1.1. Geology setting: Barrow subbasin

The field information consists of geophysical logs and seismic information from a study area in deep water to the NW of Australia knowns as the Carnarvon Basin (Figure 1), which is a marine oil and gas-producing basin containing up to 10 km thicknesses of predominantly Mesozoic deltaic siliciclastics (^{Ementon et al., 2004}).

Source: Generated from ^{Gavin (2015)}.

Figure 1 The geographic location of the study area. 

The Barrow Sub-basin is geologically an elongated marine basin trending NNE to SSW, which forms part of the North Carnarvon Basin in Northwest Australia. The Barrow Sub-basin is a deep syncline graben that forms a depocentre of about 10 km of predominantly Mesozoic and Cenozoic sequences, which are flanked to the east and west by shallow faulted terraces containing more than 5 km of Paleozoic strata from Cenozoic (^{Ementon et al., 2004}).

The development of the basin began in the Paleozoic, generating structural changes due to intense opening processes during the Late Triassic to Early Jurassic that occurred between Australia and the Burma Block to the West of India. During the Late Triassic to Early Jurassic, deltaic sand reservoirs of fluvial and coastal origin occurred. Later a transgressive phase of marine clastic sedimentation occurred in the Early to Late Jurassic. Subsequent extension events occurred during the Middle-Late Jurassic and Early Cretaceous. In contrast, for the Late Cretaceous, an inversion process led to compression until the Miocene, creating several structural traps within the Barrow Sub-basin. As a result, the Barrow Delta was prograded northward through the Barrow Sub-basin during the Early Cretaceous. It was followed by a transgressive phase of marine clastic sedimentation until the Middle Cretaceous (^{Ementon et al., 2004}).

The main exploration plays in the Barrow Sub-basin comprise anticlines and fault-limited structures, with the regional seal formed by Early Cretaceous deep marine shales. The accumulations of hydrocarbons are found predominantly in the Lower Cretaceous in siliciclastic rocks of continental origin of the volcanic type; due to this, in the storage rock of the study reservoir, we find the presence of sandstones with the content of Potassium Feldspars as well as a variety of clays among the main ones: kaolinite, smectite, and illite. Most oil and gas accumulations come from Upper Jurassic marine shales (source rock). The main pulse of hydrocarbon generation in the Upper Jurassic source rocks occurred during the Early Cretaceous and continued throughout the Late Cretaceous to the Cenozoic.

1.2. Case study: Stybarrow field

To deploy a field application of the proposed methodology, we used available stacked seismic information of the Stybarrow deepwater field, as well as the information of four wells (Stybarrow-1, Stybarrow-2, Stybarrow-3, and Stybarrow-4), from which the geological formations overlying and underlying the Lower Cretaceous reservoir were selected. The measured depths of wells range from the sea level to 2,100 m and 2,400 m approximately (Figure 2).

Figure 2 The correlation of the wells in the study area, represented from left to right, is the Gamma Ray (GR) log of the Stybarrow-2, Stybarrow-1, Stybarrow-4, and Stybarrow-3 wells. 

1.2.1. Lithology units

A selection of homologated stratigraphic peaks determined from qualitative analysis of the gamma-ray log was made for the four wells used in the proposed methodology (Figure 2). It derived from the little homogeneity in the information available. The available well logs of the Stybarrow-1, Stybarrow-2, Stybarrow-3, and Stybarrow-4 wells in the study area with which the proposed methodology are shown in Table 1.

In Figure 2, the GR log of the four wells in the Stybarrow field is displayed according to their geographic position from SW to NE, Stybarrow-2 (Sty-2), Stybarrow-1 (Sty-1), Stybarrow-4 (Sty-4), and Stybarrow-3 (Sty-3). Also, the reservoir zones penetrated by each well were highlighted with orange shading, highlighting that the Sty-3 no longer cut the reservoir, becoming a delimiting well of the field. As the Sty-1 well has complete information, it was from the results obtained for this well that the results for the other wells and seismic data of the area were calibrated.

1.2.2. Well and seismic inversion data

Herein, we used well data and volumes of inverted elastic parameters provided by Stanford University. The detailed process of seismic inversion modeling is described in ^{Arévalo-López (2017)}. The inversion process by which the volumes of Ip, Is, and RHOB were obtained is called simultaneous impedance inversion, performed using a constrained sparse-spike inversion algorithm based on the optimization of the L1 norm. This algorithm creates an ensemble of elastic models using multiple partial angle stacks of seismic data (^{Areválo-López and Dvorkin, 2017}).

A set of elastic models using multiple seismic partial angle stacks were generated in ^{Arevalo-López (2017)}. To achieve this, Aki-Richards equations were used to compute the reflection coefficients at seismic scale, while at well scale (low frequency), P and S velocities are used for the elastic model. Inversion process usually includes QC on input data, cross correlation between the angle stacks for obtaining time-aligned stacks, seismic-to-well tie, wavelet extraction for each angle gathers, horizon interpretation based on the near-stack amplitude, well- and horizon-based low frequency models of Ip, Is, and density of the earth model, inversion parameter optimization, and quality control of the inversion results.

Inversion parameters were optimized to obtain the best fit between the seismically derived values and the data from Well Sty-1. Subsequently, these parameters were used to obtain the simultaneous impedance inversion for the entire seismic cube. The most critical optimization parameter was the "contrast mismatch" that controls the variance of the elastic parameters between the inversion results and the low-frequency well data (Figure 3).

Figure 3 The sequence used to obtain volumes of elastic properties through simultaneous Seismic Inversion (^{Arévalo-López, 2017}). 

According to ^{Arévalo-López (2017)}, the key to obtaining an acceptable seismic inversion is to match the seismically derived Poisson's ratio with the same parameter calculated from well-log data.

2.1. Petroelastic model

This study extended the 1D methodology to implement self-consistent models SCM proposed by ^{Nicolás-López and Valdiviezo-Mijangos (2016)} to a 3D petroelastic workflow for lithotype interpretation. SCM was applied to obtain rock's elastic effective properties considering a heterogeneous medium of inclusions (minerals, fluids, or organic matter) for different porosity scenarios. The self-consistent method equations introduced by Sabina and Willis (1988) are non-linear; ^{Valdiviezo-Mijangos and Nicolás-López (2014)} solved them with the fixed-point method. When obtaining the solution of the equations, it is assumed that the properties of the rock μ, κ and ρ of a homogeneous system become properties of a heterogeneous system, which are called effective properties μ0,κ0yρ0 .

Self-consistent equations for n inclusions are,

where κ0 is the effective bulk modulus, κn+1 is the bulk modulus of the matrix, κr is the bulk modulus of an inclusion, αr is the volume fraction of inclusions, μ0 is the effective shear modulus, μn+1 is the shear modulus of the matrix. μr is the shear modulus of inclusion, ρ0 is the effective bulk density, ρn+1 is the density of matrix, and ρr is the inclusion density.

The self-consistent equations, (1) to (3), were used to construct the rock physics templates RPT in terms of Mu-Rho μρ vs. Lambda-Rho λρ where λ=κ0+23μ0. These equations consider elastic contributions of rock constituents based on mineral content and pore-filling fluids.

2.2. Brittleness models

This work proposed the integrated analysis of lithology's petroelastic interpretation and brittleness evaluation. Elastic parameters μρ and λρ are convenient for construing 1D-3D lithology models, and E and v are for qualitative brittleness evaluation. The brittleness models defined by Rickman et al. (2008) and worked by ^{Lizcano et al. (2018)} considered the use of rock physics templates to define pay zones guided by mineral fractions of a lithology column. They were used to construct a novel 1D-3D brittleness and petroelastic workflow.

Brittleness index based on Young's modulus (BI_E)

This brittleness index is defined by normalization of Young's modulus, taking as its limits the maximum and minimum values of a sedimentary column. In an (E - v) cross-plot, the results usually generate horizontal straight lines because E only varies. BI_E is calculated with the following equation,

where E is Young's modulus, E_min is the minimum Young's modulus, and E_max is the maximum Young's modulus. For practical applications, higher values of Young's modulus are related to brittle formations and lower to ductile formations.

Brittleness index based on Poisson's ratio (BI_v)

The normalization considers Poisson's ratio from a sedimentary column's maximum and minimum values. Poisson's ratio always has values between 0 and 0.5. This parameter generates vertical straight lines when superimposed on (E - v) cross-plot. It is computed with the following equation,

where v is Poisson's ratio, v_min is the minimum Poisson's ratio, and v_max is the maximum Poisson's ratio. Higher values of Poisson's ratio are always linked to ductile formations and lower to brittle formations.

Average Brittleness index (BA)

As for heuristics, BI_E and BI_v must be considered to improve brittleness evaluation. Therefore, BA is determined via the linear average between the values of BI_E and BI_v. This parameter generates oblique straight lines when they are superimposed on (E - v) cross-plot and with which can be performed brittleness analysis. The average brittleness index is obtained with the following expression,

where BI_E and BI_v were defined in equations (6) and (7).

2.3. Workflow for 1D-3D lithotype interpretation

The proposed workflow for 1D-3D lithotype interpretation is based on rock physics templates RPT which have risen as efficient tools for lithotype interpretation (Nicolas-López et al., 2019). They were constructed similarly described in ^{Nicolás-López and Valdiviezo-Mijangos (2016)}. Therefore, a new workflow for 3D lithotype interpretation was set up. The lithologies were identified from their 1D-3D elastic properties computed at well and field scale. It is worth mentioning that the same calibrated μρ-λρ RPT was used for μρ and λρ calculated with geophysical logs and seismic inversion data.

In Figure 4, the steps for 1D-3D lithotype interpretation are shown. First, the geometry and consistency of data clouds of petroelastic parameters and density were validated for well and field scales. Next, supervised quality control must be conducted for seismic inversion data and well logs. Finally, the missed data were correlated, honoring the lithology reported in analog wells. Next, μρ-λρ RPT based on three dominant minerals was developed using the self-consistent equations, Eqs. (1) to (3). Elastic properties of pure dominant minerals define the vertexes of RPT; for instance: quartz, feldspar, and clay are often dominant minerals for terrigenous formations, and clay, calcite, and dolomite for carbonate formations. After RPT designing, lithotype interpretation is conducted using lithological zones related to lithotypes' elastic properties. Finally, the 1D-3D interpretation results are lithotype logs for wells analyzed and the volume of lithotypes.

Figure 4 Workflow for 1D-3D lithotype interpretation using ternary rock physics templates. Petroelastic analysis sequence to relate seismic with well logs for obtaining lithotype logs and volume of lithotypes. Ternary rock physics templates were generated considering the 1D methodology described in ^{Nicolás-López and Valdiviezo-Mijangos (2016)}. 

2.4. Workflow for 1D-3D brittleness analysis

For well and field scale, lithology interpretation was suggested by calibrated μρ-λρ RPT. Nevertheless, we have proposed novel integrated workflows to couple brittleness analysis with petroelastic lithotype interpretation to enhance robustness, Figure 5. Reservoir delineation is improved because defining higher-brittleness zones aims to plan optimized well surveys for better hydrocarbon exploitation. First, the brittleness analysis is proposed using lithology-based limits to Young's modulus E and Poisson's ratio v. The E maximum and minimum used are related to brittle rocks. Conversely, E minimum and v maximum are related to ductile rocks; next, brittleness indexes in Eqs. (6) to (8) are calculated using geophysical well logs and volumes of elastic moduli. Next, cut-off values evaluate Brittleness logs and brittleness volumes to define reservoir zones prone to hydraulic stimulation. Finally, the petroelastic interpretation and brittleness analysis results are correlated to characterize zones for well placement.

Figure 5 Novel workflow for 1D-3D brittleness analysis for identifying areas with reservoir potential using enhanced 1D brittleness methodology (^{Lizcano et al., 2018}). 

3.1. Characterization of density and rock's elastic properties

1D-3D petroelastic interpretation and brittleness workflows require a rigorous characterization of density and rock's elastic properties. The required curves to determine elastic properties that will be scaled to seismic information were mainly wet bulk density RHOB, compressional Vp, and shear Vs wave velocities. In this case study, they were available for the Sty-1 and Sty-3 wells (Figure 6). In Sty-2 and Sty-4 wells, it is solely accounted with RHOB and Vp.

Figure 6 Logs of density and elastic parameters, Vp nd Vs are shown. Wave velocities were calculated from transit time logs. Shear wave velocity Vs was solely available in Sty-1 and Sty-3. 

Therefore, it was necessary to compute Vs data for all wells involved in the novel petroelastic modeling proposed. To obtain Vs values with an acceptable accuracy to Sty-2 and Sty-4 wells, a neural networks NN methodology like the one proposed by ^{López-Aguirre et al. (2020)} was followed. Herein, Vs curves of Sty-1 and Sty-3 wells with a vertical resolution of 0.1524 m were used as data in training mode; while the wells that did not have Vs were included with a lower vertical resolution (2 m) in the data set defined like test mode. After the NN process was executed, Vs curves for Sty-1 and Sty-3 were modeled (Figure 7). Then, the analysis results for the training wells were compared with the hard data with which a high correlation was obtained. Finally, Vs curves were obtained for the wells that did not have them.

Figure 7 Analysis of the Vs results obtained from neural networks NN. The left figure shows Vs results obtained for the Sty-1, while the right figure shows the results of Vs for the Sty-3 well. 

Figure 8 shows the validation of Vs results obtained with NN against actual Vs logs for the Sty-1 and Sty-3 wells. The relationships obtained are those mentioned in the following equations,

Figure 8 Value-range-based calibration of modeled Vs well logs. Sty-1 is the correlation well for qualitative analysis in tracks. 

Next, Vs logs calculated with NN for Sty-2 and Sty-4 wells also have equivalent accuracy of around 95%. In addition, the value range of the four Vs curves estimated with NN is qualitatively placed in context using Sty-1 Vs log, Figure 8. After these realizations of Vs, we have completed the set of density RHOB, compressional Vp, and shear Vs wave velocity for petroelastic interpretation of lithology columns.

3.2. Petroelastic parameters

The elastic parameters Mu-Rho μρ and Lambda-Rho λρ to each well of the study case were calculated using well logs called bulk density RHOB, compressional Vp, and shear Vs wave velocity, Figure 9. They are central inputs to petroelastic interpretation workflows of lithologies and the presence of pore fluids. Note that Mu and Lambda are non-dependent elastic parameters, and both are impacted with bulk density to highlight the presence of pore-filling fluids. On the other hand, μρ is always less than λρ because it is only sensitive to rock matrix, leaving out the pore fluids effect.

Figure 9 The elastic parameters, Mu-Rho (μρ) and Lambda-Rho (λρ) in GPa, were obtained for the wells of the study case. Reservoir intervals reported in wells from Figure 2. 

3.3. 1D petroelastic model for lithotype interpretation

RPT-based interpretation of lithotypes usually starts with defining lithology classification from the field description reported (Nicolás et al., 2019). From the geological information of the study area, the lithologies are linked to dominant minerals, i.e., shale is related to clay minerals, and sandstone to quartz. Sandstones can also be discriminated via the content of the second dominant mineral, Table 2. Therefore, the column was primarily defined by lithologies considering mineralogy content.

After defining the lithologies that constitute the lithology column, the elastic properties of three dominant minerals were determined to build the specific RPT for the study area. First, we defined two siliciclastic minerals (quartz and potassium feldspar) and one argillaceous mineral (clay), taking information on the three types of clay reported in the case study. Next, a weighted average was calculated for illite/smectite, kaolinite, and chlorite to define clay's elastic properties. Finally, reference values were set for the dominant minerals: quartz, potassium feldspars, and clay, as shown in Table 3.

For the construction of RPT in terms of μρ and λρ, the heterogeneous isotropic method was applied for ternary plots. In Figure 10, elastic properties of pure dominant minerals (quartz, potassium feldspar, and clay), denoted by diamonds, set the ternary RPT vertices for non-porous rock, i.e., the porosity of 0%. RPT bounds were calculated with the self-consistent method described above in Eqs. (1) to (3). In dashed lines, the sides of inner triangles represent 50% content of the dominant mineral portrayed in the opposite vertex. The ternary plots in black denote porosities of 0%, 15%, and 35%; intermediate plots in red are for 5% and 25%. In other words, the larger ternary plot in black is for 0% porosity, and the smaller is for 35%. Ternary plots decrease when porosity increases due to a lesser elastic contribution of pore-filling fluid.

Figure 10 Construction of μρ-λρ RPT with vertexes defined by elastic properties of the calibrated dominant minerals, RPT bounds computed with micromechanics model. In color-filled points, elastic parameters of wells: Sty1, Sty2. Sty3, and Sty4. 

In Figure 10, the elastic responses of wells from Figure 9 are plotted together with RPT. The point cloud of each color used corresponds to the elastic information of each of the four wells in the study area. Thus, the blue points correspond to Sty-1, green points to Sty-2, red points to Sty-3 well, and yellow points to Sty-4 well. The μρ-λρ cross plot shows that well data and ternary RPT are in range. Some points often drop out of RPT bounds because they could involve more minerals than the dominant minerals considered; however, the framework of the proposed RPT-based interpretation is unaltered.

Once the characterization of the ternary RPT with elastic information of the four wells was achieved, Figure 11, the petroelastic interpretation of lithotypes was carried out. At this point, lithologies are defined by elastic responses of rock mixtures where the elastic contribution of pure minerals is considered. In Figure 11, the superimposed zones are mainly guided by the vertices of ternary plots. These are related to lithotypes described in Table 2 as follows: the yellow zone is trending to quartz vertex; therefore, quartz sandstone is straightforwardly discretized; the red zone is for the reservoir and is linked to sandstones, high porosities, and pore-filling fluid; green zone for shales is biased by clay vertex, grey zone is for k-feldspar sandstone, and the light green zone is for shaly sandstone.

Figure 11 Ternary RPT shows the resulting lithotypes for the 1D model for the four wells, which would later be scaled to seismic to obtain a 3D geological model. 

Zones portrayed in Figure 11 are used to select points and differentiate between themselves. The filtered points are separately related to each lithotype for 1D petroelastic lithology interpretation. A third fundamental property, apart from μρ and λρ, is the measured depth of elastic properties. Then lithotypes are characterized in one-step along wells depth as shown in Figure 12. It is suggested that the characterization of lithotypes must be tied up with well geology reports. For instance, herein, Sty 1 reported massive intervals of k-feldspar sandstone and a better reservoir delineation. Modeling this information of correlated wells helped validate the novel petroelastic interpretation of lithology columns.

Figure 12 1D petroelastic interpretation of lithologie columns. Cross-section correlation of lithology units colored with zones of Figure 11. 

3.4. 3D petroelastic model for lithotype volume construction

1D petroelastic interpretation of lithologies is the main foundation for 3D petroelastic interpretation because the data design and visualization are analogous. In Figure 13, the validated ternary RPT was used in the same fashion as Figure 11. However, at this stage, the input data were calibrated volumes in terms of μρ and λρ. They are cross-plotted, and RPT is superimposed to guide 3D lithology interpretation. In Figure 13, the colored zones are also related to the same lithotypes, i.e., green for shale, light green for shaly sandstone, grey for k-feldspar sandstone, yellow for quartz sandstone, and red for the reservoir.

Figure 13 Shows the resulting lithotypes for the 1D model for the four wells, which would later be scaled to seismic to obtain a 3D geological model. 

The shaded zones were used to select and characterize volumes of seismic inversion data in μρ and λρ. The result is a volume of lithotypes that quantifies each lithotype's position (x, y, z) and spatial distribution. 1D petroelastic interpretation from Figure 13 was used to tie up the lithotype volume obtained. Herein, in-line: 1,965 and x-line: 12,475 were used to show the main result of the proposed 3D petroelastic interpretation of lithologies. Figure 15 portrays the spatial distribution of shale, shaly sandstone, k-feldspar sandstone, quartz sandstone, and the reservoir. In addition, constructing sedimentary models and geobodies delineation could be its potential applications.

In Figure 15, stratal slice analysis focused on the spatial distribution of the reservoir was performed. Remember that this is a 3D visualization of the selection of intervals in red from Figure 13. These points are colored zones in red in Figures 14 and 15. Sty-1 well clearly dropped in the best zone of the reservoir because it landed in the broader area in red. Commercial output is often linked to the vast presence of pore-filling fluid. In contrast, Sty-2 and Sty-3 were drilled in a sparse area in red. Sty-4, a re-entry of Sty-3, also dropped in a scarce area in red. The proposed 1D-3D interpretation of lithologies integrated with field and production well reports can assist in defining hydrocarbon-rich zones within the target stratum.

Figure 14 Lithotypes were obtained for the study case (In-line: 1,965; X-line: 12,475). 

Figure 15 Stratal slice of reservoir identified by the rock physics templates RPT. The zone in red in Figure 13 encompasses the points distributed within the stratal slice. 

3.5. 1D brittleness model

Integrating the petroelastic model of lithotypes and brittleness modeling is the major novelty of this research. Brittleness modeling reinforces 1D-3D petroelastic interpretation by quantifying brittleness in hydrocarbon-rich zones. Brittleness evaluation carried out based on Young's modulus E and Poisson's ratio v is well accepted in the oil industry, Section 2.2. Lab tests, well logs, and seismic inversion volumes can obtain these elastic parameters. Herein, the first step is to define the maximum and minimum values of E and v. In Table 5, the corresponding values are shown. They are related to dominant minerals with the highest and lowest values for E and v.

Next, E and v curves were calculated using geophysical logs modeled in section 3.1. The elastic parameters obtained for the wells are those presented in Figure 16.

Figure 16 Elastic parameters E and v of the wells in the case study. 

With the elastic parameters defined and calculated for each well, E and v curves are normalized by applying equations 6 and 7. Brittleness indexes based on Young's modulus BI_E and Poisson's ratios BI_V are portrayed in Figure 17. They are qualitative indicators because they depend on the maximum and minimum values used. For instance, we used limits referenced in Table 4; however, the curve behavior will not change when other limits are used. The critical point is honoring the elastic properties of the most brittle and ductile formations.

Figure 17 Brittleness indices, BI_E and BI_v, were calculated for wells of the case study. 

Note that normalized profiles of BI_E are lesser than BI_V in wells evaluated. This issue is because, in the discretization, the theoretical value of quartz with porosity equal to 0 was used as the maximum value for E. Therefore, we suggest using core data when available to maximum and minimum values for E and v at target formation. Finally, computing the arithmetic average of BI_E and BI_V with equation (8), the BA brittleness index was obtained and used to discriminate intervals with greater and lesser brittleness. Reservoir intervals reported are references to the brittleness quantification considering the elastic contribution of pore-filling fluid, Figure 18. Sty-1 reported a potent reservoir interval in contrast to Sty-3, which landed out of the hydrocarbon-rich zone. These well behaviors were linked to the performance of BA curves. Based on a qualitative interpretation of BA and the oil field data, the higher values BA≫0.5 could indicate the presence of hydrocarbons upon the target stratum.

Figure 18 Brittleness average (BA) based on a simple average of BI_E and BI_v to wells in the case study. Reservoir depths were used to calibrate the ranges of BA qualitatively. 

It was obtained, and in the case of the Sty-1 well, the identification of the hydrocarbon zone. In Figure 18, the results of BA are shown.

3.6. 3D brittleness model

In the same fashion that was scaled 1D to 3D petroelastic interpretation of lithotypes, sections 3.3 and 3.4., 1D brittleness modeling was scaled to 3D brittleness modeling. Herein, 3D brittleness analysis was executed using volumes of Young's modulus and Poisson's ratio. These volumes came from seismic inversion tied with the area wells (^{Arévalo-López, 2017}). Drawbacks about using data at different scales were solved in that step. For brevity, figures of intermediate normalizations of BI_E and BI_v were skipped. Results of 3D brittleness analysis based on BA are shown at the target stratum. Reservoir zones linked to BA were also investigated as an indicator of saturated zones. The latter represents an outstanding feature of lithotypes and brittleness workflows proposed.

Figure 19 shows 1D-3D brittleness modeling conducted on in-line and x-line corresponding to the intersection with the Sty-1. Zones in red denote higher values of the BA brittleness index, consistent with the 1D brittleness analysis performed on Sty-1. The horizontal spatial distribution is related to the lateral continuity of brittle intervals at the target stratum, and vertical variations can underpin well paths to drill the most significant number of brittle intervals.

Figure 19 Brittle zones in red from 1D-3D brittleness analysis of the case study. Sty-1 is used as a correlation well. 

The same stratal slice, Figure 15, is used to portray the spatial distribution of brittle zones and landing area of the well paths, Figure 20. Regions in red were discriminated by cutting the higher values of the BA brittleness index. For example, at the target stratum, Sty-1 was placed into a sizeable brittle zone; Sty-2 and Sty-3 were drilled in poor brittleness zones; and Sty-4, a re-entry well from Sty-3, was redirected to an area with higher brittleness. This brittleness analysis carried out on wells and stratal slices, features the BA brittleness index as a new reservoir indicator.

Figure 20 Reservoir identification using the brittleness index modeling. BA index was set out as a reservoir indicator in terrigenous formations. 

Finally, in Figure 21, the integration of workflow results of both petroelastic interpretation and brittleness analysis are shown. The in-line: 1965 and cross-line: 12765 corresponds to Sty-1, and the reservoir zone in red was identified at well scale and delineated at seismic scale. 3D visualization of lithotype distribution, Figure 15, and higher values of BA(≫0.5), Figure 20, are merely depicted and correlated with Sty-1.

Figure 21 Results coupled of the brittleness analysis and lithology interpretation of the study case.