inding fractures in deep and tight

F rocks has become a high priority

among explorationists in the Middle
East. Recent discoveries have shown that
fractures can play an important role in the
productivity of low permeability
formations. This is because they form an
interface with the rock matrix which is
many times greater than that provided by
the borehole.

In this article authors from different

organizations discuss the origins of
fractures, their importance within Middle
East oil reservoirs, and outline the various
fracture interpretation techniques
currently being used.

Authors:

Schlumberger: Mahmood Akbar, Roy Nurmi,

Eric Standen and Sandeep Sharma.

Oil and Natural Gas Commission of India (ONGC):

Piyush Panwar and J.G. Chaturvedi.

ONGC-Schlumberger Wireline Research Center:

Bob Dennis.
F
ractures are 3D features - a fact
that is often neglected in the early
stages of reservoir development.
While fractures seen in the wellbore will
be analyzed to determine aperture and
probable production rates, little effort is
made to develop a detailed model of
fracture distribution. This kind of study
only takes place when the reservoir is
formed almost entirely of fracture poros-
ity (as in a fractured basement) or when
some aspect of reservoir behaviour
strongly contradicts the existing reser-
voir model - for example, in cases where
there is sudden and unexpected water
production.
In general, fractures are important
because of their influence on tight reser-
voirs, not because of their actual oil stor-
age capability. Although fracture volume
may be negligible in comparison with
the total reservoir volume, they provide
an interface with the matrix which is
much larger than the borehole. Conse-
quently, very small primary permeabil-
ity values are sufficient for production in
a fractured reservoir.
Factors controlling the occurrence of
natural, open, permeable fractures
within Middle East reservoirs are the
nature and degree of folding and/or
faulting, in-situ stresses and changes in
rock properties such as porosity, bed-
ding and lithology, especially shaliness.
Such geological factors are often
mapped in reservoir studies. By accu-
rately defining the relationship of these
factors to the fracturing in a number of
wells within a field, it should be possible
to extrapolate the fracture data through-
out the reservoir.
The major problem with large-scale
fracture studies based on either borehole
imagery or oriented cores is that the frac-
Fig. 3.2: HIDDEN
AGENDA: Surface
fractures can cause
major problems at
reservoir depths.
Unconformities in
limestones often
lead to the
development of
karst and leaching
of fractures. After
burial, these
surface weathering
effects can lead to
strange reservoir
behaviour. The
leached fractures in
this photo are from
Jebel Hafit, UAE,
with Roberta Nurmi
for scale.
26
UPLIFT
Tension
Tension Fig. 3.1: UNRELIABLE
fractures fractures OUTCROPS: As rock
layers return to the
surface, stress release
allows new fractures
to develop. These
fractures do not occur
in the same formation
at reservoir depth.
tures in the borehole wall may not be
representative of the large-scale fracture
network which controls production. It is
necessary, therefore, to relate geometri-
cal information from borehole data to the
reservoir’s geological characteristics
(structure, stratigraphy, sedimentology,
diagenesis and geostatistics). Simple
cubic models used in reservoir studies
are sometimes appropriate but often more
complex geometric models are required.
Borehole-scale characterization and
geological modelling must be fully inte-
grated with dynamic testing and produc-
tion data, not simply used for compari-
son as a type of quality control.
Using outcrop data to characterize
the fracture pattern of a reservoir is frus-
trated by the stress release which
occurs as rocks come to the surface. Fig. 3.3: The fracture orientations
Uplift and erosion of overburden often commonly found in the Middle
result in tensional breaking of brittle East’s anticlinal reservoirs are
beds due to deformation along ductile highly variable. In most cases,
bedding planes (figure 3.1). As a result, information from a single well
fracture density in reservoirs is com- cannot be used to characterize
the fractures for an entire field.
monly lower than values recorded
where the same formation outcrops.
Consequently, outcrop data is unsuit-
able for modelling reservoir fractures.
Middle East Well Evaluation Review
 A B
Cretaceous Cretaceous
Fig. 3.4: AGE GAP:
Reservoirs of the
same age in different
fields can display a
wide variety of
fracture orientations.
Patterns of fracturing Jurassic
vary through
geological time;
fracture orientation
being controlled by
the timing of
fracturing, not the
formation age.
Permian
Unfortunately, it is becoming clear
that many unconformity-related features
(figure 3.2), such as karst and exfoliation
fractures (both due to subaerial weath-
ering), are present in Middle East reser-
voirs. While such fracture systems usu-
ally occur over a very narrow depth
range, they may have greater connectiv-
ity than tectonic fractures, which are
more numerous, and so exert a strong
influence on reservoir behaviour.
How fractures are formed
Fracture orientations can vary consider-
ably between reservoirs but the orienta-
tions are neither random nor chaotic.
Early fracture studies, which were based
on poor or insufficient data, provided
very misleading information about frac-
ture distribution in reservoirs.
Fractures are usually formed during
folding or doming of a reservoir, with the
most intense fracturing being concen-
trated in low-porosity rocks. In areas
where the reservoirs have little matrix
permeability fractures are critical to pro-
ductivity. This is especially true of base-
ment reservoirs where fracture porosity
makes up most of the reservoir.
Borehole imagery and 3D seismic
surveys have improved fault mapping
and horizontal wells are providing a
new insight into the fracturing associ-
ated with reservoir faulting.
Fracturing typically occurs in one of
two ways - either parallel or perpendicu-
lar to normal or reverse faults. Oblique
orientations are associated with wrench
or shear movements.
Number 14, 1993.
Crack concentration
Fracture density, or the ‘intensity’ of frac-
turing, is defined as the number of frac-
tures per unit length inside an interval of a
defined height. This value has to be cor-
rected for the orientation bias created by
changes in angle of the fracture planes
and the borehole axis.
Anomalous increases in reservoir
tectonic fracturing are sometimes seen
in wells. This is usually associated with
faulting. Other variations in fracture den-
sity have been attributed to changes in
lithology, porosity or shaliness. Low
porosity and shale-free intervals gener-
ally contain more fractures, although
these may or may not be mineralized.
Clearing up the chaos
In the past, wide variations in fracture
orientations within anticlinal reservoirs,
coupled with poor quality orientation
data and fracture characterization, led to
widespread pessimism about the useful-
ness of fracture data in reservoir mod-
els. However, improvements in fracture
detection and analysis techniques have
shown that there is order in the appar-
ent chaos. Fracture orientations can be
related to specific geological parameters
and structural events.
The common fracture orientations
found in Middle Eastern anticlinal reser-
voirs are shown in figure 3.3. Changes in
orientations can be caused by later fault
movement associated with variations in
tectonic stress through time. In the car-
bonate reservoirs of Turkey and Iran,
the orientation of karstic fractures asso-
ciated with erosional unconformities, is
much more variable. Low-angle, stress-
relief, exfoliation fractures, which occur
sub-parallel to the unconformity surface
in these reservoirs, are particularly
important features.
Variable orientations of open frac-
tures in a reservoir are often associated
with the rotation of principal stress in
proximity to some of the observed
faults. When the principal stress is per-
pendicular to the fault strike compres-
sional forces are probably at work on
the fault.
Pass the salt
Fracture orientations in salt dome reser-
voirs depend on the shape of the struc-
ture and the nature of the regional and
doming stresses. As a result, reservoirs
of the same age in different fields can
have different fracture orientations if the
fractures formed at separate times under
different regional stress conditions (fig-
ure 3.4).
For the same reason, fracture orienta-
tions vary through geological time. In
the same field it is possible to find frac-
tures formed in the Permian with very
different orientations to those formed in
the Jurassic or in Cretaceous reservoirs.
In addition, fracture orientations can dif-
fer between reservoirs of identical age in
adjacent fields if doming and fracturing
did not occur at the same time in the
two reservoirs.
Early fracturing around The Gulf was
affected by extension stresses during the
Triassic which were later followed by
compressional forces from the east due to
thrusting associated with obduction
(pushing of ocean crust onto continental
crust) in Oman. In contrast, orientations of
later fractures were determined by com-
pressional forces of the Zagros Orogeny.
These are generally oriented NNE.
Definition of reservoir fracturing in
Mesozoic salt structures is important in
The Gulf. These fractures rarely dominate
productivity as their apertures and densi-
ties are usually much lower than those
associated with tectonic fractures. (Pres-
sure and flow tests indicate that intergran-
ular pore systems in both shoal and reefal
facies generally contribute more flow).
However, it is clear that the fractures,
which are often characterized by thin
apertures, are connecting heteroge-
neously distributed porosity (Nurmi et al.,
1990, WER Structural Geology Supplement).
27
 Listric growth with rollover Normal Reverse Thrust

W E W E W E W E
Cross section

N E S W N N E S W N N E S W N N E S W N
Image

0 90 180 270 360 0 90 180 270 360 0 90 180 270 360 0 90 180 270 360

Fig. 3.5: FAMILY OF FAULTS: The four major fault types (top row) can be distinguished using borehole imagery (bottom row).
Knowing the precise geometry of the fault plane where it intersects the well allows us to project the fault away from the well and so
predict its effect in other locations.

Recognising faults
The presence of a fault is often indicated
by rapid increases in fracture intensity or
spacing. This means that once the relative
intensity of fracturing has been deter-
mined for each reservoir zone any
increase in permeability, especially in
more porous intervals, can usually be
attributed to faulting. In such situations, a
focused search using borehole imagery
can reveal the actual fault plane or zone
affecting the fractures (figure 3.5).
Fault detection and analysis have only
recently been recognized as a critical
component in the characterization of frac-
tured reservoirs. In early borehole
images few faults were recognized
because none of the available tools were
and below the feature. However, even if Natural fractures
the bedding is not recognised, the geome-
try of the fault plane can be measured Natural fractures are usually assumed to
where it intersects the well. This informa- have been created by tectonic stresses.
tion can be used to project the fault plane They are more common in carbonate
away from the well and through any for- rocks than in sandstones and typically
mation above or below. Unfortunately, occur in specific directions which are dic-
this is not as simple as it may seem (fig- tated by the regional tectonic stresses.
ure 3.6). Induced fractures associated with natural
Fig. 3.6: THINGS
ARE NOT ALWAYS (a)
AS SIMPLE AS
THEY APPEAR:
These diagrams
illustrate situations
where normal
faulting has
developed on the
N or

specifically designed to find them. Frac- hanging wall of a

ma

tures related to faults are much less abun- major reverse fault lt
fau
l fa

dant than fractures related to the folding (a) and where small
rse
u lt

reverse faults have ve

or doming of a field, but the vertical conti-
been created on a Re
nuity of faults often results in dramatic large, normal fault
water movements through fault-associ- structure (b). The
ated fracture systems. interpretation of
A major fault can be a single, high-angle either structure
plane cutting a reservoir horizon and may would depend
not be intersected by wells. However, primarily on the
there are associated smaller faults and location of the well
and its depth of (b)
fractures which formed at the same time
penetration.
and it is these zones which are more
likely to be crossed by wells.
Indications of faulting can arise dur-
ing drilling. These may take the form of
missing rock sections in areas of normal No
faulting, repeated sections in areas of rm
al
thrust or reverse faulting or loss circula- fa
ul
tion material in areas affected by wrench t
faulting or any other type of open frac-
tures associated with faults.
Recognising faults depends on our
ability to detect the bedding plane above Reverse fault

28 Middle East Well Evaluation Review

 Fig. 3.7 (Left):
UP THE WALL: (a)
Drilling-induced
tu re zone fractures are common
d frac
ce but not easily
du distinguished from
in

natural fractures
g-
llin

occurring in the
Dri

wellbore.

Borehole Fig. 3.8: (Right):

PRINCIPAL In-situ stresses normal and
PATTERNS: along the borehole axis.
Extension The pattern for
fractures, drilling-induced
hydraulic fractures depends on
fractures the orientation of (b)
Mixed extension principal in-situ
and shear fractures, stresses relative to the
borehole. These
incipient breakouts
stresses may be
parallel (a, b) or
oblique (c, d) to the
borehole axis.
θ2 θ3

θ1

Principal stresses
normal and parallel
Borehole Borehole to borehole axis.
wall view wall view

fractures might also be expected to have
a preferred orientation, although this
would not necessarily be parallel and
could even be perpendicular to the nat-
ural fractures.
Drilling-induced fractures
Borehole electrical images, especially
those which have been computer
enhanced, provide useful information
about the borehole wall and the reser-
voir rock.
Borehole enlargement is a common
feature and often occurs in the same
direction in a number of wells. The
enlargement is related to stress release
failure and is usually greatest parallel to
the least principal horizontal stress
direction. A complex mixture of induced
fracturing with both shear and extension
cracks is often present (figure 3.7).
Repeat logging of special process
wells has shown that borehole failure
and breakouts occur within days of
drilling. Large fractures induced perpen-
dicular to the direction of borehole
enlargement are usually long straight
cracks in an axial position lying on oppo-
site sides of the borehole. These induced
cracks are extensional fractures which
form and are propagated in front of the
bit during drilling. Cores taken from such
zones often reveal these fractures.
Axial drilling-induced fractures have
a modified appearance when the axis of
the borehole is not parallel to any of the
(c)
principal stress directions. This kind of
crack has a jagged appearance, resem-
bling a lightening bolt, in contrast to the
long and straight cracks which are per-
fectly parallel to the axis of the borehole
(figure 3.8). Such cracks are most com-
mon in horizontal or highly deviated
wells where the orientation of the bore-
hole with respect to the stress field was In-situ stress not normal and
not accounted for in positioning the well. non-axial to the borehole.
Recognizing and analyzing induced
fractures is valuable in determining the
orientation of the principal horizontal
stress which may vary within a reservoir.
(d)
The orientation of ‘mini-frac’ jobs has
also been ascertained and confirmed by
logging before and after ‘fracing’.
β
Enhanced fractures
θ2
In addition to natural and induced frac- θ3
tures, there are pre-existing fractures
which are extended or opened in the θ1
borehole by drilling. These have been
called ‘enhanced fractures’ (Standen,
1991). It has also been observed and Principal stresses
reported that even totally mineralized acting normal but
fractures can be re-opened during the not parallel to axis.
course of drilling.
The enhanced fractures are usually
oriented sub-parallel to the principal
horizontal stress as are drilling-induced
fractures. Generally, drilling-enhanced
fractures do not seem to affect the pro-
ductivity of an interval as they usually
have very small apertures in the undis-
turbed state.

Number 14, 1993. 29

Understanding India's Jordat
fractures r
a Rive
Brah m aputr
Borholla-Changpang

In the mid-1960s, a basement high was

mapped near Borholla, Assam in India.
The first well, drilled on the highest part Golaghat
of the antiformal feature was dry. How-
ever, a second well, drilled several Borholla
years later, located oil in a Palaeocene
sandstone overlying the granitic base- Changpang
ment.
Further ‘wildcat’ drilling in the plains
around Borholla was unsuccessful but a
last ditch attempt was made to drill a st

u
test hole in the nearby Naga Hills. This Mokokchung

hr
Mikir Hills

T
Naga Hills ‘guess’ was rewarded when

ga
Na
the first well, drilled close to Chang-

lt
Be
Uriamghat
pang, in Nagaland, struck oil in fractured

st
en

ru
granite rock (Middle East Well Evalua-

pp

Th
hu
tion Review Number 7 1989, India - 100

ng
Area of study

Sc

sa
Years of Oil ). This discovery proved
ey

Di
that commercial accumulations of oil all
iV
could be found in fractured basement
sir

rocks and had a radical effect on explo-

an
Dh

ration in India.
Conventional seismic investigations Dimapur
carried out around Borholla in the 1960s
were followed by six-fold CDP (Com- N
mon Depth Point) surveys in the early
1970s and 12- and 24-fold CDP surveys
about five years later. These studies Kohima
helped to clarify regional tectonics, but
the low quality of seismic images did
not throw light on structures within the
fractured basement.
The inadequacy of seismic data Geology - the first analysis Fig. 3.9: TWIN PEAKS: The ‘twin’
meant that the only way to decide on oilfields of Borholla-Changpang in
the location for a delineation well was to Assam and Nagaland states of
The explorationists began detailed cor- northeastern India were
analyze data from previous wells. As relation of micro-features on electrologs, discovered in the late 1960s.
more wells were drilled a structural in order to improve their understanding Located on NE-SW trending
model emerged. of the faults. By analyzing well data for basement highs, the fields produce
This approach allowed the basement sand thicknesses in the Sylhet and oil from fractured granite. 3D
structure to be described as a single Kopili formations, it was possible to seismic surveys were carried out
dome near Borholla. However, as infer the location of faults by their effect to define the details of reservoir
clearer information was gathered, the on the sediment. This approach pro- faulting.
model was refined to show two separate vided an indication of the underlying
culminations in the basement, one at basement structure, but not a detailed
Borholla and another at Changpang. The image.
contour maps for the fields became
increasingly detailed and eventually
contained faults with throws of less than
25 m. At this stage the structure began to
be redefined as a mosaic of several fault
blocks and sub-blocks (figure 3.10).

30 Middle East Well Evaluation Review

3D seismic survey -
the final analysis N

2825

0
277
3D seismic data was gathered between

275
75
25

0
2800

50

285
1987 and 1988. Difficult terrain, poor

25

2825
75

25
logistics and subsurface complexity 28

25
combined to produce lower quality data 50

2750

2675
28

2775
than had been anticipated. However,

5
277

2650
50
the high density of 3D data enabled

27

25
interpreters to produce a much better

27

2650

2700
2575

5
image than would have been possible

2875
272

2850
2550

0
270
using 2D techniques.

2725
0
267

260
25

5
Unfortunately, the reservoir map pro-

262
29
2800

50
2875
duced from the 3D survey was viewed

29
5

2700
282

5
75

267
with reservations because it contained

29

5
no reflection event corresponding to the 0

272
285

28
basement’s top surface. This deficiency

00

00

25
5
287

26

27
was explained as absorption and dissi-

25
pation of seismic energy by the frac- 290
0

26
tured basement. 5
292
More importantly, the fault pattern
from the 3D survey did not match the
pattern which had been inferred from Fault
well data and the 2D survey. 0 0.5 1km Drilled well
Mechanical contouring performed
using GEOPIC failed to provide a logical
structure over one producing field. The Fig. 3.10: BASE MODEL: This structure contour map represents the top of the basement in the
3D seismic showed the structure to be Borholla-Changpang area. This model is based on data gathered from 2D seismic surveys and
much steeper than had been previously numerous delineation wells. At this stage the structure appears to be a mosaic of a few major
fault blocks, each comprising many sub-blocks.
believed, which meant that there would
be a marked reduction in closed area.
As a result of these apparent incon-
sistencies, the 3D survey was treated Fault
with scepticism and the earlier map was Drilled well
used for delineation purposes.
However, the 3D seismic data indi-
cated a fault wherever the sand
isopachs showed reduced thicknesses
or missing stratigraphic units. Subse-
quent wells confirmed the pattern
revealed by the 3D survey (figure 3.11)
and prompted its use for fault and struc-
ture delineation. It was concluded that,
in the absence of a strong reflection
from the top of the basement, a good
reflection of the overlying Sylhet Lime-
stone would be sufficient to deduce the
basement fault pattern as well as depth.
This model proved to be highly suc-
cessful. Instead of two distinct culmina-
tions seen at the basement level, the
structure now appears to consist of a
large number of discrete antiformal
cumulations separated by basement
lows. These lows are the sites of major
normal faults typically trending NNE-
SSW which generally avoid the antifor- 0 400 800 m
mal highs.

Fig. 3.11: BASEMENT JIGSAW: A 3D survey revealed that the basement structure did not contain
two large oil-bearing structures, but many small antiformal ‘highs’ separated by basement ‘lows’.
These lows correspond to major normal faults which trend NNE-SSW.

Number 14, 1993. 31

 x823 m More than one way to image
a reservoir
From a detailed examination of all the
new data which emerged from the 3D
survey it became apparent that the max-
imum fracture intensity would be
encountered on the flanks of the antifor-
mal culminations. Production tests of
wells located on the flanks confirmed
this model and showed that these are
the most prolific oil producers in the
field. These wells have very low initial
gas-oil ratios and a negligible decline in
pressure during production. Some wells
had the capacity to sustain self-flow with
more than 60% water cut - proof that
they operate under active bottom-water
drive conditions.
In contrast, wells drilled in the syn-
formal lows often registered good daily
flow rates but produced only moderate
amounts of oil. Recent wells in the lows
have either ceased production as a
result of water loading or produced
water containing only traces of oil.
3 These observations support the idea
that oil accumulation in the basement is
controlled not only by fractures but also
by structural constraints. Below a cer-
tain level virtually all fractures are
water-bearing and the aquifer has gener-
ally flooded the troughs in the vicinity of
good producers. This is the reason why
peripheral wells have short production
lives and wells in the lows surrounding
the main field produce little or no oil.
Reservoir models evolve as the tech-
niques used to investigate them change.
2 A major obstacle, in this case, was the
uncertainty about replacing or modify-
ing an older model with the contradic-
tory information derived from a 3D sur-
vey. Perhaps the most important benefit
of the 3D survey has been in develop-
ment drilling. In situations where previ-
ously there had been uncertainty, the
x825m
model provides clear targets in the
quest to extend the field.
In the Borholla-Changpang area, a con-
tinuous compressive regime has resulted
in extensive basement fracturing. In an
attempt to define the tectonic framework
1
Fig. 3.12: This FMS image highlights the
fractures in the Changpang reservoir. The
natural open fractures (1) are seen to
extend across the diameter of the wellbore.
Drilling enhanced fractures (2) have the
same orientation but are open only in the
direction of the principal horizontal stress.
The drilling-induced fractures (3) are near-
vertical cracks which are found on
opposite sides of the wellbore.
32
of the region, 3D seismic surveys were
planned and carried out in 1987 and 1988.
A recent reservoir study highlighted
the possibility of using horizontal wells to
exploit this fractured basement reservoir
(figure 3.11). But before any decision
could be made on developing the field a
number of parameters had to be deter-
mined - fracture orientation, dip, spacing
and fracture density.
The replacement ratio Fr can be cal-
culated by taking into account the differ-
ences in drainage areas of horizontal
and vertical wells. This ratio represents
the number of vertical wells which
would be required to produce at the
same rate as a single horizontal well. In
this case, the Fr value increased with
length of horizontal section and the com-
puted value was never more than 4.8.
Production performance is a function
of the number of intercepted fractures.
Since the major fracture trends are ori-
ented NW-SE and E-W (figure 3.13), the
horizontal section of each well had to be
drilled in NNE-SSW orientations to maxi-
mize production. Environmental factors
such as land acquisition and logistics
also contributed to the argument that
horizontal, rather than vertical wells
were more suitable for developing this
field.
Oil from Bombay High
Drilling in the giant Bombay High Field,
offshore India, has revealed large vol-
umes of oil and gas in basement rocks
composed of basaltic and granitic
gneiss. The field, discovered in 1974, is
situated approximately 150 km off India's
western coast. Most of the oil and gas is
contained in Miocene limestone reser-
voirs and a smaller proportion in the
basal clastic sandstone reservoirs which
overlie the fractured metamorphic
rocks.
The main limestone reservoir is
encountered at a depth of approxi-
mately 1300 m but the basement hydro-
carbons occur at 1900 m, on the crest of
the structure.
Middle East Well Evaluation Review
 350 0 10
340 20
330 30
320 40

310 Fault directions in Borholla- 50 0

350 10
Changpang fields. 330
340 20
30
300 60 Fig. 3.13: DIRECTIONS FOR BORHOLLA-CHANGPANG:
320 40

290 70
310 Borholla 26 50
These stereonets indicate the fault and fracture directions
300 60 in the Borholla-Changpang fields of Assam, India. The
280 80 290 70
predominant fault orientation (large stereonet) is E-W,
280 80
while the fractures recorded in three separate wells
270 90
(small nets) show variable orientations but are
270 90
260 100
concentrated in E-W and NW-SE directions. These
260 100
350 0 10
results were obtained using a
340 20
250 110 250 110 330 30 Stratigraphic High Resolution
240 120
320 40 Dipmeter (SHDT)* Tool.
240 120 230 130
310 Borholla 22 50

230 130 220 140 300 60

210 150
290 70
220 140 200
190 170
160
180
210 150 280 80

200 160
190 170 270 90
180
260 100

Formation MicroScanner (FMS*) imag- Away from the well- 250 110
0
350 10
ing was used to determine orientation, bore, characterization may 240 120 330
340 20
30

dip and density of the fractures. In addi- also be divided into static 230 130 320 40

tion, acoustic waveform data was used to and dynamic approaches. The 220 210 140 310 Changpang 8 50
150
300
delineate open fractures by means of the static method consists of project- 200
190 180 170
160
60

290 70
Stoneley-derived Permeability Index. The ing fracture systems laterally, away from
280 80
study indicates that the open fractures are, the wellbore, guided by input from high-
in general, high angle and striking in a N-S resolution seismic data or offset VSPs. 270 90

direction. Low-angle fractures recorded in Dynamic characterization beyond the 260 100

the field are generally closed and filled wellbore requires well test data to deter- 250 110

with calcite and quartz. Fracture density mine fracture length, boundary condi- 240 120

230
seems to be higher in granitic basement tions, vertical communication and 130
220 140
than in rocks with a basaltic composition. extended flow capacity for reserve cal- 210 150
200 160
Four wells (BH-36, BH-19, SY-5 and II-7) culations. 190 180 170

were drilled to depths of 200 m in the Core studies, borehole images and
basement. Only SY-5 failed to produce oil Stoneley acoustic permeability logs all
or gas and required hydrofracturing. After show the presence of a high-angle (75°)
N
initial stimulation the other wells pro- open fracture set and a low-angle (20°)
duced oil and/or gas at a rate exceeding healed fractures in the study wells. The
1000 bpd. high-angle fractures strike NNW, parallel
Acoustic and petrophysical logs, core to the adjacent faults.
and well test data were integrated to The low-angle fractures, which are W E
locate open fractures and determine generally filled with either calcite or
their relative flow capability. quartz, have apertures between 5 mm
and 10 mm. These fractures may have
Fracture characterization resulted from a ‘rebounding effect’ and Depth 1
expansion in the basement with subse- S
techniques quent fluid movement. N

Fracture characterization at the wellbore

is a two-step process, requiring static Fig. 3.14: LOOKING A LITTLE DEEPER:
and dynamic approaches. The static These recent analyses of basement fracturing W E
characterization involves determining at Borholla-Changpang reveal a clear
in-situ fracture location, density and ori- variation of fracture orientation with depth.
entation using a range of wellbore imag- The diagrams illustrate dominant strike
ing tools, petrophysical anomalies, core orientations, at various depths, derived from
and drilling information. The dynamic FMS image analysis.
Depth 2
approach makes use of acoustic data S
(SDT- and LSS-derived Stoneley) for rela- N
tive fracture conductivity at the well-
bore. This can be verified with produc-
tion logs from flow and injection tests.
W E

Depth 3
S

Number 14, 1993. 33


Egypt’s basement bonus
The fractured granite reservoirs of the
Gulf of Suez, Egypt, are flanked by
porous and permeable reservoir sands
and carbonates. The permeability of the
granite horst blocks and the fact that
they are in hydraulic communication
with these flanking sands and carbon-
ates makes them ideal drainage systems.
The Precambrian-age granites were
probably first fractured in Paleozoic
times as the Suez and Aqaba shear
zones were developing (figure 3.16).
Porosity system studies of these granite
34
Photograph: Courtesy of NASA
Fig. 3.15: Shear fractures, which affect
rocks at the southern end of the Gulf of
Suez, have produced a number of large
reservoirs indicate the development of granite reservoirs. In some fields,
secondary porosity along fractures and a granites produce more hydrocarbon
pervasive leaching of feldspars. The than adjacent sandstone or carbonate
basement complex was later covered by reservoir rocks.
the Nubia sand sequence and a thick
succession of Cretaceous sediments,
prior to the start of rifting in the Gulf of
Suez (figure 3.17).
Middle East Well Evaluation Review
 Fracture spacing
Gulf of Sinai
data also shows varia-
Suez Gulf of tions which are useful
Aqaba in predicting expected
ranges of values
throughout the reser- G
ra
be
voir. However, this kind n

nd
of data does not give

a tre
Su
absolute values at

ez

Aqab
any location. Data

tre
60/70°
R ift g r a

nd
sets from the frac- ben
Eastern
tured granite reser-
Desert voirs of the Gulf of
Suez show a partic-
ularly interesting

Su
Spre

ez
variation of fracture a b

tre
aperture with depth.
adim

nd
Fig. 3.16 : SPLIT The fracture data
g ax

PERSONALITY: Fractures in the revealed up to 4000 frac- Fig. 3.17: PULLING FRACTURES APART:
tures in each 1000 ft. Frac- Fractures often occur in situations where
is

Zeit Bay area of the Gulf of Suez

show two dominant orientations - one
parallel to the Suez spreading axis and the
other parallel to the Gulf of Aqaba axis.
Fractal fractures
Statistical analysis of fracture data in the
Gulf of Suez indicates a fractal or fre-
quency relationship between the com-
puted fracture apertures. Figure 3.18
shows mean aperture of fractures in a
section of basement granite reservoir.
The increase in aperture towards the
top of the granite and the variation of
aperture at all vertical scales are particu-
larly important features.
Mean aperture (1:1000) Mean aperture (1:200)
x400 m
x400 m
x450 m
x600 m
x500 m
x800 m
tures were selected by hand and a
computer program was then used to cal-
culate mean aperture and mean
hydraulic aperture for each one. Fracture
density and porosity were then calcu-
lated using an averaging technique with a
fixed sample rate.
Several wells were analyzed in this
manner, with the results showing a simi-
lar range of porosity values but different
fracture densities and average aperture
values for each well. This variation in
average aperture was further corrobo-
rated by production results which
showed that density of fracturing and
width of fracture aperture were the
main criteria for initial production
(Taleb et al EGPC 1990).
Mean aperture (1:40) Mean aperture (1:5)
x420 m
x430 m
x430 m
x432m
x440 m
x434m
there is a shear component of extension
(a).This is the case in Yemen’s Marib
Graben and Egypt’s Gulf of Suez where
shear movements produce non-
perpendicular fractures. The simple, ‘no-
shear’ model (b) is unusual. In this model
fracture orientations are parallel and
perpendicular to the rift.
Fracture orientation
A consistent relationship could not be
found between orientation and aperture
for the granite reservoirs. In certain cases,
where a strong, principal, far-field stress is
present in the rocks, the fractures seemed
to open parallel to the principal stress.
This was usually accompanied by drilling-
induced fractures in the section. In one
granite reservoir, flowmeter data
recorded in the barefoot completed sec-
tion of the well, correlated with a pseudo-
flow profile from the cumulative fracture
porosity results. This, however, was only
after the drilling-induced fracture porosity
data was added, indicating a contribution
to the flow by the drilling-induced frac-
tures (Taleb et al EGPC 1990).
Production changes (coning of water
and gas) with respect to fracture orienta-
tion have also been noted in the granite
reservoirs. In general, there is a higher
vertical permeability when a single set of
high-angle fractures is encountered. Oil
production is better with a mixed orienta-
tion of fractures such as at the margins of
the granite blocks (El Wazeer et al EGPC
1990).

Fig. 3.18: This plot shows the mean aperture of fractures in part of a basement
granite reservoir in Egypt. It should be noted that aperture size increases
towards the top of the granite and that aperture is variable at all vertical scales.
Cyclic variations can be seen at each scale of examination.

Number 14, 1993. 35


X600 ft
X650ft
X700 ft
Fig. 3.19: Stoneley wave data gives clear indications of fractures. This example is taken from a
fractured basement in the Ashrafi Field, Gulf of Suez, Egypt.
Only the Stoneley
Analysis of acoustic Stoneley waves
offers a way of assessing the permeabil-
ity of fractures and porous beds. The
dispersive Stoneley waves move along
the interface between the borehole and
the formation. While travelling along
the borehole wall, the wave propagates
without much energy loss. However,
the wave decays when it encounters a
permeability change or a break in the
wall.
In a cylindrical plane, such as a
borehole, the detailed borehole geome-
try becomes a very significant factor in
wave propagation. At very short wave-
36
lengths, the borehole’s influence on the
wave is similar to that of a flat surface.
However, at an operating frequency of
approximately 10 kHz, the Stoneley
wavelength is approximately 6 inches,
roughly comparable to the diameter of a
borehole. In these circumstances the
wave decays only slightly as it crosses
the borehole and is referred to as the
Tube Wave. The low-frequency Stoneley
mode of the Dipole Sonic Imager* (DSI)
tool operates under these conditions.
At low frequencies, the tube wave
may be considered as a simple pressure
pulse propagating in a cylindrical bore-
hole. When it intersects a permeable
fracture crossing the borehole, pressure
is released into the fracture. This pres-
sure drop causes an attenuation of the
direct arrival and produces a secondary
(or reflected) Stoneley wave. The
reflected Stoneley wave may be
regarded as being generated by a sec-
ondary source located where the frac-
ture crosses the borehole. The strength
of this secondary source (reflection) is
dependent on the amount of energy lost
by fluid displacement inside the fracture
and, therefore, the permeability of the
fracture (figure 3.19).
The Stoneley waves generated by the
DSI tool have frequencies which have
been selected for maximum sensitivity to
fluid motion. In fast formations the Stone-
ley wave propagates faster than the fluid
slowness. In slow formations the wave is
more strongly coupled to the formation
and propagates at slownesses greater
than that of the shear slowness.
Stoneley attenuation and
permeability
At low frequencies, propagation of the
Stoneley wave actually causes fluid flow.
This makes it an ideal technique for esti-
mating permeability (figure 3.20). The
Stoneley wave can be thought of as a
‘dynamic micro-drill stem test’. Acoustic
pressure in the borehole forces fluid into
the formation. The volume of fluid flow,
and the size of the Stoneley wave attenua-
tion which it causes, is directly controlled
by permeability.
Combining the low-frequency mono-
pole Stoneley with dipole shear from the
DSI has greatly enhanced acoustic wave-
form logging. Low-frequency Stoneley
wave data can now be analyzed in terms
of the formation’s dynamic permeability
response. An additional benefit is that
the high-quality shear data obtained
with the dipole measurement offers a
way to account for non-permeability
effects recorded in the Stoneley wave
velocity. This is possible even in very
slow formations and allows accuracy far
beyond that achieved by standard sonic
tools.
Fractures derived from the DSI analy-
sis can then be compared and combined
with images from the FMS tool to improve
our understanding of the reservoir.
Middle East Well Evaluation Review
Stoneley slowness and
permeability
Stoneley slowness can also be used to
derive permeability measurements.
Although less sensitive to changes in
permeability than attenuation data, slow-
ness can be measured more precisely
and is less sensitive to mudcake and
lithology. Data gathered using this tech-
nique compares well with information
taken from cores and RFT/MDT tool data.
When using Stoneley wave reflec-
tions to detect fractures it is important to
first identify thin shales in the sequence,
as these can create reflections and atten-
uations similar to those from fractures.
Integrated fracture interpretation
One way of improving reservoir models
involves combining FMI/FMS images
with Stoneley permeability profiles and
comparing the estimates from both tech-
niques:
(1) Where the fracture aperture esti-
mates from both techniques are in close
agreement the fractures are likely to be
isolated planar features of large extent
(ie greater than Stoneley wave penetra-
tion away from the borehole).
(2) When fracture aperture measured
on electrical images is larger than frac-
ture aperture from the Stoneley waves it
generally implies that the typical frac-
ture is shorter than the wavelength for
Fig. 3.20: LATE
ARRIVAL: Stoneley
waves arrive later than
the compressional and
shear waves and have
lower frequency and,
usually, higher
amplitude. Stoneley
waves can be thought
of as a 'dynamic micro-
drill stem test' since the Stoneley
volume of fluid flow
and the size of
Time
Stoneley wave
attenuation is directly Shear
controlled by
permeability.
Compressional
Number 14, 1993.
the Stoneley wave (approximately 10 ft
at 500 Hz). They are also likely to
include features such as vugs and shal-
low drilling-induced fractures.
(3) If the fracture aperture recorded
on electrical borehole images is less
than fracture aperture from reflected
Stoneley waves then the borehole frac-
tures are probably connected to a net-
work of fractures which extend far
enough from the borehole to be beyond
the scope of the FMS.
Marked lithology changes, which
cause significant variations in formation
shear modulus (eg the interbedding of
calcite with soft shale) can give measur-
able reflected Stoneley wave responses
but will have little effect on FMS tool
response. The common occurrence of
thin, washed shales below the resolu-
tion limits of most logging calipers, often
causes false increases in recorded
Stoneley permeability. However, their
bedding geometry, seen with borehole
imaging techniques, indicates that they
are not permeable fractures.
From this it should be clear that no
single technique should be employed for
fracture evaluation. Borehole imaging,
high-resolution calipers, lithology indica-
tors and Stoneley wave data should be
used in conjunction in order to discrimi-
nate against environmental effects and to
arrive at a reliable interpretation of frac-
ture properties.
R
Compressional
Stoneley
Shear
Fracture
Sounding out
Bombay High
An Array Sonic (SDT*) tool was evalu-
ated on well BH-19 in the Bombay High
Field. The SDT has a single, low-fre-
quency transmitter with an array of
eight receivers spread 6 inches apart.
This configuration allows a better deter-
mination of the Stoneley event than data
gathered using the Long Spaced Sonic
(LSS*) tool.
This technique was used at two verti-
cal resolutions. The first, based on the
stacked Stoneley energies over the
transmitter-receiver spacings offers reso-
lution of approximately 10 ft and can be
used to evaluate gross fractured inter-
vals. Resolution of approximately 2 ft
can be achieved by examining differen-
tial Stoneley energies between selected
adjacent receivers.
Since most of the wells logged in the
basement have only LSS data, the SDT
Stoneley technique was adapted for LSS
waveform processing. Further work
resulted in the development of a Stone-
ley Permeability Index from different
energies. Results on BH-36 compared
well with the actual flow entries in the
production logs.
The Stoneley signal from the LSS
waveforms has less energy than the SDT
waveforms and contains broad-spectrum
pseudo-Rayleigh guided waves. At first,
these effects appeared to completely
mask the permeability variations. How-
ever, the introduction of a select band-
pass filter allowed the Stoneley waves to
be isolated and interpreted.
37
Fig. 3.21: OPEN OR
SHUT?: The
Appraising apertures
apparent reduction
in aperture at the Computer modelling has shown that the
top of this section width of a fracture is proportional to the
(left) is due to the fracture conductivity on an electrical
reduced volume of
x000 ft x000 ft
borehole image (Luthi and Souhaite,
water in the oil 1990). Fracture character depends pri-
zone. The plot on marily on the drilling fluid which
the right shows the invades the crack.
same apertures after
correction.
The resistivity of the drilling mud
within the borehole should be mea-
sured directly and accurately. Miscalcu-
lations can occur due to changes in the
x200 ft x200 ft conductivity of the fracture-filling fluid
or because of conductive material along
the fracture plane. Overall, the data cal-
culated in the modelling tests suggests
that reasonable fracture aperture val-
ues are obtained from water-filled frac-
tures in a wide range of lithologies.
Fracture apertures often increase in
x400 ft x400 ft
size along paleo-unconformities within
Middle East reservoirs in Turkey, Iran
and the United Arab Emirates. In Iran
the karstic fractures in Asmari carbon-
ate reservoir sequences were detected
using both imagery and core. Core
examination showed that most of these
carbonate fractures were filled with
x600 ft x600 ft porous red sandstone.

Fig. 3.22: High fracture

densities are often
associated with faults
but can be caused by a
number of geological
features. Basement
reservoirs are frequently
cut by smaller igneous
bodies such as this dyke
which is characterized
by a higher density of
fracturing than the
surrounding granite.

38 Middle East Well Evaluation Review

 Fig. 3.23: ALTERED
APERTURE: This
output shows a
reduction in
apparent aperture
in the oil zones of a
carbonate reservoir
in Turkey (left
track). Test results
x500 ft x500 ft confirmed the oil-
water contact and
correction of the
aperture data was
performed by
increasing the Rma
in the FracView*
program. The
corrected results are
shown in the right
track.
x600 ft x600 ft

x700 ft x700 ft

FracView oil-
water contact

Fig. 3.24: Aperture uncertainty in highly

x800 ft x800 ft
fractured core.

The variation of the apertures of base-
ment fractures in Egypt's Gulf of Suez
shows an unusual cyclic nature. Each
sequence of downward increasing aper-
ture widths is followed by another cycle
with a similar range of aperture varia-
tion. Even more surprising is the fact that
these cycles have a fractal nature. Each
cycle is composed of smaller cycles of
downward increasing apertures which
are in turn composed of still smaller
cyclic variations of aperture.
The presence of hydrocarbons within
reservoir fractures can lower apparent
fracture apertures in some wells (figure
3.21). This shift is due to a reduction in
the volume of water within the fracture
which is not taken into account by com-
puter software during the initial aperture
calculation. However, the location of the
shift in apparent aperture seems to
occur at the oil-water contact within the
fracture network even in cases where
there is little to no oil within the sur-
rounding rock matrix. In these cases,
reservoir fracture porosity can be cor-
rected empirically. Careful analysis of
aperture variation indicates whether oil
or water will be tested or produced.
The correction to the data is found by
dividing the input mud resistivity (Rm)
by the average calculated water satura-
tion (Sw) from standard log interpreta-
tion. This gives an estimate of the correct
input apparent mud resistivity (R ma )
which can be used in the workstation
fracture analysis program (figure 3.23).
Production information
Massive mud losses are generally
caused by open fractures being encoun-
tered or created during drilling. Some
reservoirs in the Middle East have even
been discovered as a result of severe
mud losses. In the Ain Zalah Field in
Iraq, the mud losses typically correlated
with the productive potential of devel-
opment wells. Since mud losses can
result from either induced or natural
fractures, it is recommended that bore-
hole imagery be used to analyze the
nature of fracturing and the geometry of
the fracture system. Careful monitoring
of mud losses may reveal intervals
which ought to be logged that might not
otherwise be scheduled for logging.
However, during normal monitoring
conditions only the largest fractures will
be detected, and mud losses are not
necessarily related to fracturing.

Number 14, 1993. 39

High hopes in the
horizontal
The growing use of highly deviated wells The productivity of horizontal wells orientations because of the greater num-
has thrown new light on the distribution can be optimized by drilling the well in ber of fractures encountered in highly
of fracture planes in reservoirs. There the most effective orientation - parallel deviated wells. This has shown that
are obvious limitations in using a vertical to the major fractures. Horizontal drilling while fracture spacing may be fairly con-
wellbore as a method for sampling the appears to be most beneficial when sistent in some fields it is highly variable
density of vertical or sub-vertical frac- there is little communication between in others, with the greatest single cause
tures and fracture spacing. A well’s ori- fracture systems, for example when of variation being faulting.
entation to the major fracture sets can be there is a single fracture orientation
controlled to avoid or to intersect the and/or when the fractures are not highly Fracture spacing
maximum number of fractures, depend- interconnected.
ing on the well’s intended role in any Borehole imagery in horizontal wells Fracture spacing has been one of the
given field. (Middle East Well Evaluation (figures 3.25 and 3.26) has provided most elusive parameters for reservoir
Review Number 8, 1990 Putting a bit direct evidence of variations in fracture modellers. This is mainly because of the
aside.) small number of fractures intersected by
vertical wells. As petroleum companies
Fig. 3.25: come to appreciate the improvement in
HORIZONTAL fracture detection and accept that frac-
HEADACHE 1: ture systems can be modelled, it is cer-
The three vertical tain that highly deviated development
fractures shown wells will be used to increase the statisti-
in this block all cal sampling of fracture spacing.
have very The information will also improve
different our understanding of fracture-related
appearances on
production effects (figure 3.27). In 1987
the FMS image
(immediately Nolan-Hoeksema and Howard (AAPG
below) of this Bulletin) suggested a method for com-
horizontal well. puting optimal drilling direction in reser-
The actual voirs where there are a number of frac-
orientation of the ture sets or populations.
fractures with Numerous outcrop studies have
respect to the revealed a statistical relationship
0°(360°) Top wellbore can be between fracture spacing and thin beds.
determined from
This has led many people to expect a
the shape of the
traces. similar relationship in the subsurface but
180° Bottom examination of imagery and core sug-
gests that such a relationship is not as
common as occurs in outcrops. This may
0° Top be because uplift and overburden ero-
Horizontal FMS image sion cause stress relief and tensional
breaking of brittle beds which deform
along ductile bedding planes.
Fig. 3.26:
HORIZONTAL
HEADACHE 2:
In this case, three
dipping fractures
cross the
3-D block borehole. Once
again the
orientations can
Top
be determined
from their trace
on an FMS image
(below).
Bottom Explorationists
accustomed to
examining
fracture data in
0°/360° vertical wells
Top
often find the
fracture traces in
180° horizontal wells
Bottom difficult to
visualize.

0° Top
Horizontal FMS image

40 Middle East Well Evaluation Review

 400 m 500 m 600 m Fig. 3.27: WATER IN
THE WELL: Vertical
fractures are very
Potential water entry zones important in
horizontal wells. The
first example (a)
shows a sub-
horizontal well
passing through a
karst (deeply
weathered)
unconformity surface
with high aperture
zones which are
potential water-
producing intervals.
Section (b) illustrates
a case where high
fracture densities
associated with faults
led to severe water
(a) production problems.

Oil-water contact
(b)
Shale

Improved interconnectivity -
hydraulic fracturing
The interconnectivity of a fracture sys-
tem can be further increased hydrauli-
cally by creating deeply penetrating
fracs. Research by Schlumberger Dowell
has led to the development of tech-
niques designed to obtain minimum
leak-off and deeper penetrating acid fracs
(>200 ft).
The intervals most suitable for ‘fracing’
and the best methods for fracture propaga-
tion within a reservoir have been discov-
ered by modelling the fracture process.
Tests of a new fluid-loss-controlled acid
fracturing system indicate that it is possi-
ble to connect more natural fractures to
the borehole.
Information concerning the vertical dis-
tribution of natural fractures is important
when planning a hydraulic fracture job.
Even the recognition of induced fractures
and their orientation from imagery is valu-
able for frac planning. In some cases, it is
not possible to generate fracs from perfora-
tions that are not parallel or subparallel to
the principal horizontal stress. Moreover,
fractures induced from perforations which
are not parallel to the principal horizontal
stress have a tendency to lose frac effec-
tiveness as a result of later closure.
Recent developments in
fracture detection
New tools have been developed to com-
plement the FMI. Recently, two new imag-
ing devices have been introduced. The
Azimuthal Resistivity Imager* (ARI) tool
appeared in 1992, and the Ultrasonic
Borehole Imager* (UBI) tool entered ser-
vice in 1993.
The main area of overlap between
these tools occurs in fracture and thin-
bed analysis. The techniques they
employ to detect fractures or bed bound-
aries are different and the images pro-
duced by each tool may or may not be
similar. Clearly, the interpreter must
know when and why the images will be
similar and the reasons for any discrep-
ancies when they occur (figure 3.28).
ARI: A second opinion
The ARI tool is a standard dual laterolog
modified by the addition of 12 azimuthal
electrodes. This electrode array provides
a 360°, quantitative and calibrated resis-
tivity image of the formation.
In common with other resistivity tools,
the ARI tool response is strongly affected
by fractures filled with conductive fluids.
The ARI tool provides the best indication
of fracture porosity in partially mineral-
ized fractures and is capable of distin-
guishing shallow, drilling-induced frac-
tures from tectonic fractures.
Tight
Porous
Non-
reservoir
The ARI tool also provides a link
between geological data and traditional
formation evaluation of fractures and
thin beds.
Numerical models have been con-
structed to show that fracture aperture
can be determined from tool response.
Individual fractures, which are more than
8 inches apart, can be distinguished
using the ARI tool. In cases where frac-
ture separation is less than 8 inches, the
computed aperture value will be the sum
of the individual fracture apertures.
Providing estimates of fracture and
structural dip data will be valuable in
wells or zones where FMI/FMS logs will
not be run.
UBI: A third opinion
The UBI tool was developed from the
Ultra Sonic Imager* (USI) tool and is suit-
able for open hole use. The UBI tool has
a revolving transducer which emits ultra-
sonic pulses and receives returning
echoes from the borehole wall. Two-way
transit time and echo amplitude can be
obtained and, with a high sampling rate,
borehole images can be rendered using
either time or amplitude.
If the velocity of ultrasound in the
borehole fluid is measured, the transit

Number 14, 1993. 41

 FMI image ARI image UBI amplitude time can be converted to borehole
radius. Either transit time or radius mea-
surements give a high-resolution, quan-
titative and continuous scan of borehole
shape. If the borehole wall is disrupted,
by fractures or features which are large
compared to the spot size for the tool
(about 0.25 inches for the 250 kHz trans-
ducer), they can be seen in the transit
time values. The main limitation is that
the UBI tool cannot provide quantitative
aperture information. The amplitude
measurement facility available with UBI
offers a more detailed image than transit
x40 m
time, but is difficult to analyze quantita-
tively.
Under normal logging conditions
there are many factors which influence
amplitude. These include the angle of
incidence, scattering by irregularities on
the borehole wall and the acoustic
impedance contrast and loss between
mud and formation.
However, fractures and vugs should
be visible in the amplitude image pro-
vided that losses in mud and those from
high angles of incidence do not elimi-
nate a particular echo. UBI data is ana-
lyzed using the FracView utilities devel-
oped for the FMI/FMS image analysis
techniques. Dip and strike are easily
defined in FracView. However, it should
be recognized that accurate measure-
ments of borehole radii are necessary
for good fracture orientation analysis
and identification of bed boundaries.
Field tests using the UBI tool have
shown a dramatic improvement in bore-
hole acoustic imaging. The introduction
x45 m of advanced tools has helped to remove
most of the low-quality images and
broad black bands that have frustrated
geological interpreters over the years.

Pictures in oils

The UBI tool’s acoustic technology

provides high-quality borehole imaging
in oil-base muds. While this represents
a significant step forward, there are a
few limitations. For example, the very
small fractures which can be resolved
using electrical imagery cannot be
detected by the UBI tool or any other
acoustic imagery tools. This presents a
problem because small fractures are
important for statistical analyses of
Fig. 3.28: A SECOND (AND THIRD) OPINION: In addition to the FMI and ARI tools, the fracture- fracture geometry. Acoustic images
seeking geologist or engineer can now call upon the UBI tool for locating fractures. The UBI tool contain less bedding information than
has been developed for wells with oil-base muds where the FMI tool is not suitable. The ARI
tool is being used to help identify deep open fractures and the fluid they fractures contain. their electrical counterparts. This
bedding data is critical for differentiating
fault types. Therefore, in wells with oil-
base mud the Oil-Base Dipmeter*
(OBDT) tool is sometimes run along
with the UBI tool to ensure that
fractures and their associated structures
can be defined.

42 Middle East Well Evaluation Review

 Amplitude Borehole radius Fig. 3.29: This UBI log Borehole side view
shows a combination of
breakouts and natural 1
Conductivity from 2.94 to 22.67 Depth Conductivity from 4.46 to 4.86 fractures in a Cretaceous Fault plane
(feet) reservoir in the
Middle East.

x200

Borehole plan view

Fig. 3.30: ON THE
SLIDE: During drilling ,
movements can occur
Breakout along fault or fracture
planes which may
damage or even
2 1
destroy production
x210 tubing or casing.
Evidence for this
movement can be seen
on many UBI images.

Natural fractures

Slip sliding away References

Cross-sectional plots of wellbore shapes
(figure 3.29) show the detailed geometry
of various types of borehole damage;
including stress release, borehole break-
outs and key seat wear by drill pipe. A
careful examination of borehole shapes
by Schlumberger Etudes et Production
in Paris has revealed slippage or shear
displacement along fault and/or fracture
planes in response to drilling (figure
3.30). This slippage, which is often
imaged on UBI tool surveys, may result
from reduced friction along the fault
planes, with drilling fluids acting as a
lubricant.
In a few cases, the slippage may be
so large that production tubing or casing
will be damaged. Vincent Maury of Elf
suggests that some wells in southeastern
France may have been completely lost
as a result of this movement. This fault
slippage data may prove to be an impor-
tant source of tectonic information.
R Dennis, V Saxema. and A Rajvanshi (1991):
Fracture Characterization in the Basement of the
Bombay High. ONGC-Schlumberger Wireline
Research Centre Report.
F El Wazeer, F Ismail and E Standen. (1990):
Fracture Geometry and Hydrocarbon Productivity in
the Basement Rocks of the Zeit Bay Field - Gulf of
Suez, Egypt, presented at the 10th EGPC Exploration
and Production Conference, Cairo, Egypt.
O Faivre (1993): Fracture Evaluation from
Quantitative Azimuthal Resistivities, presented at the
SPE Fall Annual Meeting (SPE Paper 26434).
SM Luthi and P Souhaite (1990): Fracture Apertures
from Electrical Borehole Scans. Geophysics (55) pp
821-833.
V Maury and A Etchecopar (1992): Shear Induced by
Oil and Gas Wells Drilling and Production Along
Faults and Discontinuities, presented at the World
Geological Congress, Japan.
BB Neogi, BS Josyulu and KVB Singh (1991):
Hydrocarbon Detection through Seismic Attribute
Parameters of Fractured Basement, Borholla-
Changpang Fields, Assam. ONGC Bulletin.
E Standen (1991): Tips for Analysing Fractures on
Wellbore Images. World Oil (212) pp 99-118.
E Standen, R Nurmi, F El Wazeer and M Ozkanli,
(1993): Quantitative Applications of Wellbore Images
to Reservoir Analysis presented at the Annual
SPWLA Meeting.
HA Taleb, I Helal and E Standen (1990): Granitic
Basement Fracture Study in the Geisum Field - Gulf of
Suez, presented at the 10th EGPC Exploration and
Production Conference, Cairo, Egypt.

Number 14, 1993. 43