Abstract In the Appalachian basin, the Middle Devonian organic-rich shale interval, including the Marcellus Shale, is an important target for exploration. This unconventional gas reservoir is widespread across the basin and has the potential to produce large volumes of gas (estimated to have up to 1,307 trillion cubic feet of recoverable gas). Although the Middle Devonian organic-rich shale interval has significant economic potential, stratigraphic distribution, depositional patterns and petrophysical characteristics have not been adequately characterized in the subsurface. Based on log characteristics, tied to core information, the lithostratigraphic boundaries of the Marcellus and associated units were established and correlated throughout West Virginia and southwestern Pennsylvania. Digital well logs (LAS files) were used to generate estimates of lithology and to identify zones of higher gas content across the study area. In addition, a lithologic solution was calibrated to X-ray Diffraction (XRD) data. Using previous studies on organic shale, relationships between the natural radioactivity (as measured by the gamma-ray log) were incorporated with techniques to identify gas-prone intervals. The comparison between the Uranium content and the measured bulk density identified intervals in the Marcellus having high gas saturations and were used to generate an approach to correct water saturations. These techniques of identifying lithology and potential gas in the Marcellus are useful to identify areas of higher exploration potential and to target zones for fracture stimulation or to land a horizontal leg.

Abstract Middle Pennsylvanian coal units in eastern Kansas produce commercial quantities of coal-bed natural gas. Annual coal-bed natural gas production in 2008 was 49.1 billion cubic feet (Bcf) (13% of state output); cumulative production since 2000 is 165 Bcf. Coal beds are commonly less than two feet thick and are mostly produced by vertical wells at 80- to 160-acre spacing. Wells usually have comingled gas production from several coal beds. The main producing region is a fourcounty area (Labette, Montgomery, Neosho, and Wilson counties) in southeastern Kansas immediately north of the Oklahoma state line. Most wells are not prolific; their average maximum production rate is approximately 67 mcf/day, peaking about 14 months after initial production. Decline rates are low, as some coalbed natural gas wells have produced 15 years and beyond. North-northwest–south-southeast trending production fairways can be defined by mapping maximum production rates. These fairways generally correlate to where coal beds are individually and compositely thick. The most prolific wells in the thickest coal units record maximum production rates as great as 615 mcf/day. The median as-received gas content for coals in southeastern Kansas is 139 scf/ton, with maximum gas content of approaching 400 scf/ton. Gas content in eastcentral and northeastern Kansas coal beds generally runs half that of southeastern Kansas, indicating economics of coal-bed natural gas production are harsher northward. Coals increase in depth westward at a rate of approximately 20 feet per mile. Their gas content commensurately increases by 10 to 20 scf/ton for each 100 feet of burial. Thin (<4 foot) black shale beds interbedded with the coal units may have commercial potential, for their as-received gas content can be great as 65 scf/ton, but 20 scf/ton is the median of all shale samples assayed.

Abstract The widespread Western Interior Plains aquifer system of the central United States provides a significant potential for sequestration of CO 2 in a deep saline formation. In Kansas, several severely depleted Mississippian petroleum reservoirs sit at the top of this aquifer system. The reservoirs are primarily multilayered shallow-shelf carbonates with strong water drives. Fluid flow is strongly influenced by natural fractures, which were solution enhanced by subaerial karst on a Mississippian–Pennsylvanian regional unconformity. We show that three-dimensional (3-D) seismic volumetric reflector curvature attributes can reveal subtle lineaments related to these fractures. Volumetric curvature attributes applied to a 3-D seismic survey over a Mississippian oil reservoir in Dickman field, Ness County, Kansas, reveal two main lineament orientations, N45°E and N45°W. The northeast-trending lineaments parallel a down-to-the-north fault at the northwestern corner of the seismic survey and have greater length and continuity than the northwest-trending lineaments. Geologic analysis and production data suggest that the northeast-trending lineaments are related to debris-, clay-, and silt-filled fractures that serve as barriers to fluid flow, whereas the northwest-trending lineaments are related to open fractures that channel water from the underlying aquifer. The discrimination of open versus sealed fractures within and above potential CO 2 sequestration reservoirs is critical for managing the injection and storage of CO 2 and for evaluating the integrity of the overlying seal. Three-dimensional seismic volumetric curvature helps to locate fractures and is a potentially important tool in the selection and evaluation of geologic sequestration sites.

Abstract Many essential aspects are involved in quantitative characterization of oolite carbonate reservoirs. Rock-facies classification, external facies geometry, and internal rock-property distribution are fundamental to characterization for reservoir simulation and prediction of future hydrocarbon recovery. The typical challenge for small Midcontinent fields in the U.S is absence of high-resolution seismic data capable of resolving relatively thin reservoir intervals. An integrated geostatistical approach is presented that uses available well data from the St. Louis Limestone in the Archer Field, southwestern Kansas, to improve oolitic reservoir modeling and corresponding streamline simulation. The proposed approach uses neural network and stochastic methods to integrate different types of data (core, log, stratigraphic horizons, and production); at different scales (vertical, horizontal, fine-scale core data, coarse-scale well-log data); and variable degrees of quantification (facies, log, well data). The results include: three-dimensional stochastic simulations of facies distribution of St. Louis oolitic reservoirs; improved reservoir framework models (lithofacies) for carbonate shoal reservoirs; increased understanding of spatial distribution and variability of petrophysical parameters within carbonate shoal reservoirs; quantified measures of flow-unit connectivity; 3D visualization of the St. Louis carbonate reservoir systems; streamline simulations of the static geostatistical models to rank and determine the efficacy of the geological modeling procedure; and better understanding of key factors that control the facies distribution and the production of hydrocarbons within carbonate shoal reservoir systems. Geostatistical 3D modeling methods are applicable to other complex carbonate oolitic reservoirs or siliciclastic reservoirs in shallow-marine settings.

Abstract Horizontal drilling has become a key technology used to reduce costs and enhance recoveries from producing reservoirs. Through 2001, commercial databases contained records on 34,777 horizontal wells from 72 countries. Canada (18,005 wells) and the United States (11,344 wells) were the leading countries for horizontal drilling. More than 5400 horizontal wells were recorded outside of North America. Russia, Venezuela, Oman, United Arab Emirates, Nigeria, Saudi Arabia, and Indonesia were the leading countries in terms of numbers of wells. Although the concept of horizontal drilling emerged in the 1920s, economic viability was not demonstrated until the 1980s, when pilot projects at Rospo Mare field in Italy (1982) and Prudhoe Bay field (1984) and in the Austin Chalk of Texas (1985–1987) achieved three- to fourfold productivity increases with less than twofold cost increases. From a base of 51 wells in 1987, horizontal drilling increased rapidly; it expanded to the world’s active producing provinces and peaked during 1997 with 4990 wells. Horizontal drilling, which increases wellbore exposure to the reservoir, has delivered multiple benefits. Operators have used horizontal wells to revive economic production, to increase and speed recoveries, to reduce costs, and to increase rate of return. These benefits are critical for operators that must cope with increasing competition and volatile oil prices. The objective of this paper is to characterize the global geographic and geologic distribution of horizontal wells and to illustrate some of the benefits of horizontal drilling with examples from key fields and trends.

Abstract The deepest onshore horizontal well in California is the ARCO-DOE 91X-3, which was drilled at Yowlumne field in the San Joaquin Basin, California, to exploit the thinning, distal margin of a fan-shaped, layered turbidite complex. Yowlumne is a giant oil field that has produced more than 17.2 million m 3 (108 million bbl) of oil from the Stevens Sandstone, a clastic facies of the Miocene Monterey Shale source rock. Most Yowlumne production is from the Yowlumne Sandstone, a layered, fan-shaped, prograding Stevens turbidite complex deposited in a slope-basin setting. Well-log, seismic, and pressure data indicate seven depositional lobes with both left-stepping and basinward-stepping geometries. To facilitate cost-effective exploitation of remaining field reserves, a 3-D model of the reservoir architecture was constructed from log-derived petrophysical data, constrained by core analyses. This model indicates concentration of channel and lobe facies along the axis and west (left) margin of the Yowlumne fan to result in average net/gross sandstone ratios of 80%, porosity (Ø) of 16%, and liquid permeability (K liquid ) of 10–20 md. By contrast, more abundant levee and distal margin facies along the east margin result in shale-bounded reservoir layers with higher clay contents and lower net/gross sandstone ratio (65%), porosity (12%), and permeability (2 md). Thus, the distal fan margin is not an attractive place to drill for oil. However, modeling indicates that most remaining field reserves exist along the east fan margin. Although a waterflood during the last 20 years will enable recovery of 45% of original oil in place along the fan axis, about 480,000 m 3 (3 million bbl) of oil trapped at the thinning fan margins will be abandoned with the current well distribution. The ARCO 91X3 was a Department of Energy-funded well to test economic recovery of this bypassed oil by utilizing a high-angle deviated well with three hydraulic fracture stimulations to provide connectivity between reservoir layers, thereby providing the same productive capacity as three vertical wells. The well was drilled along strike to a measured depth of 4360 m (14,300 ft) to tangentially penetrate at angles up to 85° as much as 600 m (2000 ft) of the distal margin of the Yowlumne fan. Because of drilling and completion difficulties, the proposed multiple fracture stimulations were not attempted. Nonetheless, use of highly deviated to horizontal wells with multiple fracture stimulations remains an economically viable option for maximizing productivity from the thinning, distal margins of layered, low-permeability turbidite reservoirs.

Abstract The giant Wilmington oil field of Los Angeles County, California, on production since 1932, has produced more than 2.6 billion barrels of oil from basin turbidite sandstones of the Pliocene and Miocene. To better define the actual hydrologic units, the seven productive zones were subdivided into 52 subzones through detailed reservoir characterization. The asymmetrical anticline is highly faulted, and development proceeded from west to east through each of the 10 fault blocks. In the western fault blocks, water cuts exceed 96%, and the reservoirs are near their economic limit. Several new technologies have been applied to specific areas to improve the production efficiencies and thus prolong the field life. Tertiary and secondary recovery techniques utilizing steam have proved successful in the heavy oil reservoirs, but potential subsidence has limited their application. Case history 1 involves detailed reservoir characterization and optimization of a steamflood in the Tar zone of Fault Block II. Lessons learned were successfully applied in the Tar zone, of Fault Block V (4000 m to the east). Case history 2 focuses on 3-D reservoir property and geologic modeling to define and exploit bypassed oil. Case history 3 describes how this technology is brought deeper into the formation to capture bypassed oil with a tight-radius horizontal well.

Abstract Ahorizontal-well redevelopment drilling program around the flanks of the South Pass 62 salt-dome field resulted in significant successes and costly failures. Successful wells exploited thin, oil-filled shoreface sandstones; partially depleted zones; and massive, sand-filled channels. Failures were those wells that attempted to connect multiple fault blocks and drain low-resistivity/laminated-sandstone reservoirs. This paper reviews the field history; describes the geologic setting, including a summary of significant structural features and producing-sandstone depositional environments; discusses the horizontal-well strategy; and examines successful and unsuccessful wells. South Pass 62 field lies 50 km (30 mi) east of the Mississippi River delta in 104 m (300 ft) of water. The field was discovered in 1965, developed with 61 directionally drilled wells from three platforms in the late 1960s, redeveloped from 1986 to 1988 with 31 wells from a fourth platform, and redeveloped again from 1994 to the present with horizontal and directionally drilled slim-hole sidetracks. A 3-D seismic-based field study completed in 1994 identified reservoir targets for the horizontal-drilling program. Nearly 60 stacked, variable pay sandstones combine with steep formation dips and extensive faulting to create a complex field with hundreds of reservoirs. The field lies on the north flank of a mushroom-shaped, south-leaning salt dome that rises from below 8000 m (25,000 ft) to within 200 m (656 ft) of the seafloor. Typical formation structural dips decrease from 70° adjacent to the salt to 10° off structure. Several generations of faults exist, with throws ranging from centimeters to more than 100 m. Approximately 60 Pliocene and Miocene deltaic and turbidite pay sandstones ranging in depth from 1158 to 5791 m (3800 to 19,000 ft) onlap the salt.

Abstract The Andrew field is a small oil and gas field with a 58-m oil column, a 66-m gas cap, and a simple dome structure, producing entirely from horizontal wells. It has been a successful development for BP and the Andrew field partners, with plateau oil production extending 18 months beyond the predicted onset of field decline. Development success has been helped substantially by focusing presanction activity on key reservoir uncertainties and business decisions. The decisions that resulted were to drill all horizontal producers to optimize low gas-oil-ratio (GOR) oil recovery, to closely manage the reservoir under production, to delay gas coning and water breakthrough, and to collect sufficient surveillance data to allow regular updating of the reservoir management plan. The objective of the Andrew development is to maximize oil recovery before going to gas-cap blowdown. The challenge is to manage the GOR throughout the life of the field. Central to this are well design, location, numbers, and the drawdown strategy. The horizontal wells produce at higher rates (average 10 MBOPD) and at relatively lower drawdown pressures (100 psi). They recover increased reserves per well (13 MMBO per well), compared with a conventional well. Project economics were improved as well numbers were reduced from 24 in a conventional well case to 10 horizontal producers. Low GOR oil production has been maximized by well positioning relative to the gas-oil contact (GOC) and oil-water contact (OWC); by drilling long wells that enter the reservoir on the crest and exit through the flank of the field; and by completion design, perforation strategy, careful well management, and drilling two additional infill wells. As a result, the recovery factor has risen from 45% at sanction in July 1996 to 49% by the end of 2002. The final field recovery factor is expected to rise to 53% by sidetracking low-rate producers and continuing to manage the reservoir drawdown. Oil reserves also have increased from 132 to 154 MMBO from 1996 to 2002 as the result of an increase in field STOOIP (stock tank original oil in place) and better-than-expected reservoir and horizontal-well performance.

Abstract R am Powell is one of the major tension-leg platform (TLP) developments in the eastern deep-water Gulf of Mexico. The three main turbidite reservoirs are producing from a total of five horizontal, open-hole, gravel-packed wells. The high rate and high ultimate recovery from these horizontal wells have reduced substantially the development well count (in one reservoir by 50%). Well A-3ST1 at one time held a Gulf of Mexico (GOM) record of highest rate for a single well—40,900 BOE per day. Understanding reservoir architecture is key to execution of a trajectory plan. This knowledge is best obtained through use of pilot wells near the horizontal-well path. In using exploratory or appraisal wells as pilots, survey accuracy can be critical to success. Horizontal-well trajectories can be optimized in reservoirs (such as levees or sheet sandstones) in which lateral variations are well understood. At Ram Powell, horizontal wells in these depositional systems were drilled oblique to strike, so as to transect the entire section. In more laterally variable systems, such as channels, more well control or geosteering is required. New applications of petrophysical tools that are unique to horizontal wells aided in drilling and reservoir analysis. Density images from a 121-mm (4¾-in.) OD logging-while-drilling (LWD) density tool were used to calculate the dip of the thin-bedded L sandstone. This tool was also used to evaluate the geometry of calcite-cemented zones in an unconsolidated sandstone. Quantitative fluorescence technique (QFT), a patented Texaco process, quantifies the fluorescence in cuttings for pay evaluation. Measurements were used to distinguish mudstone, thin-bedded sandstone, and massive sandstone.