Since fault permeability is principally controlled by the internal structure of the fault, it is necessary to improve our understanding of its governing processes. For each reservoir it is possible to define a tens of factors that may influence the fault zone structure. However, field observations suggest that only a small number of the many "conceptual" factors have a critical influence on fault structure and corresponding permeability, and it may be possible to define a limited number of "fault zones types" and describe their associated dynamic properties. In every case, lithology is a fundamentally important parameter.

We propose to establish a database of fault zone characteristics for key reservoir lithological sequences.

We will concentrate on outcrop scale descriptions because this is the scale at which the internal and external characteristics of the fault zone can be described in detail, the rheology is well constrained and the mechanisms of the fault and associated damage zone formation can be investigated.

Integration of outcrops into their regional context is essential since structures observed at outcrop scale may depend on larger structures.
 

Normal fault in Triassic sandstones, Southern High Atlas, Morrocco

Fracture corridors are zones of fracture clustering dominated by highly persistent parallel joints. Fracture corridors can be easily distinguished from other types of joint sets and are often well marked because of their effect on morphology (localisation of small valleys, bays along coastlines).

It is well known that in fractured reservoirs fluids arrival may occur in a spatially discontinuous and yet systematic manner with spacing ranging from several meters to several tens of meters. Fractured corridors are thought to promote fluids arrival, and an ability to predict the existence and properties of fractured corridors is clearly of benefit in the assessment of fractured reservoirs.

We aim to define the geometrical properties of fracture corridors from outcrop to kilometric scale, and derive a classification of fracture corridors as a function of lithology and major structures. The mechanical interpretation will be mainly based on field data analysis. Mechanical modelling dealing with physical conditions of mode I fracture clustering in bounded multi-layered analogs will also be performed.
 
 

Fracture corridor in the Old Red Sandstones at Clardon Head (Scotland U.K.)

Fracture corridors in the Aalenian limestones, Mende area, Southern France.

Fault permeability is usually considered to depend on the structure and content of the fault zone. One common application of this idea is the determination of the "Shale Gouge Ratio" for the evaluation of fault permeability in sandstone reservoirs. This ratio is computed using the clay content of the host rock and the throw of the fault, and is believed to be a reliable indicator of fault permeability. But, as is often the case, the results are not always as reliable as one would hope.

Part of the inaccuracy in the shale gouge ratio is due to the potential presence of drag-folds adjacent to a fault, leading to an overestimation of the faults throw. In order to improve the prediction of fault permeability, we can incorporate various models of drag-folding when calculating the shale gouge ratio. In particular, an improved understanding of the mechanical factors controlling drag-fold development is necessary. This would also be helpful to improve the prediction of sub-seismic faults from dipmeter log analysis.
We plan to address the problem of drag-folding by using field analysis and new analogue experiments in multi-layered media with the aim of improving drag fold and hence fault permeability prediction and characterisation.
 

Curvature analysis is a tool now routinely used for predicting the location of the areas with the highest density of potentially productive fractures. Despite some very remarkable results, it is often an inaccurate approach that achieves varying degrees of success in predicting relative fracture densities. We think that part of this inaccuracy arises because the curvature analysis is usually performed on an interpretation of the present day geometry of the reservoir, ignoring any previous curvature history.

The conditions under which fracture apertures can be permanently maintained in the same zones of maximum curvature during the entire evolution of the trap have to be carefully examined. It is important to study these conditions using field data interpreted in conjunction with results from specifically tailored experiments in multi-layered materials submitted to cylindrical and non-cylindrical folding.
 

Fractured anticline in the Devonian sandstones of Southern Anti Atlas (Morocco)

Complexity of joint pattern  in the limb of a non cylindrical fold (Lavernock Point, Wales, U.K.)

Study of sub-seismic features can be done by field analysis of high quality outcrops analysed in their structural and stress history context and by comparison with 3D attribute maps in collaboration with companies.

The organisation and internal structure of sub-seismic structures may vary because of lithology, geological history, and many other factors with major implications for permeability. The classification of sub-seismic fault types in combination with physical modelling will allow to define key parameters and to establish the tectono-mechanical rules controlling their genesis.
 
 

A truly multi-disciplinary approach in which field data collection and innovative modelling will provide tectono-mechanical rules as guidelines for describing the geometrical and hydrological behaviour of the sub-seismic array. This approach will also help to improve the geometrical/hydrodynamical fracture system models, which must rely on realistic physical basis or at least on mechanically reasonable concepts taking real world constraints into account.
 

Image of plastic shear bands in a PVC plate branched on an open oblique defect under uniaxial compression. Possible model for secondary fault propagation.

View of a fracture surface which propagated under mixed mode conditions in a transparent organic material. The fracture propagated from the oblique, pre-existing, defect seen as the dark region on the right and terminiated at the irregular crack front on the left.

Incipient fault illustrating the effect of contrasting lithologies on fault characteristics (Lodève Basin, South France.

Fractures and other structural features commonly observed in wells do not have a large observed extent. Hence, extrapolation of this structural data to the reservorir scale remains a major challenge. One sensible approach to this upscaling problem is to find relationships between the limited structural data and other types of data, the characteristics of which are better known at the reservoir scale. The extrapolation of the structural data can then be performed using the guides provided by the intermediary data.  Using this approach, we think that investigating the relationships between lithology and microstructures would make it possible to predict small scale structures for the following reasons:    microstructures depend on both rheology and deformation  each lithology is likely to have a different rheology, and therefore to develop specific "microstructural facies" in response to the local structural context  the distribution of the different lithofacies is generally well known in 3D at the reservoir scale.  Therefore, we want to establish a database on the relationships that exist between structures and lithology by studying outcrops and cores featuring a great range of structural and sedimentological settings and by trying to find the major factors controlling the development of the principal "microstructural facies".

Traditional structural analysis of core attempts to list, as accurately as possible, every fracture on the core and its geometry, but often with no connection to the other core analyses performed. As a result, the structural data is often difficult to extrapolate from the borehole to the reservoir scale, and therefore difficult to integrate into the reservoir model.

We propose a new approach to the core and borehole data analysis, adapted to reservoir modelling and focusing on:

• understanding the organisation of the structures observed in cores in both the structural and stratigraphic frameworks. This will provide constraints to their extrapolation from the well to the reservoir. Analysis of the relationships between lithology and microstructural facies has to be carefully addressed with that aim.
• estimating the potential influence of the observed microstructures on the permeabilities at the bed scale combining the structural analysis with all other relevant borehole analyses, and take them all into account to perform a single, synthetic multidisciplinary reservoir model, in terms of rock types, potential reservoirs, seals, conduits and barriers.

Published work addressing this problem is very scarce. However, where attempts have been made, the results suggest that it is possible to classify which fractures are conductive within the entire fracture network and that there are underlying mechanical reasons controlling the distribution of these conductive fractures.

A thematic study on this subject, conducted by combining in situ observations in mines or quarries with production data and specific analogue and numerical modelling will enable improvement in our understanding of this problem and thereby improvement of hydrodynamic modelling.
 
 

Flow of water from a joint in a uranium mine ( Lodeve, S. France)

.

Fluid flow paths in fractured layered rock from the Lodève Basin (Southern France). Inspired from quarry and mine observations and well data.