Merging chronostratigraphic modeling and global horizon tracking

Abstract

We have determined to combine the automatic interpretation of horizons in a seismic cube with a space/time framework to construct a chronostratigraphic model that matched seismic events and could be used later, without having to rework it in reservoir modeling and seismic characterization. A large number of single seismic events were automatically extracted from the cube as horizon patches. Each patch was associated to an individual isogeologic time constraint. An optimization process then proposed a geologically coherent model in the volume, filling the gaps in the area where no patch was extracted, and taking into account additional geologic information, such as unconformities, fault displacement, or well information. Furthermore, this process generated a seismic flattened volume used to check the quality of the model and revealed some geologic features such as unpicked faults, channels, lobes, and splays. We evaluated a use case in which this method was successfully tested on a complex faulted data set.

Building 3D subsurface models conforming to seismic structural and stratigraphic features

“…In this workflow, I first convert the seismic and well-log data from the original time (or depth) domain to the Wheeler domain (Wheeler, 1958), then I interpolate subsurface models following stratigraphic features in the Wheeler domain, and finally I convert the interpolated models back to the original domain. This workflow is similar to previous volume-based methods (Jayr et al, 2008;Souche et al, 2013Souche et al, , 2014Mallet, 2014;Labrunye and Carn, 2015), but it is implemented in a simpler and more efficient way in this paper. Instead of using unstructured meshes to compute the domain transformation and specific grids to interpolate subsurface rock properties, I compute the domain transformation directly on a seismic volume based on seismic structural features (folding and faulting).…”

“…Naeini and Hale (2015) provide a way to improve this method by using interpreted horizons or unconformities as additional controls to guide the interpolation. Benefiting from recent progresses of high-resolution interpretation of seismic data, Jayr et al (2008), Souche et al (2013Souche et al ( , 2014, Mallet (2014), and Labrunye and Carn (2015) propose volume-based techniques to compute subsurface models. In these methods, densely interpreted seismic horizons, faults, and unconformities are used to build subsurface models conforming to the interpreted structural and stratigraphic framework.…”

“…The uvt-transform method, proposed by Mallet (2004Mallet ( , 2014, is a general space-time mathematical framework for the domain transform. This method has been applied to remove folding and faulting (Labrunye et al, 2009;Mallet et al, 2010) and to handle manually interpreted unconformities (Mallet, 2014;Labrunye and Carn, 2015) in a seismic image. Stark (2004Stark ( , 2005aStark ( , 2005bStark ( , 2006 propose to use the phase-unwrapping method to construct an relative geologic time (RGT) volume and convert 3D seismic data to the Wheeler domain using the RGT volume.…”