Decline curves are the simplest type of model to forecast production from oil and gas reservoirs. Based on a selected decline model and observed production data, a trend is projected to predict future well performance and reserves. Despite capturing general trends, these models are not sufficient at describing the underlying physics of complex multiphase porous media flow phenomena and at explaining variations in production due to changes in operational conditions. The application of these models within a Bayesian framework is a feasible alternative to mitigate this issue and obtain more robust forecasts by considering a range of possible results. However, one important aspect that conditions the production forecasts and their uncertainty is the design of a suitable prior distribution, which can be subjective.
To address the aforementioned issue, this paper presents a workflow for the development of a localized prior distribution for new wells drilled in shale formations which combines production data from preexisting surrounding wells and geospatial data, specifically well surface and bottom coordinates. This workflow aims to establish engineering criteria to reduce the subjectivity in the design of a prior distribution, reducing and reliably quantifying the uncertainty while assuming spatial continuity of decline curve parameters. A case study of 898 gas wells in the Barnett shale is presented, and several maps are generated for analysis of important properties to be considered during field development.
As shale fields are developed in the United States, a massive amount of data is being collected by the operators, service companies and governmental agencies. Monthly production data can be easily obtained from public and commercial databases, and are essential information to assess the performance of operators across different fields. Additionally, analyzing the production of the wells via maps is helpful in the visualization of general aspects of the field geology and potential identification of sweet spots. In this context, decline curve models are a feasible choice to process the available production data, because this is the only data required by these models, they have a reduced number of parameters which can be promptly history matched and analyzed, and they provide production forecasts and estimated ultimate recoveries (EUR's).
