Detailed Seismic Mapping and Time-Lapse Analysis of a Fault Network in the Chalk
Donatella Astratti¹, Victor Aarre², Ole V. Vejbaek³, and Gillian White³
¹Schlumberger Stavanger Research, Stavanger, Norway
²Schlumberger Information Solutions, Stavanger, Norway
³Hess DK, Copenhagen, Denmark
Detailed mapping of a fault network from seismic data is critical for reservoir modeling for it either establishes compartmentalization or enables inference of the subseismic fracture network distribution. However, manual picking of fault surfaces is time-consuming and often inaccurate, and capturing the complete set of structural lineaments can be impossible without semi-automated techniques.
The superposition of two or more tectonic events complicates the task. In addition, the seismic data do not give a precise image of the underburden and fluids distort the signal: faults can leak and gas clouds can obscure the image of part or all of the field. Production introduces further complexity. On the other hand, changes observed in time-lapse seismic data often carry useful information on the connectivity of the fracture network.
The case we present is from a chalk reservoir and it contains all these challenges. One example is the coexistence of an initital fault and fracture system assumed to have developed during Late Cretaceous or Early Tertiary, coeval or just after the chalk deposition. It is associated to pulse growth of the structure, which caused slumping and sliding of the semilithified chalk, and to a newer set developed during the Tertiary by inversion tectonic (Evans et al. 1999) that might have partially reactivated the older system. The geologic section of interest is relatively thin on seismic sections and partially obscured by a gas cloud.
The two structural sets can have a dramatically different impact on fluid flow depending on the diagenesis the rock was subjected to from the time of development to the time of the oil charge. Consequently, their responses to stress changes resulting from current production can be dissimilar and possibly reflected on the seismic data. Reactivation cannot be excluded.
Our challenge is to separate seismic faults of different ages and characteristics for later validation by dynamic simulation and geomechanical modeling, which is not a part of this study. To reduce the uncertainties in the prediction of the effective fluid pathways, we integrate published research components and techniques still under development with common tools used by seismic data interpreters.
AAPG Search and Discovery Article #120034©2012 AAPG Hedberg Conference Fundamental Controls on Flow in Carbonates, Saint-Cyr Sur Mer, Provence, France, July 8-13, 2012