The comprehension of the fracture network at different scales is mandatory for understanding and developing a sub-surface fractured reservoir. The different scales of fractures are investigated using several approaches and then integrated in the final fracture distribution model.
For example the micro-fractures (from cm to meter length) are analyzed, for characteristics and distribution, from oriented cores and well logs, instead faults (macro-fractures, more than hundred meters length) are interpreted and mapped from seismic. The meso-fractures (tens to hundred meters length) are the most problematic since it is impossible to analyze them with direct tools as they are hard to be detected and described in the seismic volumes. This type of fractures, also called sub-seismic fractures, is extremely important for fractured reservoir permeability characterization.
We have developed a workflow which integrates data from analogous outcrops/cores and an automatic detection of structural features in seismic volumes enhanced by attributes as continuity or positive/negative curvature.
The seismic attribute volume is scanned and lineaments are detected and collected. The resulting dataset is statistically analyzed for length distribution (histograms), for strike (rosette diagrams and fracture sets detection) and for density (P21 and P20 maps or 3D grids).
The comparison with the fault network coming from seismic interpretation allows the detection of different fractured domains or fractures corridors related or not to the faults and to better constrain the structural evolution of the studied reservoir.
It is also possible to use these lineaments to compute probabilistic fracture surfaces that can be easily imported in a DFN (Discrete Fracture Network) model for the reservoir characterization.
The results are finally checked taking into account the stress/strain evolution through time of the studied area.
This workflow was successfully applied on the main fractured reservoir operated by eni as: Kashagan field (Kazakhstan), Perla field (Venezuela), Val d’Agri field (Italy).
The description and detection of the real fracture network in a sub surface reservoir is an extremely complex task as it isn’t possible to use a direct and unique approach. The fractures network is always related to the interaction of fractures and faults of different scales (from millimetric to kilometric scale) not uniformly distributed in the rocks volume. There are different and well known approaches to investigated fractures at each particular scale: for example the micro-fractures (from cm to meter length) are analyzed from cores, oriented or not, and well image logs, instead faults (macro-fractures, more than hundred meters length) are interpreted and mapped from seismic. Nevertheless there is a problematic scale to be investigated: the meso-fractures (tens to hundred meters length). It is impossible to analyze them with direct tools as they are hard to be detected and described in the conventional seismic volumes. This type of fractures, also called sub-seismic fractures, is extremely important for fractured reservoir permeability characterization.
The final fracture network models must be reached by the integration of all the available analysis and approaches, at the different scales, with any other useful information about geology, drilling and reservoir engineering. So integration is the only possible solution to the fracture network characterization problem.
We have developed an integrated workflow for collecting all the available data and approaches and to characterize the final fracture network. This workflow essentially focuses on the sub-seismic fracture detection and characterization. Here we present mainly this last point.