Abstract

SACROC unit forms the largest producing unit of the Pennsylvanian-age Horseshoe Atoll play in the foreland of the Fort Worth Basin. Since discovery in the 50’s, primary, secondary, and tertiary recovery activities have been extensive, with over 1500 wells in this, the first CO2 flood in West Texas. In spite of this history, only the basics of the stratigraphic and petrophysical architecture are understood. Here I contrast the fundamentally different reservoir styles of layered Canyon shelf strata with the fractured and possibly karst-modified Cisco section.

This study is based on data from well log and core-based examination of the northern third of the unit including 550 wells and 3500 ft of core. The 700 ft thick reservoir column consists of Canyon and Cisco carbonates that change from layered cyclic open-shelf subtidal cycles with minimal diagenetic overprint (early and mid Canyon) to high-energy shoal-related cycles with frequent exposure surfaces (late Canyon-early Cisco) and increased evidence for cycle and sequence-scale erosion. Early Cisco deposition was characterized by dramatic changes in depositional style including growth of pinnacle reefs and the formation of complex fractured muddy crinoid-dominated facies that resemble Waulsortian deeper-water buildups.

The transition from shoal/exposure-capped Late Canyon depositional cycles to mud-dominated crinoidal facies with associated pinnacle development is consistent with the gradual backstepping and eventual drowning of the Horseshoe Atoll in the SACROC area. This is consistent with several studies that have suggested that the irregular upper surface topography of the platform is not erosional. However, the extensive fracturing and newly defined abrupt facies changes are best viewed in terms of subaerial erosion. These different interpretations yield important differences in reservoir property distribution and must be solved before confident 3D models are constructed.

Abstract

High-relief (> 100 m), well-exposed Pennsylvanian (Virgilian) algal mounds of the western Orogrande basin (San Andres Mountains, New Mexico) consist of stacked high-frequency sequences bounded by surfaces of paleo-subaerial exposure. Facies within and proximal to this complex include (1) boundstones (cement-, algal-, peloidal- and foraminiferal-rich variants) within mound-core regions, (2) packstones (skeletal-, foraminiferal-, algal-, and peloidal-rich debris) within mound-flank regions, and (3) auxiliary facies (oncoidal wackestone, algal bindstone, carbonate mudstone) formed in low-relief off-mound regions. Sequence stacking in this system was the result of high-frequency, high-amplitude glacioeustasy that prevailed during late Paleozoic time. Paleorelief on subaerial exposure surfaces records glacioeustatic amplitudes in excess of 80-100 m; preserved paleoslopes locally exceeding 40° (compactionally corrected) indicate that the mounds were cemented geologically-instantaneously and not easily eroded, even at lowstands.

Timing of stratal accretion within a given sequence varied significantly as a function of both spatial and temporal position within a biohermal complex. Stratal accretion in mound-nucleation stages produced late-falling-stage (“catch-down”) sequences both on- and off-mound. During the “acme” phase of mound growth, accretion occurred during sea-level rise and maximum inundation phases, nearly pacing production of short-term (glacioeustatic) accommodation and producing an anomalously thick, near keep-up sequence in the mound-core region. Subsequent sequences of the mound core nucleated atop significant paleobathymetric relief, grew during maximum inundation to incipient fall and therefore display thicknesses that likely reflect long-term (tectonic) accommodation potential. In a shallow-water biohermal system capable of accretion rates commensurate with production of short-term accommodation space, very thick near keep-up sequences are laterally offset from one another in progradational, retrogradational, or random shifting patterns owing to the limits of short-term (glacioeustatic) accommodation production in any given locality. Consequently, successive sequences of very thick mound-core facies are laterally offset, resulting in marked lateral reservoir heterogeneity where porosity is developed in mound-core facies.

In the algal bioherms of the western Orogrande basin, reservoir geometry is further affected by dolomitization: stratiform dolomitization that preferentially affects peritidal facies and strata proximal to sequence boundaries is widespread in this system. Field, petrographic, and geochemical analyses reveal two replacive dolomite phases, and several dolomite cements. The facies-selective (FS) replacive phase affects peritidal facies, and consists of micritic to finely crystalline, fabric-preserving dolomite with high Sr and Na concentrations and low to moderate Fe and Mn. The non-facies-selective (NFS) phase affects facies proximal to sequence boundaries, and consists of finely to medium crystalline dolomite that is both fabric-selective and commonly fabric-destructive. This phase has moderate to locally elevated Sr and Na concentrations, and moderate to high Fe and Mn. Dolomite cements line and occlude pores and molds, and consist of medium to coarsely crystalline rhombs with moderate Sr and Mn, low Na, and high Fe. Stable oxygen isotope values for bulk dolostones exhibit a wide range and include relatively low values, suggesting precipitation and recrystallization over a wide range of conditions.

FS dolomitization occurred penecontemporaneously with peritidal deposition at glacioeustatic fall to lowstand. NFS dolomitization occurred early postdepositionally, by reflux of mesosaline to hypersaline brines during early–middle stages of each glacioeustatic transgression. Virgilian lowstand gypsum deposits of the southwestern Orogrande provided the Mg source for the transgressive brines. Dolomite cements precipitated largely under reducing conditions of the burial environment. Dolomitization appears to be associated with porosity preservation in this system, such that reservoir potential is particularly heightened in sequence-bounding, mound-flank strata.

Abstract

The Strawn Formation is one of the more enigmatic hydrocarbon producing formations on the Eastern Shelf of the Midland Basin. Of the Strawn fields, St. Lawrence trend in Glasscock County, Texas is one of most prolific. In order to gain a better understanding of the nature of the reservoir properties of the Strawn a continuous core was taken by Texaco. The core is 120 ft thick and includes both the upper and lower contacts with the overlying Canyon sand/shale sequence and the underlying “Atoka” Formation.

The core is composed of 22 high-frequency shallowing-upward cycles. Each cycle begins with a black, fissile shale or black argillaceous lime mudstone containing an open marine fauna. These shaly and argillaceous carbonates represent the initial relative rise in sea level. Overlying the initial muddy part of the cycles is a succession of carbonate facies indicating a continuous shallowing, reflecting a reduction in accommodation space due to sedimentation, as well as an increase in water energy. Each cycle is capped by grain-rich packstones or skeletal grainstones. These grain-rich facies represent the primary reservoir facies in the Strawn Formation. The high-frequency cycles can be grouped into three composite sequences, which likely represent three fourth-order sequences, and are correlatable with cycle data from additional cores on the Eastern Shelf. Each of these cycles and the composite sequences are interpreted as representing fluctuations in sea level driven by continental glaciation.

The tops of cycles exhibit signs of exposure, including karstification and soil development. Diagenetic processes associated with sub-aerial exposure in some cases acted to enhance porosity and in other cases to reduce porosity through pervasive fresh-water cementation. Porosity enhancement was through the dissolution of unstable carbonate grains resulting in the development of moldic and vuggy porosity. Maximum porosity measured in core samples is on the order of 10%, and correlates well with log porosity. Zones of high porosity represent the best reservoir development, and can be correlated over a wide distance on logs.

Abstract

Wolfcampian skeletal, pelletal and oolitic packstones/grainstones and skeletal buildups are productive reservoir facies along much of the eastern margin of the Central Basin Platform. Cored wells from the Phillips Edwards West Field, Ector County, Texas, are described in this paper. These Wolfcamp shelfal carbonates display numerous shallowing upward depositional cycles, typically capped by exposure surfaces and abundant karstification. These depositional cycles are commonly truncated with only minor development and preservation of intertidal and supratidal facies. Porosity and permeability are low and discontinuous, as widespread meteoric cementation and chemical compaction have plugged much of the early interparticle pore networks. Where present, interparticle, vuggy, fracture, and stylolitic porosity are developed near cycle tops. They have been locally enhanced by dissolution associated with exposure and karstification. Reservoir permeability is dominated by fracture permeability. Matrix permeability is almost always less than one (1) millidarcy. Permeabilities up to 10 darcies have been measured in fractured whole cores associated with karst zones. Reservoir fluid flow is thought to be strongly controlled by the distribution of stacked karst zones. These reservoir zones are interpreted to be poorly interconnected resulting in a zoned reservoir system.

Abstract

The large subsurface feature termed the Horseshoe Atoll, located in the Midland Basin of west Texas, is a series of primarily Missourian and Virgilian aged carbonate reservoirs developed atop the poorly understood Desmoinesian capped Garza Platform. Carbonate deposition in these reservoirs occurred during sea level high-stands while erosional modification of these carbonates occurred during ensuing sea level low-stands. The existence of icehouse conditions and the accompanying numerous and significant fluctuation of sea level created reservoirs of various complexity. Carbonate deposition was also heavily influenced by the location of the “Atoll” in regard to the paleoequator and the direction toward North at the time of deposition. The tectonic history of the Midland Basin and the Garza Platform in particular also plays a part in the present day configuration of the reservoirs around the “Atoll”.

Abstract

A reservoir model for the South Wasson Clear Fork (SWCF) field, Gaines County, Texas has been constructed using information from an outcrop analog in Apache Canyon, Sierra Diablo Mountains, and from the subsurface. The SWCF field contains two reservoirs, the lower Clear Fork and the middle Clear Fork. There is a seal between the middle and upper Clear Fork, and the upper Clear Fork apparently contains residual oil resulted from oil remigration. Permeability and initial water saturation are calculated using simple relationships to total porosity despite the wide diversity in rock fabrics. The large volume and patchy distribution of anhydrite apparently reduces porosity with little change in pore size, resulting in a grouping of the various rock fabrics into one rock-fabric petrophysical class. Porosity values were calculated from wireline logs, and vertical profiles of permeability and initial water saturation were calculated using a single porosity-permeability transform.

A sequence stratigraphic framework was used to distribute porosity and permeability throughout the reservoir. The framework was based on outcrop studies, seismic response, and core descriptions. The cyclic nature of the Clear Fork is illustrated by outcrop descriptions, and this character has been applied to the subsurface using seismic data and core descriptions. The basic correlation unit is the high-frequency cycle (HFC). The HFC’s are subtidal cycles defined by caps of grain-dominated fabrics. One group has silt bases and can be easily correlated using acoustic and porosity logs. The second group is correlated on the basis of statistical data that shows that grain-dominated dolostones tend to be more porous than mud-dominated dolostones. Each HFC is divided into a grain-dominated and a mud-dominated flow layer for reservoir simulation studies. Statistical analysis of porosity and permeability data from the outcrop shows poor small-scale spatial correlation within one rock-fabric flow layer but a larger scale of variability that may have more spatial correlation and impact on fluid flow.

The rock-fabric reservoir model contains 44 layers in the middle Clear Fork and 44 layers in the lower Clear Fork. Permeability is distributed within these layers using various statistical methods, and simulation results are compared. The results are also compared with a proportional method of building the reservoir model. The impact of fractures on performance is examined.

Abstract

Medium and short radius horizontal drilling has provided new life to mature fields in the Permian Basin. Many properties have been rejuvenated utilizing laterals from existing well bores targeting unswept hydrocarbons in secondary projects or untapped hydrocarbons due to heterogeneity in the reservoirs. BP (former ARCO Permian) has taken advantage of this technology and efficiency in three of their old properties, giving them new life.

BP has re-completed 25 wells with horizontal laterals in Grayburg, San Andres, and Holt reservoirs. The existing vertical wells averaged three-fold increases in rate after re-completion. Utilizing a modified completion rig minimized re-completion costs. Laterals were horizontal within approximately 150’ from the vertical well bore. To capture the heterogeneity in the reservoirs these wells slanted through the 50-100’ target both top-down and bottom-up.

Today, horizontal drilling has done to our mature fields what 3D seismic did to exploration in the Permian Basin several years ago - provide life to properties that were barely breathing. This technology is not for every field and every reservoir. There is a learning curve for each area. With the proper reservoir description and clear objective(s) horizontal drilling can be utilized with a high degree of success and profitability.

Abstract

The Halley and Emperor fields, Winkler County, west Texas, are located along the western margin of the Central Basin Platform (CBP), a late Paleozoic fault-bounded structural high in the Permian Basin. Well data, regional 2D seismic lines, and a 3D seismic data set were used to develop a detailed structural and stratigraphic interpretation for the area. Variance volume attributes were derived from the 3D seismic data, which improved imaging of subsurface features.

The Halley and Emperor fields are situated over asymmetric anticlines with short steeper limbs that are faulted by steeply dipping reverse faults with a component of right-lateral strike-slip displacement. The orientation of the fold axes and faults is NNW-SSE, which is parallel to the overall trend of the CBP’s western margin. Deformation occurred during early Pennsylvanian to early Leonardian time.

The CBP can be subdivided into two major blocks or tectonic domains: the Andector Block to the north and the Fort Stockton Block to the south. These blocks were located between an inferred right-lateral mega-shear, which forced the Andector and Fort Stockton blocks to undergo clockwise rotation. A left-lateral shear zone must have existed along the E-W fault boundary between the Fort Stockton and Andector blocks. The Halley structure is situated at the southwestern corner of the Andector Block and shows evidence of younger (middle Pennsylvanian to middle Leonardian?) left-lateral strike-slip deformation. In contrast, deformation along the Emperor structure to the north had ceased by late Pennsylvanian time, as indicated by the age of strata that onlap the structure.

Abstract

The late Middle to Late Ordovician Montoya Group (130-160 m thick) of southern New Mexico and west Texas formed a gently dipping ramp on the southern Laurentia passive margin. A detailed sequence stratigraphic study of the Montoya Group based on measured sections, hand-held gamma-ray logs and conventional wireline logs indicates a single 2nd-order (10-15 m.y. duration) supersequence composed of 3 to 5 regionally correlative 3rd-order (2-5 m.y. duration) sequences. Smaller-scale 4th- and 5th-order parasequence sets and parasequences are discernible at outcrops but do not appear to correlate regionally.

The basal Montoya supersequence unconformity, a composite karst surface representing a 10-30 m.y. hiatus, varies from a sharp planar surface to irregular karst sinkholes (from a few meters to a few 10’s of meters deep) filled with blocks of the underlying El Paso Group in a sandstone matrix. The 2nd-order Lowstand Systems Tract is not present in the study area. Updip the lower part of the 2nd-order Transgressive Systems Tract consists of carbonate-cemented, burrowed, coarse granule sandstone (Cable Canyon Sandstone) that passes basinward and upward into burrow-mottled skeletal packstone with abundant hardground surfaces (Upham Formation). The skeletal packstone is abruptly overlain by cherty, laminated calcisiltite of the lower Aleman Formation. A 2nd-order maximum flooding surface is not discernible in these deepwater facies, but it is a maximum flooding “zone” evident on the gamma-ray logs almost immediately above the packstone/calcisiltite contact. The 2nd-order Highstand Systems Tract is composed of subtidal cherty dolostone (upper Aleman Formation) that passes upward into peritidal carbonate (Cutter Formation) capped by a regional disconformity.

Abstract

The number and location of layers have a major impact on reservoir simulation results. In the past, computer capacity has limited the number of layers that could be used, and scale-up from numerous layers to a few layers was a major issue. With the improvement in computer technology, however, more layers can be accommodated and the number and location of the layers have become a major issue. Two layering methods used to describe the South Wasson Clear Fork reservoir (SWCF) are compared: rock-fabric flow layers and proportional layering. The rock-fabric method divides the high-frequency cycles (HFC) into two flow layers, an upper grain-dominated layer and a lower mud-dominated layer. HFC’s are based on upward-shallowing successions observed in core and outcrop and correlated using established sequence boundaries as a guide and porosity as an indication of grain content. The proportional method is based on dividing the intervals between five major correlation markers into a fixed number of layers on the basis of a combination of vertical porosity variograms and cycle thickness.

Maintaining high and low permeability is a major goal when defining layers; high permeability controls the flow rate, and low permeability controls cross flow. In the SWCF reservoir, permeability is a simple function of porosity because the effect of grain/crystal size and sorting has been reduced by a diagenetic overprint of poikilotopic anhydrite. Fractures (modeled as a type of touching vug) probably provide a small improvement in permeability over the matrix, although many fractures observed in cores are filled with anhydrite. Therefore, vertical porosity profiles from wireline logs can be converted to permeability profiles using a single porosity-permeability transform. A comparison of permeability values by layer shows that the proportional layers tend to group high and low permeability, whereas the rock-fabric flow layers tend to maintain high and low permeability. This difference results in significantly different predictions of reservoir performance.

Abstract

Since the early 1980’s the evaluation of dip meter data as a series of data points has become more and more common, with slow development of the method to include curvature analysis, Cumulative Dip, and more recently Synthetic Deviation. Most of these methods are helpful in determining locations of features in the borehole but require good quality dip data. Most of this data is collected from oriented cores, borehole imaging, or dipmeter data. Each depends on a single data set that is of reasonable quality the must be acquired at the time the well is drilled. The solution to this limitation is to simply expand the concepts used in cumulative dip analysis to the analysis of not just bed dip magnitude but also to bed dip direction, wellbore deviation direction, and wellbore deviation from vertical, in the same sample number domain as simple cumulative dip analysis. The wellbore deviation direction and wellbore deviation from vertical can be measured even after casing of the well, making identification of faults, unconformities, and sequence boundaries possible even after the well was drilled, by simply running a well bore deviation log. In addition the use of residuals of this cumulative data analysis can add extra detail to interpretation of these features and making it easy to compare each independent data set.

Abstract

The Glorieta formation (Permian, Leonardian) is a shallow water carbonate unit that occurs between the prolific Clear Fork (Leonardian) and San Andres (Leonardian-Guadalupian) formations on the Central Basin Platform and northwestern and eastern shelves of the Permian Basin. Along the eastern side of the Central Basin Platform, the Glorieta interval is often a productive reservoir in its own right, as is the case in the Texaco-operated Robertson, Wharton and South Harris units in Gaines County, Texas.

In Robertson, Wharton and South Harris units, the Glorieta formation occurs at an average depth of 5800 feet and is roughly 500 feet thick. Examination of cores from numerous wells in the Wharton and South Harris units reveals that the Glorieta consists of several successions of shallowing upwards cycles culminating in tidal flat capped facies. These peritidal deposits are characterized by laminated, stromatolitic dolowackestones and dolopackstones containing pisolites, intraclasts, pellets, and fenestral textures. Tidal flat bases are generally argillaceous and sandy and pores in the dolopackstones are plugged with anhydrite. Thus, these tidal flat intervals are tight and are typically barriers to fluid flow.

Subtidal cycles beneath the tidal flat capped intervals consist of skeletal, pelletal mud-rich to grain-rich dolopackstones. Skeletal grains in the deeper parts of the cored Glorieta interval include numerous crinoids, bryozoa, and scattered brachiopod and sponge (?) fragments, suggestive of deposition in an open marine environment. Cycles are typically extensively burrowed. The subtidal cycles immediately beneath the tidal flat caps are characterized by abundant pellets, peloids and sparse bivalve fragments, suggesting deposition in a more restricted environment.

Individual cycles in the cored wells have been grouped into 4 high frequency sequences that can be traced across the Robertson-Wharton-South Harris area. The younger two sequences (HFS 1 and HFS 2) display tidal flat caps in both dip and strike directions across most of the study area. The older (deeper) sequences (HFS 3 and HFS 4) are capped with tidal flats on the western side of the study area. However, the eastern side of the area displays subtidal capped cycles in these deeper HFS’s. The best reservoir intervals occur in these subtidal capped cycles.

The Glorieta formation produces from various subtidal facies exhibiting a wide range of reservoir characteristics; however, the best reservoir facies consists of grain-rich dolopackestones. Porosities in subtidal mud-rich dolopackstones average 7.7% and permeabilities average .76 md. In grain-rich dolopackstones, porosities average 10.5% and permeabilities average 3.0 md. Pore types include interparticle, moldic, fracture and some intercrystalline. Variations in reservoir quality are mappable and may follow trends of depositional facies patterns inherited from paleotopography.

The Robertson-Wharton-South Harris Glorieta trend was discovered in 1942. Through primary and waterflooding activities, the units have produced over 115 million barrels of oil to date. Significant amounts of movable oil remain to be produced, but enhanced production requires detailed geologic modeling and understanding of the controls on the reservoir heterogeneity resulting from depositional and diagenetic variations.

Abstract

Numerous Permian Basin gas/oil fields contain residual oil intervals beneath producible hydrocarbon columns. Residual oil intervals were formed by massive recharge of meteoric water into the Permian Basin in the middle Tertiary. The source of this water was the east-dipping limb of the Rio Grande Rift. This rift developed in three phases: First was initial uplift of a large, tilted, slowly extending landmass (Late Oligocene-Early Miocene); Second was rapid extension and normal faulting (Middle-Late Miocene); Third was slow extension (Pliocene-Recent).

The east limb of the Rift covered an area from its center (present-day Rio Grande River) to the western edge of the Central Basin Platform. The first phase of initial uplift, tilting and slow extension of the rift recharged massive volumes of meteoric water into the subsurface. This recharging water had enough hydrostatic head (energy) to partially- to completely sweep oil columns to residual oil saturation (Sorw). The second phase of rapid extension broke the eastern limb of the rift into numerous horsts and grabens. As soon as classic Basin and Range province horsts and grabens developed, only small landmasses were available to recharge much smaller volumes of meteoric water into the Permian Basin. Many oil columns re-saturated, some with gas instead, while others stayed at residual oil saturation (Sorw). The third phase of slow extension recharged only small volumes of meteoric water into the Permian Basin.

Though residual oil intervals may not have any primary/secondary recovery potential, they may be additional targets of opportunity for tertiary CO2 floods.

Abstract

The lower Devonian Thirtyone Formation of West Texas and New Mexico is the largest chert reservoir succession in the world, having produced more than 750 million barrels of oil. As much as 650 million barrels of additional mobile oil remains in these reservoirs, making this play an important target for further exploitation.

All Thirtyone chert reservoirs have much in common, including relatively high porosities and low permeabilities and a bimodal pore distribution containing abundant microcrystalline pores. However, distinct differences in depositional geometries and styles of reservoir heterogeneity are apparent between reservoirs in proximal and distal settings. Proximal reservoirs are composed of a single, thick (up to 100 ft), sheet-like chert unit that extends for hundreds of square miles. Heterogeneity in these reservoirs, which were deposited on a gently sloping outer platform during regional transgression, is a function of faulting, fracturing, and dissolution of associated carbonate along unconformities.

By contrast, distal reservoir successions comprise thin, vertically stacked and laterally discontinuous chert intervals whose origin is a function of transport and deposition of siliceous sediments as debris flows and turbidites. Flow units in these reservoirs are thin (1020 ft), spatially limited in size, and separated vertically and laterally from one another by low-permeability mud-rich, siliceous sediments and hemipelagic deposits. Flow unit distribution is the result of both paleotopography and sea-level cyclicity. Cherts are most abundant in transgressive and early highstand legs of sea-level rise-fall cycles and display offset stacking suggestive of topographically controlled reciprocal sedimentation. Faults and fractures appear to be less significant contributors to reservoir heterogeneity in distal reservoirs.