SPE 102289
An integrated approach to obtain reliable permeability profiles from logs in a carbonate
reservoir
P.Balossino, and F.Pampuri ENI E&P, C.Bruni, and K.Ebzhasarova KPO b.v.
Copyright 2006, Society of Petroleum Engineers
This paper was prepared for presentation at the 2006 SPE Russian Oil and Gas Technical
Conference and Exhibition held in Moscow, Russia, 3–6 October 2006.
This paper was selected for presentation by an SPE Program Committee following review of
information contained in an abstract 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 an abstract of not more than
300 words; illustrations may not be copied. The abstract 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
Permeability is one of the key petrophysical parameters in
reservoir evaluation. Information about permeability is
commonly derived from cores and test data, that generally
cover only part of the reservoir section, but can also be derived
from logs, and then extrapolated to uncored intervals and
wells. Two logs provide such information: acoustic and
nuclear magnetic resonance. In the Karachaganak Field, an
approach based on the acoustic tool was preferred because of
the textural characteristics of the vuggy carbonate reservoir.
The approach relies also on the use of image logs to obtain a
detailed description of reservoir rocks texture and discriminate
between rocks with primary interparticle porosity or very
small vugs and lithotypes with multi modal distribution of
pores, enlarged and touching vugs.
More than 900 meters of core have been used to validate the
permeability log derived from the analysis of Stoneley waves
in 25 wells from this field. A correlation between the validated
log derived permeability and the textural facies from image
logs, has allowed the relationship between permeability
variations and the geological framework to be established.
The results have been compared with dynamic data from
production logging through the definition of flow units from
Stoneley wave-derived permeability and porosity log data and
the use of a Stratigraphic Modified Lorenz Plot to identify
possible fluid entry points.
Three main permeability trends have been identified:
• for undolomitised or patchily dolomitised biohermal
deposits.
• for pervasively dolomitised lithologies characterised by
lower mean permeability values.
• for facies characterised by well developed vuggy
porosity with enhanced dissolution phenomena
(touching vugs, microfracturing etc.)
Besides, the occurrence, in some wells, of very high K values
in the biohermal deposits, that are clearly unmatchable by log
derived matrix permeability, has been highlighted by the well
test results. Open fractures have been observed in these wells
on the image logs, thus suggesting that the enhanced
permeability and the improved well performances have to be
related to their presence.
The comparison of log derived permeability and well test
results has given a new perspective in the interpretation of
well test results with respect to the geological framework and
a well defined sequence stratigraphic model. The extension of
this methodology to new wells could improve the knowledge
of the petrophysical characteristics of the reservoir and allow
better prediction of reservoir productivity.
Introduction
This paper presents the interpretative methodology, integrating
wireline and image logs, cores and tests data, that has been
followed to describe vertical trends of the main petrophysical
parameters and correlate them throughout a carbonate
reservoir. The study has been mostly focused on
understanding the complex variations in the porosity-
permeability relationship and made use of the flow unit
concept and partially of the graphical method described by
Gunter et al. (Ref. 1).
Porosity and permeability are key petrophysical parameters for
a detailed reservoir description. The first one can be
confidently obtained from cores and logs. But, especially in
complex reservois, image logs can add valuable information
about texture and porosity types distribution. Permeability is
usually derived from core measurements and from the
interpretation of production tests. As is well known, there are
no wireline logs that directly measure permeability. However
continuous permeability profiles can be extracted from the
analysis of the Stoneley waves acquired by an acoustic tool or
from NMR T2 distribution and the use of the Coates equation.
In this latter case a good correlation must exist between pore
size and pore connectivity.
In the study field, Karachaganak (Western Kazakhstan), the
main reservoir consists of platform carbonates and biohermal
deposits where multiple diagenetic events have created and
destroyed porosity and generated complex porosity-
permeability relationships (Ref. 2). Due to the widespread
presence of vuggy and moldic porosity and the overall textural
characteristics of the reservoir, the approach based on
Stoneley waves attenuation has been considered as more
2 SPE 102289
suitable to derive reliable permeability profiles from logs.
Moreover acoustic waveform data were available together
with image logs in at least 25 wells thus providing a good
coverage in all the zones of the field.
More than 900 meters of cores from 8 different wells were
also available and have been used to ascertain the validity of
the computed permeability. Further discrimination of
unreliable intervals has then been performed by integrating
textural facies data.
Once the permeability and porosity data have been validated,
flow units have been defined by analyzing the vertical
variations in storage capacity and flow capacity on a
Stratigraphic Modified Lorenz Plot (SMLP) and correlated
throughout the field and with the results of production tests.
A good correlation has been established between flow units
and the textural facies from image logs that show vuggy
porosity distribution, being the best flow units associated with
the most heterogeneous facies while a unique correlation
between flow units and lithofacies is often not verified.
Stoneley permeability computation
Stoneley waves are guided waves, travelling along the
formation/borehole interface, and therefore sensitive to the
movements of the fluids within the formation very close to
the borehole. At low frequency the Stoneley mode becomes
the tube wave and propagates as a piston-like compression of
the borehole fluid in the well (Fig.1). When the formation is
porous and permeable, the Stoneley wave pressure at the
borehole wall induces pore fluid motion in the rock, which in
turns affects Stoneley wave propagation (Ref. 3).
Consequently, permeability can be obtained from the effects
on the monopole Stoneley wave propagation (Ref. 4). The
measured parameters do not really describe the formation
permability but the fluid mobility (k/μ).
When the Stoneley waves propagate in a porous and
permeable formation they lose energy moving the fluid in the
formation and this results in an increase of wave attenuation
and a decreasing of wave velocity.
Increasing permeability increases both attenuation and travel
time and the lower the frequency of the monopole the higher is
the attenuation and the dispersion. These two effects will
always occur simultaneously whenever the formation is
permeable. The correspondence, or correlation, between the
two wave effects provides a QC indicator of the Stoneley
permeability results.
The sensitivity analysis defines the importance of a parameter
in affecting the Stoneley wave propagation (Ref. 5). Analysing
the effects on the Stoneley wave of a number of controlling
parameters (borehole fluid velocity, formation horizontal and
vertical shear velocity) has pointed out that the effects of the
borehole fluid and pore fluid are always the largest (mud
slowness and attenuation and pore fluid compressibility.).
The presence of a logging tool also affects the Stoneley waves.
The tool effect is governed by its diameter and rigidity.
As Stoneley waves travel along the formation/borehole
interface, borehole shape and borehole rugosity can strongly
affect the Stoneley response.
When the formation porosity is low (below around 6 pu) the
noise is greater than the signal response and can affect the
mobility results.
All these effects on the monopole Stoneley waveforms can be
analysed and a permeability curve is calculated.
The module utilized in the present study to compute
permeability from Stoneley waveforms, makes use of
equations describing the fluid movements in the porous rocks
according to Biot’s theory (Ref. 6,7). The formation is
assumed to be a two components system: solid and fluid, and
the total fluid displacement into the rocks due to wave
compression, is the sum of the elastic deformation of the solid
phase and the dynamic flow of the fluid in the pores.
Most of the input parameters of the modelling are obtained
from logs others from existing reports or calibrations (Ref. 8).
They include formation properties (slowness, density,
porosity, elastic moduli), mud properties (slowness,
attenuation, mud density) and pore fluid properties (density
and elastic modulus).
The output of the modeling is represented by the fluid mobility
with its associated error at in–situ conditions (Fig.2). Given
the limited depth of investigation of the Stoneley waves (15-
30cm), it essentially represents the mobility of the mud
filtrate. The vertical resolution of the log is related to the array
length and is around 1 meter (i.e. it describes the heterogeneity
of the reservoir at a metric scale).
Textural facies from acoustic images
The acoustic image tool uses a rotating transducer to scan the
borehole wall with ultrasonic pulses. The amplitude of the
reflected echo and the time of arrival are measured and used to
provide images of the borehole wall. Generally the image
based on amplitude variations contains more information and
is used in geological studies to provide a detailed description
of the reservoir.
The borehole shape, the rugosity and the tool centralization
strongly affect the quality of the images and, as a
consequence, the reliability of the subsequent analysis. In the
Karachaganak field fair to good borehole conditions are
generally found in the vertical or slightly deviated recently
drilled wells, while the wells drilled in the 80s and the
horizontal wells frequently show higher rugosity and more
irregular borehole shapes.
To achieve the target of a textural facies classification a two
step processing has been used:
• The first step is an automatic zonation of the log based
on amplitude values
• The second step is a manual subdivision of the first
zonation based on the presence and characteristics of
heterogeneities (such as vugs or porosity patches)
Three main classes have been separated at the end of the first
step of the analysis, characterized by prevailing high
amplitude values (light colours in the images) (class_1),
intermediate amplitude values (orange colour on the images)
(class_2) and low amplitude values (brown colour on the
images) (class_3) respectively.
This subdivision matches quite well the porosity profile from
logs, the light colours being associated with low porosity, the
brown colour with the highest porosities and the orange with
intermediate porosity values.
SPE 102289
The second zonation is based on the identification of
heterogeneities defined as anomalies that do not correlate
across the borehole: they correspond to vugs and more porous
patches that appear as dark spots on the images and, as such,
have low amplitude values. For all the three main classes,
zones can be identified where heterogeneities are absent or
very rare (homogeneous facies) (classes 1.1, 2.1 and 3.1),
more abundant but mostly patchy and of small size at the log
scale (slightly heterogeneous facies) (classes 1.2, 2.2 and 3.2),
or very frequent and from small to very big in size
(heterogeneous to strongly heterogeneous facies) (classes 1.3,
2.3 and 3.3). Subclasses definition has been performed by
setting appropriate cutoffs on the low amplitude curve.
(Fig.3).
As the acoustic images devices provide a full coverage of the
borehole but have no depth of investigation, they map acoustic
impedence variations at the borehole wall and are affected by
the presence of the mud cake. In the Karachaganak wells a
sized calcium carbonate mud has been used during drilling, to
form a thin filter cake at the surface of the exposed formation
and provide a quick leak-off control in front of the most
permeable layers. So the homogeneous textures observed
within the porous facies are likely to be images of the mud
cake formed in correspondence of permeable layers whose
pores and heterogeneities are smaller in size than the
carbonate particles of the mud. The integration with the
petrographic study and the pore network characterisation
highlighted that they mostly correspond to rocks with primary
interparticle porosity or very small (micro) vugs. The slightly
heterogeneous facies has multi-modal distribution of pores and
heterogeneities, some of them being bigger in size than the
carbonate particles of the mud. Finally in the most
heterogeneous facies big heterogeneities, not fully coated by
the mud, prevail. They correspond to vuggy facies with
enlarged pores.
To complete facies classification a new sub-class needs to be
separated within class_1, where images show the presence of
two black stripes roughly 180° apart, and representing
borehole breakouts, i.e. hole enlargements related to the
present in situ stress field. This class had a strong impact in
the Stoneley permeability validation phase allowing the
discrimination of zones where the log response is affected by
borehole rugosity and permeability is likely overestimated.
Core data analysis
Six of the analyzed wells have cored intervals of significant
lenght and two have single cores randomly distributed along
the sequence for a total thickness of over 900 meters.
Available core data include probe permeameter K
measurements and plug K, Φ, grain density data at ambient
conditions.
The probe permeameter measures a very localised value of
absolute permeability commonly on uncleaned cores. The flow
geometry is assumed to be hemispherical if the rock is
homogeneous at the scale of the measurement and the
sampling rate is approximately 1 inch thus providing a high
resolution continuous permeability profile essential to quantify
the local reservoir heterogeneity.
Routine core plug data provide an estimation of porosity and
permeability for 1 inch cylinders of rock generally spaced 1
3
foot along the core. So they do not represent a continuous
measurement of petrophysical properties. Core plug are
generally suitable for evaluating K in relatively homogeneous
reservoirs but may be not frequent enough to capture the
whole variability of petrophysical characteristics in strongly
heterogeneous deposits.
As previously mentioned the log derived permeability
represents the average value of a volume of rock around 1m
high and 0.6 m wide and the log sampling rate is 0.1524 m.
The available measurements thus have different resolutions
and investigate volumes that vary over several order of
magnitude. This must be taken into account when the data are
compared in a reservoir characterisation.
Plug and permeameter data distribution has been analysed
using vertical profiles, histograms and cross-plots (Fig.4). In
most cases the permeability profile shows that although the
permeameter data have a more defined trend while plug data
are more scattered they both move within a similar range.
The cross-plot of miniperm K vs plug K indicates that most of
the samples fall along a diagonal. At low permeability values
data appear much more scattered, and miniperm K is often
higher than plug K, but both measurements have a high degree
of uncertainty. The analysis of histograms allowed the
definition of an average matrix permeability that is clearly
pointed out by minipermeameter data that always show a well
defined mode. Obtained values range between less than 0.1mD
to around 1mD. On the contrary plug histograms are in most
cases characterized by multi modal or scattered distributions.
Core plug data distribution has also been analysed on the K vs
Φ xplots that have been done first for each cored well
separately and then with all the available data together (Fig.5).
In all the wells a main linear trend of increasing matrix
permeability vs increasing effective porosity can be identified.
Above and below this trend two main areas are present; one
characterised by relative low permeability and medium to high
porosity, the other by high to very high permeability and low
to medium porosity. In the first area rock types characterised
by microporosity with connate water and unconnected vugs
are present. The area above the main matrix trend includes, on
the contrary, all the rocks where microfractures, at the thin
section or sample scale, contribute to connect vugs and
increase permeability.
Log derived permeability validation
The results of the processing of the Stoneley waveforms have
then been compared for validation with the probe permeameter
and plug measurements, using histograms, cross-plots and
plots.
The validation phase produced the following conclusions:
• The overall shapes of the Stoneley K curves and the
range of variation show generally a fair to good match
with core values.
• Average matrix permeability values from histogram
modes are quite similar.
• Matrix trends of increasing permeability with
increasing porosity highlighted by core measurements
can be observed also on the log data.
4 SPE 102289
• Zones of mismatch have been locally pointed out and
appear to be related to the presence of borehole
rugosity and breakouts, low total porosity (below 6%)
and a layer thickness less than 1m.
These zones of mismatch could be further filtered out using
the textural facies obtained from image log analysis and, in
particular, that representing breakout zones and the tight
homogeneous or slightly heterogeneous facies (1.1 and 1.2).
The remaining intervals represent zones of reliable log derived
permeability that have been used confidently to extend the
analysis to uncored intervals and wells (Fig. 6).
At the end of the validation phase, 25 permeability profiles
were available to study petrophysical characteristics
distribution and define the K/Φ trends in the reservoir. This
target has been achieved by integrating validated log K values
with all the other available well data and the definition of flow
units (group of rocks with the same K/Φ relationship).
Available geological and engineering data include:
stratigraphy, depositional lithofacies from cores, textural
facies from image logs, CPI computation results (for lithology
and porosity), petrographical analysis and pore network
characterisation and production test data.
Flow units definition
The term flow unit was introduced to define the representative
elementary volume of a reservoir “within which geological
and petrophysical properties that affect fluid flow are
internally consistent and predictably different from properties
of other rock volumes” (Ref. 9). Since then many authors have
used core and logs data to identify zones with similar fluid-
flow characteristsics. Gunter et al. (Ref. 1) proposed the
application of a graphical method “to transform rock-type
based zonations into petrophysically based flow units”. This
method was based on the use of four graphical tools. In the
present study two of these plots have been used in the flow
units identification phase: Stratigraphic Modified Lorenz Plot
(SMLP) and Modified Lorenz Plot (MLP).
The Lorenz plot represents, in a graphical form, the
relationship between transmissivity (permeability thickness
product kh) and storativity (porosity thickness product Φh).
To create a Lorenz plot the log or core data are conventionally
ordered by decreasing K/Φ values. In a homogeneous
reservoir the flow and the storage capacity are in balance and
the data plot along the diagonal. Departure of the obtained
curve from the diagonal is a measure of the reservoir
heterogeneity.
If, on the contrary, data are plotted preserving their
stratigraphic order a Stratigraphic Modified Lorenz Plot is
obtained. Significant inflection points on this plot correspond
to changes in flow capacity or storage capacity that can be
used to define flow units and to identify the flowing intervals
in a reservoir. The interpretation is performed by drawing
straight lines segments on the SMLP. The angular coefficient
of these segments is given by the ratio between flow capacity
and storage capacity. Data having the same ratio define the
same flow unit. Ratio values higher than 1 represent rock
types with a predominant flow capacity over the storage
capacity.
Storage and flow capacity distribution within the reservoir
interval were first analyzed by plotting their cumulative curves
versus depth (Fig.7). In most of the wells a constant increase
in the cumulative storage capacity has been observed within
the biohermal deposits. Exceptions are related to the presence
of crinoidal deposits characterized by very low porosity and
pervasively dolomitised intervals showing a porosity higher
than undolomitized layers.
Where storage capacity is increasing at a constant rate
significant inflexion points in the cumulative flow capacity
indicate the most promising flow intervals.
Flow units have been identified in the cumulative flow
capacity plots first and then verified on the SMLP for each of
the 25 wells for which a continuous porosity and permeability
profile was available (Fig.8). Then an average value of the
ratio between flow capacity and storage capacity was
computed for each flow units and a Modified Lorenz Plot was
created by ordering in each well the flow units by decreasing
ratio values. The MLP provides quick information about flow
units performance and indicate which interval will be more
likely producing if tested. As the plots preserve the
stratigraphic information all the identified flow units were
analyzed with respect to the sequence model defined for the
field.
Finally most of the studied wells have production tests results
available; so the flow unit zonation has then been compared
with PLT data indicating flowing intervals and the
contribution of each layer.
Analysis of the results
When all the flowing zones have been obtained from the
analysis of the SMLP and the MLP and separated according to
the stratigraphic scheme, a set of K/Φ cross plots has been
created for each of the main litho-stratigraphic units: the
layered platform of Late Serpukhovian age, the biohermal
deposits of Upper Visean to Lower Serpukhovian age and the
Devonian to Lower Carboniferous platform deposits.
The analysis was mostly focused on the biohermal deposits
which represent the main target for oil production. Within
these layers the possible flowing zones identified from log
data indicate the presence of three main porosity/permeability
trends (Fig.9).
A first trend appears to be related to undolomitised or patchily
dolomitised biohermal deposits that show log K values
between 0.5 and 5 mD when Φ is ranging between 7 and 13
p.u.. Mean permeability values derived from well tests of a
group of wells mostly located in the western and southern part
of the field plot along this trend.
Biohermal deposits that underwent pervasive dolomitisation
show an increase in total porosity that does not correspond to a
similar increase in permeability; they are therefore
characterized by a different K/Φ relationship and by
petrophysical characteristics poorer than undolomitized layers
showing an average permability of 0.4-0.5mD with a porosity
of around 12 p.u.. Dolomitized bioherms are much less
frequent than undolomitized deposits in the studied wells; this,
coupled with a higher data scattering at low permeability
values, results in a poorly defined trend.
Intervals characterized by a well developed vuggy porosity
SPE 102289
with frequent large and touching vugs, corresponding to
strongly heterogeneous textural facies, plot in the field of
enhanced permeability, and form the flow units with the
highest ratio values. DST data indicate that permeability can
reach some tens of mD and when plotted on the K/Φ xplot
from logs they generally fall close to the best matrix trend
(vuggy facies trend in Fig.9).
Finally, in four of the analyzed wells, DST results highlighted
the presence of very high values of K in the biohermal
deposits, that are clearly unmatchable by log derived matrix
permeability. The enhanced K in these wells has been related
to the presence of fracturing, observed on image logs,
improving the well performances in the volume investigated
by the test.
Conclusions
An integrated approach has been applied to obtain
permeability profiles and K/Φ relationships from logs in 25
wells of the Karachaganak field.
This method is based on the integration of wireline and image
logs, cores and tests data and makes use of the flow unit
concept.
It allowed to obtain reliable matrix K profiles estimation from
Stoneley waves, provided that the results of log processing are
carefully validated against cores and further filtered using
textural facies from image logs, to remove zones of low total
porosity or where breakouts or borehole rugosity are present.
In this approach image logs represent a key tool not only in the
Stoneley derived permeability validation phase but also to
obtain detailed descriptions of reservoir rocks texture. They
help to discriminate between homogeneous facies
corresponding to rocks with primary interparticle porosity or
very small (micro) vugs and heterogeneous and strongly
heterogeneous facies with multi modal distributions of pores,
enlarged and touching vugs.
The distribution of the petrophysical characteristics and the
definition of the K/Φ trends in the reservoir have been studied
by graphically identifying flow units on the Stratigraphic
Modified Lorenz Plot created for each well using log data.
Three main matrix permeability trends have been identified:
• for undolomitised or patchily dolomitised biohermal
deposits characterized by mean K values between 0.5
and 5 mD when Φ is ranging between 7 and 13 p.u..
• for pervasively dolomitised lithologies characterised by
lower mean permeability values.
• for facies characterised by well developed vuggy
porosity with enhanced dissolution phenomena
(touching vugs, microfracturing etc.) corresponding to
strongly heterogeneous textural facies
A good match has been observed between log derived matrix
K trends and DST results in most of studied wells, within
biohermal deposits and vuggy facies. Only in four of them
well tests highlighted the presence of very high values of K in
the biohermal deposits, that are clearly unmatchable by log
derived matrix permeability. In these wells open fractures
have been observed on the image logs, thus suggesting that the
enhanced permeability has to be related to their presence.
Finally the study has provided some key benefits for further
5
developments in the reservoir characterization:
• improvement of the reservoir petrophysical properties
knowledge
• better prediction of the reservoir productivity and more
reliable identification of the most promising intervals
for test
• a new perspective in the interpretation of well test
results with respect to the geological framework and a
well defined sequence stratigraphic model
• improvement in the cut off determination for net pay
definition
References
1. Gunter, G.W.: “Early determination of reservoir flow units
using an integrated petrophysical method” paper SPE 38679
presented at the 1997 SPE Annual Technical Conference and
Exhibition, San Antonio, Oct. 5-8
2 Hendry, J. et al.: “Geological Characterization of upper
Devonian to lower Permian intervals in the Karachaganak
reservoir, Kazakhstan: technical challenges and initial
conclusions”, paper presented at the SEPM-IAS Research
Conference on Permo-Carboniferous Carbonate Platforms and
Reefs, El Paso 15-16 May 2000
3. Rosembaum, J.H.: “ Synthetic micro-seismograms: logging
in porous formation”, Geophysics,39, (1974) 14-32.
4. Williams, D.M. et al.: “The long spaced logging tool”,
paper T presented at the 1984 SPWLA 25th Annual Logging
Symposium, New Orleans, June 10-13.
5. Brie, A. et al.: “Quantitative Formation Permeability
Evaluation from Stoneley Waves”, paper SPE 49131
presented at the 1998 SPE Annual Conference & Exibition,
New Orleans, Sep..
6. Biot, M.A.: “Theory of propagation of elastic waves in a
fluid saturated porous solid, I. Low frequency range”, J.
Acoustic, Soc.Am.28, (1956) 168-178.
7. Biot, M.A.: “Theory of propagation of elastic waves in a
fluid saturated porous solid, II. High frequency range”, J.
Acoustic, Soc.Am.28, (1956) 179-191.
8. Pampuri F. et al. : “Effective Evaluation of Fluid Mobility
from Stoneley Wave Using Full Biot Model Inversion: Two
cases histories”, paper SPE 49132 presented at the 1998 SPE
Annual Conference & Exibition, New Orleans, Sep.
9. Ebanks, W.J. Jr.: “Flow unit concept – Integrated approach
to reservoir description for enginnering projects” (abstract
only) AAPG bulletin v.71 (1987) 551-552
6 SPE 102289

Receivers
Stoneley
waves
travel Fig. 1 Stoneley are guided waves travelling along the
direction
formation/borehole interface and propagate as a piston-like
Stoneley compression of the borehole fluid in the well
waves
displacement

Source Formation fluid

movement

Fig.2 Output of the permeability processing from Stoneley waves: the fluid mobility is displayed in track 3 together with its associated error.
Logs of attenuation (frequency shift) and time delay used in the modeling are presented in track 2
SPE 102289 7

Fig. 3 Subclasses definition based on the presence of heterogeneities highlighted by the low amplitude curve. Low values in the curve
correspond to zones where heterogeneities are absent or very rare (classes 1.1, 2.1 and 3.1); intermediate values indicate facies with small
sized heterogeneities (classes 1.2, 2.2 and 3.2); high values correspond to heterogeneous to strongly heterogeneous facies with touching
vugs and enlarged pores (classes 1.3, 2.3 and 3.3).
8 SPE 102289

A D

Fig.4 Plug and permeameter data distributions have been analyzed using vertical profiles (A) histograms (B and C) and cross-plots (D). In
most cases permeability profiles show that permeameter (dots) and plug (squares) data move within a similar range (A). Miniperm data
histograms always show a well defined mode indicating average matrix permeability (B). Plug histograms are, on the contrary, in most cases
characterized by multi modal or scattered distribution (C). On the cross-plot of miniperm K (Y axis) vs plug K (X axis) most of the samples fall
along the diagonal; at low permeability values data are much more scattered and affected by a higher degree of uncertainty.
SPE 102289 9

Fig. 5 The K vs Φ cross plot of core plug data show the presence of a main linear trend of increasing matrix permeability with increasing
porosity. Above this trend a zone of high to very high K and low to medium porosity characterizes rocks where microfractures connect vugs
and increase permeability. Below the main trend an area with low permeability and medium to high porosity characterizes rock types with
microporosity and unconnected vugs.

Fig.6 Composite plot showing the results of the validation phase. Intervals with reliable log permeability values are shaded in green in track 9
(first from the right). In track 8 log permeability and its associated error is compared with plug K (dots) and mperm averaged data. In track 6
the textural facies from acoustic image logs are displayed; track 7 contains miniperm K raw (dots) and averaged data. Zones with total
porosity from log higher than 6 p.u. are shaded in track 4.
10 SPE 102289

Fig. 7 Plots of cumulative storage and flow capacity vs depth. In most of the wells a constant increase of cumulative storage capacity has
been observed in the biohermal deposits (left). An exception is represented by crinoidal deposits (lower part of the cumulative storage plot)
that are characterized by very low porosity values. Where storage capacity is increasing at a constant rate significant inflexion points in the
cumulative flow capacity indicate the most promising flow intervals (right)

Fig. 8 Flow units interpretation has been performed by drawing straight lines segments on the SMLP. The angular coefficient of these
segments is given by the ratio between flow capacity and storage capacity. Unlike standard SMPLT flow capacity has been plotted on the X
axis of the plot so that the high transmissive intervals can be more directly correlated with flowmeter curve.
SPE 102289 11

Fig. 9 Three main porosity / matrix permeability trends have been identified within the biohermal deposits: one characterizes pervasively
dolomitised lithologies, one is related to undolomitised or patchily dolomitised layers and the third, with the best k/Φ relationship, is found in
facies with a well developed vuggy porosity and frequent large and touching vugs.