2019 AAPG Annual Convention and Exhibition:

Datapages, Inc.Print this page

Complex Geological Modeling and Fault Seal Using Unstructured Grids


Complex geological modeling and fault seal using unstructured grids

S D Harris, R K Davies, S Santoshini, S J Grenfell

Conceptual limitations of existing gridding technologies often lead to undesirable simplifications in the modeling of structurally complex areas and, consequently, poor fault seal and reservoir simulation predictions. The new approach described here enables the creation of complex and accurate 3D models and grids without any structural compromises. The unstructured grid and calculated grid properties are the basis of fault seal workflows for exploration and production scenarios; the association of the grid to a flattened space provides some benefits to both the grid and fault seal property analyses.

A volume-based structural model is constructed from fault/horizon input data that have undergone initial quality checking. The structural consistency of the input data and modeling must be validated at this early stage to enable an ‘accurate’ transformation of the structural model (under mechanical constraints) to a flattened space. The regular gridding and grid property population occur in this flattened space. The 3D unstructured grid is generated by precisely ‘cutting’ this property-populated grid by the faults and unconformities in the flattened space, and then reverse transforming the grid cells into geological space. A variety of metrics applied to the flattened/geological grid geometries provide spatial measures of the ‘quality’ of this transformation, and the ultimate validation of the structural quality of the input data.

The definition of the flattened space is a critical step in the grid creation. The tectonic consistency and preservation of geodetic distance and cell layer-orthogonality are inherent to this transformation. The flattened space is therefore ideal for a range of grid property modeling approaches and the grid is more suited to numerical simulation approximations than existing technologies. The calculation of fault seal properties takes place over the accurate unstructured grid representation of the fault geometry; the relationship between the geological and flattened-space grids can be used as the basis for the fault seal calculations once the structural consistency of the grid has been ensured.

We present several case studies to demonstrate the quality assurance and improved fault seal prediction workflows on high-quality unstructured grids. The fault seal properties and uncertainties are critical components of either an exploration prospect analysis or as the basis for a history matching approach utilizing a next generation simulator.