Summary. The ability to predict incremental rates and reserves reasonably from infill development drilling in tight-gas reservoirs is enhanced through use of a three-dimensional (3D), multiwell, dry-gas model. Single-well [two-dimensional (2D)] and multiwell [three-dimensional (3D)] input and results are compared, and examples of both are presented. Measured initial reservoir pressures on an eight-well infill program compared favorably with the 3D model predictions. Introduction Upon the Federal Energy Regulatory Commission's establishment of the Sec. 107 incentive pricing, development of the Cotton Valley tight-gas-sand resource base expanded significantly, with more than 1,100 wells drilled during 1977–82. The soft gas market, coinciding with gas deregulation, slowed further development between 1983 and the present (Fig. 1). As of Oct. 1985, there were approximately 1,230 Cotton Valley sand producers on Texas Railroad Commission records for the area encompassing Harrison, Panola, and Rusk counties. Amoco Production Co. has drilled or participated in the drilling of more than 170 of these tight-gas-sand wells, situated primarily in the Blocker, Carthage, Dirgin, Henderson North, Tatum, and Woodlawn areas of the Cotton Valley field (Fig. 2). Most of this development has been on 640-acre [260-ha] density. The Cotton Valley (Jurassic) sandstone of east Texas is a series of marine and lagoonal deposits. Diagenesis in the form of calcite cementation and quartz overgrowth, combined with overburden pressure, has reduced its porosity and permeability. With permeabilities in the microdarcy range, massive hydraulic fracture (MHF) stimulations are usually required to make a commercial completion. Gas production from the east Texas Cotton Valley sands has been at depths ranging from 9,000 to 10,500 ft [2700 to 3200 m]. The gross thickness of the Cotton Valley sand/shale sequence averages 1,500 ft [460 m]. This paper discusses modeling associated with only the lowermost Yellow, paper discusses modeling associated with only the lowermost Yellow, or Taylor, zone as shown in Fig. 3, a type log from the Blocker Cotton Valley field. The model work presented here was undertaken to determine the incremental rates and reserves associated with infill drilling of existing units. Two types of reservoir models were used. The first, a 2D model, contains a single well located in the center of a rectangular, homogeneous drainage area. This model is appropriate for minimum-well-density situations. This single-well model, however, fails to account for interference from other wells, which was suspected after early infill drilling (160 to 320 acres/well [65 to 130 ha/well]) was analyzed. The second model has a 3D capability that allows areal and vertical variations in reservoir properties. Most important, it can model several wells at once, thereby providing the opportunity to determine realistic estimates of incremental production associated with new well drilling. The 3D model results were initially validated through the measurement of pressures on eight infill wells drilled during 1985. Discussion The following discussion details the basic data requirements for the model work, describes both the single-well (2D) and multiwell (3D) models, and provides an example analysis of an infill well location using both models. The single-well model was used to provide historical performance matches of existing wells for later use in the multiwell model. The single-well model can also be used to predict reserves of a proposed infill well. The modeling comparison predict reserves of a proposed infill well. The modeling comparison section of this discussion illustrates the benefits of using the multiwell model vs. the single-well model for the determination of infill-well reserves. Gas Properties. Separator tests from wells in each field area were used to estimate an average Yellow zone well-stream composition. Gas viscosity was estimated by Thodos-type correlations. Other gas properties were determined with the modified Redlich-Kwong equation properties were determined with the modified Redlich-Kwong equation of state. These field average properties were used in all pressure-transient and modeling efforts. pressure-transient and modeling efforts. Initial Reservoir Pressure. Estimates of initial reservoir pressure in the various field areas were determined from early well-pressure bomb measurements (greater than 140-hour shut-ins). The average pressure gradient in this Cotton Valley area equaled 0.55 psi/ft [12.4 kPa/m] (15 wells). Previous literature has documented the overpressured nature of the Yellow zone as compared with the uphole Cotton Valley sand intervals. Porosity, Water Saturation, Net Pay. Yellow zone volumetric gas Porosity, Water Saturation, Net Pay. Yellow zone volumetric gas in place was determined through detailed well log and core analyses. Some of the techniques used to estimate net pay in the Cotton Valley have been presented previously. Field average porosities and water saturations for the pay intervals ranged from 5.7 to 7.9% and from 27 to 49%, respectively. Formation Flow Capacity. Documented tight-gas-sand prefracture pressure-buildup techniques were used to analyze pressure-buildup pressure-buildup techniques were used to analyze pressure-buildup data from more than 100 Yellow zone producers. Average field values of formation flow capacity ranged from 0.39 to 0.92 md-ft [1.9 × 10(-1) to 2.8 × 10(-1) mdm]. Real-gas pseudotime and pseudopressure were used in conjunction with the equivalent-time function for the analysis. Refs. 5 and 6 detail the benefits associated with using these modified time and pressure functions in tight-gas reservoirs. Fracture Flow Capacity and Half-Length. Postfracture analysis of all available test data was attempted. Two techniques were used:the constant-rate andthe constant-pressure type curves developed by Agarwal et al. for use in finite-flow-capacity fractures. With these techniques, values for fracture half-length, Lf, and dimensionless fracture capacity, FCD, were estimated. A third method, the scalar technique discussed by Tison et al. was also tested and found to provide some assistance in generating estimates for Lf. JPT P. 881

Summary Production was dramatically improved by cavitating (i.e., intentionally producing sand) an offshore gas well on the Gulf of Mexico shelf. The cavity completion was conducted over a period of 10 days and involved producing the well with temporary equipment to safely produce, collect, and dispose of the produced sand. No hydrocarbon emissions resulted because production could be routed back through the permanent facilities. Following the operations, the well was returned to flow through the permanent facilities at a much higher producing rate, with much lower sand production. Before the cavity-completion operations, the well had been choked back to a low gas rate (<1 MMscf/D) because of sand production. The low gas rate was insufficient to continuously lift liquids, causing the well to become liquid loaded and/or slug. The well was evaluated for a sand-control completion, but this type of repair was believed to entail too much economic risk because of significant reservoir unknowns. Further evaluation suggested that the well had the attributes for successful cavity completion. This paper discusses the cavity-completion evaluation process employed and the operational results and learnings; a similar process may be applicable to other wells. Introduction A cavity-like completion has been used in other areas to attempt to reduce the amount of solids produced and improve the productivity of wells.1 For this offshore-well application, a cavity-like completion was believed to provide a viable option once it was determined that the well produced too many solids as originally completed. A conventional repair to install a sand-control completion was evaluated but was believed to entail too much economic risk. While there were significant reservoir unknowns, a cavity-type completion could be attempted at a relatively low cost. Following the cavity completion, the well was returned to production at a higher rate (four-to five-fold increase), with minor amounts of sand production, until the reservoir watered out. A cavity-like completion involves purposely producing sand with the intention of creating an enlarged void or enhanced-permeability area around the wellbore. Sand production can entail risks (discussed below), so this method may not be appropriate in other situations.

TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractProduction was dramatically improved by cavitating (intentionally producing sand) an offshore gas well on the Gulf of Mexico Shelf. The cavity completion was conducted over a period of 10 days and involved producing the well through temporary equipment in order to safely produce, collect, and dispose of the produced sand. No hydrocarbon emissions resulted since production could be routed back through the permanent facilities. Following the operations, the well was returned to flow through the permanent facilities at a much higher producing rate, with much lower sand production.Prior to the cavity completion operations, the well had been choked back to a low gas rate, <1 MMscfd, due to sand production. The low gas rate was insufficient to continuously lift liquids, causing the well to become liquid loaded and/ or slug. The well was evaluated for a sand control completion, but this type of repair was felt to entail too much economic risk due to significant reservoir unknowns. Further evaluation suggested that the well had the attributes for a successful cavity completion. This paper will discuss the cavity completion evaluation process employed and operational results and learnings. A similar process may be applicable to other wells.