A study was done to characterize deep, prolific Ellenburger gas reservoirs at Lockridge, Waha, West Waha, and Worsham‐Bayer fields in Pecos, Ward, and Reeves counties in West Texas. A major effort of the study was to interpret a 176-mi23-D seismic data volume that spanned these fields. Well control defined the depth of the Ellenburger, the principal interpretation target, to be 17 000–21 000 ft (5200–6400 m) over the image area. Ellenburger reflection signals were weak because of these great target depths. Additionally, the top of the Ellenburger had a gentle, ramp‐like increase in acoustic impedance that did not produce a robust reflection event. A further negative influence on seismic data quality was the fact that a large portion of the 3-D seismic area was covered by a variable surface layer of low‐velocity Tertiary fill that was, in turn, underlain by a varying thickness of high‐velocity salt/anhydrite. These complicated near‐surface conditions attenuated seismic reflection signals and made static corrections of the data difficult. The combination of all these factors has caused many explorationists to consider this region of west Texas a no‐record seismic area for deep drilling targets. Although the 3-D seismic data aquired in this study produced good‐quality images throughout the post‐Mississippian section (down to ∼12 000 ft, or 3700 m), the images of the deep Ellenburger targets (∼20 000 ft, or 6100 m) were limited quality. The challenge was to use this limited‐quality 3-D image to interpret the structural configuration of the deep Ellenburger and the fault systems that traverse the area so that genetic relationship could be established between fault attributes and productive Ellenburger facies. Two techniques were used to produce a reliable structural interpretation of the 3-D seismic data. First, log data recorded in 60-plus wells within the 3-D image space were analyzed to determine where there was evidence of overturned and repeated units caused by thrusting and evidence of missing sections caused by normal faulting. These petrophysical analyses allowed reliable fault patterns and structural configurations to be build across 3-D seismic image zones that were difficult to interpret by conventional methods. Second, cross‐section balancing was done across the more complex structural regimes to determine if each interpreted surface that was used to define the postdeformation structure had a length consistent with the length of that same surface before deformation. The petrophysical analyses thus guided the structural interpretation of the 3-D seismic data by inferring the fault patterns that should be imposed on the limited‐quality image zones; the cross‐section balancing verified where this structural interpretation was reliable and where it needed to be adjusted. This interpretation methodology is offered here to benefit others who are confronted with the problem of interpreting complex structure from limited‐quality 3-D seismic images.

A multidisciplinary team, composed of stratigraphers, petrophysicists, reservoir engineers, and geophysicists, studied a portion of Boonsville gas field in the Fort Worth Basin of north‐central Texas to determine how modern geophysical, geological, and engineering techniques can be combined to understand the mechanisms by which fluvio‐deltaic depositional processes create reservoir compartmentalization in a low‐ to moderate‐accommodation basin. An extensive database involving well logs, cores, production, and pressure data from more than 200 wells, [Formula: see text] [Formula: see text] of 3-D seismic data, vertical seismic profiles (VSPs), and checkshots was assembled to support this investigation. We found the most important geologic influence on stratigraphy and reservoir compartmentalization in this basin to be the existence of numerous karst collapse chimneys over the [Formula: see text] [Formula: see text] area covered by the 3-D seismic grid. These near‐vertical karst collapses originated in, or near, the deep Ordovician‐age Ellenburger carbonate section and created vertical chimneys extending as high as 2500 ft (610 m) above their point of origin, causing significant disruptions in the overlying clastic strata. These karst disruptions tend to be circular in map view, having diameters ranging from approximately 500 ft (150 m) to as much as 3000 ft (915 m) in some cases. Within our study area, these karst features were spaced 2000 ft (610 m) to 6000 ft (1830 m) apart, on average. The tallest karst collapse zones reached into the Middle Pennsylvanian Strawn section, which is some 2500 ft (760 m) above the Ellenburger carbonate where the karst generation began. We used 3-D seismic imaging to show how these karst features affected the strata above the Ellenburger and how they have created a well‐documented reservoir compartment in the Upper Caddo, an upper Atoka valley‐fill sandstone that typically occurs 2000 ft (610 m) above the Ellenburger. By correlating these 3-D seismic images with outcrops of Ellenburger karst collapses, we document that the physical dimensions (height, diameter, cross‐sectional area) of the seismic disruptions observed in the 3-D data equate to the karst dimensions seen in outcrops. We also document that this Ellenburger carbonate dissolution phenomenon extends over at least 500 mi (800 km), and by inference we suggest karst models like we describe here may occur in any basin that has a deep, relatively thick section of Paleozoic carbonates that underlie major unconformities.

Abstract In this paper, we present a new approach to software development for solving engineering problems. The paper presents the possible solutions to the problems that we encounter during the development of engineering software where we incorporate different techniques and tools such as a database, case base, knowledge base, expert help, on-line help, multiple-applications, and guidance board/map. The paper also discusses the reusability of code that is derived for a different application. Finally, the paper illustrates how we use this approach to develop a comprehensive software system. The paper concludes that engineering software must be versatile in nature. The software should be a working tool for industry experts and experienced engineers. The software should also be a training and learning tool for young and inexperienced engineers.

A multidisciplinary team, composed of stratigraphers, petrophysicists, reservoir engineers, and geophysicists, studied a portion of Boonsville gas field in the Fort Worth Basin of North‐Central Texas to determine how modern geophysical, geological, and engineering techniques could be combined to understand the mechanisms by which fluvio‐deltaic depositional processes create reservoir compartmentalization in a low‐ to moderate‐accommodation basin. An extensive database involving well logs, cores, production, and pressure data from 200‐plus wells, [Formula: see text] [Formula: see text] of 3-D seismic data, vertical seismic profiles (VSPs), and checkshots was assembled to support this investigation. The reservoir system we studied was the Bend Conglomerate, a productive series of gas reservoirs composed of Middle Pennsylvanian fluvio‐deltaic clastics 900 to 1300 ft (275 to 400 m) thick in our project area. We were particularly interested in this reservoir system because evidence suggested that many of the sequences in this stratigraphic interval were deposited in low‐accommodation conditions (that is, in an environment where there was limited vertical space available for sediment accumulation), and our objective was to investigate how fluvio‐deltaic reservoirs were compartmentalized by low‐accommodation depositional processes. Using an extensive well log database (200 plus wells) and a core‐calibrated calculation of rock facies derived from these logs, we divided the Bend Conglomerate interval into ten genetic sequences, with each sequence being approximately 100 ft (30 m) thick. We then used local VSP and checkshot control to transform log‐measured depths of each sequence boundary to seismic two‐way time coordinates and identified narrow seismic data windows encompassing each sequence across the [Formula: see text] [Formula: see text] 3-D seismic grid. A series of seismic attributes was calculated in these carefully defined data windows to determine which attributes were reliable indicators of the presence of productive reservoir facies and which attributes could, therefore, reveal distinct reservoir compartments and potentially show where infield wells should be drilled to reach previously uncontacted gas reservoirs. Our best success was the seismic attribute correlations we found in the Upper and Lower Caddo sequences, at the top of the Bend Conglomerate. These sequences were deposited in a low‐accommodation setting, relative to other Boonsville sequences, and we found that reflection amplitude and instantaneous frequency, respectively, were reliable indicators of the areal distribution of reservoir facies in these low‐accommodation sequences.

The objective of this work was to learn more about the reservoir characteristics in the Barnett Shale. Specifically, from an analysis of pressure, production, interference, and fracture treatment data in three Mitchell Energy Corporation Cough area wells, the authors can infer the relationship between the induced hydraulic fractures and the natural fracture system in the reservoir. The authors are learning something about drainage area size, shape, and orientation.

The objectives of this project are to define undrained or incompletely drained reservoir compartments controlled primarily by depositional heterogeneity in a low-accommodation, cratonic Midcontinent depositional setting, and, afterwards, to develop and transfer to producers strategies for infield reserve growth of natural gas. Integrated geologic, geophysical, reservoir engineering, and petrophysical evaluations are described in complex difficult-to-characterize fluvial and deltaic reservoirs in Boonsville (Bend Conglomerate Gas) field, a large, mature gas field located in the Fort Worth Basin of North Texas. The purpose of this project is to demonstrate approaches to overcoming the reservoir complexity, targeting the gas resource, and doing so using state-of-the-art technologies being applied by a large cross section of Midcontinent operators.

ABSTRACT The feasibility of detecting and accurately estimating gas desorption parameters from a history match of Devonian shale well-test pressure data is examined. Both drawdown and buildup tests are analyzed, and based on the results of these analyses a desorption-specific well-test design is proposed. The results from a simulated desorption-specific test suggest that it is possible to characterize gas desorption from a well test with reasonable accuracy, even when the effects of desorption are partially masked by wellbore storage and skin effects.

ABSTRACT Hydraulic fracturing is one of the most effective stimulation methods for increasing oil and gas recovery. There are several powerful 3-D (and pseudo 3-D) numerical simulators available in the industry for designing hydraulic fracture treatments. With each model, the accuracy of the simulation results depends heavily on the input data used. Thus, it is important to prepare a complete and accurate input data set for these simulators. This can be difficult to do, however, not only because lots of data items are required, but also because many decisions must be made, such as selecting fracturing fluids, selecting proppants, and determining the optimal pumping schedule and technique. Using artificial intelligence methods, we have developed a powerful, interactive software application that uses a series of friendly and intelligent interfaces used to acquire the large amount of data needed by fracture simulators. In these interfaces, the user is guided through a set of screens specific to his problem. More importantly, the system helps the user make many decisions, such as selecting fracturing fluids and additives, selecting proppants, and selecting pumping schedules and pumping techniques. The system includes several databases from which a considerable amount of information can be accessed automatically, such as typical formation data, fluid rheology, and proppant conductivities. The system also provides a powerful expert help facility. In addition, based on the fracture simulation results, the system produces data sets that can be used to run reservoir performance and economics software. This paper presents the methodology we used to implement the system and fully describes the different parts of the system, including the knowledge bases, fuzzy logic evaluators, data acquisition blocks, and databases. The paper also describes the expert help facility in the system. A companion paper1 describes how we designed the system.

ABSTRACT There are several powerful 3-D (and pseudo 3-D) numerical simulators available in the industry for designing hydraulic fracture treatments. With each model, the accuracy of the simulation results depends heavily on the input data used. Thus, it is important to prepare a complete and accurate input data file for these simulators. This can be difficult to do, however, not only because a lot of data items are required, but also because many decisions have to be made. Using artificial intelligence methods, we have developed a powerful and interactive software application that uses a series of friendly and intelligent interfaces to acquire the large amount of data needed by fracture simulators. In these interfaces, the user is guided through a series of screens specific to his problem. More importantly, the system helps the user make many decisions, such as selecting fracturing fluids and additives, selecting proppants, and selecting pumping schedules and pumping techniques. The system includes several databases from which a considerable amount of information can be accessed automatically, such as typical formation data, fluid rheology, and proppant conductivities. The system also provides a powerful expert help facility. In addition, based on the fracture simulation results, the system produces data sets that can be used to run reservoir performance simulators and economics software. This paper is the first of two papers which will summarize our work – how to design and implement the intelligent interface system that we call STIMULATION EXPERT. This paper analyzes the requirements of different users, outlines the purpose and the system structure, and describes the system. The second1 of the two papers will describe the implementation and application of the system in detail. This system can be used to help engineers design hydraulic fracture treatments. The methodology described in both papers can be used to develop other similar software for different types of simulators in the industry.

Analytical techniques that have grown out of the Gas Research Institute's Appalachian basin research effort are being successfully applied to the Mississippian Barnett shale of Texas' Fort Worth basin. Analysis has shown that well performance can be explained by a layered reservoir description, productive pay may be overstated by log analysis, productivity is enhanced by natural fractures, and long, bounded, high conductivity fractures are indeed being created and propped. As with the Appalachian shales, the key to fracture treatment optimization appears to be a better characterization of the Barnett shale via an integration of log and test data. The paper describes the geologic setting, drilling since 1981, a summary of the cooperative work performed on the 2 T.P. Sims well, the shale-specific log, core analysis fracture orientation, frac treatment analysis, and production data analysis.

Abstract This paper presents a summary of the complete integration of geological, geophysical, geochemical, and reservoir engineering analyses into a single reservoir model that accurately simulates observed production and pressure transient data fa-a gas well completed in the Devonian Shales of the Appalachian Basin in Pike Co., KY. This well was part of a three-well research program sponsored by the Gas Research Institute (GRI). Data collected and evaluated include a detailed openhole logging suite (including the formation microscanner), 240 ft of oriented whole core, open- and cased-hole stress tests, nitrogen slug tests, fracture treatment pressure data, and pre- and post-fracture pressure buildup tests. The purpose of this paper is to discuss the data collected on the well, to present the data analyses, and to demonstrate how the various analyses are being integrated to form a better understanding of the physical properties controlling well performance in the Devonian Shales. This paper also provides intrinsic reservoir properties, such as matrix permeability, desorption isotherms, and natural fracture spacings from the Devonian Shales that were measured in this well.

ABSTRACT The Gas Research Institute (GRI) has been sponsoring a three-well cooperative program with Ashland Exploration, Inc. (ΑΕΙ) during the past two years evaluating the Devonian Shale and Berea sandstone formations in Pike County of eastern Kentucky. The COOP 2 was targeted to evaluate the Berea and was GRI's first comprehensive tight gas sand research project in the Appalachian Basin. The objective of the research was to use and apply technologies developed previously in other GRI research. This paper will summarize the integrated reservoir and hydraulic fracture descriptions determined from analyzing the data collected in this project. The results presented can be applied by operators to better understand the productive mechanisms in the Berea reservoir, to predict well performance, and to design completion procedures and stimulation treatments. The methodology can also be applied to other tight gas sand formations.

The report summarizes recent activities and results from a new, three-well, field research project being conducted by the Gas Research Institute in eastern Pike County, KY. The field research project is focusing on improving gas production from the Devonian Shales and the Berea Sand. Two wells, the Ashland Ford Motor Company (FMC) 69 and 80, have already been drilled; research activities are underway on both. In the report, the authors' summarize the operations and all results and conclusions from the research to date on both wells. Two new methods for measuring the permeability of the Devonian Shales using core samples in the laboratory are also described. One is modified pulse testing technique, and the other is x-ray computed tomography. Both appear to hold considerable promise for quantifying matrix and fracture properties in the Shales. Finally, the report also describes the results of cooperative research activities conducted with Mitchell Energy Corporation (MEC) in the Barnett Shale in the Fort Worth Basin of north-central Texas.

ABSTRACT This paper describes the evaluation, completion, stimulation, and testing of Barnett Shale wells operated by Mitchell Energy Corporation (MEC) in the Fort Worth Basin of north-central Texas. In particular, the paper presents a detailed analysis of data collected from a Gas Research Institute/MEC cooperative research well that was used to gain a better understanding of the mechanisms controlling gas production from the Barnett Shale. The Barnett Shale covers a large geographic area of north-central Texas; thus far, most of the activity is centered around Wise and Denton Counties, TX. On the basis of the data analyzed, the Barnett Shale appears to be characterized best with a layered reservoir description where most of the well deliverability is associated with thin, higher permeability, naturally fractured zones, while most of the gas-in-place is confined to thicker, extremely low permeability layers. Gas-in-place in the Barnett Shale may average 10 to 12 Bscf per 160 acres, but the better wells are expected to recover only 1 to 1.5 Bscf in a twenty-year well life. Approximately 20% of the gas-in-place in the Barnett Shale is adsorbed gas; however, desorption appears to become important only after the reservoir pressure falls below 1,000 psia (original reservoir pressure is about 4,000 psia). Evaluation of fracturing pressures and post-fracture well tests suggests long, propped hydraulic fractures are being achieved and that the hydraulic fractures are typically contained within the Barnett Shale by limestone formations above and below.

The report summarizes recent work in GRI's Comprehensive Study Well (CSW) program in the Devonian Shales of the Appalachian Basin. It focuses particularly on recent post-fracture diagnostic experiments and efforts to improve reservoir description in the Devonian Shales. The authors have learned that single fracture treatments pumped commonly over 500- to 700-ft intervals in the Shales do not stimulate all zones effectively. They have also found that overall reservoir quality, permeability anisotropy, two-phase flow effects, and particularly, the distribution of permeability within the Shales can significantly impact stimulation effectiveness. They also show that stimulation treatment designs similar to those recommended and pumped in the Shales are being used successfully by Appalachian Basin operators in other low permeability reservoirs, particularly tight sands.

In the report, the authors present the methodology for calculating sandface rates from pressure data obtained during buildup tests, eliminating the need to physically measure the rates. The techniques for calculating sandface rates and applying deconvolution using these calculated, rather than measured, rates are the key accomplishments of the research. The application of deconvolution using calculated afterflow rates permits analysis of data distorted by wellbore storage; therefore, it permits shorter, less costly tests to be used to determine reservoir properties. This may lead to more routine testing of low permeability wells, especially prior to stimulation.