Abstract
A new data-driven method for production forecasting and estimating ultimate recovery (EUR) of tight-gas sand and unconventional-shale wells is proposed as an alternative to empirical decline-curve analysis (DCA) methods and industry standard type-well curves. The proposed method requires generating a dimensionless recovery profile (DRP) from the production history of mature wells producing from an analogous reservoir where ultimate recovery is determinable with a high degree of confidence. The daily and/or monthly production data of a newer well selected for analysis is made dimensionless by using an initial estimate of EUR and observed maximum production rate. The best dimensionless match to the DRP is obtained by adjusting the EUR and maximum rate the well could have potentially produced if unencumbered by operational constraints.
The proposed method does not utilize pressure data, so it is suitable for use with data available from public sources. Often, only 60 to 90 days of daily data is required to provide a good estimate of future production performance and EUR. One capability of the proposed method is the performance characteristics of the ratios of the secondary phases to the primary fluid phase can be simultaneously examined to better define the expected production performance and to identify wells with uncharacteristic behavior.
EnerQuest Oil & Gas, L.L.C. has successfully applied this method in its producing property acquisition and drilling evaluation work for the last 25 years and the biggest revelation is that the resulting DRP of the primary phase, for most plays, is the same regardless of a well's geographic location within the play, generation of its completion design or length of lateral. The derivation of the DRP using individual well production histories across an entire play has the advantage/benefit of a larger population of data to work with than the common industry standard method of developing a type-well curve that is specific to a much smaller population of wells sharing similar horizontal landing targets and completion designs.
The new method will be demonstrated utilizing production data from the Barnett Shale (N Texas), Austin Chalk (S Texas), Eagle Ford Shale (S Texas), Codell Sand (SE Wyoming), Woodford Shale (SE Oklahoma), and Utica Shale (Ohio).
