Interpolating high-resolution well log volume using seismic dip vectors and control grids
Seismic-guided volumetric well log interpolation is a unified solution for inversion, interpolation, and integration problems. For inversion, it uses both the direct measurement from the borehole and the indirect measurement from seismic data to derive the parameters of geological formations; for interpolation, it employs the sparsely distributed well logs to interpolate a full 3D volume of well logs; for integration, it uses a combination of the information from boreholes, structural models (horizons, grids, faults, etc.), and/or seismic surveys.
There are multiple documented approaches to generate a volume of pseudo well logs including geo-statistical, image-guided, and neural-network-based. In this paper written for and presented in San Antonio, Texas at URTec 2016, we present a unique way of interpolating high-resolution well log volumes which combines the model-driven and data-driven approaches. The model-driven approach is interactive and allows input from human interpreters. The data-driven approach enables interpolation following the local geological structural patterns from the seismic dip vectors. In the experiments, we demonstrated that the approach can be successfully applied to both synthetic and real-world datasets with models of control grids, and/or seismic dips.
Read the full paper, authored by Yingwei Yu, as presented at URTeC.
Hank Chambers is Director Product Management Geological Software at IHS Markit
Posted 29 September 2016
This article was published by S&P Global Commodity Insights and not by S&P Global Ratings, which is a separately managed division of S&P Global.