Estimation Of Reservoir Properties From Seismic Data By Smooth Neural Networks
Traditional joint inversion methods reqnire an a priori prescribed operator that links the reservoir properties to the observed seismic response. The methods also rely on a linearized approach to the solution that makes them heavily dependent on the selection of the starting model. Neural networks...
| Main Authors: | , , |
|---|---|
| Other Authors: | |
| Format: | Technical Report |
| Published: |
Massachusetts Institute of Technology. Earth Resources Laboratory
2012
|
| Online Access: | http://hdl.handle.net/1721.1/75457 |
| _version_ | 1826194581370175488 |
|---|---|
| author | Saggaf, Muhammad M. Toksoz, M. Nafi Mustafa, Husam M. |
| author2 | Massachusetts Institute of Technology. Earth Resources Laboratory |
| author_facet | Massachusetts Institute of Technology. Earth Resources Laboratory Saggaf, Muhammad M. Toksoz, M. Nafi Mustafa, Husam M. |
| author_sort | Saggaf, Muhammad M. |
| collection | MIT |
| description | Traditional joint inversion methods reqnire an a priori prescribed operator that links the reservoir properties to the observed seismic response. The methods also rely on a linearized approach to the solution that makes them heavily dependent on the selection of
the starting model. Neural networks provide a useful alternative that is inherently nonlinear and completely data-driven, but the performance of traditional back-propagation
networks in production settings has been inconsistent due to the extensive parameter
tweaking needed to achieve satisfactory results and to avoid overfitting the data. In
addition, the accuracy of these traditional networks is sensitive to network parameters,
such as the network size and training length. We present an approach to estimate the
point-values of the reservoir rock properties (such as porosity) from seismic and well
log data through the use of regularized back propagation and radial basis networks.
Both types of networks have inherent smoothness characteristics that alleviate the nonmonotonous generalization problem associated with traditional networks and help to
avert overfitting the data. The approach we present therefore avoids the drawbacks of
both the joint inversion methods and traditional back-propagation networks. Specifically,
it is inherently nonlinear, requires no a priori operator or initial model, and is not
prone to overfitting problems, thus requiring no extensive parameter experimentation. |
| first_indexed | 2024-09-23T09:58:28Z |
| format | Technical Report |
| id | mit-1721.1/75457 |
| institution | Massachusetts Institute of Technology |
| last_indexed | 2024-09-23T09:58:28Z |
| publishDate | 2012 |
| publisher | Massachusetts Institute of Technology. Earth Resources Laboratory |
| record_format | dspace |
| spelling | mit-1721.1/754572019-04-12T20:32:24Z Estimation Of Reservoir Properties From Seismic Data By Smooth Neural Networks Saggaf, Muhammad M. Toksoz, M. Nafi Mustafa, Husam M. Massachusetts Institute of Technology. Earth Resources Laboratory Saggaf, Muhammad M. Toksoz, M. Nafi Traditional joint inversion methods reqnire an a priori prescribed operator that links the reservoir properties to the observed seismic response. The methods also rely on a linearized approach to the solution that makes them heavily dependent on the selection of the starting model. Neural networks provide a useful alternative that is inherently nonlinear and completely data-driven, but the performance of traditional back-propagation networks in production settings has been inconsistent due to the extensive parameter tweaking needed to achieve satisfactory results and to avoid overfitting the data. In addition, the accuracy of these traditional networks is sensitive to network parameters, such as the network size and training length. We present an approach to estimate the point-values of the reservoir rock properties (such as porosity) from seismic and well log data through the use of regularized back propagation and radial basis networks. Both types of networks have inherent smoothness characteristics that alleviate the nonmonotonous generalization problem associated with traditional networks and help to avert overfitting the data. The approach we present therefore avoids the drawbacks of both the joint inversion methods and traditional back-propagation networks. Specifically, it is inherently nonlinear, requires no a priori operator or initial model, and is not prone to overfitting problems, thus requiring no extensive parameter experimentation. Massachusetts Institute of Technology. Borehole Acoustics and Logging Consortium Massachusetts Institute of Technology. Earth Resources Laboratory. Reservoir Delineation Consortium Saudi Aramco 2012-12-13T17:22:27Z 2012-12-13T17:22:27Z 2000 Technical Report http://hdl.handle.net/1721.1/75457 Earth Resources Laboratory Industry Consortia Annual Report;2000-02 application/pdf Massachusetts Institute of Technology. Earth Resources Laboratory |
| spellingShingle | Saggaf, Muhammad M. Toksoz, M. Nafi Mustafa, Husam M. Estimation Of Reservoir Properties From Seismic Data By Smooth Neural Networks |
| title | Estimation Of Reservoir Properties From Seismic Data By Smooth Neural Networks |
| title_full | Estimation Of Reservoir Properties From Seismic Data By Smooth Neural Networks |
| title_fullStr | Estimation Of Reservoir Properties From Seismic Data By Smooth Neural Networks |
| title_full_unstemmed | Estimation Of Reservoir Properties From Seismic Data By Smooth Neural Networks |
| title_short | Estimation Of Reservoir Properties From Seismic Data By Smooth Neural Networks |
| title_sort | estimation of reservoir properties from seismic data by smooth neural networks |
| url | http://hdl.handle.net/1721.1/75457 |
| work_keys_str_mv | AT saggafmuhammadm estimationofreservoirpropertiesfromseismicdatabysmoothneuralnetworks AT toksozmnafi estimationofreservoirpropertiesfromseismicdatabysmoothneuralnetworks AT mustafahusamm estimationofreservoirpropertiesfromseismicdatabysmoothneuralnetworks |