SPE 92980

Controls on Fracturing in Carbonate Rocks

D.C.P. Peacock and A. Mann, Fugro Robertson
Copyright 2005, Society of Petroleum Engineers Inc.
This paper was prepared for presentation at the 14th SPE Middle East Oil & Gas Show and
Conference held in Bahrain International Exhibition Centre, Bahrain, 1215 March 2005.
This paper was selected for presentation by an SPE Program Committee following review of
information contained in a proposal submitted by the author(s). Contents of the paper, as
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to
correction by the author(s). The material, as presented, does not necessarily reflect any position of the Society of Petroleum Engineers, its officers, or members. Papers presented at SPE
meetings are subject to publication review by Editorial Committees of the Society of Petroleum
Engineers. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to a proposal of not more than 300 words;
illustrations may not be copied. The proposal must contain conspicuous acknowledgment of
where and by whom the paper was presented. Write Librarian, SPE, P.O. Box 833836,
Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.

Abstract
Sedimentology (lithology, sedimentary structures, bed thickness, mechanical stratigraphy, the nature of bedding planes),
structural geology (tectonic setting, palaeostresses, subsidence/uplift history, proximity to faults, position in a fold,
timing of structural events, mineralisation, the angle between
bedding and fractures) and present-day factors (orientations of
in situ stresses, fluid pressure, perturbation of in situ stresses,
depth) can all control the nature and distribution of fractures
within carbonate rocks. Understanding these factors requires
traditional geological skills, including the analysis of core,
borehole images and exposed analogues.
Introduction
More than half of the Worlds hydrocarbon reserves are in
carbonate rocks. As these reservoirs become mature, it is being increasingly recognised that fractures play a significant
role in controlling production and water breakthrough. Specific objectives of a fracture study can include the following:
Understand the fractures in terms of the tectonic history of
the region.
Analysis of core, borehole image data, well test data, etc.,
to determine the nature of fractures in reservoir formations.
Determine the role of fractures on hydrocarbon migration,
entrapment and production.
Predict optimum well locations and orientations.
Provide inputs into reservoir simulation models.
Determining the factors that control the nature and distribution of fractures is a key to the accomplishment of these
objectives. This paper reviews the factors that control the nature and distribution of fractures within carbonate rocks. Examples are presented from Mesozoic limestones of southern
England. It is shown that traditional geological skills are of
vital importance in determining the sedimentological, structural and present-day factors that control fractures.

Definitions
Fracture is a general term for a discontinuity, with fractures in
rock including faults, veins and joints. Veins are mineral-filled
fractures [e.g. 1], are commonly mode I (extension) fractures,
and usually form as en echelon segments in a vein array.
Joints (or open fractures) are mode I fractures, usually with
apertures of less than a few millimetres, across which there
has been no measurable shear displacement [e.g. 2], and that
are not mineral-filled. Veins and joints are mechanically similar [e.g. 3], and so it may be appropriate to consider them together in mechanical analyses. There are, however, important
geometric, scaling and genetic differences between veins and
joints [e.g. 4], so it is commonly unhelpful to group them together or to ignore their differences [5]. This paper deals with
the factors controlling veins and joints, although the same
techniques can be applied to faults.
Factors Controlling Fractures In Carbonate Rocks
Various factors have been shown to control the frequency,
distribution and nature of fractures in rock. This section lists
these factors and gives key references. It is shown how the
influence of these factors can be determined using careful geological interpretation of core, borehole image or exposed analogue data. It must be noted that there are particular problems
in using fracture frequencies from wells that are at a low angle
to fractures (e.g. vertical wells intersecting vertical fractures),
because there are biases in the sampling of the fracture population [e.g. 6].
Sedimentological factors. The nature of the rocks is a primary control on fracturing. Sedimentological factors include
the following:
Lithology. Lithological competence can control the nature
of fractures, with more fractures tending to occur in more brittle beds (Figure 1) [7]. The control of lithology on fracturing
can be determined by using well data (core, borehole images)
or scanlines measured in exposed analogues to determine the
number of fractures per metre for different lithologies.
Sedimentary structures. Fractures can initiate at such anisotropies as bedding plane irregularities and fossils [e.g. 2, 8].
Such effects can be difficult to see in core or borehole images
because identification relies on the well intersecting an initiation point. Such effects are commonly seen, however, in exposed analogues.

Bed thickness. Fracture frequencies tend to be higher in

thinner beds than in thicker beds (Figure 2) [e.g. 7]. The effects of bed thickness can be tested by plotting a graph of fracture frequency against bed thickness, taking care that similar
lithologies are compared. Such data can be hard to obtain from
wells because fractures are commonly at a high angle to bedding. For example, bed thickness can be measured easily in a
vertical well, but the vertical fractures are poorly sampled. It is
relatively easy, however, to measure fracture frequencies in
different thickness beds in exposed analogues.
Mechanical stratigraphy. Sequences of beds can act as a
single mechanical unit, with throughgoing fractures developed
[e.g. 9]. This appears to be common in rocks that have been
deeply buried or affected by compaction-related stylolites, so
the beds have been welded together. Fractures tend to be more
widely spaced in such mechanical units than would be expected in the component beds. Such effects are indicated in
well data by a lack of correlation between fractures and lithology, and by fractures crossing bedding planes. Mechanical
stratigraphy can be seen most easily, however, in exposed analogues (Figure 3).
Nature of bedding planes. The nature of the bedding
planes control fracture propagation and mechanical stratigraphy. For example, fractures tend to stop at bedding planes that
mark boundaries between beds of very different competences
(Figure 1), or at bedding planes across which slip can occur
[e.g. 10]. The termination of fractures, or propagation of fractures across, bedding planes can commonly be seen in core
and borehole images.
Structural factors. These include the following:
Tectonic setting. The tectonic setting can control the nature of fracturing, especially because the stress regime controls
fracture orientations. Extension fractures (veins and joints)
usually form perpendicular to the least compressive stress [e.g.
2]. If, for example, the least compressive stress is vertical,
there will be a tendency for thrust faults to develop [e.g. 11]
and extension fractures will tend to be sub-horizontal [e.g. 12].
The tectonic setting can be determined from earthquake data
[e.g. 13] and from the regional geological setting.
Palaeostresses. The palaeostress history controls the sequence of fracture development and therefore the geometry of
the fracture network [e.g. 14]. Different sets of fractures can
develop during the stress history of a rock mass, and it is important to determine the chronology of structures [e.g. 5]. Palaeostress orientations can be determined using the orientations of fault slickenside lineations [e.g. 15] and extension
fractures, and the chronology of palaeostresses can be determined by the cross-cutting relationships of fracture sets.
Subsidence/uplift history. Subsidence and uplift are important aspects of the palaeostress history. For example, open
fractures near the Earths surface are commonly caused by
stress release during uplift [e.g. 14]. Caution is therefore
needed when directly comparing subsurface data with exposed
analogues.
Proximity to faults. It is commonly assumed that fracture
frequency increases close to faults [e.g. 16]. Care is needed,
however, because open (non-mineralised) fractures commonly
do not cluster around faults [e.g. 17]. Figure 4 shows an ex-

SPE 92980

ample of such behaviour. The relationship between faults and

other fractures can be determined by analysing the frequencies
of fractures along wells that intersect faults or from appropriate use of exposed analogues.
Position in a fold. It is also commonly assumed that fracture intensity is directly and simply related to curvature or
strain in a fold [e.g. 18], even though few field-based studies
have been carried out on the relationship between fractures
and folds. As with faults, care is needed in assuming such a
simple relationship, especially where the fractures formed before or after folding [e.g. 19]. The relationships between fractures and folds can be determined by analysis of fracture frequencies and orientations in wells at different locations in a
fold, or from appropriate use of exposed analogues.
Timing of structural events. The age of fractures in relation to other structures is important in controlling their distribution. For example, fractures can be clustered around faults if
they formed during faulting, but will not cluster around faults
if they formed before or after faulting [17]. This seems to be
especially common for open fractures, with the fractures
around faults commonly being mineralised (Figure 4). The
relative timing of structures can be determined by analysing
cross-cutting relationships in core or borehole images, and by
careful use of information about the tectonic history of a region.
Mineralisation. The nature of the fractures is influenced
by phases of mineralisation, with pre- or syn-mineralisation
fractures being closed or partly closed veins, and postmineralisation fractures tending to be open (Figure 4). It is
therefore important to determine the mineralisation history [5]
using thin sections from core.
Angle between bedding and fractures. Fractures are commonly developed approximately perpendicular to bedding,
especially where bedding planes are free surfaces. The dip of
bedding therefore commonly controls the dip of fractures. The
angles between beds and fractures can be determined from
core and borehole images.
Evolution. A set of fractures is unlikely to form instantaneously, but will have evolved. For example, Rives et al. [20]
found an evolution of joint spacing from initially negative
exponential to log-normal and normal. Figure 5 shows how
such an evolution may occur with the gradual infilling of
spaces between fractures. Interpretations about developmental
stage of fracture sets based on well data should be treated with
caution, but variations in fracture characteristics at different
locations and in different beds can be related to the evolution
of the fractures.
Present-day factors. Current activity that can control fracturing include the following:
Orientations of in situ stresses. The apertures of open
fractures are commonly influenced by their relationship to the
in situ stresses. For example, there is a tendency for fractures
that are normal to the maximum compressive stress to be
closed, while fractures that are normal to the least compressive
stress are more likely to be open (Figure 6). The orientations
of in situ stresses can be determined using the orientations of
borehole breakouts and induced fractures, and the orientations
of open fractures can be measured from borehole images.

SPE 92980

Fluid pressure. The relative magnitudes of fluid pressure

and in situ stresses can control the initiation of fractures and
which fractures are open. If fluid pressure exceeds the minimum compressive stress, then the effective stress [21] is such
that effective tension exists in the direction of the least compressive stress. Extension fractures (vein and joints) can initiate and remain open perpendicular to the direction of the least
compressive stress if the rock is in effective tension. The relationship between in situ stresses and fluid pressure is therefore
of crucial importance in controlling whether fractures are open
or closed (Figure 6). Fluid pressure can be determined from
reservoir engineering data, with the orientations of open and
closed fractures giving an indication of the orientations and
magnitudes of the in situ stresses.
Perturbation of in situ stresses. Local variations in
stresses occur, for example around folds or faults [e.g. 20], and
such stress perturbations can control the orientations and geometries of fractures. Figure 7 shows and example of joints
that are perturbed around a fault, these reflecting stress perturbation. Stress perturbation in a reservoir is indicated by variations in the orientations of natural or induced fractures seen in
core or borehole images.
Depth. Variations in stresses and fluid pressure occur with
depth [e.g. 22], and so there can be related variations in the
nature of the open fractures. Such effects are indicated in well
data by variations in the nature of open fractures with depth.
The Importance Of Geological Interpretation
Understanding the factors that control fracture geometry, orientation, spacing, etc., helps to reduce risk and to improve
predictions and modelling. Such knowledge can, for example,
be used to plan optimum well locations and orientations. Good
data and careful geological interpretation are two vital components of any useful fracture study. Core and borehole image
data provide direct information about the fractures in a reservoir, but useful information can also be obtained from reservoir engineering and seismic data. Statistical analysis of fracture data is important, but full interpretation requires traditional geological interpretations, such as the determination of
age relationships between structures. Modelling methods that
do not use data but rely on simulations (such as populating a
reservoir with a fracture distribution controlled by curvature),
or that are not geologically sensible or realistic, are unlikely to
be successful.
Exposed analogues can be useful in developing models,
especially if well data are limited. Care is needed, however,
because an exposed analogue will have a different geological
history and will be under different conditions than a reservoir.
Use of analogue data requires geological knowledge. For example, Figure 8 shows a zone of closely spaced joints in Cretaceous Chalk. This example is not a good analogue for the
North Sea Chalk fields, which tend to have much higher porosities [e.g. 23]. The location can be used, however, to gain
insights into aspects of the carbonate reservoirs on the Middle
East, where such zones of closely spaced joints are common.
For example, this location can be used to test fracture frequencies inside and outside of the joint zones, and to measure the
spacing between swarms. The careful use of analogues can
therefore provide important insights into subsurface geology.

Conclusions
The nature and distribution of fractures within carbonate rocks
can be controlled by sedimentology (lithology, sedimentary
structures, bed thickness, mechanical stratigraphy, and the
nature of bedding planes), structural geology (tectonic setting,
palaeostresses, subsidence/uplift history, proximity to faults,
position in a fold, timing of structural events, mineralisation,
and the angle between bedding and fractures) and present-day
factors (orientations of in situ stresses, fluid pressure, perturbation of in situ stresses, and depth). Traditional geological
skills are of great importance in determining the factors that
control fractures and in providing realistic inputs into reservoir
simulation models. The careful geological analysis of core,
borehole images and exposed analogues can be used to determine the relative timing of events, and this plays a crucial role
in determining the nature and distribution of fractures. For
example, open fractures tend to be clustered around faults only
if the open fractures and faults formed at the same time. Clustering does not occur if the open fractures pre-date or post-date
the faults.
Acknowledgements
Fugro Robertson is thanked for permission to publish this
work.
References
1. Davis, G.H. and Reynolds, S.J.: Structural Geology of Rocks and
Regions, Wiley, New York (1996).
2. Pollard, D.D. and Aydin, A.: Progress in understanding jointing
over the past century, Geological Society of America Bulletin
(1988) 100, 1181.
3. Thomas, A.L. and Pollard, D.D.: The geometry of echelon fractures in rock: implications from laboratory and numerical experiments, Journal of Structural Geology (1993) 15, 323.
4. Gillespie, P.A. et al.: Scaling relationships of joint and vein
arrays from The Burren, Co. Clare, Ireland, Journal of Structural
Geology (2001) 23, 183.
5. Peacock, D.C.P.: Differences between veins and joints using the
example of the Jurassic limestones of Somerset, In: Cosgrove,
J.W. and Engelder, T. (Eds.) The Initiation, Propagation, and Arrest of Joints and Other Fractures, Geological Society of London, Special Publication (2004) 231, 209.
6. Terzaghi, R.D.: Sources of errors in joint surveys, Geotechnique (1965) 15, 287.
7. Ladeira, F.L. and Price, N.J.: Relationship between fracture
spacing and bed thickness, Journal of Structural Geology (1981)
3, 179.
8. McConaughy, D.T. and Engelder, T.: Joint interaction with embedded concretions: joint loading configurations inferred from
propagation paths, Journal of Structural Geology (1999) 21,
1637.
9. Finn, M.D. et al.: Kinematics of throughgoing fractures in
jointed rocks, Tectonophysics (2003) 376, 151.
10. Cooke, M.L. and Underwood, C.A.: Fracture termination and
step-over at bedding interfaces due to frictional slip and interface
opening, Journal of Structural Geology (2001) 23, 223.
11. Anderson, E.M.: The Dynamics of Faulting, Oliver and Boyd,
Edinburgh (1951).
12. Sibson, R.H.: Earthquake faulting as a structural process, Journal of Structural Geology (1989) 11, 1.
13. Reinecker, J et al.: The 2004 release of the World Stress Map.
Available online at www.world-stress-map.org (2004).
14. Rawnsley, K.D. et al.: Jointing in the Mesozoic sediments

around the Bristol Channel Basin, Journal of Structural Geology

(1998) 20, 1641.
15. Angelier, J.: Tectonic analysis of fault slip data sets, Journal of
Geophysical Research (1984) 89, 5835.
16. Pohn, H.A.: Joint spacing as a method of locating faults, Geology (1981) 9, 258.
17. Peacock, D.C.P.: The temporal relationship between joints and
faults, Journal of Structural Geology (2001) 23, 329.
18. zkaya, S.I.: CURVAZa program to calculate magnitude and
direction of maximum structural curvature and fracture-flow index, Computers and Geosciences (2002) 28, 399.
19. Silliphant, L.J. et al.: The state of stress in the limb of the Split
Mountain anticline, Utah: constraints placed by transected
joints, Journal of Structural Geology (2002) 24, 155.
20. Rives, T. et al.: Joint spacing: analogue and numerical simulations, Journal of Structural Geology (1992) 14, 925.
21. Hubbert, M.K. and Rubey, W.: Role of fluid pressure in mechanics of over-thrust faulting. Parts I and II, Bulletin of the
Geological Society of America (1959) 70, 115.
22. Seplveda, N. and Zack, A.L.: The effects of overburden stress
on the specific storage and hydraulic conductivity of artesian aquifers, Journal of Hydrology (1991) 128, 305.
23. Mallon, A.J. and Swarbrick, R.E.: A compaction trend for nonreservoir North Sea Chalk, Marine and Petroleum Geology
(2002) 19, 527.

SPE 92980

(a)

(b)

Figure 2. Examples of joints on Liassic limestone bedding planes

from Somerset, showing the influence of bed thickness. (a) Bed
80 mm thick. (b) Bed 300 mm thick. Joint frequencies tend to be
higher in thinner beds than in thicker beds [e.g. 7], with different
patterns of joint occurring in the different beds.

Figure 1. Example of the lithological control on fractures, from the

Liassic limestones and shales of Somerset, U.K. More veins and
joints occur in the more brittle limestones than in the less brittle
shales.

SPE 92980

Figure 3. Gently dipping Jurassic limestones at Osmington, Dorset, U.K. The steeply dipping joints cut across bedding planes,
and so there does not appear to be a lithological control on fracturing.

Figure 4. Oblique view of a Liassic limestone bedding plane cut

by a dextral strike-slip fault, Somerset. Calcite veins are clustered
within the fault zone, but the later joints are not clustered. This
example illustrates the importance of the timing of different fractures [17].

Figure 5. Schematic diagram of the evolution of fracture spacing.

(a) Initial fracture development, with clustered fractures showing
a negative exponential distribution. (b) New fractures infill the
spaces, producing a log-normal distribution. (c) The wider spaces
are filled by fractures, producing an approximately equal spacing
of fractures and a normalised distribution. (d) Graphs of the fracture spacings shown in (a) to (c).

SPE 92980

Figure 6. Relationships between the stress system and open fractures. (a) All of the fractures will tend to be closed if fluid pressure
(Pf) is less than the least compressive stress (3). (b) One set of
fractures will tend to be open if the intermediate compressive
stress (2) > Pf > 3, these being perpendicular to 3. (c) Open fractures will tend to be perpendicular to 3 if the maximum compressive stress (1) > Pf > 2, but can be of an orientation parallel to 1.
(d) Any orientation of fracture can be open if Pf > 1.

Figure 8. Photograph of a 0.6 m wide zone of closely spaced

joints in Cretaceous Chalk, at Walpole Bay Cliffs, Margate, U.K.
The joints strike ~ 140, dipping at 79 towards 230. This exposure does not provide a good analogue for the North Sea Chalk
fields, which tend to have much higher porosities and are much
less brittle [e.g. 23]. This may, however, provide a good analogue
for aspects of the carbonate reservoirs on the Middle East, where
such zones of closely spaced joints are common. The Chalk cliffs
around Margate provide an opportunity for analysing the nature
and distribution of such joint swarms.

Figure 7. Photograph of joints curving into a NNE-trending sinistral strike-slip fault zone that has tens of millimetres displacement, Somerset. Beds have a separation across the fault zone
because the slip vector is slightly oblique to the gently dipping
beds [17].