Big seismic data: improvement of processing efficiency

Improvement of seismic data processing efficiency is often associated with economic indicators only, such as reduced time and lower cost, while maintaining quality. However, processing quality standards raise constantly, and the need for more complex tasks that focus on preserving the dynamic features of seismic records during processing is becoming more and more important. The authors discuss an integrated approach to efficiency improvement based on the development of specialized processing sequences aimed at solving specific geological problems, as well as application of new resource-intensive algorithms and efficient computing technologies. Examples of processing based on the Generalised Layered Model of subsurface, the use of kinematic and dynamic transformation aimed at time field parametrization, and tomographic algorithms designed for reconstruction of individual reservoirs taking into account anisotropy are presented. In addition, ways of applying wavefield modelling to the seismic processing sequence and the use of inverse migration transformation are discussed. Reduction in time and optimization of costs are associated with the extensive use of cloud technologies shown by the example of the Prime Cloud system, which allows the Cloud to perform a full data processing cycle, thus achieving both high performance and optimizing computational costs.