Abstract The Arbuckle Group of the midcontinent comprises the mid-southern part of the great American carbonate bank (GACB) and consists mostly of carbonates with a few laterally consistent sandstones. The Arbuckle Group is found in the Anadarko, Ardmore, and Arkoma Basins and surrounding environs in the Texas panhandle, Oklahoma, and Arkansas. These basins represented a significant downwarp associated with early rifting in the area now located in the southern one half of both the states of Oklahoma and Arkansas. Similar to other parts of the GACB, the thick widespread Cambrian–Ordovician Arbuckle Group was deposited as mostly restricted shallow-water marine carbonates. The Arbuckle is a cyclic carbonate dominated by intertidal and shallow subtidal facies. In some areas, supratidal or deeper subtidal facies are observed. The depositional model is represented byan extensive, dominantly regressive, tidal flat with persistent peritidal facies across much of the GACB. These peritidal cycles shallow upward with significant variation in thickness from as thin as 4 ft (1.2 m) to more than 110 ft (>33.5 m) thick. Large-scale regional changes in relative sea level may have had a large influence on the type of cycles and sequences that formed during Arbuckle deposition. Arbuckle strata, especially within third-order sequence boundaries, are correlatable across the basin. Within the sequence boundaries, cycles can be further grouped into packages of sequences that are composed mostly of either intertidally or subtidally dominated cycles. Detailed local to regional correlation of the facies bundles can be made with gamma-ray and resistivity logs; however, facies are commonly obscured by a strong diagenetic overprint that makes detailed correlation difficult. Reservoirs in the Arbuckle are complex, and porosity is controlled by original depositional fabric, diagenesis, paleokarst, and fracture overprint. Upper subtidal and lower intertidal facies typically have the depositional fabric most conducive to reservoir development. Dia-genetic changes are a continuum that begins with early diagenesis, including hypersaline or evaporative conditions as well as vadose and phreatic conditions, and followed by deep phreatic to late thermal diagenesis. Evidence that porosity formed during multiple diagenetic phases exists. Dolomitization and precipitation events are also evidenced at various levels of the profile. Dolomite is the most abundant mineral and can be subdivided into early (syn-genetic to penecontemporaneous) hypersaline dolomite, shallow burial mixed-water (phreatic) dolomite, and deeper burial to thermal (baroque and xenotopic) dolomite. The super-Sauk unconformity is recognized as evidence of a eustatic sea level drop and has been used to mark the boundary between the Sauk and Tippecanoe depositional mega-sequences. The Arbuckle Group contains multiple unconformities at major sequence boundaries. Paleokarst is especially prevalent beneath the super-Sauk unconformity, especially along major sequence boundaries with related unconformity surfaces. Paleokarstic features in the Arbuckle Group have been identified in outcrop in the Arbuckle Mountains of southern Oklahoma and in the southern Ozark uplift in northeastern Oklahoma. Numerous cores and logs indicate collapse breccias that are interpreted to have formed in response to karst conditions. The Arbuckle Group is an important petroleum reservoir in the midcontinent, and has great potential especially for natural gas. Exploration is enhanced by understanding the complex relationships of depositional processes, stratigraphic relationships, paragenesis, and structural overprints. Reservoir development is typically along sequence boundaries, especially where facies have strong diagenetic overprints from dolomitization and dissolution associated with paleokarstic events. No major source rocks exist below or within the Arbuckle Group, so the best reservoirs are structurally related with strong fracture overprints and juxtaposed with source rocks or are along migration pathways.

Abstract Karst is the product of subaerial (terrestrial and coastal) exposure of carbonate rocks, recognizable by features produced by dissolution, precipitation, erosion, sedimentation and collapse in a variety of surface and subsurface landforms, and cave deposits consisting of both cements and sediments. Natural karst constitutes a drainage unit (Fig. 1) consisting of: (1) input of meteoric waters, (2) pre-existing permeability pathways enhanced or reduced by karst flow, and (3) output of resurgent waters with transported sediments and solute. There are various types of karst depending on rock types, insurgence and flow patterns, climate, and etc., corresponding to different modes of porosity creation and destruction. Lithologies can be: (a) tight (dense) with bedding plane control, (b) tight, with fracture control, and (c) porous, with intergranular porosity control. Flow patterns can be diffuse, confluent, allogenic (water collected from non-karst drainage) or authigenic (catchment surface is karst) (Fig. 2). Most diagenetic models for subaerial exposure in carbonate rocks have been developed in Holocene Caribbean carbonates, forming a karst in very porous carbonates with diffuse recharge and flow, short exposure times, low relief and interaction with coastal exposure environments. This is but one of the many types of karst and these diagenetic models cannot be applied in many of the cases encountered by explorationist. Karst systems present zoned flow patterns, normally with many anomalies in the distribution of the hydraulic potential (hydraulic traps, confined flow) and resulting thermal and chemical zonation The level of regional groundwater saturation (water table, piezometric level) separates the infiltration or

Abstract Cores of Ordovician-age Arbuckle Group carbonates from Oklahoma were examined for evidence of paleokarst. The depositional and diagenetic fabric of the rock was analyzed and compared with outcrop analogs to illustrate the nature of sedimentary, karstic, and diagenetic facies. Burial diagenesis and hydrothermal alteration have in many cases obscured the original fabric of these rocks. Arbuckle rocks in different tectonic settings and stratigraphic intervals in the subsurface of south-central and north-central Oklahoma display surprisingly similar suites of karstic and diagenetic phenomena. Dissolution cavities, solution-enlarged fractures, collapse breccias, and vugular porosity are present in many cores and attest to the predominance of fabric-destructive processes in the development of Arbuckle paleokarst Collapse breccias and sediment-filled solution features bear striking resemblance to outcropping analogs. Primary speleothemic precipitates were not readily observed; either they were not precipitated or were obscured by later dolomitization. Phreatic cements were more commonly encountered than vadose cements. A complex history of exposure, subsidence, and diagenesis is recorded in these rocks. Although the actual physical manifestations of paleokarst are not difficult to identify, interpretation of the genesis and age of these features is decidedly problematic. Arbuckle carbonates have been exposed to surficial weathering for periods of variable intensity and duration numerous times in geologic history. Paleokarst horizons may have developed subjacent to disconformities within and between formations of the Arbuckle Group and where these rocks subcrop beneath regional unconformities. This complex hierarchy of unconformities can produce numerous porous horizons whose preservation potential may depend on subsidence rates rapid enough to prevent extensive low-temperature phreatic cementation, thereby preserving the open pore network of the karst profile. Burial diagenesis is evidenced by the multi-event dolomitization of these rocks. Ferroan and nonferroan "growth-zoned" baroque and limpid phreatic dolomite cements commonly occlude vugular and fracture porosity. Host-rock carbonates have been extensively replaced or neomorphosed. Cathodolumincscent microscopy and chemical staining indicate that "growth-zoned" baroque dolomite is commonly uniform in composition and was precipitated under mildly reducing conditions. Dolomite cementation was arrested by the migration of oil into the remaining pore space.

Abstract In the Val Verde basin area, the Lower Ordovician Ellenburger Group represents two third-order sea-level fluctuations as determined from facies architecture, stacking patterns, and accommodation plots based on subsurface core studies. These third-order cycles are superimposed on part of a second-order sea level rise and fall of the upper Sauk Sequence. Ellenburger Group deposition ceased in response to a major sea-level fall represented by a second-order unconformity at the end of the Sauk Sequence. This upper bounding surface is characterized by an extensively developed karst profile, indicative of prolonged subaerial exposure. Carbonate dissolution and cave formation were most significant at and along major block boundaries and resulted in the generation of vugs, caverns, caves, and solution-enlarged fractures and joints. The roofs of larger caves were brecciated as the caves were buried and subjected to static loading by flooding of the platform and deposition of the Simpson Group. The fracture and breccia porosity found in the cave roof portions of these karst profiles accounts for much of the regionally significant porosity developed within the Ellenburger Group. A detailed upper Ellenburger Group isopach of the Brown-Bassett/JM fields area illustrates a linear trend of isopach thins coincident with the crest of the present structural trend indicating that, not only were the structures active during Ellenburger time, but that the entire structural trend was regionally high. A well-developed paleokarst system was described from cores taken from the Ellenburger interval in this area. This karst system has a distinctive log signature characterized by elevated gamma ray response that has been interpreted to represent more radioactive clay-rich sediment deposited as cave-fill material. Correlation and isopach mapping of the cave-fill portion of the cave zone, shows the main portion of the paleo-cave network to extend across the entire Brown-Bassett/JM trend in a west-northwesterly direction, paralleling the principal bounding faults. The caves are thickest (up to 70') and best developed adjacent to, but not necessarily coincident with the crests of the structures. It is interpreted that the maximum cave development was localized along the main basement fault zones which acted as secondary conduits for fluid flow.

Abstract Garland field is an asymmetric anticlinal trap located in the north-central Big Horn basin, Wyoming. The field produces hydrocarbons from interlayered, fractured limestones and dolomites of the Madison Limestone (Mississippian). Significant karstification occurs in the form of field-wide intraformational breccias and locally developed cavernous porosity. Most breccias and caverns apparently formed during prolonged post-Madison exposure, prior to deposition of the overlying Darwin Sandstone. Three types of karst breccia occur: (1) red, siltstone-matrix breccias, (2) clay-matrix breccias, and (3) dolomicrite-matrix breccias. Red, siltstone-matrix breccias occur in the upper 30 ft (9 m) of the Madison, and are related to the exposure event at the top-of-Madison unconformity. Clay-matrix breccias form a regionally correlatable layer which is about 50 ft (15 m) thick in the Garland field area. These breccias, which occur roughly 200 ft (60 m) below the top of the Madison, probably formed by evaporite dissolution and subsequent collapse. Dolomicrite-matrix breccias occur at the tops of shallowing-upward sequences at several levels within the Madison, and they apparently pre-date clay-matrix breccias. Dolomicrite-matrix breccias may have formed during periodic intraformational exposure events.

Abstract Brecciation is a common diagenetic fabric in subsurface dolomitized sequences of the Upper Elk Point Group in western Canada. While not generally associated with hydrocarbon production from these sequences, breccias were a product of the same deep-burial diagenetic processes responsible for creating other secondary pores from which production occurs. Several relationships demonstrate conclusively that brecciation and other associated styles of dolomite dissolution were deep-burial in origin, having formed coincident with, or after, pressure solution in these rocks. These breccias, therefore, are an example of deep-burial "karstification." Upper Elk Point breccias are invariably associated with fractures and burial replacement anhydrite, both of which were related to local faulting. They are always associated with dolomites and show no preference for development along depositional cycle breaks or formation tops. The common presence of stylolitic clasts, rotated at all angles to each other and the horizon, demonstrates that solution collapse occurred after the onset of pressure solution at depth. Contrary to popular models, brecciation is not unique to near-surface processes such as freshwater karstification or leaching of evaporites. For the Upper Elk Point Group, to invoke these processes as explanations for the observed brecciation is to totally ignore the stratigraphical, petrographical and geochemical attributes of these sequences. Our case study shows that given the right tectonic and diagenetic settings, impressive deep-burial dissolution can occur in buried carbonate sequences, resulting in creation of substantial secondary porosity and brecciation.

Abstract The top surface of the Trenton Limestone and equivalent carbonate units has be variously described as a subaerial exposure surface (paleokarst), a submarine erosion surface, and a submarine hardground. Detailed study of the contact between the carbonates and overlying shale in outcrop and core and regional stratigraphic analysis indicate that the surface represents a drowning unconformity on the Galena and Lexington carbonate platforms in Ohio and Indiana. This unconformity also appears within the Sebree Trough in Indiana between the platforms, but it is within the overlying shale section rather than at its base. The unconformity has not been recognized in the Point Pleasant Basin in central and southern Ohio. Paleokarst may locally exist on this surface in southern Ontario.

Abstract The Lower Ordovician Ellenburger Group of West Texas is a prolific oil and gas producer in the Permian Basin of West Texas. Regional analysis of depositional and diagenetic fabrics within the Ellenburger show reservoir facies to be dominated by a variety of breccia fabrics. A descriptive classification of Ellenburger breccias, including fracture, mosaic, clast-supported chaotic, siliciclastic-matrix-supported chaotic, and carbonate-matrix-supported chaotic types, allows simplified but genetically significant characterization of these highly varied breccia types. Although undoubtedly Ellenburger breccias are of diverse origins, vertical sequences from unbrecciated Ellenburger upward through chaotic, mosaic, and fracture breccia types in sub-Simpson Group Ellenburger reservoirs are best interpreted in terms of a karst model of cave formation, infill, and collapse. Roof and lower collapse portions of these sequences form the best reservoir intervals with siliciclastic-rich cave fill sediments commonly acting as baffles to fluid migration. Core from the Gulf McElroy St. No. 1 well illustrates the characteristic succession of breccia fabrics used to develop the karst model. Karst facies recognized in the McElroy St. No. 1 core are lower collapse, cave-fill, and cave-roof, fracture breccias are not well developed in this core. Additional core material from producing zones in other wells is also displayed to illustrate pore types and breccia fabrics most commonly associated with producing intervals.

Abstract Casablanca Field, offshore Spain, produces oil from karsted Jurassic - Cretaceous carbonates. Subaerial exposure that produced the paleokarst was significant and affected up to 386 meters of section. Locally, karst dissolution was extensive enough to form large, solution-enhanced fractures or small, probably horizontal, caves. Multiple phreatic zones that developed during regional uplift probably produced the various cave levels recognized in cores. Cores contain representative and distinctive attributes of paleokarst including breccias, cave-fill sediment, and fractures. Fitted, mosaic, and rubble breccias which are distributed throughout the cored interval formed in part during cave-roof collapse and compaction of cave-fill sediments. The cave-fill is principally dolomitized carbonate mud or clast-supported sediment that is red in the upper portions of the cored interval and green in the lower portions. Fractures, in which a significant volume of the reservoir pore volume is contained, formed during both karst-collapse and tectonism.

Abstract The Devonian Chert of the Black Warrior and Arkoma basins is part of a regional chert accumulation across the southern North American Continent. In the Arkoma Basin the Devonian is called the Penters Chert in the subsurface and the Sallisaw Formation on outcrop (Figure 1). In the Black Warrior Basin the Devonian Chert is an unnamed formation (Figure 2). The Arkansas Novaculite and the Caballos Formation are equivalents found in the Ouachita and Marathon region, respectively. Other shelfward equivalents of the novaculites are the Thirty-one Formation of the Permian Basin, Camden Chert of Tennessee, and Clear Creek Chert of Illinois. The Devonian Chert of the Black Warrior and Arkoma basins has a greater affinity towards its shelfal equivalents to the west, east, and north than to Ouachita equivalents to the south. Whereas there has been little published on the shelfal equivalents, a large amount of data exists on the Arkansas and Caballos novaculites. The source of silica and the depositional environment for the novaculites and cherts in general have been debated for years. A biogenic source for the silica has been accepted by most workers. Other theories for the silica source advocate alteration of volcanic ash and volcanism promoting growth of siliceous organisms and the production of siliceous sediments; however the evidence of volcanism in the Ouachitas is scarce (limited to scattered tuffs in the Mississippian). General agreement on the depositional environment has not been reached for the novaculites with McBride (1989), Thomas (1988), Sholes (1978), McBride and Folk (1977), Folk

Abstract Rocks of the Cambro-Ordovician systems are exposed in the Appalachian fold belt in the states of Tennessee, Virginia, Georgia, and Alabama, and extend westward from the southern Appalachians under the Black Warrior Basin of Alabama-Mississippi. These strata are practically all carbonate and were deposited on the eastern and southern passive margins of the North American craton. The great thickness of strata (up to 7000 ft. for the Lower Ordovician to Middle Cambrian) constitutes what R.N. Ginsburg has termed the Great American Bank, a vast Early Paleozoic carbonate platform extending from Newfoundland down the Appalachians, completely across the southern part of North America, and part way up the western part of North America, thus encircling more than half of the craton. The typical bank facies extends far up on the eastern craton, its characteristic strata being present as far north as the Canadian border. Throughout this extensive area the Cambro-Ordovician facies are very similar (Figs. 1 and 2). The Middle and Upper Cambrian develop sandy "pinch out" edges along the Transcontinental Arch and are onlapped by the Lower Ordovician. Practically all of the Lower Ordovician carbonate facies consist of shallow, restricted marine, dolomitized, upward shoaling, subtidal to tidal flat cycles. These occur over a broad area of more than 1000 miles (e.g., from the Oneota-Shackopee dolomites of Wisconsin to the Knox Group of the Warrior Basin in Alabama (Fig. 3 and 4).

Abstract This volume is a compilation of papers relative to paleokarst and associated reservoirs. The examples illustrate many of the rock types, and stratigraphic, structural, and paleotopographic features of carbonate strata which result chiefly from solution and collapse due to ingress of meteoric waters at and below unconformities. Examples presented here range from settings with considerable dissolution and collapse to those with significant unconformities but little evidence of meteoric alteration. It is estimated that 20–30% of recoverable hydrocarbons are in some way related to unconformities. Paleokarst reservoirs may also be important future reservoirs for application of horizontal drilling technology.

The Trenton and Black River Formations of the Michigan Basin have been diagenetically altered by a complex sequence of events related to both the stratigraphic and structural history of the basin. The physical distribution and chemical composition of dolomite in the Trenton and Black River Formations are variable and suggest multiple episodes of dolomitization. The most extensive diagenetic alteration of both Trenton and Black River limestones has occurred in fracture-controlled hydrocarbon reservoirs. Within reservoirs several stages of dolomitization were followed by carbonate and sulfate cementation, and sulfide mineralization. Although the general patterns of reservoir alteration have been recognized for some time, possible causes of such alteration have not been adequately addressed. Several lines of evidence indicate that mineralization and hydrocarbon migration are related and occurred in the late Paleozoic, perhaps in response to compressional deformation caused by Appalachian tectonism. During such episodes, fluids were mobilized and channeled vertically through preexisting fracture zones. This fluid migration also served to drive maturing hydrocarbons out of Trenton–Black River source beds and into previously dolomitized, high-porosity intervals. This general mechanism could be applied to other fracture-related reservoirs in the Michigan Basin area based on the regional distribution of Mississippi Valley–type (MVT) reservoir alteration and compressional stress fabrics.

Abstract Tethyan carbonate platforms are relatively short-lived depositional systems that were born, developed, died, and were resurrected in a tectonically active area. In fact, they experienced the entire geodynamic spectrum (rifting, drifting, transtension, transpression, and collision) of the Wilson cycle. Many of their peculiar features, such as lack of ramps and common occurrence of faulted boundaries and megabreccia wedges, suggest tectonics as a primary control, with eustatic sea-level oscillations being only a secondary overprint. During the rifting stages, many platforms drowned or underwent tectonic retreat, while others were able to survive the pulsating Liassic subsidence. Shedding of ooid sands during sea-level highstands, and subaerial exposure with karstification and bauxite formation during lowstands, are some typical events that occurred during the Middle Jurassic-Middle Cretaceous period of drifting and convergence. Beginning from Late Jurassic, the Tethyan platforms were progressively involved in the Alpine-Himalayan orogenic systems, and, after an early uplift, were generally buried under thick piles of siliciclastics. Some, however, were resurrected before the final collisional stages.

Abstract Flanks of carbonate platforms steepen as the platform rises higher above the basin floor. Furthermore, platform slopes are steeper on average than siliciclastic slopes. Termination of platform growth through rapid submergence or suffocation by siliciclastics produces an unconformity, because the clastics cannot assume the steep carbonate slope angle and because they are shed from different directions. This "drowning unconformity" resembles the unconformity produced by a lowstand of sea level, even though it is associated with a rise or a highstand of sea level. Examples of drowning unconformities include the mid-Cretaceous unconformity in the Gulf of Mexico, the unconformities on the flanks of the Wilmington platform, the Lahave platform, and the platforms off Morocco, all drowned in the earliest Cretaceous, and the mid-Jurassic unconformities on certain platforms of the High Atlas. Drowning unconformities are best developed on platforms that rise 800 m or more above the basin, have concave upper flanks of 6° or more, and commonly possess an elevated rim. Drowning and burial of smaller platforms with gentle flanks still produce unconformities, because the pattern of sediment input and dispersal is different for carbonates and siliciclastics.

Abstract A deterministic computer program has been developed to simulate the stratigraphic evolution of two-dimensional transects across sedimentary basins. Clastic, carbonate, and mixed clastic-carbonate systems can be simulated. The main application of the program is to simulate basin stratigraphy using all available data as constraints. In this paper we concentrate on the simulation of carbonate systems and illustrate the following two specific applications of the program. (1) The history of sea-level fluctuations is reconstructed using the stratigraphy and geometry of carbonate systems as constraints. In a first example, the architecture of isolated Miocene carbonate buildups is simulated using sea-level curves similar to published eustatic charts. In a second example, the stratigraphic patterns at the margin of a Mesozoic cratonic carbonate basin are used to derive an estimate of the sea-level history. (2) Our understanding of the controls on carbonate platform architecture is improved by isolating individual processes. In this respect, we investigate the possible significance of isostasy in controlling the formation of intra- and inter-platform basins and in the initiation of carbonate pinnacles.

Abstract Modeling of early Paleozoic passive margins in the Cordilleran and Appalachian orogens indicates that factors controlling growth of early Paleozoic passive-margin carbonate platforms were thermally controlled subsidence, time-dependent flexure of the lithosphere, and at least two orders of eustatic sea-level changes. Initiation of the carbonate platforms in Middle Cambrian time followed a marked reduction in supply of Lower Cambrian coarse siliciclastic material to the passive margins. Two-dimensional modeling of palinspastically restored cross sections implies that the reduction in relief of onshore sediment sources resulted mainly from increased time-dependent flexural rigidity and extension of the area of subsidence into the craton. Continued increase in rigidity and bending of the craton edge, combined with a long-term eustatic sea-level rise, further reduced the supply of siliciclastic material to the carbonate platforms, resulting in a progressive cratonward shift of the siliciclastic shoreline and cratonward expansion of the carbonate platforms. Additional evidence of eustatic controls on growth of the platforms is obtained from one-dimensional analyses of post-rift subsidence of the platforms. The effects of sediment loading and lithification are removed from cumulative subsidence curves, producing reduced cumulative curves, designated R1 curves. The first-order form of the R1 curves is exponential, matching closely the form of theoretical curves calculated from cooling plate models for passive margins. After subtracting best-fit model cooling curves from the R1 curves, the residual curves, designated R2 curves, contain evidence of two orders of "events" superimposed on the thermally controlled subsidence of the margins. One event is the long-term rise and fall of sea level observed in the two-dimensional modeling. The long-term event coincides temporally with the Sauk transgression-regression on the craton. The other consists of repeating short-term sea-level changes with wave lengths of 2 to 6 Ma. The short-term sea-level events have similar timing in the southern Canadian Rockies, in the Great Basin, and in the Virginia-Tennessee Appalachians, suggesting a eustatic control. These inferred eustatic events appear to have exerted a major influence on the lithologic framework of the carbonate platforms. The long-term eustatic fall in Late Cambrian and Ordovician time augmented the reduction in rate of net subsidence of the platforms resulting from decay of the thermal anomaly. The much slower subsidence probably was the principal cause of the marked expansion in Late Cambrian and Ordovician time of carbonate shoal facies within the platforms. The short-term eustatic events produced distinct cycles composed of fine-grained shaley material in their lower halves and coarser grained shoal facies in their upper halves. Apparently, each short-term sea-level rise reduced the rate of carbonate production sufficiently to allow widespread deposition of subtidal facies with large amounts of interbedded siliciclastic mud. During each short-term fall, rates of carbonate production increased and led to expansion of shoal facies across the platforms.