Tesla
Printer Friendly
 
 
Oil and Gas
SeismicCity-logo
 

Challenge


Seismic illumination and RTM PSDM.  Regional Gulf of Mexico model.
Seismic illumination and RTM PSDM. Regional Gulf of Mexico model.

In the current environment of low oil prices, oil companies are seeking ways to reduce exploration and development costs. The highest cost in exploration for new oil and gas reservoirs and development of discovered fields is in well drilling. Having better and more accurate seismic data reduces the cost of well drilling, resulting with a lower cost in producing oil and gas. This will enable producers to operate in the current market.

A key recent development in seismic imaging that is helping the industry to obtain more accurate seismic data is the application of depth imaging based on ‘Reverse Time Migration’ prestack depth migration (RTM PSDM) algorithms. This unique algorithm was introduced to the industry more than 30 years ago but became operational only in the past several years due to the recent developments in computer hardware technology.

RTM PSDM has been proven to provide the seismic data quality that the oil and gas industry desires. However, due to computational costs, it still proved too costly to use the full frequency of the seismic data when using the algorithm. This led to depth-migrated seismic data that did not present the optimal fidelity of the seismic data which, in turn, resulted in less accurate interpretations of the data.

In order to provide a solution to the industry, SeismicCity has developed a new mathematical formulation enabling high frequency application of RTM PSDM in an industrial setting.

 

Solution


Transverse Anisotropy (TTI) model and RTM PSDM.  The exploration objectives are the sand layers around the salt and the 4-way structure beneath the salt body.
Transverse Anisotropy (TTI) model and RTM PSDM. The exploration objectives are the sand layers around the salt and the 4-way structure beneath the salt body.

The computational cost of RTM PSDM is very high. In order to be able to migrate a broader frequency range, a small cell size needs to be used as the numerical grid. This results in large computer memory requirements as well as an enormous number of floating point operations. These two limitations very often lead to the use of a band limited selection of the input seismic data resulting in low frequency RTM PSDM. The motivation to overcome these limitations in the implementation of RTM PSDM led SeismicCity and NVIDIA to work together to achieve two goals.

The first goal is to be able to migrate the full frequency range without creating significant numerical dispersion of the propagated wavefield. A solution to that was accomplished by using newly developed recursive operators numerical solution that are able to accurately compute spatial derivatives without the need to use a very small operational grid.

The second goal in the implementation of the new generation of RTM PSDM is to adapt the software to a specific hardware configuration capable of performing a very high number of floating point operations in a short time. While searching for the right hardware platform, SeismicCity found the new K80 GPU to be the right solution for the newly developed algorithm.

SeismicCity’s research and development staff collaborated with NVIDIA experts to adopt and optimize the new algorithm for a CPU / GPU hardware where the computational kernel is done on the K80 GPU. The high throughput that the K80 provides enabled SeismicCity to offer this new generation of RTM PSDM to its clients, providing images of the subsurface more accurate than ever before possible.

For more information, visit:  http://www.seismiccity.com/RTM.html




 
 
LinkedInTwitterGoogle+FacebookReddit