Online Journal for E&P Geoscientists

Jeganathan, Pradeep ^*1
(1) Exploration Technical Services Department, Saudi Aramco, Dhahran, Saudi Arabia.

In this case study we will present a series of workflows where geostatistics was particularly effective. Geostatistics consists of a number of interpolation techniques, which are used to analyze and model the spatial relationship between variables. In this paper, a description is given regarding the application of an interpolation technique called kriging with external drift (KWED). This technique is based on the assumption that there is a spatial dependency between the variables being analyzed.

Generating reliable depth maps on a regional scale typically requires merging large datasets. This can often lead to mis-ties between the merged datasets and becomes a problem during depth conversion. These mis-ties can be due to a number of reasons such as different vintages of near-surface models, different survey acquisition azimuths and parameters, and artifacts caused by different seismic data processing sequences or parameters (e.g., different stacking and migration velocities). To minimize the effects of these mis-ties, a series of workflows were developed incorporating geostatistical techniques, to condition the various data inputs from the different surveys.

We first show that a robust relationship exists between the floating datum time static corrections from two overlapping surveys in our case study. We then demonstrate how KWED was a practical tool, for creating a continuous floating datum time static correction grid, by variogram modeling of the spatial relationship between these variables.

Second, as a good relationship was found between horizon-based RMS velocities from the two overlapping seismic velocity datasets, a continuous RMS velocity grid was constructed by exploiting the spatial dependency between these two datasets using the KWED technique. The output of this workflow was then used as external drift velocity control.

Finally, a reasonable relationship was observed between the seismically derived external drift velocity control and the well derived velocities, which supported the application of KWED during the velocity modeling stage. In this workflow, an average velocity field was built at the target level by gridding the well-derived average velocities. The velocity trend away from well control was calculated by using the seismic based external drift dataset. This workflow generated an average velocity grid calibrated to wells, which was used to depth convert the target horizon.