Seismic attributes have been widely used in seismic object detection. While no unique attribute is expected to perfectly identify the targeted object, various attributes contributing to the same purpose should be utilized simultaneously when performing detection. Artificial neural network has been successfully applied in seismic object detection by combining multiple attributes into a single object-sensitive attribute. An optimized neural network fault detection approach was introduced by a case study from Stratton field, south Texas, United States. Results indicate that the new produced fault probability attribute could suppress the surrounding noises and highlight the faults. Application of ant-tracking to the fault-cube shows more convincible results than to individual attributes. In addition, several potential cross-strike fractures increase the structural complexity. The neural network-based fault detection contributes to better structural interpretation and is of great significance to hydrocarbon exploration, especially in area with complex fault system.

The URTeC Technical Program Committee accepted this presentation on the basis of information contained in an abstract submitted by the author(s). The contents of this paper have not been reviewed by URTeC and URTeC does not warrant the accuracy, reliability, or timeliness of any information herein. All information is the responsibility of, and, is subject to corrections by the author(s). Any person or entity that relies on any information obtained from this paper does so at their own risk. The information herein does not necessarily reflect any position of URTeC. Any reproduction, distribution, or storage of any part of this paper without the written consent of URTeC is prohibited.

Naturally fractured reservoirs such as the Marcellus shale require an integrated reservoir modeling approach to determine well spacing and well-to-well interference. The Marcellus Shale Energy and Environment Laboratory (MSEEL) is a joint project between universities, companies, and government to develop and test new completion technologies and acquire a robust understanding of the Marcellus shale. The study presented in this paper aims to reveal an approach to determine reservoir depletion with time through coupled geological modeling and geomechanical evaluation followed by completion and well performance history matching for a multiwell pad in the Marcellus shale. The geomechanical model was prepared with interpreted vertical log data. A discrete natural fracture (DFN) model was created and used to determine the complexity of hydraulic fracture geometry simulated through complex fracture models on a two well pad. The microseismic data obtained during the hydraulic fracture simulations served as a constraining parameter for the hydraulic fracture footprint in these wells. Sensitivity to the DFN is realized by parametric variations of DFN properties to achieve a calibrated fracture geometry. Reservoir simulation and history matching the well production data confirmed the subsurface production response to the hydraulic fractures. Well spacing sensitivity was done to reveal the optimum distance that the wells need to be spaced to maximize recovery and number of wells per section. Hydraulic fracture geometry was found to be a result of the calibration parameters, such as horizontal stress anisotropy, fracturing fluid leakoff, and the DFN. The availability of microseismic data and production history matching through integrated numerical simulation are therefore critical elements to bring unique representation of the subsurface reaction to the injected fracturing fluid. This approach can therefore be consistently applied to evaluate well spacing and interference in time for the subsequent wells completed in the Marcellus. With the current completion design and pumping treatments, the optimal well spacing of 990 ft was determined between the wells in this study. However, wells to be completed in the future needs to be modeled due to the heterogeneity in the reservoir properties to ensure that wells are not either underspaced to cause well production interference or overspaced to create upswept hydrocarbon reserves in the formation. By adopting the key learnings and approach followed in this paper, operators can maximize subsurface understanding and will be able to place their wellbore in a nongeometric pattern based on reservoir heterogeneity to optimize well spacing and improve recovery.

No abstract

No abstract