Genetic Classification of Natural Fractures: A Key to Understand Flow in Kashagan Field

Abstract

Fractures are one of the main features that impact reservoir quality in carbonates. They record the history of the chemical and physical processes undergone by the reservoir from deposition through burial. These processes dictate the distribution and intensity of different fracture types and their effective flow properties. Understanding fractures from a genetic perspective helps develop better conceptual models that compensate for the often skewed distribution of wells, and the lack of uniformity in the data. Three distinctive fracture generations can be distinguished in the Kashagan core. (1) During deposition processes associated with exposure and gravitational instability along the margin of the build-up result in early fracturing (Generation A). A-fractures are filled by carbonate, volcanic debris or early marine cements. (2) During burial fractures formed as a result of vertical loading and compaction (Generation B), and mostly in association with pressure dissolution surfaces (stylolites). Both types of structural features were effectively closed and unable to transmit fluids at the time of their formation. (3) During the reservoir charge elevated fluid pressure led to breaching of the top seal. A fine network of small hairline fractures (Generation C) developed in the tightly cemented rocks, while existing fractures (A and B) were enhanced and in cases reactivated. Through integration of core and image logs a set of guiding principles is defined to identify fracture generations in wellbore images away from where there is core control. Interpretation results show that syndepositional fractures are more abundant along the rim, while burial fractures are ubiquitous in platform and rim. Core-to-log integration also proves that the resistivity signal of a fracture is no guarantee for its dynamic potential. Instead losses experienced while drilling are used as a proxy for excess permeability, in the absence of well test or production data. Lost circulation events and PLTs are evaluated using observations from core and image logs to determine the causes. Results suggest that A-generation drives the large scale fracture permeability in the rim. In the platform, losses are mostly related to B-fractures. This provides a modeling approach more closely tied to geological processes and consistently integrated among data types. It can be used to distribute fractures based on the inferred spatial variability of causative processes within the reservoir over time.

AAPG Datapages/Search and Discovery Article #90259 ©2016 AAPG Annual Convention and Exhibition, Calgary, Alberta, Canada, June 19-22, 2016