on topset geometry, shelf-edge and onlap-point trajectory, internal seismic facies, and interpreted bottomset deposits: type A has moderate topset aggradation, ascending shelf-edge trajectory, and mass-transport bottomset deposits; type B has eroded topset, descending shelf-edge trajectory, and bottomset distributary channel-lobe complexes; and type C has maximal topset aggradation, ascending shelf-edge trajectory, and concordant bottomsets. Type A and C clinothems exhibit reduced sediment bypass and delivery to the basin, whereas type B clinothems are associated with short intervals of increased sediment export from the shelf to deeper water. Clinothems individually span a range of 0.4–4.7 k.y., contemporaneous with significant eustatic and climate changes, but their stacking patterns resemble those found in ancient successions and ascribed to significantly longer durations, indicating that (1) the response time of ancient continental margin–scale systems to high-frequency variations in accommodation and sediment supply could be as short as centuries, (2) even millennial- to centennial-scale stratal units can record substantial influence of allogenic controls, and (3) sandy deposits can be compartmentalized even in a short-duration lowstand systems tract.

Accommodation succession (δA/δS) sequence stratigraphy: observational method, utility and insights into sequence boundary formation

The future of sequence stratigraphy depends on stratigraphers making observations with a common method so that physical frameworks can be clearly separated from interpretations of driving mechanisms. Depositional sequence boundary selection is a well-known controversy that could be resolved with objective recognition criteria. Accommodation succession sequence stratigraphy refines traditional methods, using sedimentary facies, facies associations, vertical stacking, stratal geometries and stratal terminations as the objective record of competing rates of accommodation change and sediment fill through time. Observations are placed in context of lateral (transgression and regression) and vertical (aggradation and degradation) movement of shoreline through time, across multiple timescales in hierarchal stacks. The repeating motif consists of a subaerial unconformity and its correlative subaqueous surface overlain in coastal settings by a basinward shift in coastal onlap and strata with progradational to aggradation stacking, then retrogradation and aggradation–progradation–degradation stacking. These stacking patterns are bounded by key surfaces, recognized by stratal terminations and characteristic vertical successions of facies. This pattern is independent of time duration or position on a sea-level curve, but incorporates data resolution, regional extent and hierarchal stacking. Examples from multiple datasets show the utility and objectivity of the method and provide insights into sequence boundary formation.

Vertical and lateral distribution of lacustrine carbonate lithofacies at the parasequence scale in the Miocene Hot Spring limestone, Idaho: An analog addressing reservoir presence and quality

Lacustrine carbonate lithofacies in the Hot Spring limestone vary systematically at meter to decameter scales and record paleobathymetry, limnologic conditions, and paleogeographic influences. This unit accumulated during the late Miocene in a lake system in an extensional basin complex, closely associated with lava flows and volcaniclastics. Sedimentology and sequence stratigraphy enable understanding and prediction of the occurrence, distribution, and character of lacustrine carbonates. Occurrence of lacustrine carbonates is a function of lake-basin type, in this case a volcanically mediated balanced-filled lake basin that contains various lithofacies: microbialite, grainstone, packstone, wackestone, and carbonate mudstone. Distribution of lithofacies is strongly controlled by depositional subenvironment and gradient (through water depth, bottom energy, circulation, and accommodation). Internal character of microbialites (i.e., type, size, porosity, vertical and horizontal permeability, associated lithotypes, diagenesis) is influenced by stratal position (at the parasequence scale) and location along the depositional profile. Depositional reservoir quality characteristics, such as microbialite porosity and thickness, grainstone size and sorting, and overall carbonate continuity and connectivity peak in the medial sublittoral zone. Conversely, secondary diagenetic effects, such as dissolution, carbonate and quartz cement, and possible authigenic clays, are highest in the updip lake-plain–littoral zones and are less common lakeward. Thus, the medial sublittoral zone has the optimum potential, where primary depositional characteristics are best developed, and negative secondary diagenetic effects are minimal. These strata share many attributes with the Cretaceous presalt systems of the south Atlantic and can provide insights about controls on potential reservoir character, distribution, and connectivity for exploration and development.