Abstract
The objective of the current study is to integrate dynamic well performance data with static reservoir volumetric constraints to obtain an improved characterization of the transient drainage volume and predicted decline of a well. The specific application is to unconventional reservoirs where the time to reach boundary dominated flow is large compared to the period of production and so classical decline curve analysis is not directly applicable.
We have previously introduced the concept of the diffusive time of flight and the transient drainage volume of a well for infinite acting systems. It has been applied to field performance analysis and, for instance, used to rank well re-fracturing candidates. These same concepts may be extended from production analysis to decline prediction for a bounded reservoir through the definition of a transient well productivity. The result is a first order ordinary differential equation that may be integrated numerically to predict future production decline. This approach allows us to integrate static reservoir volume constraints (treated as an uncertainty) to obtain a probabilistic decline curve and probabilistic EUR estimates, conditioned to the well performance data. Implementation has required the development of improved field data manipulation techniques compared to our earlier studies. The steps of outlier identification, noise reduction, and regularization of the inversion for the underlying drainage volume geometry are also included in this study.
The methods are tested on approximately two dozen Eagle Ford wells, providing data driven determination of their transient drainage volumes and decline curve predictions. The decline curves are similar in character to a stretched exponential or Duong’s decline curve for early time. However, the new results differ in mid to late time as the use of a static reservoir volume constraint removes the need for an arbitrary cutoff when calculating the ultimate recovery and/or the EUR.
A novel methodology has been developed based upon the diffusive time of flight, which extends the earlier infinite acting analysis for the transient drainage volume to the transient well productivity for a bounded reservoir. This has allowed us to obtain a new class of decline curves that are driven by well performance data, while providing finite production predictions at late time, removing the need for arbitrary cutoffs.
The methodology provides a means of integrating static reservoir volumes (a geologic constraint) with well production data so that the predicted decline is not purely extrapolated from the well data but is instead a transient interpolation that honors both the production and static data.
