Utilizing Well Interference Data in Unconventional Gas Reservoirs

Abstract

This paper describes the use of a dual porosity radial gas flow simulator for estimating values for the key parameters controlling gas production from the Devonian shale or similar unconventional resource. Using best estimates from the literature on shale fracture and matrix properties as a starting point, the model is used to history match production decline and cumulative production data from a Devonian shale well in Meigs County, Ohio. Three sets of parameters were obtained which gave good agreement with the production data. Utilizing simulated pressures at a production data. Utilizing simulated pressures at a distance of 96 feet (29.3 m) as a third matching parameter, it was possible to distinguish between the parameter, it was possible to distinguish between the three cases and thus obtain a "more unique" match. From the results obtained it is concluded that interference data can provide additional information needed for determining the key parameters that control shale well deliverability. Introduction Devonian shale wells have produced in the Appalachian Basin for 50 to 100 years with very low but sustained productivity. In recent years there has been a large effort by government and private industry to develop and implement improved techniques for recovering a greater percentage of gas from this important resource. A part of this effort involves the development of mathematical models with predictive capability so that various alternative production schemes can be investigated and accurate forecasts of additional recovery can be made. Most Devonian shale reservoirs are believed to consist of a very tight, low porosity shale matrix with high storage capacity and low flow capacity coupled with a relatively high flow capacity natural fracture system of very low storage capacity. As pointed out by Smith et al the weight of all pointed out by Smith et al the weight of all available evidence favors the position that the shale matrix provides the major contribution of the recoverable gas. The Kucuk-Sawyer model was used in this study in an effort to approximate the key parameters controlling gas production from a particular shale well in Meigs County, Ohio. The model was used in the history mode so that observed and simulated data were plotted for immediate comparison. Parameters were plotted for immediate comparison. Parameters were adjusted manually until good agreement between observed and simulated production decline and cumulative production curves was obtained. Good matches were obtained for three sets of the following parameters:fracture porosity,fracture permeability,fracture spacing,matrix porosity, andmatrix permeability. In an attempt porosity, and (5) matrix permeability. In an attempt to obtain "more uniqueness" the simulated pressures at a distance of 96 feet (29.3 m) were compared for three "matches". It was found that, for each of the three sets, pressures were significantly different. Thus the possibility of utilizing interference data from offset wells to delineate storage and flow properties of the Devonian shale is inferred. Details properties of the Devonian shale is inferred. Details of the matching procedure and the sensitivity to the various parameters are presented. Background Warren and Root and Kazemi have presented models and techniques for analysis of naturally fractured reservoirs for the flow of compressible liquids. Kazemi et al also studied the pressure behavior of an observation well in a naturally fractured reservoir with an adjacent well produced at a constant rate. It was found that the early time response was substantially different from that of an equivalent homogeneous reservoir. Crawford et al analyzed more than twenty field-measured pressure buildup curves in a reservoir know to be naturally fractured and concluded that the Warren and Root model adequately described the buildup response and is therefore useful in determining effective (fracture system) permeability. In 1976 DeSwaan presented an analytical model for a naturally fractured reservoir which does not involve the Warren and Root parameters. P. 137

A Quantitative Approach to Analyze Fracture Area Loss in Shale Gas Reservoirs

“…Meehan et al (1989) extended the analysis to multiple wells including finite conductivity hydraulic fractures in their interference analysis. Some historical work has even been performed on shale interference tests such as Frohne and Mercer (1984) and Sawyer et al (1980). Frohne and Mercer's (1984) work did not include any hydraulic fractures; the shale wells were vertical wells that intersected large amounts of natural fractures resulting in a reservoir similar to the Warren and Root model.…”