Integrated reservoir characterization for unconventional reservoirs using seismic, microseismic and well log data

This study is aimed at an improved understanding of unconventional reservoirs which include tight reservoirs (such as shale oil and gas plays), geothermal developments, etc. We provide a framework for improved fracture zone identification and mapping of the subsurface for a geothermal system by integrating data from different sources. The proposed ideas and methods were tested primarily on data obtained from North Brawley geothermal field and the Geysers geothermal field apart from synthetic datasets which were used to test new algorithms before actual application on the real datasets. The study has resulted in novel or improved algorithms for use at specific stages of data acquisition and analysis including improved phase detection technique for passive seismic (and teleseismic) data as well as optimization of passive seismic surveys for best possible processing results. The proposed workflow makes use of novel integration methods as a means of making best use of the available geophysical data for fracture characterization. The methodology incorporates soft computing tools such as hybrid neural networks (neuro-evolutionary algorithms) as well as geostatistical simulation techniques to improve the property estimates as well as overall characterization efficacy. The basic elements of the proposed characterization workflow involves using seismic and microseismic data to characterize structural and geomechanical features within the subsurface. We use passive seismic data to model geomechanical properties which are combined with other properties evaluated from seismic and well logs to derive both qualitative and quantitative fracture zone identifiers. The study has resulted in a broad framework highlighting a new technique for utilizing geophysical data (seismic and microseismic) for unconventional reservoir characterization. It provides an opportunity to optimally develop the resources in question by incorporating data from different sources and using their temporal and spatial variability as a means to better understand the reservoir behavior. As part of this study, we have developed the following elements which are discussed in the subsequent chapters: 1. An integrated characterization framework for unconventional settings with adaptable workflows for all stages of data processing, interpretation and analysis. 2. A novel autopicking workflow for noisy passive seismic data used for improved accuracy in event picking as well as for improved velocity model building. 3. Improved passive seismic survey design optimization framework for better data collection and improved property estimation. 4. Extensive post-stack seismic attribute studies incorporating robust schemes applicable in complex reservoir settings. 5. Uncertainty quantification and analysis to better quantify property estimates over and above the qualitative interpretations made and to validate observations independently with quantified uncertainties to prevent erroneous interpretations. 6. Property mapping from microseismic data including stress and anisotropic weakness estimates for integrated reservoir characterization and analysis. 7. Integration of results (seismic, microseismic and well logs) from analysis of individual data sets for integrated interpretation using predefined integration framework and soft computing tools.