{short description of image}
 Fault damage zone scaling and spatial organisation of granulation
 seams in porous sandstone: Implications for flow properties 

Back to R and D programme
Back to Fault properties evaluation
 
This page contains some background information on granulation seams and the effects they have on reservoir permeability. It is intended to give a background understanding of the technical issues and introduce some of the solutions we have developed. A brief reference list is included at the bottom of this page.

Brittle faults in the Earth's crust are generally not simple planar surfaces of detachment and slip, but are typically composed of a heterogeneous volume of small-scale structures which can include granulation seams, open and mineralized fractures, slip surfaces, fault gouge zones and shale smears (e.g. Burhannudinnur and Morley, 1997; Knott et al., 1996; Berg and Avery, 1995; Antonellini and Aydin, 1994). At the centre of the fault zone, often termed the fault core or master fault zone, most of the displacement is accommodated along a primary slip surface which may contain associated fault gouge or breccia, shale smear, cataclasites and/or mylonites. Surrounding this fault core, and mechanically related to growth of the master fault, is an associated damage zone. Whilst damage zones in heterogeneous sediments can contain fractures, small faults, granulation seams, veins and small folds, in porous reservoir sandstones they are dominated by slip surfaces and granulation seams (e.g. Antonellini and Aydin, 1994; 1995; Underhill and Woodcock, 1987).

Granulation seams are individual quasitabular bands of crushed rock, commonly less than 1 cm in thickness, which are characterised by intense grain size reduction, grain rotation and compaction (Fowles and Burley, 1994). They are thought to form under conditions of high deviatoric stress, low confining pressure (i.e. at relatively shallow depths in the crust), and low temperature (Mitra, 1988) and typically accommodate small shear offsets which are generally less than a few millimetres in dimension (Antonellini and Aydin, 1995).

It is widely reported that the grain-size reduction and granulation caused by cataclasis in granulation seams can significantly reduce their permeability by up to four orders of magnitude with respect to the undeformed host rock (e.g Burhannudinnur and Morley, 1997; Freeman et al., 1995; Edwards et al., 1993; Hippler, 1993; Mitra, 1988). That is, typical host-rock permeabilities of 1000mD can be reduced to a little as 0.1 mD within granulation seams. Our own laboratory rock-deformation experiments in pristine reservoir sandstone confirm this permeability reduction and strongly suggest that distributions of granulation seams have the potential to trap hydrocarbons across fault damage zones and to significantly affect the recovery performance of faulted reservoirs. Whilst, permeability reduction is also accompanied by a marked increase in capillary entry pressure within granulation seams, we will concentrate on the permeability reduction in this paper.

From our extensive field work within the damage zones of reservoir scale faults it is clear that granulation seams typically exhibit several distinct characteristics: (1) they tend to form in conjugate sets with orientations parallel to the strike of local (extensional) faulting - leading to high connectivity in a plane perpendicular to strike, (2) they cluster in space - leading to relatively high structure densities, and (3) they tend to anastomose along strike producing splays and links with other seams rather than tipping out - leading to high connectivity along strike. All these factors can lead to a highly connected and compartmentalized (Edwards et al., 1988) damage zone. Importantly, these isolated compartments occur at a range of scales, from sub-millimetre size compartments within dense clusters of seams to compartments several tens of metres in dimension formed between dense clusters (figure 1.0). Here, we assess the affect of these complex structures on fluid-flow within deformed reservoirs.
 

{short description of image}

Figure 1.0 Schematic diagram depicting the typical 3D geometries exhibited by granulation seams at outcrop level. Note the conjugate orientations and compartmentalisation on a wide range of scales.



Results from our field work within fault damage zones in porous reservoir-quality sandstones suggests that the spatial clustering of granulation seams within fault damage zones can be described using a simple statistical model. Using this model, we have constructed synthetic distributions of granulation seams as inputs to a 1D code for fluid flow to calculate effective permeability (figure 2.0) for a range of granulation seam densities and permeabilities.
 

{short description of image}

Figure 2.0 Effective permeability (Keff) as a function of distance across a zone of clustered synthetic granulation seams (the clusters of vertical bars along the x-axis) with a mean density of 6.3 seams/m. Flow is from the right. Host rock and granulation seams are assumed to have a permeability of 1D and 1mD respectively. The entire damage zone has an effective permeability of 59 mD.
 
Using these simulations, we have investigated the dependence of effective permeability on granulation seam density and clustering along 1D transects through fault damage zones. Our initial results suggest that effective permeability is independent of the degree of clustering, but is related to spatial density through a simple scaling relationship, where Keff is a function of seam density, host-rock permeability and granulation seam permeability. Using this relationship and digitized images of real granulation seam networks, we can calculate density (figure 3.0a), and therefore effective permeability (figure 3.0b), as a function of orientation along 1D transects in complex 2D networks of granulation seams for a range of host-rock and granulation seam permeabilities.
{short description of image}
 Figure 3.0 Using the method described here, we can calculate granulation seam density (a) and effective permeability (b) as a function of orientation in such networks.


Our results suggest that the 2D distribution of effective permeability in such systems is highly anisotropic and is controlled, not only by spatial density, but by the connectivity and compartmentalisation of the network. Initial results, using a 2D finite element code for fluid flow support our findings using the 1D method. This suggests that the gross flow properties of complex networks of granulation seams in 2 or 3 dimensions can be modelled effectively using the approach described here.

 In relatively unconnected networks we find that flow is preferentially channelled into a direction parallel to the mean structure orientation (i.e. parallel to fault strike). Whereas, in connected networks, such as in figure 3.0a, flow is severely reduced in all orientations and the direction of maximum flow is controlled by the geometry of the connected network. For example, in figure 3.0a the direction of maximum flow is strongly controlled by the aspect ratio of elongate compartments which have a dominant orientation from upper left to lower right
 
In summary, the compartmentalisation of porous reservoir sandstone by granulation seams in fault damage zones is likely to lead to severe permeability reduction and flow anisotropy on the production time scale. Importantly, since effective permeability appears to be independent of spatial clustering in our simulations, the flow properties of granulation seam networks can be easily up-scaled from the field outcrop to the reservoir scale.

From our research and development program we have the capability to provide the following solutions right now!
  1. Field studies to assess the impact of granulation seams (density, geometry, compartmentalisation and connectivity).
  2. Studies of image logs (FMI, UBI, STAR etc.) to assess the impact of granulation seams (density, geometry, compartmentalisation and connectivity).
  3. An assessment and scale-up of effective permeability along 1D line-samples, including boreholes and field traverses.
  4. Assessment of damage zone effective permeability (based on granulation seams) and its variation laterally along individual faults.
  5. The production of maps highlighting the likely distribution of damage expected in a reservoir (density, orientation and connectivity), based on fault throw vs damage zone width relations and modelling.
  6. The production of maps showing structural density as a function of orientation in grid cells.
  7. The production of maps showing up-scaled effective permeability in faulted reservoirs.

All of these practical approaches can be applied to a reservoir to help identify the effects of granulation seams on hydrocarbon recovery.
 
References:
Antonellini, M. and Aydin, A. (1994), Effect of faulting on fluid flow in porous sandstones: Petrophysical properties, Bull. Am. Ass. Petrol.Geol. 78, pages 355-377.

Antonellini, M. and Aydin, A. (1995), Effect of faulting on fluid flow in porous sandstones: Geometry and spatial relations, Bull. Am. Ass. Petrol. Geol. 79, pages 642-671.

Berg, R. R. and Avery, A. H. (1995), Sealing properties of Tertiary Growth Faults, Texas Gulf Coast, Bull. Am. Ass. Petrol. Geol. 79, pages 375-393.

Burhannudinnur, M. and Morley, C. K. (1997), Anatomy of growth fault zones in poorly lithified sandstones and shales: implications for reservoir studies and seismic interpretation: part 1, outcrop study, Petrol. Geoscience 3, pages 211-224.

Edwards, H. E., Becker, A. D. and Howell, J. A. (1993), Compartmentalization of an aeolian sandstone by structural heterogeneities: Permo-Triassic Hope Sandstone, Moray Firth, Scotland, in North, C. P. and Prosser, D. J. (eds.) Characterization of Fluvial and Aeolian Reservoirs, Geological Society Special Publication No. 73, pages 345-371.

Fowles, J. and Burley, S. (1994), Textural and permeability characteristics of faulted, high porosity sandstones, Mar. Petrol. Geology, 11, pages 608-623.

Freeman, B., Yielding, G. and Needham, T. (1995), Predicting fault seal potential in hydrocarbon reservoirs, in Structural Geology and Reservoir Characterisation (abstract). Hippler, S. J. (1993), Deformation microstructures and diagenesis in sandstone adjacent to an extensional fault: Implications for the flow and entrapment of hydrocarbons, Bull. Am. Ass. Petrol. Geol. 77, pages 625-637.

Knott, S. D., Beach, A., Brockbank, P. J., Lawson Brown, J., McCallum, J. E. and Welbon, A. I. (1996), Spatial and mechanical controls on normal fault populations, J. Struct. Geol 18, pages 359-372.

Mitra, S. (1988) Effects of deformation mechanisms on reservoir potential in Central Appalachian overthrust belt, Bull. Am. Ass. Petrol. Geol. 72, pages 536-554.

Underhill, J. R. and Woodcock, H. N. (1987), Faulting mechanisms in high-porosity sandstones; New Red Sandstone, Arran, Scotland, in Deformation of sediments and sedimentary rocks, Jones, M. E. and Preston, R. M. F. (eds), Geol. Soc. Spec. Publ. 29, pages 91-105.