Loading...
Shallow seafloor characterization using deep-water ocean-bottom-cable data: Jubarte field, offshore Brazil
de Souza Bezerra, Moacyr
de Souza Bezerra, Moacyr
Citations
Altmetric:
Advisor
Editor
Date
Date Issued
2023
Date Submitted
Collections
Research Projects
Organizational Units
Journal Issue
Embargo Expires
Abstract
Multi-component seismic acquisitions are commonplace in the oil and gas exploration industry, seeking to better characterize hydrocarbon accumulations in the subsurface. On offshore environment, this type of acquisition is normally carried using ocean bottom cables (OBC) or ocean bottom nodes (OBN). When these acquisitions are repeated at different times over the same area, time lapse (4D) datasets are generated.
Although the techniques to process and image compressional wavefield (P-wave) data are well established, converted wave (C-wave or PS-wave) data processing remains a very challenging topic. All efforts aiming to improve C-wave data reliability and turn around time aid in the task of incorporating it into the standard workflow.
In this project, I analyze some of the issues that arise at the very beginning of the multi-component data processing flow, such as C-wave noise on vertical component, poor receiver coupling, velocity model building for the shallow part of the seafloor and elastic wavefield separation, focusing primarily on improving C-wave data quality.
I found that it is possible to improve the C-wave noise model on the vertical by using an interleaving process combine both vintages of the 4D available data, yielding improved repeatability of the datasets. I also found that poor coupling of the sensors to the seafloor generate effects that are similar to C-wave splitting, which can lead to inaccurate interpretation of subsurface anisotropic properties.
By adopting a gradient model to describe the behavior of the elastic properties (density, P- and C-wave velocities) in the shallow part of the seafloor, I apply a machine learning engine to extract those properties directly from the seismic data. With the velocity models for both P- and C-waves determined, I perform a wavefield separation and compare the results with previous processing efforts.
All this leads to a C-wave dataset that is more amenable to processes such as imaging or joint inversion, using tools that can be easily applied to any OBC or OBN data with minor modifications, helping to reduce the overall processing time and enhancing data quality.
Associated Publications
Rights
Copyright of the original work is retained by the author.