DEVELOPMENT OF A VIRTUAL INTELLIGENCE TECHNIQUE FOR THE UPSTREAM OIL INDUSTRY

The objective of the research and development work reported in this document was to develop a Virtual Intelligence Technique for optimization of the Preferred Upstream Management Practices (PUMP) for the upstream oil industry. The work included the development of a software tool for identification and optimization of the most influential parameters in upstream common practices as well as geological, geophysical and reservoir engineering studies. The work was performed in cooperation with three independent producing companies--Newfield Exploration, Chesapeake Energy, and Triad Energy--operating in the Golden Trend, Oklahoma. In order to protect data confidentiality, these companies are referred to as Company One, Two, Three in a randomly selected order. These producing companies provided geological, completion, and production data on 320 wells and participated in frequent technical discussions throughout the project. Research and development work was performed by Gas Technology Institute (GTI), West Virginia University (WVU), and Intelligent Solutions Inc. (ISI). Oklahoma Independent Petroleum Association (OIPA) participated in technology transfer and data acquisition efforts. Deliverables from the project are the present final report and a user-friendly software package (Appendix D) with two distinct functions: a characterization tool that identifies the most influential parameters in the upstream operations, and an optimization tool that seeks optimization by varying a number of influential parameters and investigating the coupled effects of these variations. The electronic version of this report is also included in Appendix D. The Golden Trend data were used for the first cut optimization of completion procedures. In the subsequent step, results from soft computing runs were used as the guide for detailed geophysical and reservoir engineering studies that characterize the cause-and-effect relationships between various parameters. The general workflow and the main performing units were as follows: (1) Data acquisition. (GTI, OIPA, Participating producers.) (2) Development of the virtual intelligence software. (WVU, ISI); (3) Application of the software on the acquired data. (GTI, ISI); (4) Detailed production analysis using conventional engineering techniques and the DECICE neural network software. (GTI) and (5) Detailed seismic analysis using Inspect spectral decomposition package and Hapmson-Russell's EMERGE inversion package. (GTI) Technology transfer took place through several workshops held at offices of the participating companies, at OIPA offices, and presentations at the SPE panel on soft computing applications and at the 2003 annual meeting of Texas Independent Producers and Royalty Owners Association (TIPRO). In addition, results were exhibited at the SPE annual meeting, published in GasTips, and placed on the GTI web page. Results from the research and development work were presented to the producing companies as a list of recommended recompletion wells and the corresponding optimized operations parameters. By the end of the project, 16 of the recommendations have been implemented the majority of which resulted in increased production rates to several folds. This constituted a comprehensive field demonstration with positive results