Thesis-Petrophysical and Empirical and Theoretical Rock Physics Modeling
Thesis-Petrophysical and Empirical and Theoretical Rock Physics Modeling
Thesis-Petrophysical and Empirical and Theoretical Rock Physics Modeling
2016
Recommended Citation
Yener, Deniz, "PETROPHYSICAL AND EMPIRICAL AND THEORETICAL ROCK PHYSICS MODELING FOR
THE MOKI FORMATION IN TWO FIELDS IN THE TARANAKI BASIN, NEW ZEALAND", Open Access
Master's Thesis, Michigan Technological University, 2016.
https://doi.org/10.37099/mtu.dc.etdr/117
By
Deniz Yener
A THESIS
MASTER OF SCIENCE
In Geophysics
2016
Acknowledgements ............................................................................................................. v
Abstract .............................................................................................................................. vi
1. Introduction .................................................................................................................. 1
2. Geology ........................................................................................................................ 3
3. Methodology................................................................................................................. 4
5. Conclusion .................................................................................................................. 40
6. References .................................................................................................................. 42
7. Appendices ................................................................................................................. 44
iii
7.1 Appendix (I) ........................................................................................................ 44
7.9 Appendix (IX): Lithologs of Moki-1 Well for Whole Depth Range .................. 69
iv
Acknowledgements
Being in the last step of having my Master’s degree, I owe a debt of gratitude to those who
have helped me through my education in Michigan Tech. Especially, I would like to thank
to the person who has never stopped supporting me and has the biggest contribution in my
career. I would not be able to improve my research skills in such extends. Thanks for all of
your efforts and patience Dr. Wayne D. Pennington. I also would like to thank my
committee members, Professor Mir Sadri-Sabet and Dr. Roohollah Askari, for their
register group) in order to conduct this research, so I would like to thank it as well.
Finally, I must express my gratitude to my friends and family. All my friends, especially,
Fatma Sinem Boyaci and Timucin Cakir, I appreciate your support and help for my
blessed to have them. Before finishing my sentences, I would like to thank to my boyfriend,
Mustafa Aggul. I have infinitely many reasons to thank you. I could not finish my work
v
Abstract
determination from empirical relations, rock physics modelling, and fluid substitution for
six wells in the Moki formation of New Zealand’s Taranaki Basin. The project composed
empirical and theoretical approaches, and finally rock-physics modeling for shaley sands
made between results for the different wells, based on their depths and mineralogy.
The first part of this study is presented as a petrophysical analysis, including crossplot
saturation profile, applied to several wells. After petrophysical analysis, we concluded that
the Moki Formation may have well sorted grain size distribution or/and may be less
compacted in the Tui Field than in the Maari Field although the formation is located at
greater depths in the Tui Field. Furthermore, crossplot analysis indicates that these two
fields have different types of clay; this may help explain the different compaction trends.
In the second part of this study, empirical relations and rock physics modelling were used
for sonic response prediction. The P-wave velocity was predicted from the empirical
Wyllie’s Time Average equation and from the Kuster-Toksoz rock physics model. The S-
wave velocity was predicted from the empirical Greenberg-Castagna relation and from the
vi
After P-wave velocity predictions, we concluded that Kuster-Toksoz model works better
in shale than Wyllie’s Time Average equation due to the bound water in shale structure.
This situation causes to have higher density response and to predict higher P-wave
velocities calculated from Wyllie’s Time Average equation in shale. Furthermore, less
compaction in deeper depths causes to have higher porosity response; therefore, predicted
P-wave velocities from Wyllie’s Time Average equation are lower than P-wave velocities
derived from sonic log. The Kuster-Toksoz model optimizes parameters in order to obtain
best fit with the observed P-wave velocity derived from sonic log since the P-wave results
from Kuster-Toksoz fit well for each field. We note that the Greenberg-Castagna model
predicts a lower S-wave velocity than the Kuster-Toksoz model, expect in the highest-
velocity streaks where limestone is likely present. Because there is no recorded an S-wave
In the last part of this study, Gassmann’s equation was modified by using effective porosity
for the sand-shale mixture. In order to estimate sonic velocities for different saturations,
two different fluid substitution approaches were used: Dvorkin et al. (2007) and Simm
vii
1. Introduction
In the discovery of petroleum and natural gas resources, well-logging is an essential tool.
formations. Most of the properties of interest to development team are not measured
directly. However, they are obtained from other measured such properties. Seismic velocity
is one of these properties derived from sonic log. In the event of the unreliable or absent
sonic logs, many approaches have been published manipulating other available data. Log
data, petrophysical features of core data, and rock physics models have been commonly
In this study, sonic response prediction was attempted on the basis of petrophysical
analysis. In order to estimate the sonic response of Moki Formation, six different wells
were selected from two different fields, Maari Field and Tui Field, in Taranaki Basin. These
two fields are 40km away from each other, shown in Figure 3.1. The Moki Formation is
found at different depth ranges in these fields and consists of sandstones with interbedded
Sonic velocities are affected by many factors other than simple porosity and lithology,
including pore type. According to the Kuster-Toksoz (Kuster and Toksoz, 1974) method,
pore types can be determined as penny-shaped ellipsoids of revolution, with aspect ratios
representing ranging from spherical pores to fine-crack pores. Different minerals in the
matrix exhibit different pore aspect ratio behavior. Xu and White (Xu and White, 1995)
1
In order to specify the mineral composition, porosity, and water saturation profile of the
response, empirical relationships and rock physics model, inspired by the mixed-lithology
studies by Xu and White (1995), approaches were computed by using lithology, porosity,
and other physical properties of the rock. Then the results obtained from empirical relations
and rock physics model were compared with the observed P-wave velocity derived from
sonic log and each other. The last part of this study, Gassmann’s (1951) equation was
modified for shaley sand to predict sonic velocities. At the end, we compare the all results
for the Maari Field and the Tui Field based on the differences and similarities of
petrophysical properties.
2
2. Geology
King (1988a, 1988b) introduced “Waiti Group” as a name for most of the Miocene
sedimentary succession in the Taranaki Basin. The Waiti Group is a regressive and
dominantly clastic succession. Due to the tectonism of the Neogene converging plates, a
large volume of rock was deposited in the Taranaki Basin during the Miocene age. The
Messenger, Urenui, and Ariki formations. The Moki formation is the subject of interest in
The Moki formation is a turbidite complex in the southern and central parts of the Taranaki
siltstone and limestone stringers. The grain size of the sandstones are predominantly fine
to very fine. The formation is located between the lower and middle Manganui formations
which form a bathyal succession. Because of the nature of the formation, gamma ray
readings are lower than those of the lower and upper Manganui formations, and readily
In this thesis, Maari Field and Tui Field were chosen for the analysis of the Moki formation.
These two fields are 40km away from each other and the Moki formation is found at
different depth ranges in the two fields, at 1300-1600m in the Maari field and 2700-2800m
3
3. Methodology
In order to determine sonic velocities of Moki Formation and apply fluid substitution on
its log data, it is necessary to know its constituent mineral types, porosity, and fluid types
in pores. These lithological properties are sensitive to locations of wells and depth range
of the formation. In order to observe differences and similarities of such properties, we will
consider two different fields, Maari Field and Tui Field, in Taranaki Basin. From each
field, shown in Figure 3.1, we select three wells with regular suite of logs: gamma ray
(GR), density (RHOB), neutron porosity (NPHI), sonic (DT), photoelectric absorption
(PEF), deep lateralog (LLD), and shallow lateralog (LLS). In this section, conventional log
analysis techniques were applied to the selected wells. Moki-1 well from Maari Field and
Amokura-1 well from Tui Field were chosen to provide examples, and the remaining wells
were presented in the Appendix (VIII). In Figure 3.2 and 3.3, well logs are displayed for
4
Figure 3.2. Litholog for Moki-1 well. The first track shows depth range. The second track
displays gamma ray (red), caliper (black), and spontaneous potential (blue) logs. The third
track represents resistivity logs: deep lateralog (blue), shallow lateralog (red), and micro-
resistivity log (green). The fourth track identifies neutron (green), density (red), sonic
(blue), and photoelectric absorption (black) logs. The orange circle in third track shows the
potential hydrocarbon area.
5
Figure 3.3. Litholog for Amokura-1 well. The first track shows depth range. The second
track displays gamma ray (red), caliper (black), and spontaneous potential (blue) logs. The
third track represents resistivity logs: deep lateralog (blue), shallow lateralog (red), and
micro-resistivity log (green). The fourth track identifies neutron (green), density (red),
sonic (blue), and photoelectric absorption (black) logs.
6
3.1. Petrophysical Analysis
Conventional petrophysical analysis was applied to the selected wells to identify the
mineral composition, porosity, and water saturation of Moki Formation. Shale fraction was
calculated by using linear relationship between index of sand and shale volume from
gamma ray log (Appendix (I)). Sand and shale values were estimated by visual inspection
of the gamma ray log. In order to determine fractions of the other minerals, the gamma ray,
density, neutron, and photoelectric absorption logs were used. Porosity was calculated
using average of the neutron and density logs (Appendix (I)). Neutron porosity had been
use in the calculations. Finally, the deep lateralog, shallow lateralog, neutron log, and
Crossplots provide a good visual description of lithology and help to determine some
required parameters (e.g., GR values for sand and shale, and NPHI, RHOB, PEF values for
shale). Three different crossplots were prepared for the selected wells: NPHI-RHOB, UMA
Typical NPHI-RHOB crossplot, with the color bar showing GR log, is shown in Figure
3.4.a-b. These crossplots indicate high concentration of shale in the formation. Because of
the high content of shale, all points tend to move towards the shale point. Radioactive
elements are more concentrated in shale as compared to other common rock types. Because
gamma ray tools record the natural radioactivity of the formation, higher gamma ray
7
readings are recorded in shale. Additionally, high neutron response is recorded in shale
Moki Formation is located at greater depths in the Tui Field than in the Maari Field. Due
to this difference, it is expected that lower porosity readings should be recorded in the Tui
Filed than in the Maari Filed. However, the Amokura-1 well, shown in Figure 3.4.b,
exhibits higher neutron porosity values than the Moki-1 well, shown in Figure 3.4.a. This
In order to examine this contrast, we prepared a composite log that shows gamma ray,
neutron, density, and sonic logs of these two wells together, shown in Figure 3.5.
Compaction trends in the two wells are somewhat similar up to 1500m. After that depth,
they start to diverge. The trends in both wells do not suggest overpressure in the depth
ranges made consideration (see Figure 3.5), since the trends are generally consistent and
do not reserve. Having higher porosity readings in deeper depths may be explained by well
sorted grain size distribution in the Tui Field. The formation may have larger grain size
distribution in the Maari Filed than in the Tui Filed. Furthermore, this contrast may be
explained by subsidence rates. That is the deeper Moki formation in the Amokura-1 well
seems to be less-compacted than in the Moki-1 well, even though it is 1100m deeper. The
shaliness of the Amokura-1 well, as indicated by the GR log, suggests that the bound water
in the shales has not been released with subsidence, and remains in the formation. It is
possible that the part of the basin containing the Amokura-1well subsided much more
8
quickly than mean the Moki -1 well, retaining its higher porosity to greater depth, but an
analysis of sedimentary and diagenetic history is beyond the scope of this study.
Figure 3.4. Neutron-Density Crossplots of Moki-1 well (a) and Amokura-1 well (b)
indicate that the Moki Formation consists of considerable amount of shale. The color bar
shows the gamma ray log. Each data point refers to different depth in the Moki Formation.
The orange circles in crossplots indicate “pure” shale point locations.
9
Figure 3.5. Composite litholog of Moki-1 well (green) and Amokura-1 well (red). The first
track shows depth range. The second, third, fourth, and fifth tracks display gamma ray,
neutron porosity, density, and sonic logs, respectively.
10
For accurate determination of mineral composition, UMA-DGA (obtained by using NPHI,
RHOB, and PEF logs, Appendix I), shown in Figure 3.6.a-b, and PEF-RHOB, shown in
Figure 3.7.a-b, crossplots have also been prepared. These crossplots indicate that Moki
Formation does not consist of pure sandstone. Because of the considerable amount of shale,
the points on the crossplots tend to locate between sandstone and shale points.
The points in Figure 3.6.a and Figure 3.7.a (for the shallower Maari Filed) are closer to
sandstone point than in Figure 3.6.b and Figure 3.7.b (for the deeper Tui Field). Due to the
higher PEF and UMA values in the Tui Field, we may conclude that these two fields have
different types of clay. While the points in the Maari Field tend to move towards illite, the
points in the Tui Field tend to move towards chlorite; this may help explain the different
compaction trends. These crossplots also indicate the presence of small amount of
11
Figure 3.6. UMA-DGA Crossplots of Moki-1 well (a) and Amokura-1 well (b) indicate
that the points in the crossplots are located between sand and shale points due to the high
content of shale. In addition, these crossplots indicate that the Moki Formation comprises
of small amount of limestone. The color bar shows the gamma ray log.
12
Figure 3.7. PEF-RHOB Crossplots of Moki-1 well (a) and Amokura-1 well (b) indicate
that each well comprises of different types of clay. The color bar shows the gamma ray log.
13
According to these three crossplots, both the Maari Field and the Tui Field present the same
kind of trend and show that the Moki Formation consists of sandstone and shale with a
small amount of limestone. Moreover, there are some distinctions in crossplots because of
the depth, distance, and variations of clay types in these two wells.
identification approach was applied to calculate the fractions of these three minerals. The
gamma ray, density, neutron, and photoelectric absorption logs were used, as explained in
Appendix II with all the parameters and equations. The results for reference wells are
shown in Figure 3.9-10, along with the saturation results obtained in the following section.
3.1.2. Saturation
The dual-water model was used to calculate water saturation profile by using the deep
lateralog, shallow lateralog, neutron, and density logs. The model is explained in Appendix
III, along with the parameters and equations. Formation water resistivity (Rw), cementation
factor (m), and saturation exponent (n) are some of these required parameters for
computing the model. These Archie parameters were obtained from Pickett plot.
For Moki-1 well, 1327m was chosen as water-oil contact as suggested by the character of
the resistivity curves. To determine the Archie parameters, Pickett plot shown in Figure
3.8.a was prepared from the top of the formation at 1303m depth, through the water-oil
14
For Amokura-1 well, there is no distinctive depth to determine water-oil contact in
resistivity curves. Because of the absence of water-oil contact, all depths in Moki
The cementation factor is ranging from 1.8 to 2 for consolidated sandstone (Archie (1942)).
Because of the high cementation in the studied wells, we took higher cementation factor
value than 2. The factor is related to the slope of the curve in the Pickett plot. When we
examine the Pickett plots, there is a small change in the oil saturation curves even if the
After applying dual-water model, saturation profiles were obtained and results were shown
in Figure 3.9 for Moki-1 well and in Figure 3.10 for Amokura-1 well.
15
Figure 3.8. Pickett plots of Moki-1 well (a) and Amokura-1 well (b) help to identify some
required parameters for saturation profile calculations. The color bar shows the gamma ray
log. The Moki-1 well includes only data points from the uppermost portion of the Moki
formation, while the Amokura-1 well displays points from all depths.
16
Figure 3.9. Litholog for Moki-1 well. The first track shows depth range. The second track
displays gamma ray (red), caliper (black), and spontaneous potential (blue) logs. The third
track represents resistivity logs: deep lateralog (blue), shallow lateralog (red), micro-
spherically focused (green), and true resistivity (light blue) derived from formula. The
fourth track identifies neutron (green), density (red), and photoelectric absorption (black)
logs. The last track shows the lithology obtained from a 3-mineral identification approach
and dual-water model. In order to provide high resolution, just 75 meters of the formation
was shown in here. The entire formation was plotted in the Appendix (IX).
17
Figure 3.10 Litholog for Amokura-1 well. The first track shows depth range. The second
track displays gamma ray (red), caliper (black), and spontaneous potential (blue) logs. The
third track represents resistivity logs: deep lateralog (blue), shallow lateralog (red), and
true resistivity (light blue) from formula. The fourth track identifies neutron (green),
density (red), and photoelectric absorption (black) logs. The last track specifies the
lithology found by applying 3-mineral identification approach and dual-water model.
18
3.2. Sonic Response Prediction
In this section, empirical relationships and rock physics model approaches were applied to
the selected wells to predict sonic response: Wyllie’s Time Average equation, Greenberg
and Castagna’s relation, and Kuster - Toksoz rock physics model. The lithology, porosity,
and the fluid content in pores are required parameters for applying these approaches.
Lithology determination was described in the previous section, and the average of the
density and neutron porosities was used for the porosity estimation. These approaches were
applied by assuming water-saturated matrix. Finally, the results were compared with
First, Wyllie’s Time-Average equation, which is one of the empirical relations (Wyllie et
al., 1956), is used to estimate the compressional transit time by using the information of
lithology, porosity, and fluid content in pores (Appendix (IV)) and is compared with the
recorded DT values. Table 3.1 shows the travel times of each mineral used in the
calculations. The 5th tracks in Figure 3.15-16 (presented following the Kuster-Toksoz
discussion) show the P-wave velocity results obtained from Wyllie’s Time Average
equation.
Table 3.1 The matrix travel times of each mineral used in the calculations.
Travel Time
(Slowness) 56 47 64 55.5 189
(usec/ft)
19
The second empirical relation is used to predict shear wave velocity: Greenberg and
Castagna’s relation (Greenberg and Castagna, 1992). This is the relationship between
compressional and shear wave velocities depending on the lithology (Appendix (IV)).
Predicted sonic response was compared only with the rock physics model result, which
will be discussed in the following section, because none of our wells has recorded shear
sonic response. The 6th tracks in figure 3.15-16 show the S-wave velocity results obtained
theoretical models have been developed. One of them is Xu and White (1995) model which
is based on the Gassmann (1951, 1965) model and Kuster-Toksoz model (1974). The
model divides pore space into two parts. One of them is associated with sand and the other
one is associated with clay. These parts have different pore aspect ratio, which is a ratio of
short axis to long axis. The Xu and White model associates the effect of clay content and
In this study, the Xu and White model was applied to selected wells to predict sonic
velocities and compare them with observed P-wave velocity derived from DT log and the
predicted P-wave and S-wave velocities obtained from empirical relations. The model was
explained in Appendix VI with the parameters and equations. Instead of using dry-rock
frame properties, water-saturated rock properties were simply employed for calculations in
here; therefore, the Gassmann model aspect of the Xu and White model was not performed.
20
The optimal aspect ratios for sand and for clay were chosen on a RMS (Root Mean Square)
error basis, shown in Appendix V, by using a range of pore aspect ratios and finding the
minimum RMS error (agreement with observed P-wave velocity). We prepared a contour
map to pick the minimum RMS value that gives the best sonic prediction fit with the
observed P-wave velocity. Figure 3.11.a-b show the contour map for the reference wells,
respectively for Moki-1 well and Amokura-1 well. Both wells exhibit the same sort of
trend. According to Figure 3.11.a-b, the lower RMS errors can be found using the aspect
ratios between 0.15 and 0.2 for sand-related pores and between 0.10 and 0.15 for clay-
related pores, but the values are poorly constrained due in large part to the limited range of
21
Figure 3.11. RMS contour maps of Moki-1 well (a) and Amokura-1 well (b) help to identify
the aspect ratios for sand and clay and those which give the minimum RMS error.
22
In Figure 3.15-16, the first tracks show depth range. The second tracks display the gamma
ray, caliper, and spontaneous potential logs. The third tracks identify resistivity logs: deep
lateralog, shallow lateralog, micro-spherically focused, and true resistivity obtained from
the formula. The fourth tracks represent neutron, density, and photoelectric absorption logs.
The compressional wave velocities that are observed from sonic log and predicted from
Wyllie’s Time Average equation and Kuster-Toksoz model are shown in the fifth tracks.
The shear wave velocities predicted from Greenberg-Castagna equation and Kuster-
Toksoz model are presented in the sixth tracks. The last tracks in these figures display the
lithology log which was attained from 3-mineral identification approach and dual-water
model.
crossplots between observed P-wave velocity derived from sonic log and predicted P-wave
velocity from Wyllie’s Time Average equation and Kuster-Toksoz model for each well,
shown in Figure 3.12-13. According to Figure 3.12.a-b, Kuster-Toksoz model works better
in shale than Wyllie’s Time Average equation. Due to the bound water in shale structure,
higher density response is recorded in shale. This situation causes to predict higher P-wave
velocities calculated from Wyllie’s Time Average equation in shale. For the Amokura-1
well, Figure 3.13.a implies that Wyllie’s Time Average equation works better in shale than
in sandstone. Because of the less compaction or/and well-sorted grain size distribution in
deeper depths, higher porosity response are recorded; therefore, predicted P-wave
velocities from Wyllie’s Time Average equation are lower than P-wave velocities derived
from sonic log. In Kuster-Toksoz model, we optimize parameters in order to obtain best fit
23
with the observed P-wave velocity derived from sonic log and Figure 3.12.b and Figure
While we do not have an S-wave log for comparison, we note that the Greenberg-Castagna
model consistently predicts a lower S-wave velocity than the Kuster-Toksoz model, expect
in the highest-velocity streaks where limestone is likely present. We may conclude that by
taking saturation profile of Moki-1 well into consideration, shown in Figure 3.9, because
we employed water-saturated rock properties for Kuster-Toksoz model, the model predicts
higher S-wave velocity than the Greenberg-Castagna model. For Amokura-1 well, there
are no significant differences between the results from empirical relations and rock physics
model. Both predictions gave similar results for this well. But this comparison is not
24
Figure 3.12. The crossplot of observed P-wave velocity derived from sonic log and
predicted P-wave velocity from Wyllie’s Time Average equation (a) and Kuster-Toksoz
model (b) of Moki-1 well. The color bar shows the gamma ray log.
25
Figure 3.13. The crossplot of observed P-wave velocity derived from sonic log and
predicted P-wave velocity from Wyllie’s Time Average equation (a) and Kuster-Toksoz
model (b) of Amokura-1 well. The color bar shows the gamma ray log.
26
Figure 3.14. The crossplot of predicted S-wave velocity derived from Greenberg-Castagna
relation and predicted S-wave velocity from Kuster-Toksoz model for Moki-1 well (a) and
Amokura-1 well (b). The color bar shows the gamma ray log.
27
Figure 3.15. Sonic wave prediction logs for Moki-1 well. The fifth track shows the P-wave
velocities: observed from sonic log (blue), predicted from Wyllie’ Time Average Equation
(pink) and Kuster-Toksoz model (light blue). The sixth track displays the S-wave
velocities: predicted from Greenberg-Castagna relation (pink) and Kuster-Toksoz model
(light blue). Moki formation is located between 1303 and 1630 meters in Moki-1 well. For
displaying with high resolution, just 75 meters of Moki Formation is shown in here. Whole
formation is plotted in the Appendix (IX).
28
Figure 3.16. Sonic wave prediction logs for Amokura-1 well. The fifth track shows the P-
wave velocities: observed from sonic log (blue), predicted from Wyllie’ Time Average
Equation (pink) and Kuster-Toksoz model (light blue). The sixth track displays the S-wave
velocities: predicted from Greenberg-Castagna relation (pink) and Kuster-Toksoz model
(light blue). Moki formation is located between 2687 and 2797 meters in Amokura-1 well.
29
3.3. Fluid Substitution for Velocity Predictions
One of important benefits of petrophysical analysis is fluid substitution analysis for seismic
velocities, used for the interpretation of both logging and seismic data. It is also a useful
tool for the fluid identification and quantitative reservoir research. Gassmann’s (1951)
equation is the most common method for applying fluid substitution in porous rocks. The
equation is used here to predict sonic velocities of rocks saturated with one pore fluid with
The substitution supposes that the pore fluid attains instantaneous hydraulic equilibrium
throughout the pore space. However, this supposition is not applicable for shaley sediment
because impermeable shale structure contains bound water (Dvorkin et al., 2007). In order
to apply Gassmann’s equation in shaley sediment, the equation can be modified by taking
In this section, for shaley sediment, two different alternative fluid substitution methods
were applied to the selected wells to predict sonic velocities and compare the methods with
each other. In this study, since there are no recorded S-wave logs, we used an approximate
Vp-only fluid substitution method, from Mavko et al. (1995), that uses the compressional
modulus; the equations shown in Appendix VII for the shaley-sand methods described in
the following sections reflect this modification. For the calculations, the water saturation
profile which was computed by applying the dual-water model, explained in Chapter 3.1.2,
was used. To perform the substitution, we replaced the oil with gas for the first exercise
30
3.3.1 Fluid Substitution in Shaley Sediment using Effective Porosity
Dvorkin et al. (2007) developed an alternative fluid substitution method for shaley sand by
modifying Gassmann’s equation. The method introduces porous wet shale as a new mineral
phase in the matrix. According to this method, composite mineral phase consists of porous
wet shale and nonporous minerals. Therefore, elastic modulus are calculated by taking this
new mineral phase, porous wet shale, into account. Furthermore, in the calculations,
effective porosity was used instead of total porosity. The method was explained in
Appendix VII with all parameters and equations. Table 3.2 shows the modulus and density
values of each component used in the calculations. Porous wet shale parameters were
Figure 3.17 shows the effects of shale volume, effective porosity, and effective water
saturation on substituted sonic velocities and densities for two wells by applying Dvorkin
et al’ s approach. In Amokura-1 well, since the oil saturation is not enough to observe the
differences between measured and substituted sonic velocities clearly, the range of 2760–
2780 meters is modeled under the assumption that the original log was made under 50%
oil saturation, and the fluid substitution is applied to this depth range, for demonstration
purposes only.
31
Figure 3.17. Substitution logs obtained from Dvorkin et al.’s approach for Moki-1 well (a)
and Amokura-1 well (b). The first tracks show the depth range. The second, third and fourth
tracks display respectively shale volume, effective porosity, and effective water saturation
profile. The densities are shown in fifth tracks: observed (blue) and substituted assuming
gas (pink) and water (light blue) instead of oil. The sixth tracks display sonic velocities:
observed (blue) and substituted assuming gas (pink) and water (light blue) instead of oil.
As expected, when oil is replaced with gas, substituted P-wave velocity and density
decreased. On the other hand, when oil is replaced with water, substituted P-wave velocity
32
3.3.2 Practical Gassmann Fluid Substitution in Sand/Shale Sequences
The other fluid substitution method examined here for shaley sediment is developed by
Simm (2007). According to his model, there are two important aspects of fluid substitution
on velocity determination: the role of pore space stiffness and the effect of gas saturation.
The pore space stiffness is associated with porosity, elastic modulus of the minerals, and
𝜙
𝐾𝑃𝐻𝐼 =
1 1
𝐾𝑑 − 𝐾0
where 𝐾𝑃𝐻𝐼 is the pore space stiffness, 𝐾𝑑 is the dry rock bulk modulus, 𝐾0 is the mineral
modulus, and 𝜙 is the porosity. By taking the pore space stiffness into account,
Although the paper includes both effective and total porosity approaches, in here, we just
used the effective porosity approach. The method was explained in Appendix VII with all
parameters and equations. Table 3.2 shows the modulus and density values of each
The results of this approach for two wells were presented in Figure 3.18 by displaying the
effects of shale volume, effective porosity, and effective water saturation on substituted
sonic velocities and densities. In Amokura-1 well, since the oil saturation is not enough to
observe the differences between measured and substituted sonic velocities clearly, the
range of 2760-2780 meters is assumed 50% oil saturated and the fluid substitution is
This approach also produced expected results. The P-wave velocity and density decreased
upon replacing oil with gas and they increased with water.
34
3.3.3 Comparison of the Approaches
According to Figure 3.17-18, there are no significant differences between the results
obtained from Dvorkin et al.’s approach (first) and from Simm’s approach (second). Both
substitutions gave similar and expected results for both field. In order to compare these two
approaches, the crossplots of the substituted P-wave velocities derived from Dvorkin et
al.’s approach against from Simm’s approach were prepared for replacing oil with gas for
each reference well. The crossplots are shown in Figure 3.19-20, respectively for Moki-1
Figure 3.19. The crossplot of the substituted P-wave velocities derived from Dvorkin et
al.’s approach against from Simm’s approach of Moki-1 well. The color bar shows the
gamma ray log.
35
Figure 3.20. The crossplot of the substituted P-wave velocities derived from Dvorkin et
al.’s approach against from Simm’s approach of Amokura-1 well. The color bar shows the
gamma ray log.
According to these figures, there are no significant differences between the results from
Dvorkin et al.’s and Simm’s approaches. Both predictions gave similar results and good fit
36
4. Results & Discussion
Petrophysical analysis and rock physics modelling were applied to the Moki formation in
Taranaki Basin. This work consists of conventional petrophysical analysis and sonic
response determination from empirical relations, rock physics modelling, and fluid
substitution.
crossplots reveal that sandstone, shale, and small amount of limestone are presented
together in the Moki formation. Shale is composed of clay and silt. Although these
crossplots gave similar trend for each field, there were some distinctions. When we
compared the crossplot results of each field based on the locations of the wells, depth range
of the formation in the wells, and the points in the crossplots, we concluded that Tui Field
is in an area where the subsidence occurred at faster rates or/and has a well-sorted grain
size distribution; therefore, porosity remains high in the greater depths in there.
Furthermore, different types of clay are presented in these two fields: illite in the Maari
Field and chlorite in the Tui Field. In order to calculate the fraction of the constituting
minerals, conventional 3-mineral identification approach was performed. The last step in
the petrophysical analysis is determination of water saturation profile which was computed
by using dual-water model. In this thesis, the water saturation profile was used as an input
37
Sonic responses were predicted from two different approaches: empirical relations and rock
physics model. The first empirical relation is Wyllie’s Time Average equation which was
used for the prediction of P-wave velocity. The equation was modified by taking into
account shale and calcite as a separate matrix in addition to quartz. The P-wave velocity
was calculated by using transit times of each constituting minerals and their fractions.
Wyllie’s Time Average equation is affected by shaliness. The equation predicts higher P-
wave velocity in the layer containing significant amount of shale. Furthermore, porosity
has an essential role on the equation. Higher porosity records in the Tui Field causes to
predict lower P-wave velocity from the equation. The second empirical relation is
Greenberg-Castagna relation which was used for the prediction of S-wave velocity. In this
study, due to the absence of the recorded S-wave log, the relation results were compared
with only rock-physics model results; therefore, it is not beneficial for lithology
identification. The Kuster-Toksoz rock physics model was used to predict both P-wave and
S-wave velocities of the Moki formation. The model is associated with the effect of clay
content and pore aspect ratios of sand and clay part in the mixture. We tried a range of pore
aspect ratios for sand and clay part, separately. Then we chose the couple of the aspect
ratios that gave the best sonic prediction fit with the observed P-wave velocity based on
the minimum RMS error. The minimum RMS error was found using the aspect ratios
between 0.15 and 0.2 for sand-related pores and between 0.10 and 0.15 for clay-related
pores for all wells. Kuster-Toksoz model optimizes parameters to obtain best fit the
observed P-wave velocity derived from sonic log; therefore, the model works better in
38
Gassmann’s equation was modified by using effective porosity to apply fluid substitution
and estimate sonic velocities of it in shaley sediment. Two different alternative fluid
substitution methods were performed to predict sonic velocities and compared the methods
with each other. Since there is no recorded S-wave logs, an approximate Vp-only fluid
substitution method was used. The method uses the compressional-wave modulus instead
of bulk modulus. Both approaches reveal similar results. When we replaced the
hydrocarbon with gas, as expected, substituted P-wave velocity and density decreased
because gas has lower bulk modulus and density value than oil. On the other hand, when
we replaced the hydrocarbon with water, substituted P-wave velocity and density increased
because water has higher bulk modulus and density value than oil. The results from
Dvorkin et al.’s and Simm’s approaches show similar trends for Moki formation.
39
5. Conclusion
In order to obtain the mineral composition and their fractions and explain the lithology
changes through the depth differences, three different crossplots were prepared for the
indicates that Moki Formation consists of sandstone and shale which reflects combination
of clay and silt with a small amount of limestone in both fields. Secondly, crossplot analysis
reveals that Moki Formation is less compacted or/and well-sorted grain size distribution in
the Tui Field than in the Maari Field although the formation is located at greater depths in
the Tui Field. Thirdly, we may conclude that these two fields have different types of clay
In the Maari Field, while Wyllie’s Time Average equation gives good results in shale-free
sediments, it predicts higher P-wave velocities in shale because of the bound water in shale
structure. In the Tui Field, contrarily, Wyllie’s Time Average equation predicts lower P-
wave velocities than P-wave velocities derived from sonic log because of the higher
porosity readings. On the other hand, Kuster-Toksoz model works better in shale and shale-
free sediments because the model optimizes parameters to obtain best fit the observed P-
wave velocity derived from sonic log; therefore, the model works better in both fields.
In the Maari Field, the Kuster-Toksoz model consistently predicts a higher S-wave velocity
than the Greenberg and Castagna’s relation except in the highest-velocity streaks where
limestone is likely present because water-saturated rock properties were used for Kuster-
Toksoz model. In the Tui Field, there are no significant differences between the results
from empirical relations and rock physics model. Both predictions gave similar results for
40
this well. In this study, because there is no recorded S-wave velocity, the comparisons are
not reliable.
There are no significant differences between the results from Dvorkin et al.’s and Simm’s
approaches for fluid substitution in shaley-sand. These two predictions gave similar results
and good fit with each other for Moki formation in both field.
41
6. References
Adcock, S. (1993), In search of the well tie; what if I don’t have a sonic log?, The leading
Edge, 12(12), 1161-1164, doi: 10.1190/1.1436929
Archie, G.E. (1942), The electrical resistivity log as an aid in determining some reservoir
characteristics, Petroleum Transactions of AIME, 146: 54–62, doi: 10.2118/942054-g
Dvorkin, J., G. Mavko, and B. Gurevich (2007), Fluid substitution in shaley sediment using
effective porosity: Geophysics, 72, O1-O8
Gassmann, F. (1951), Uber die Elastizität Poröser Medien: Vier. der Natur. Gesellschaft in
Zürich 96, 1-23
Hill, R. (1952), The elastic behavior of a crystalline aggregate: Proceedings of the Physical
Society, Section A, Volume 65, Issue 5, pp. 349-354 (1952).
Jordan J. R. and F. L. Campbell (1986), Well Logging II – Electric and Acoustic Logging:
SPE Monograph Series, SPE, Dallas, TX
King, P.R., and G.P. Thrasher (1996), Cretaceous-Cenozzoic geology and petroleum
systems of the Taranaki Basin, New Zealand. Institute of Geological and Nuclear Sciences
monography 13.243pp
Kuster, G. T., and M. N. Toksöz (1974), Velocity and attenuation of seismic waves in two-
phase media: Part1: Theoretical formulation: Geophysics, Geophysics 39, 587-606
42
Mavko G, T. Mukerji, and J. Dvorkin (2003), The rock physics handbook: Tools for
seismic analysis in porous media. Cambridge University Press,
Miller, S. L. M. and R. R. Stewart (1990), Effect of lithology, porosity and shaliness on P-
and S-wave velocities from sonic logs: Canadian Journal of Exploration Geophysics, Vol.
26, Nos 1&1, P. 94-103
Xu, S., and R. E. White (1995), A new velocity model for clay-sand mixtures: Geophysical
Prospecting, 43, no. 1, 91–118
43
7. Appendices
𝑅𝐻𝑂𝑚𝑎 − 𝑅𝐻𝑂𝐵
𝑃𝐻𝐼𝐷 =
𝑅𝐻𝑂𝑚𝑎 − 𝑅𝐻𝑂𝑓
𝑃𝐻𝐼𝐷 + 𝑁𝑃𝐻𝐼
𝑃𝐻𝐼𝐴 =
2
𝑅𝐻𝑂𝐵 + 0.1883
𝑈 = 𝑃𝐸𝐹 ∗
1.0704
𝑈 − 𝑈𝑓 ∗ 𝑃𝐻𝐼𝐴
𝑈𝑀𝐴 =
1 − 𝑃𝐻𝐼𝐴
Where
𝐺𝑅𝑐𝑙 = Gamma ray value for shale free lithology (We used 15 API for Moki-1 well,
𝐺𝑅𝑠ℎ𝑙 = Gamma ray value for 100% shale (We used 75 API for Moki-1 well, 105
44
𝑃𝐻𝐼𝐷 = Density porosity (v/v)
𝑅𝐻𝑂𝑓 = Pore fluid density (We used 1.00 g/cm3)by assuming water in pores)
𝑈𝑓= Pore fluid volumetric factor (We used 0.398 b/ cm3 by assuming water in pores)
45
7.2. Appendix (II)
identification method is used. For the method, the gamma ray log, density log, neutron
𝑈𝑀𝐴 = 𝐹1 ∗ 𝑈1 + 𝐹2 ∗ 𝑈2 + 𝐹3 ∗ 𝑈3
𝐷𝐺𝐴 = 𝐹1 ∗ 𝐷1 + 𝐹2 ∗ 𝐷2 + 𝐹3 ∗ 𝐷3
𝐹1 + 𝐹2 + 𝐹3 = 1
These three equations are mutually solved to determine the fractions of the constituting
volumetric factor, dry grain density, and fractions of the constituting minerals with
respectively.
D3 = Clay’s dry grain density (2.92 g/cm3 for Moki-1 well, 3.05 for Amokura-1 well)
In order to calculate water saturation profile, the dual-water model was performed by
using the deep lateralog, shallow lateralog, neutron log, and density log.
𝑉𝑆𝐻𝐺𝑅 ∗ (1 − 𝑃𝐻𝐼𝐴)
𝐶=
1 − 𝑃𝐻𝐼𝑐𝑙𝑎𝑦
𝑅𝑤𝑎 = 𝑅𝑡 ∗ 𝑃𝐻𝐼𝐴𝑚
𝑃𝐻𝐼𝑐𝑙𝑎𝑦 ∗ 𝐶
𝑆𝑊𝐵 =
𝑃𝐻𝐼𝐴
𝑅𝑤
1−
𝑏 = 𝑆𝑊𝐵 ∗ 𝑅𝑤𝑏
2
𝑅𝑤 1/𝑛
𝑆𝑊 = 𝑏 + 𝑎 ∗ (𝑏 2 + )
𝑅𝑤𝑎
𝑆𝐻 = 1 − 𝑆𝑊
1 − 𝑆𝑊
𝑆𝑊𝑒𝑓𝑓 = 1 − 𝑃𝐻𝐼𝐴 ∗
𝑃𝐻𝐼𝐸
𝑆𝐻𝑒𝑓𝑓 = 1 − 𝑆𝑊𝑒𝑓𝑓
47
Where
Rt = True resistivity (ohm.m) (The formula was obtained from well reports)
48
7.4. Appendix (IV)
304.8
𝑉𝑝 =
𝐷𝑇
49
Greenberg and Castagna’s Relation
In order to determine shear wave velocity derived from compressional wave velocity,
∗ 𝑉𝑙𝑖𝑚𝑒𝑠𝑡𝑜𝑛𝑒
+ (𝑉𝑝2 ∗ 𝑎1 + 𝑉𝑝 ∗ 𝑎2 + 𝑎3 ) ∗ 𝑉𝑠ℎ𝑎𝑙𝑒
a1, a2, a3 = Regression coefficients which vary with lithology, shown in Table 7.1
Table 7.1. The regression coefficients for compressional-wave velocities (in km/s) and
shear-wave velocities (in km/s) in pure porous matrix.
Lithology Sandstone Limestone Shale
a1 0 -0.05508 0
50
7.5. Appendix (V)
For the whole depth range, RMS error was calculated to find minimum one.
51
7.6. Appendix (VI)
Kuster-Toksoz Model (1974)
In order to predict sonic velocities, Kuster-Toksoz model uses the elastic properties,
the porosities, and the aspect ratios of two different mineral components in the matrix.
The elastic properties of the matrix were calculated by using Voigt-Reuss-Hill. The
average porosity of the neutron and density porosities was used in the calculations. The
method.
(4⁄3) 𝜇 (𝐾 ′ − 𝐾) 𝑇
𝐾+ 3𝐾 + 4𝜇
𝐾𝑒𝑓𝑓 =
(𝐾 ′ − 𝐾) 𝑇
1 − 3𝐾 + 4𝜇
(𝜇 ′ − 𝜇) (9𝐾 + 8𝜇) 𝐹
𝜇+
25 (3𝐾 + 4𝜇)
𝜇𝑒𝑓𝑓 =
6 (𝜇 ′ − 𝜇) (𝐾 + 2𝜇) 𝐹
1−
25 𝜇 (3𝐾 + 4𝜇)
𝜇𝑒𝑓𝑓
𝑉𝑠 = √
𝑅𝐻𝑂𝐵
52
Where
𝐹 = ∑ 𝑃𝐻𝐼𝑙 ∗ 𝐹(𝛼𝑙 )
𝑙=𝑠,𝑐
𝑃𝐻𝐼𝐴
𝑃𝐻𝐼𝑠 = 𝑉𝑠 ∗
1 − 𝑃𝐻𝐼𝐴
𝑃𝐻𝐼𝐴
𝑃𝐻𝐼𝑐 = 𝑉𝑐 ∗
1 − 𝑃𝐻𝐼𝐴
53
Where
3𝐹1
𝑇𝑖𝑖𝑗𝑗 (𝛼) =
𝐹2
2 1 𝐹4 ∗ 𝐹5 + 𝐹6 ∗ 𝐹7 − 𝐹8 ∗ 𝐹9
𝐹(𝛼) = + +
𝐹3 𝐹4 𝐹2 ∗ 𝐹4
Where
3 3 5 4
𝐹1 = 1 + 𝐴 [ (𝑔 + 𝜃) − 𝑅 ( 𝑔 + 𝜃 − )]
2 2 2 3
3 𝑅
𝐹2 = 1 + 𝐴 [1 + (𝑔 + 𝜃) − (3𝑔 + 5𝜃)] + 𝐵(3 − 4𝑅)
2 2
𝐴
+ (𝐴 + 3𝐵)(3 − 4𝑅)[𝑔 + 𝜃 − 𝑅(𝑔 − 𝜃 + 2𝜃 2 )]
2
𝐴 1 + 𝛼2
𝐹3 = 1 + [𝑅(2 − 𝜃) − ∗ 𝑔 ∗ (𝑅 − 1)]
2 𝛼2
54
𝐴
𝐹4 = 1 + [3𝜃 + 𝑔 − 𝑅(𝑔 − 𝜃)]
4
4
𝐹5 = 𝐴 [𝑅 (𝑔 + 𝜃 − ) − 𝑔] + 𝐵𝜃(3 − 4𝑅)
3
𝐴
𝐹7 = 2 + [9𝜃 + 3𝑔 − 𝑅(5𝜃 + 3𝑔)] + 𝐵𝜃(3 − 4𝑅)
4
𝑔 𝜃
𝐹8 = 𝐴 [1 − 2𝑅 + (𝑅 − 1) + (5𝑅 − 3)] + 𝐵(1 − 𝜃)(3 − 4𝑅)
2 2
𝜇′
𝐴= −1
𝜇
1 𝐾 ′ 𝜇′
𝐵= ( − )
3 𝐾 𝜇
3𝜇
𝑅=
3𝐾 + 4𝜇
𝛼2
𝑔= (3𝜃 − 2)
1 − 𝛼2
𝛼
𝜃= 3 [cos−1(𝛼) − 𝛼 √1 − 𝛼 2 ]
(1 − 𝛼 2 ) ⁄2
55
7.7. Appendix (VII): Joint work with Karaman (2016)
𝑉𝑝 = √𝑀𝑠𝑎𝑡 ⁄𝑅𝐻𝑂𝑠𝑎𝑡
Where
𝑀𝑤𝑒𝑡 = Compressional moduli of the composite saturated with one fluid, obtained from
𝑀𝑠𝑎𝑡 = Compressional moduli of the composite saturated with another fluid (GPa)
𝐾𝑓 = Bulk moduli of the fluid mixture, obtained by using saturation profile, explained
below (GPa)
𝐾𝑓𝑒 = Bulk moduli of the new fluid mixture, explained below (GPa)
56
𝑃𝐻𝐼𝑒𝑓𝑓 = Effective porosity calculated in Appendix III (v/v)
𝑅𝐻𝑂𝑠𝑎𝑡 = Bulk density of the composite saturated with another fluid (g/cm3)
𝑅𝐻𝑂𝑓1 = Density of the fluid mixture, obtained by using saturation profile, explained
below (g/cm3)
𝑅𝐻𝑂𝑓𝑒 = Density of the new fluid mixture, obtained by using saturation profile,
Where
2 2 −1
1
𝐾𝑠𝑒 = ∗ [∑ 𝑓𝑠𝑖 𝐾𝑠𝑖 + (∑ 𝑓𝑠𝑖 𝐾𝑠𝑖−1 ) ]
2
𝑖=1 𝑖=1
2 2 −1
1
µ𝑠𝑒 = ∗ [∑ 𝑓𝑠𝑖 µ𝑠𝑖 + (∑ 𝑓𝑠𝑖 µ−1
𝑠𝑖 ) ]
2
𝑖=1 𝑖=1
57
4
𝑀𝑠𝑒 = 𝐾𝑠𝑒 + µ𝑠𝑒
3
𝑓𝑠1 = 1 − 𝑓𝑠2
𝑉𝑆𝐻𝐺𝑅
𝑓𝑠2 =
(1 − 𝑃𝐻𝐼𝑐𝑙𝑎𝑦 + 𝑃𝐻𝐼𝑐𝑙𝑎𝑦 ∗ 𝑉𝑆𝐻𝐺𝑅)
Where
58
Calculations for fluid mixtures
−1
𝐾𝑓𝑒 = [𝑆𝑊𝑒𝑓𝑓 ∗ 𝐾𝑤−1 + (1 − 𝑆𝑊𝑒𝑓𝑓 ) ∗ 𝐾𝑔−1 ]
Where
59
7.1.2. Correction for Fluid Substitution in Shaley Sediment using Effective
Porosity
The saturated compressional moduli equation in Dvorkin et al.’s paper was corrected
course notes of Dr. Wayne Pennington. The correction starts with the following
Pennington’s equation and after seven steps the corrected equation was attained:
(1 − (𝑀𝑑𝑟𝑦𝑒 /𝑀𝑠𝑒 )2
𝑀𝑠𝑎𝑡 = 𝑀𝑑𝑟𝑦𝑒 + 2
𝑃𝐻𝐼𝑒𝑓𝑓 /𝐾𝑓𝑒 + (1 − 𝑃𝐻𝐼𝑒𝑓𝑓 )/𝑀𝑠𝑒 − 𝑀𝑑𝑟𝑦𝑒 /𝑀𝑠𝑒
(1 − (𝑀𝑑𝑟𝑦𝑒 /𝑀𝑠𝑒 )2
𝑀𝑠𝑎𝑡 = 𝑀𝑑𝑟𝑦𝑒 + 𝑀𝑠𝑒
𝑃𝐻𝐼𝑒𝑓𝑓 𝑀𝑠𝑒 /𝐾𝑓𝑒 + (1 − 𝑃𝐻𝐼𝑒𝑓𝑓 ) − 𝑀𝑑𝑟𝑦𝑒 /𝑀𝑠𝑒
𝑀𝑠𝑎𝑡
2 2
𝑀𝑑𝑟𝑦𝑒 𝑃𝐻𝐼𝑒𝑓𝑓 𝑀𝑠𝑒 + 𝑀𝑑𝑟𝑦𝑒 𝐾𝑓𝑒 − 𝑀𝑑𝑟𝑦𝑒 𝑃𝐻𝐼𝑒𝑓𝑓 𝐾𝑓𝑒 − 𝑀𝑑𝑟𝑦𝑒 𝐾𝑓𝑒 /𝑀𝑠𝑒 + 𝑀𝑠𝑒 𝐾𝑓𝑒 − 2𝑀𝑑𝑟𝑦𝑒 𝐾𝑓𝑒 + 𝑀𝑑𝑟𝑦𝑒 𝐾𝑓𝑒 /𝑀𝑠𝑒
=
(𝑃𝐻𝐼𝑒𝑓𝑓 ∗ 𝑀𝑠𝑒 ) + (1 − 𝑃𝐻𝐼𝑒𝑓𝑓 ) ∗ 𝐾𝑓𝑒 − 𝑀𝑑𝑟𝑦𝑒 ∗ 𝐾𝑓𝑒 ⁄𝑀𝑠𝑒 )
60
𝑀𝑑𝑟𝑦𝑒 𝑃𝐻𝐼𝑒𝑓𝑓 𝑀𝑠𝑒 − 𝑀𝑑𝑟𝑦𝑒 𝐾𝑓𝑒 − 𝑀𝑑𝑟𝑦𝑒 𝑃𝐻𝐼𝑒𝑓𝑓 𝐾𝑓𝑒 + 𝑀𝑠𝑒 𝐾𝑓𝑒
𝑀𝑠𝑎𝑡 =
(𝑃𝐻𝐼𝑒𝑓𝑓 ∗ 𝑀𝑠𝑒 ) + (1 − 𝑃𝐻𝐼𝑒𝑓𝑓 ) ∗ 𝐾𝑓𝑒 − 𝑀𝑑𝑟𝑦𝑒 ∗ 𝐾𝑓𝑒 ⁄𝑀𝑠𝑒 )
There is sign differences between Dvorkin et al.’s equation and our equation. Below
equations show the differences. First equation belongs to Dvorkin et al. and the second
equation is corrected by using the equation from Seismic Petrophysics course notes of Dr.
Wayne Pennington.
61
7.1.3. Practical Gassmann Fluid Substitution in Sand/Shale Sequences
𝑉𝑝 = √𝑀𝑠𝑎𝑡2 ⁄𝑅𝐻𝑂𝑠𝑎𝑡
𝑃𝐻𝐼𝑒𝑓𝑓 1
𝑀𝑃𝐻𝐼𝑒𝑓𝑓 = −
1 1 1 1
− −
𝑀𝑠𝑎𝑡1 𝑀𝑠𝑒 𝐾𝑓1 𝑀𝑠𝑒
1
𝑀𝑠𝑎𝑡2 =
1 𝑃𝐻𝐼𝑒𝑓𝑓
+ 𝑀𝑠𝑒 ∗ 𝐾𝑓2
𝑀𝑠𝑒
𝑀𝑃𝐻𝐼𝑒𝑓𝑓 + 𝑀 − 𝐾
𝑠𝑒 𝑓2
Where
𝑀𝑠𝑎𝑡1 = Compressional moduli of the composite saturated with one fluid, obtained from
𝑀𝑠𝑎𝑡2 = Compressional moduli of the composite saturated with another fluid (GPa)
𝐾𝑓1 = Bulk moduli of the fluid mixture, obtained by using saturation profile, explained
below (GPa)
𝐾𝑓2 = Bulk moduli of the new fluid mixture, explained below (GPa)
62
𝑅𝐻𝑂𝐵 = Bulk density from logging data (g/cm3)
𝑅𝐻𝑂𝑠𝑎𝑡 = Bulk density of the composite saturated with another fluid (g/cm3)
𝑅𝐻𝑂𝑓1 = Density of the fluid mixture, obtained by using saturation profile, explained
below (g/cm3)
𝑅𝐻𝑂𝑓2 = Density of the new fluid mixture, obtained by using saturation profile,
𝑀𝑠𝑎𝑡1 and fluid mixture calculations are the same with the ‘Fluid Substitution in Shaley
Sediment using Effective Porosity’ method; therefore, they will not rewritten in here.
𝑀𝑠𝑒 is calculated from Reuss-Voigt-Hill by using modulus of shale and quartz, shown
in Table 3.3.
2 2 −1
1
𝐾𝑠𝑒 = ∗ [∑ 𝑓𝑠𝑖 𝐾𝑠𝑖 + (∑ 𝑓𝑠𝑖 𝐾𝑠𝑖−1 ) ]
2
𝑖=1 𝑖=1
2 2 −1
1
µ𝑠𝑒 = ∗ [∑ 𝑓𝑠𝑖 µ𝑠𝑖 + (∑ 𝑓𝑠𝑖 µ−1
𝑠𝑖 ) ]
2
𝑖=1 𝑖=1
4
𝑀𝑠𝑒 = 𝐾𝑠𝑒 + µ𝑠𝑒
3
63
Where
𝑓𝑠2 = The volume fraction of shale obtained from gamma ray log (v/v)
64
7.8. Appendix (VIII)
Kea-1 Well
Figure 7.1.The litholog and crossplots of the Kea-1 well in Maari Field and were used to
identify the mineralogy of the Moki formation in the Taranaki Basin.
65
Maari-1 Well
Figure 7.2.The litholog and crossplots of the Maari-1 well in Maari Field and were used
to identify the mineralogy of the Moki formation in the Taranaki Basin.
66
Tui-1 Well
Figure 7.3.The litholog and crossplots of the Tui-1 well in Tui Field and were used to
identify the mineralogy of the Moki formation in the Taranaki Basin.
67
Pateke-2 Well
Figure 7.4.The litholog and crossplots of the Pateke-2 well in Tui Field and were used to
identify the mineralogy of the Moki formation in the Taranaki Basin.
68
7.9. Appendix (IX)
Figure 7.5. Litholog for Moki-1 well. The first track shows depth range. The second track
displays gamma ray (red), caliper (black), and spontaneous potential (blue) logs. The third
track displays resistivity logs: deep lateralog (blue), shallow lateralog (red), micro-
spherically focused (green), and true resistivity (light blue) from formula. The forth track
displays neutron (green), density (red), and photoelectric absorption (black) logs. The last
track shows the lithology by applying 3-mineral identification approach and dual-water
model.
69
Figure 7.6. Sonic wave prediction logs for Moki-1 well. The fifth track shows the P-wave
velocities: observed from sonic log (blue), predicted from Wyllie’ Time Average Equation
(pink) and from Kuster-Toksoz model (light blue). The sixth track displays the S-wave
velocities: predicted from Greenberg-Castagna relation (pink) and from Kuster-Toksoz
model (light blue).
70