SPE 88793

Seismic Integration for Better Modeling of Rock Type Based Reservoir

Characterization : A Field Case Example
Bahar, A., SPE, Kelkar and Associates, Inc., Abdel-Aal, O., SPE, Ghani, A., Silva, F. P., SPE, ADCO, and Kelkar, M.,
SPE, The University of Tulsa

Copyright 2004, Society of Petroleum Engineers Inc.

seismic has clearly indicated the value added by seismic data,
This paper was prepared for presentation at the 11th Abu Dhabi international Petroleum namely improving the simulation result in the areas of poor
Exhibition and Conference held in Abu Dhabi, U.A.E., 10–13 October 2004.
well-coverage. This is important in order to honor the
This paper was selected for presentation by an SPE Program Committee following review of
information contained in a proposal submitted by the author(s). Contents of the paper, as
diagenesis effect that is difficult to model due to a limited
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to number of wells.
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
Introduction
for commercial purposes without the written consent of the Society of Petroleum Engineers is Simplified models for describing reservoirs with complex
prohibited. Permission to reproduce in print is restricted to a proposal of not more than 300
words; illustrations may not be copied. The proposal must contain conspicuous geology are not able to adequately represent the reservoir
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O. heterogeneities and their impact on connectivity patterns and
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.
flow mechanisms. A critical step in the construction of
reservoir simulation models is the description of the
Abstract underlying reservoir geology. This is even more mandatory in
This paper presents the result of a reservoir study where the case of carbonate reservoirs, which is frequently
seismic data have been fully utilized to obtain a better rock- characterized by an intensive diagenetic overprint of the
type based reservoir characterization model. The study was original carbonates. This process frequently includes assessing
conducted for a highly faulted carbonate reservoir in the the complex reservoir internal architecture reflected in the
Middle East. The available 3D Seismic data have been used to internal discontinuities within the lithological units, vertical
determine the structure map, fault and fracture network, and and lateral variation of rock properties and similar
porosity distribution. Additionally, two seismic attributes heterogeneities within the reservoir layers.
(Energy Half Time and Max. Peak Amplitude) have also been The state of the art technology in generating reservoir
used to assist the determination of stationarity regions used for models is to start with proper characterization of the
geostatistical simulation. underlying geology. The result of the characterization should
The reservoir characterization strategy for this study was be used in generating the 3-D model. The challenge that needs
started with the detailed development of a geological to be considered from the start is how the model can replicate
reservoir rock type (RRT) scheme. The RRT scheme was the reality, i.e., what is defined during the characterization
developed based on depositional facies, diagenetic overprints process.
and petrophysical properties, including pore throat size To cope with the geological complexity and assist in
distribution, porosity and permeability. The scheme was then understanding the property distribution in highly
used to constrain the property modeling inside a geological heterogeneous reservoirs, the concept of “Rock Types” was
framework that was built based on the seismic interpretation. introduced and has been widely used as a basis for reservoir
Prior to the development of 3D porosity model, four characterization in recent years.
porosity maps were generated based on the available Typically, once reservoir rock types were determined and
interpreted Acoustic Impedance (AI) and well log data. Each distributed in the 3D volume, the next task is to generate
map represents the differing influence of seismic data, from properties, namely porosity, permeability, and saturation,
singnificant influence to no seismic influence. One map has based on the well data (log and core). Ideally, these properties
been selected to represent the seismic derived porosity based should be generated in such a away that they are consistent
on the degree of correlation between the AI data and well-log with each other. Depending on the availability of the data,
porosity. additional data integration may be done to improve the
The integration of seismic-derived porosity into the 3D property model. The objective of this study is to provide a
simulation model was conducted using Bayessian Update consistent property model by integrating geological based rock
principle inside a conditional simulation technique. In this type, well log/core and seismic data. To satisfy this objective,
technique, 3D porosity distribution honors the underlying RRT seismic data are used to further improve porosity model,
and at the same time its vertical-column average honors which was originally constrained to rock type and well log
seismic-derived porosity. The comparison of the results data only.
between the simulation with seismic and simulation without
2 SPE 88793

Seismic data are also used to construct the structural (toward flanks and saddles), and gradually within a specific
model. We do not consider the structural modeling in this Lithofacies Unit into a lower quality rock type. The
paper; instead, we only concentrate on integration of seismic deterioration of the RRT varies over the field area. The
data for porosity modeling. occurrence of poor quality rock types is more prominent in the
northern part of the field compared with the South, and more
Reservoir Rock Type Characterization in the Western part compared with the East.
Reservoir Rock Types (RRTs) were established for the five
zones of the reservoir. This paper presents the rock type Seismic and Stationarity Regions Determination
charactersization of the most important zone only, i.e., Zone The ultimate goal in performing property modeling is to
B. Detailed description for all other zones is not presented in simulate reservoir properties at inter-well locations. To
this paper. achieve this result, we first need to assume that the data
Zone B is the top porous carbonate rock body in the collected from the wells are representative of the inter-well
Barremian-Aptian, Kharaib Formation. It was deposited as a region. This is a key assumption in any statistical model. This
major regressive phase on a broad and relatively shallow assumption is called an assumption of stationarity. This
Lower Cretaceous shelf of open to slightly restricted subtidal assumption requires that the well data are representative of the
marine environment. Four major Facies Units have been entire reservoir. Unless we can make this assumption, we will
established, each of them subdivided into several Lithofacies not be able to predict the reservoir properties at locations
Units. They extend as layer cake intervals over the entire field where we do not have any data.
area, with variable thickness and reservoir properties. The In this study, we have used preliminary statistical analysis
Upper part is mainly skelletal-peloidal packstone-grainstone, to determine the definition of region of stationarities. That is,
while the Lower is slightly fossiliferous wackstone-mudstone the goal of stationarity analysis is to determine the region
facies. where we can make assumption that the data collected from
The rocks have been subjected to several diagenetic events the wells are representative of the inter-well region. Although
throughout the geological history, e.g. recrystallization, many statistical parameters can be used to conduct a statistical
dissolution, compaction, pressure solution and cementation. analysis, we have restricted our analysis to Mean and
Generally, better porosity and permeability have been Variance. The Mean represents the central tendency of the
observed in the crest of the reservoir while the lower part values; whereas, the Variance represents the spread of the
shows degradation in reservoir properties. Degradation of the data. For geostatistical analysis, only the stationarity with
porosity in the oil-bearing rocks is directly related to subsea respect to Mean and Variance need to be satisfied.
depth, while in the water bearing rocks this trend does not Based on the result of stationarity analysis, it was found
exist. This is mainly due to the preservation of the original that there are 5 stationarity regions that need to be used for
rock properties, by inhibiting or minimizing the diagenesis in property modeling purposes (Figure 4). The boundary of each
the presence of oil. The diagenesis in the water bearing rocks stationarity region was determined based on qualitative as well
affected the original rock properties, and led to the same trend quantitative analysis. During the qualitative analysis, the
of porosity with depth across the water bearing rocks. initial boundary was drawn, whereas the final boundary was
The vertical changes in the reservoir quality are mainly a was drawn using a quantitative approach.
function of depositional facies, texture, and to some extent The initial boundary between the stationarity regions was
diagenesis. The lateral changes across the same Lithofacies drawn with the help of two seismic attributes. Two seismic
Unit, however, are mainly controlled by diagenesis attributes, namely Energy Half Time and Max. Peak
(compaction and pressure solution), which is reflected by the Amplitude were used in the qualitative analysis. Figure 5
existence of many dense intervals, some of them are stylolitic. shows the plot of those two seismic attributes. Based on these
The magnitude of these phenomena is reflected by the relative figures, an initial boundary was drawn to differentiate the 5
abundance of stylolites, which exist in all facies and reservoir regions.
units, with a variable occurrence, their frequency increases The assignment of stationarity region number for all wells
down structure toward flanks and saddles. Six main stylolitic was a simple task for the wells that were located far away
horizons are considered as continuous throughout the field, from the boundary. However, the determination for the wells
with variable thickness and properties; others are patchy. that were located close to the boundary was not easy. In this
Based on sedimentological, petrophysical and fluid flow case, the final decision was made by comparing the Mean
characteristics, ten Reservoir Rock Types (RRTs) have been value (e.g., of porosity) at that well compared to the Mean
established. Types 1 to 7 are grain-dominated lime and Types value of well’s population in each region. The sector that has
8 to 10 are mud dominated lime (Figure 1). Each rock type is closer Mean to the Mean of the boundary’s well retained the
characterized by a distinctive set of reservoir properties ownership of that well.
(Figure 2). The quality of RRT is best for RRT 1 and
deterioriates toward the higher number. Seismic-Derived Porosity
The quality of RRT is best for RRT 1 and deterioriates One of the objectives of this study is to obtain seismic-derived
toward the higher number. porosity map based on the Accoustic Impedance (AI) data.
RRTs vary vertically and horizontally within the zone; This map will be used as one of the constraints during the 3D
generally the good quality rock types exist in the upper part Property Modeling. The AI data is available for two of the five
and are downgraded into lower quality in the lower one. major reservoir units. However, the quality of the data is only
Generally, all RRTs in the crestal area change laterally good for one of the two units. Therefore, seismic-derived
SPE 88793 3

porosity can only be made for one of the units only, i.e., Zone reproduction of well data at the final result. That is, the well
B. data is honored at the well location, which is not possible for
Figure 6 shows the inverted AI map available for Zone B. the cross plot method due to the linear regression application.
This map represents the rock volume within the whole Zone B Additionally, kriging also allows the input from the spatial
with the average thickness of 158 ft from top to bottom. The relationship of the data. Therefore, the spatial relationship of
map clearly shows the presence of low AI (or high porosity) the well data may be honored in the final result. However, the
values around the crestal area and the degradation of reservoir quality of the result depends on the number of available data,
quality toward the flank (East, West and North). However, especially when extrapolation is required. In fact, for the case
compared to the other flank areas, the Southern part shows a with limited data, the result may not be reliable. Figure 10
better region. Considering that this region (i.e., Southern Part) shows the porosity map generated based on purely well log
is not yet fully developed, (the number of wells drilled is still data via kriging. The figure shows smooth changes of porosity
limited), the presence of seismic data may be expected to distribution, which is the characteristic of the kriging method.
improve the quality of reservoir description compared to using Comparing Figure 9 and Figure 10, it can be seen that
the well control only. missing information at the interwell location has significantly
The average porosity at well location based on the well-log changed the overall distribution of porosity. However, certain
data for Zone B is shown in Figure 7. These maps were global features of the map, e.g., high porosity at the crest and
obtained after calculating arithmetic average of well-log low at the North, are retained on the two maps.
porosity from top to bottom of this zone. Comparing Figure 6 One of the most important observations that was revealed
and Figure 7, it can be seen that agreement exists between from the seismic data is the presence of high porous region at
well log data and the AI. For example, higher well-log the southern part (west-flank) of the field (see Figure 9),
porosity data are observed at the crestal area which which is not detected from the current well log data (see
corresponds to the lower AI value, and vice versa, lower well- Figure 10). This result was then proven true from the newly
log porosity at the North, East and West flank regions. For the drilled well. This provides good reason for integrating seismic
Southern area, due to the sparseness of well distribution, the data into the reservoir description process.
presence of better reservoir quality is not yet as clear as In order to overcome the two extreme results presented in
indicated by the seismic data. the above paragraphs, two additional methods were
The correlation coefficient for the relationship between investigated. These two approaches represent the case where
well log porosity and inverted AI at well location is equal to – the influence from both two sources of information, i.e., well
0.81 (R2 = 0.65) (see Figure 8). The negative correlation log porosity data as the primary variable and seismic AI data
indicates that porosity is inversely correlated with the AI data, as the secondary variable, were considered. Both methods are
i.e., higher porosity corresponds to the lower AI and vice based on the kriging technique. The first method is Kriging
versa, as discussed in the previous paragraphs. Note the with External Drift and the second one is the Collocated-
relatively high correlation between the two variables. Cokriging technique. Since kriging is used, the results honor
The deterministic inversion of AI data into seismic-derived porosity at well location and replicate the spatial relationship
porosity can be done in several ways. Three methods were of the well data. Additionally, since seismic data is included,
used to generate three seismic derived porosity maps in this better interwell estimation will be obtained.
study. Additionally, porosity map based on purely well log Kriging with external drift is a variation of simple kriging
data is also generated for comparison purposes. Thus, four method where in addition to honoring the primary variable,
porosity maps are available as the outcome i.e., porosity, the overall trend of the result was captured by
The common way to generate seismic-derived porosity is using the trend information of the second variable, i.e., AI.
the cross plot method, which is based on the linear relationship Note that the influence of the secondary variable is very strong
between porosity and the inverted AI data. This is a simple because the estimation of porosity at an unsampled point is
application where the linear regression is used to convert the influenced by AI data measured at that location. The
AI data into porosity. As in any linear regression application, assumption we make in external drift method is that the
the quality of the result depends on the strong relationship overall trend in the primary variable is captured by secondary
between the two variables. In this approach, the well data may data, with small perturbations in the estimation are added with
not be fully honored, i.e., well data points that do not lie on the the help of surrounding primary data. Therefore, to apply
regression line will not be honored. Thus, once the regression external drift method, relatively high degree of correlation
line is established, the seismic data will dominate the between well log and AI data is required. Figure 11 shows the
determination of the porosity distribution and no further porosity map generated using the Kriging with External Drift
influence of the well data. Figure 9 shows the porosity map of Method. It can be seen that features from both source of data
Zone B generated using the cross-plot method. It can be seen are captured.
that map is the mirror image of the AI map (Figure 6) only The Collocated-Cokriging method is a kriging application
plotted with different scale. where the major influence comes from the primary data (well-
The result presented in the previous paragraph may be log porosity) but “assistance” from secondary variable (AI) is
considered as the one with 100% seismic data influence. The retained in two ways. The first one is from the actual value of
other extreme result may be obtained when there is no, or 0%, the AI at the unsampled location and the second one is through
influence comes from the seismic data. This is the porosity the correlation coefficient between the two variables. The
map from purely well log data that can be generated by technique has been used in the literature with a great success.1
kriging method. The application of kriging method allows the Since the secondary variable is only used at the unsampled
4 SPE 88793

location, i.e., at the collocated point only, the influence of the models. Therefore, best earth models can be created by
AI data for the final result is less compared to the previous reconciling data at seismic and well log scales. In this study,
method (kriging with external drift). Therefore, this method each vertical column of cells in a simulated 3-D model is
should be a preferred method when the degree of correlation constrained to reproduce approximately a seismic-derived
between the two variables is not very high, such as the case of average value. For these purposes a sequential simulation
Zone B data where the R2 value of the linear regression is only procedure, which is based on a Bayesian updating of point
0.65. Note that, as in the previous method, the use of kriging kriging estimates, will be used. This technique was originally
application ensures that the well-log porosity data at well derived by Doyen et al.4
location is honored. Additionally, the spatial relationship of Bayes’ Rule is used to calculate the conditional probability
the primary variable is used to estimate the results. Figure 12 of the additional information, i.e., seismic average porosities.
shows the result of the porosity map generated using the It merely states that the posterior distribution in one cell is
Collocated-Cokriging technique. obtained by taking the product of the likelihood function,
As the summary, the four results shown in the previous controlling the contribution of the seismic average
paragraphs show different levels of influence of seismic AI information, and prior conditional distribution, representing
data in controlling the final porosity map, i.e., from no- the influence of previously simulated and original cell data. In
influence at all (purely well log data via kriging method) to this case,
completely controlled by the AI data (cross-plot method). The ( ) ( ) (
p φi Si α p φi . p Si φi ) (1)
result of the collocated-cokriging technique, instead of kriging where
with external drift, has been selected to represent the seismic-
derived porosity map to be used as one of the constraint during ( )
p φi Si is the posterior probability of porosity at
the 3D-Property Modeling (see next section) based on the location i, given that seismic attribute (average porosity)
degree of correlation between AI and well log data. Note also
that the presence of high porous region at the southern part of ( )
Si is observed; p φi is the prior probability for porosity
the field (see Figure 12), which can be considered as the blind based on hard data at well locations;
test of the well-log data, is estimated very well using the
collocated-cokriging method. ( )
p S i φi is the probability of seismic attribute given
hard data, which can be obtained from the correlation of
Property Model without Seismic Data seismic attribute with well-log data.
The 3D Property Modeling process in this study is What is important to mention here is that, a simulated
performed using geostatistical methodology, specifically value for φ will be obtained by sampling at random from the
simultaneous sequential Gaussian simulation technique.2 This local posterior distribution, and the simulations will be
technique generates reservoir rock type (RRT), porosity, constrained to reproduce the vertical seismic average porosity
permeability and water saturation simultaneously according to within a tolerance ε ; that is,
their dependency level. In this case, the RRT, as the least
dependent variable, is created first, followed by porosity and
∑
S = a φ +ε
j
j j
(2)

then finally by permeability/saturation, as the most dependent where the coefficients aj represent fixed averaging weights,
variables. Using this technique the generated properties the sum is over all cells in the column containing the cell to be
preserve the well data including their variability, spatial simulated, and φj represents the simulated cell values.
distribution, and local variability. In above equation the error at each column is modeled as a
A total of 20 realizations have been generated for this realization of Gaussian white noise with zero mean and
study. Figure 13 and Figure 14 show the top view of rock constant variance. Averaging weights (aj) can be selected
type distribution from one of the realizations for one of the equal for each cell in the column. On the other hand, if the
most heterogeneous and one of the most homogeneous zones, seismic derived vertical averages are believed to be more
respectively. It is shown that even for the most heterogeneous sensitive to rock property values near the top of a layer, one
zone, continuity of the rock type, as it was predicted from can choose weights, which decrease with increasing depth in
geological interpretation point of view, has been established. the layer.
However, considering that the continuity of the rock type is The main steps of porosity simulation, including the
one of uncertain parameters in the simulation process, results Bayessian Update, can be summarized as follows,
from other realizations may give different type connectivity of 1. Select the grid block to be simulated at random.
the rock type. 2. Apply point kriging using all original hard data, and
As previously mentioned, the reservoir property generated previously simulated cell values
with this technique has been conditioned to the rock type 3. Compute the mean and variance of posterior
distribution. Figure 15 shows the example of porosity and distribution after Bayesian updating.4 (Equation 2)
permeability which are consistent with the underlying rock 4. Obtain the simulated value, which is conditioned to
type description. More detail results of this model has been seismic by sampling the posterior probability. As an
previously published in other technical paper.3 additional constraint, the final porosity value must be
within the limit of each rock type. This is to ensure
Integration of Seismic into 3D Property Modeling that consistency between porosity and rock type is
Seismic data provides a dense lateral coverage but poor maintained.
vertical coverage due to their low vertical resolution for earth
SPE 88793 5

Note that the first two steps represent the regular paper SPE 24742, presented at the 67th SPE ATCE
sequential Gaussian simulation using hard data only. The held in Washington, D.C., October 4-7, 1992.
approach described has also been implemented in other nearby 2. Bahar, A. and Kelkar, M. “Journey From Well
field study.5 Logs/Cores to Integrated Geological and
Figure 16 and Figure 17 show comparison of vertical Petrophysical Properties Simulation: A Methodology
average porosity map for the whole Zone B for the case of and Application”, SPE Reservoir Evaluation and
without and with seismic data, respectively. The map of the Engineering 3(5), October 2000.
difference between the two cases is shown Figure 18 and the 3. Silva, F., Ghani, A., Al-Mansoori, A., and Bahar, A.,
histogram of this difference is shown in Figure 19. From the “Rock Type Constrained 3D Reservoir
histogram, it can be seen that the mean of the difference is Characterization and Modeling”, SPE 78504
interval between –0.64 and 0.12. Thus, on the average, the presented at the 10th Abu Dhabi International
maps are the same, which can be seen to occur in the area of Petroleum and Exhibition Conference held in Abu
dense well location. This is not surprising since the seismic Dhabi, U.A.E., 13-16 October 2002.
porosity map was derived with collocated cokriging technique 4. Doyen, M., Psaila, D.E., Den Boer, L.D., and Jans,
that honors the well data, so some influence of seismic data on D., “Reconciling Data at Seismic and Well Log
a coarse scale is already incorporated. However, at the poor Scales in 3D Earth Modeling”, SPE 38698, presented
well coverage, the result shows the influence of seismic data at the 1997 SPE Annual Technical Conference and
in the final distribution. Therefore, the integration of seismic Exhibition held in San Antonio, Texas, 5 – 8 October
data into the modeling process has helped the model in 1997.
distributing the porosity at the poor well coverage. 5. Al-Deeb., et. al., “Fully Integrated 3D-Reservoir
Characterization and Flow Simulation Study: A Field
Conclusions Case Example”, SPE 78510 presented at the 10th Abu
The conclusions of the study can be summarized as follows: Dhabi International Petroleum and Exhibition
1. This paper presents the use of seismic data to improve the Conference held in Abu Dhabi, U.A.E., 13-16
3D property distribution of a rock type based reservoir October 2002.
characterization study. Two applications of seismic data
were presented, namely as the tool to determine the Acknowledgements
boundary of stationarity regions and to improve porosity We wish to thank ADNOC for giving the permission to
distribution publish the result of this field field study and to thank ADCO
2. Reservoir characterization of rock type has been Management for continuous support during the course of the
established for the whole zones in the reservoir resulted in study.
10 different Reservoir Rock Types.
3. RRT vary vertically and horizontally. The vertical
changes are mainly a function of depositional facies,
texture, and to some extent diagenesis. The lateral
changes are mainly controlled by diagenesis.
4. The determination of stationarity region boundaries was RT - 1 RT - 2 RT - 3 RT - 4 RT - 5

done using two seismic attributes, namely Energy Half

Time and Max. Peak Amplitude.
5. Four different porosity maps using four different methods
were derived to represent the average porosity of the RT - 6 RT - 7 RT - 8 RT - 9 RT - 10

whole Zone B. Each map represents different level of

influence of seismic data. Figure 1 Reservoir Rock Type (RRT)
6. Porosity map from collocated-cokriging technique has
been selected to represent seismic-derived porosity. The
selection is based on the level of correlation between well
data and inverted AI data.
7. The integration of seismic-derived porosity into 3D Rock-
type constrained property modeling was done using the
Bayessian Updating principle implemented within the
Sequential Gaussian Simulation system.
8. The use of seismic data has improved porosity
distribution at the poor well coverage area but
maintaining the distribution generated by purely well log
data at the dense well coverage area.

References:
1. Xu, Wenlong, Tran, T.T., Srivastava, R. M., and
Journel, A. G., “Integrating Seismic Data I Reservoir Figure 2 Permeability vs. Porosity for different RRT
Modeling : The Collocated Cokriging Alternative”,
6 SPE 88793

Energy Half Time

Zone B

3
2
1 7

4
5

6
8

10

Figure 3 RRT Distribution in Axial Section of the Field

Figure 2(a)

Max. Peak Amplitude

Zone B

1 : CREST
2 : NORTH
3 : EAST
4 : SOUTH
5 : WEST

Figure 4 Definition of Stationarity Regions

Figure 2(b)
Figure 5 Two Seismic Attributes Used for Drawing the
Initial Boundary of the Stationarity Regions.
SPE 88793 7

Well Porosity vs. AI

25

20

15

10

y = -3E-06x + 48.539
2
5 R = 0.6503

0
6000000 7000000 8000000 9000000 10000000 11000000 12000000
AI, kg/m3*m/s

Figure 8 Correlation between Well Log Porosity and

Inverted Accoustic Impedance (AI) at Well Location.

Figure 6 Accoustic Impedance (AI) Map

High porous
region

Figure 9 Seismic-Derived Porosity Map – Cross Plot

Method

Figure 7 Average Well-Log Porosity at Well Location

8 SPE 88793

Figure 10 Well-Log Based Porosity Map – Ordinary Kriging Figure 12 Seismic and Well-Log Based Porosity Map –
Method Collocated-Cokriging Method

Figure 13 RRT Distribution (Heterogeneous Layer)

Figure 11 Seismic and Well-Log Based Porosity Map –

Kriging with External Drift Method
SPE 88793 9

Figure 14 RRT Distribution (Homogeneous Layer) Figure 17 Average Porosity Distribution for Zone B as the
Result of 3D Property Modeling WITH the use of Seismic
Data

Figure 15 Consistency of RRT, Porosity and Permeability

Figure 18 Distribution of the Difference between Porosity

WITH and WITHOUT Seismic Data.

Figure 16 Average Porosity Distribution for Zone B as the Figure 19 Histogram of the Difference of Porosity WITH
Result of 3D Property Modeling WITHOUT the use of and WITHOUT Seismic Data
Seismic Data