--> Fault Seal Analysis Using Well and Seismic Data
[First Hit]

AAPG Middle East Region, Second EAGE/AAPG Hydrocarbon Seals of the Middle East Workshop

Datapages, Inc.Print this page

Fault Seal Analysis Using Well and Seismic Data

Abstract

In areas with an active tectonic past, faulting can result in very complex structural styles of hydrocarbon traps. Moreover, faults can create either migration pathways or migration barriers to fluids. Faults create compartments, and the appraisal of their connectivity is a key issue in understanding the intrinsic structure of the reservoir. The main factors, defining the sealing properties of a fault in a clastic setting, are: (1) The stress regime (extension-compression) and the timing of (re)activation; (2) The contact between different lithologies on both sides of the fault ("juxtaposition" of sand/sand, sand/shale or shale/shale); (3) The net shale content in the actual fault zone (shale gouge ratio, "SGR”). We have developed a workflow for the evaluation of these parameters from seismic data, calibrated to well data if available. The workflow comprises four main steps, resulting in an integrated evaluation of the fault seal probability: (1) Detailed fault interpretation from seismic data supported by a comprehensive paleotectonic analysis; (2) Calculation of different types of Previous HitcoherencyNext Hit and other specific seismic attributes to map potential pathways of vertical fluid migration (chimney cube); (3) Creation of a 3D lithofacies model from well and seismic data, based on the results of simultaneous inversion, from which the juxtaposition and SGR parameters can be extracted for each fault of interest; (4) Integration and interpretation, and corroboration of the conclusions using any other available information. The analysis of the tectonic pattern starts with the mapping of the fault system, including low amplitude throws, to identify the individual compartments. For this, we primarily use special seismic attributes. Local faults are interpreted with respect to their origin and in the context of regional displacements. While doing this we adjust the theoretical stress ellipsoid to local conditions, in order to understand the prevailing regional, sub-regional and local stress fields. At this stage, we get a first qualitative idea of the possible sealing capacities of the faults and the orientation of the stresses has a direct impact on this property. The analysis includes the timing of activation and reactivation, as every movement has an inherent risk of breaching an already filled trap, leading to a reduced hydrocarbon column. To assess the leakage risk, from seismic data we create a so called "chimney cube". It is assumed that fault elements most often have an angle of dip, whereas fluid migration takes place predominantly vertically, leaving traces of vertical incoherencies (chimneys). The full stack data set is transformed into tens of Previous HitattributeTop cubes, emphasizing these vertical incoherencies. A supervised neural network is trained on samples using these attributes to discriminate potential chimney cubes from non-chimney areas. The NN is then applied to the whole data set, creating a chimney probability which can be analyzed in 3D and compared to the location(s) of the fault of interest, giving an indication of possible leakage at this fault. Very often, the identified chimneys culminate in shallow gas accumulations or seabed pockmarks, which may represent near surface drilling hazards.