Abstract Several gas basins within North America are under-going a period of infill drilling in order to supply the increasing demand for natural gas. It is not uncommon for an operator to have hundreds, if not thousands of infill candidates from which to choose. Is there a practical way to high-grade infill locations within a particular basin or field? This paper presents several practical techniques to high-grade infill drilling opportunities when faced with large data sets and little time. We will present results for the Mesaverde in the San Juan Basin, the Morrow in the Permian Basin, and the Cotton Valley in east Texas. We calculated the infill potential for each of the three example formations. Based on our experience, we have found that comparisons between actual and predicted individual infill well performance can vary significantly. Therefore, ranking infill candidates on predicted individual well performance may not necessarily yield the best overall results. Poor wells, predicted to be good wells, may still be drilled and good wells, predicted to be poor wells, may not be drilled at all. Alternatively, when considering an infill drilling program as a whole, we have found that the predicted performance for the group can be quite accurate. In this paper, we will present a method to divide a basin or field into smaller areas and predict the distributions of infill performance as a group for the smaller areas, rather than individual wells. We will then present two methods to rank the small areas to achieve the most economic results. The first method ranks the areas based on the average of the distribution. The second method uses decision tree analysis to calculate a typical infill well in an area using the P10, P50, and P90 values from the distribution. We considered actual recent drilling programs in each of the three formations, used the above techniques to eliminate predicted uneconomic wells, and compared the actual to predicted drilling program results. We found that the finding and development costs were lower, return on investment was higher, and the success ratio was higher when the techniques presented in this paper were used to high-grade drilling opportunities. The results of a study applying this analysis technique can be used when budgeting and planning near- and long-term drilling programs.

Abstract This paper describes the application of a practical technique to determine infill potential when faced with little time, large data sets, and complex geology. Using this technique, we determined where newer wells are encountering potentially depleted reservoir and the infill potential for the Milk River formation within a 900-well, 200,000-acre area in the Western Canada Sedimentary Basin. We obtained these results in a minimal amount of time and used only monthly production and wellbore location data. We validated our technique by "history matching" the production performance of recently drilled wells. We correlated well quality with historical well densities in order to predict the infill well potential from 160-acre spacing to an 80-acre well spacing. We estimated ultimate recoveries for all existing wells and infill candidates and show their reserve distributions. We identified 896 infill candidates with 8.9 × 109 m3 of gas reserves. The results of this study are presented in this paper using tables, graphs, and maps. The results of a study applying this analysis technique can be used when budgeting and planning near-and long-term drilling programs. The analysis techniques described in this paper could be applied by operators in other areas and reservoirs to evaluate their own acreage position or infill drilling potential.

ABSTRACT This paper presents the application of a new analytical simulator specifically developed for multilayer reservoirs. In this paper, we demonstrate how this analytical simulator can be used, the same way as a conventional numerical simulator, to history-match field pressure data, production data, and forecast reservoir performance. Two field cases are presented in which analytical simulation results and numerical simulation results are compared.

ABSTRACT This paper summarizes the detailed analysis of geological, geophysical, geochemical, and well test data from three Devonian Shale and Berea sand wells located in Pike County, Kentucky. These three wells were part of a Gas Research Institute (GRI) project designed to gain a better understanding of the producing mechanisms of the Shales and Berea and the impact of stimulation on well performance.1 All three wells were completed in both the Shales and Berea, which is typical for the area. Detailed data were collected in both the Shales and the Berea on two of the three wells which were used to develop layered, naturally fractured reservoir descriptions for each producing horizon. Each formation was stimulated with nitrogen foam and sand fracture treatments, and in general, the pre- and post-stimulation well performance was similar to predictions made using the layered, naturally-fractured reservoir descriptions. This paper compares these reservoir descriptions and provides a technical basis for the widely varying production often observed with Shale and Berea wells in this region.

;Contents: Naturally fractured reservoir description; Geologic considerations; Shale-specific log model; Stress profiles; Berea reasearch; Benefits analysis; Summary of technologies; Novel well test methods; Natural fracture identification; Reverse drilling; Production data analysis; Fracture treatment quality control; Novel core analysis methods; and Shale well cleanouts.

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 This paper presents a new test procedure designed to measure individual layer reservoir properties (permeability, skin factor, and reservoir pressure) in a single wellbore and in a reasonable amount of time for low permeability gas wells. The method involves (1) isolating a small interval with packers, either in an open- or cased-hole, (2) injecting nitrogen into the interval at a rate such that the bottomhole pressure builds up to a pre-determined value (below fracturing pressure), and (3) shutting in the well, preferably downhole, and monitoring the pressure falloff. The pressure falloff data is then analyzed to determine reservoir properties. Essentially, this method can be described as a nitrogen slug test and can be applied to any formation where single-phase gas is produced. We have found that average reservoir properties determined with conventional analyses of a single pressure buildup test conducted over a large interval can lead to optimistic predictions of the future performance of a well producing from a layered reservoir system. Therefore, being able to quantify individual layer properties can lead to better reservoir characterization, modified completions, and improved performance projections. Existing techniques, such as multiple pressure buildup tests, are inadequate for measuring reservoir permeability in several intervals in a single wellbore because, often in low productivity gas wells, either a pre-stimulation rate is not attained or the time required to reach the correct semilog straight line for a conventional well test analysis is too long. Knowledge of layered reservoir properties impacts completion decisions (where to perforate), fracture design considerations, and post-fracture analyses.

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.

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.