Introduction

Carbonate stringers represent a major target for oil exploration and production in the sedimentary basin of South Oman. These unique geological formations, composed mainly of carbonate rocks, are found sandwiched within the Neoproterozoic Ara salt deposits. The understanding and managing of these carbonate stringers and their hosting salts is essential for minimizing risk and maximizing the potential for successful hydrocarbon discovery and extraction. Salt structures are characterized by their complex shapes of variable dipping flanks and large impedance and velocity contrasts at their boundaries. They usually reject considerable amount of seismic energy. Thus, the characterization and imaging of stringers inside salt bodies pose several challenges. These challenges have been documented in different world salt basins including Ara salt from Oman basins, Gillen salt from Amadeus Australian basins, and Zechstein salt from North Sea basins. Jones and Davison1 have reported the complexity of imaging stringers and distinguishing them from their surrounding rocks. The authors highlighted the challenges faced in acquiring and processing good quality of seismic data. Strozyk, et al.2 reported the challenges of imaging intra-salt Zechstein (Z3) carbonate stringer in offshore area of Netherland. To overcome challenges pertaining to the geological complexity of stringers, advanced geophysical and geological techniques are commonly invoked. The latter techniques help evaluate their distribution, thickness, porosity, permeability, and connectivity. Al Barawani and McClay3 have combined 2-D, and 3-D seismic data with well data to study the carbonate stringers in Thumrait area and their encasing salt bodies. The authors studied also the influence of salt movement on the stringers and the overlying formations and discussed the concept of halokinetic pods (Haima pods). The pods were stratigraphically inverted to form small turtle-structure anticlines, a process originally attributed to salt dissolution. The concept of small, variably inverted minibasins migrated from Oman to the North Sea in the early 1990s (Jackson and Hudec, 2017). Esestime and Stamp4 conducted a detailed structural and stratigraphic interpretation of about 4500Km2 3D seismic from the Blakeney group in central North Sea. Their objective was to define the sedimentary patterns and seismic facies of Zechstein group which is confined with pods or mini-basin generated during earliest phase of salt movement.

Geophysical modeling allows interpreters to simulate the behavior of geological objects and predict their seismic and elastic properties. Numerous published papers illustrate the importance of modeling in enhancing the understanding of subsurface geological formations, their geometry, structures, and elastic behaviors, etc. Jones and Davison1 used a variety of modeling approaches to help them understand and interpret salt geometry, structural, and elastic properties. The authors demonstrated how modeling can help fill gaps in data acquisition, processing, and interpretation. Farfour and Foster5 used seismic modeling to understand seismic and elastic properties of the Mid-Jurassic fluvial deltaic Plover sandstone from Offshore Australia. The modeling helped the authors interpret seismic data covering the area of the study and find new potential zones that have not been drilled. There is a few published papers that document the use of geophysical modeling to study south Oman Basin salt structures and their encased stringers. Li6 used numerical modeling to investigate the influence of salt tectonic on the deformation and displacement of stringers. Al-Obaidani and Farfour7 utilized geophysical modeling to study carbonate stringers from Southern Oman. The authors used a geological model from Al-Siyabi8 and developed different seismic and elastic models to allow them to understand the characteristics of carbonate stringers. Also, Rowan et al.9 studied the deformation of competent intra-salt beds at different phases of salt tectonics using conceptual, numerical and analog models and seismic data.

This study integrates seismic modeling with the interpretation of real seismic data to detect and characterize carbonate stringer bodies from Southern Oman Salt Basin (SOSB). A newly acquired seismic data is interpreted with a special care to identify key formations in the area, such as Amin, Top Ara, Base Ara, etc. The result of the interpretation is presented in a conceptual model of the area. From that model, different elastic properties are created and utilized to build a seismic model of the area that mimics the real seismic data. The created data is inverted to acoustic impedance model to test the capability of seismic inversion data in detecting stringers. The resulting seismic data and its inverted impedance demonstrated a promising detectability of the carbonate stringers. Next, the real seismic data is inverted to acoustic impedance using the existing well logging data. The impedance data from the inversion has supported the results from the modeling. The modeling played a key role in the understanding of the stringers and in increasing the confidence of their interpretation.

Seismic forward and reverse modeling

Understanding and characterization of reservoirs is very crucial in the field of petroleum exploration and production. Seismic forward modeling serves as a powerful tool in characterizing reservoirs, offering valuable insights into the subsurface structure and properties of reservoirs. The modeling helps simulating the propagation of seismic waves through the geological formations of interest which allows geoscientists and engineers to predict their seismic responses and interpret processed seismic data over areas of interests in the context of reservoir properties.

Some of the examples of seismic modeling include: (1) Convolutional model-based approaches where elastic properties are prepared and convolved with predefined wavelet to create seismic traces at normal incidence at the desired space locations; (2) Wave equation-based approaches where elastic models are used as input to the seismic wave equations that are solved using finite difference methods. The solution produces a seismic model with seismic responses computed at time and space.

Seismic inversion has undergone significant evolution since its inception during the 1970s. Over the years, it has become an important task in geophysical exploration, and witnessed considerable development from simple trace rotation to advanced model-based solutions. The primary aim of interpreters when conducting seismic inversion is to generate an approximation of the Earth's acoustic impedance from processed seismic data. This process entails transforming seismic data into pseudo-acoustic impedance logs at various trace locations. The derived acoustic impedance data offer several advantages over conventional seismic data. Unlike the latter, which primarily reflects interface properties approximating the convolution of a wavelet with a series of reflection coefficients, acoustic impedance data represent rock properties directly correlated with lithology, porosity, and fluid content10,11.

Geological setting and data set

The study area is located in the southwestern part of Oman within the western flank of the South Oman Salt Basin. It is interpreted to be a structurally complex area due to the western deformation front and salt tectonic processes “halknoises and withdrawal”. Several geological models that highlight the area's complexity have been published (Al-Syiabi 2008; Al Barawani and McClay3). This paper focuses on the Ara group, which consists of four formations “Birba, Al-Noor, Athel & Dahaban.” The Ara group is bounded from the bottom by the base Ara major sequence boundary and bounded from the top by the Angudan sequence boundary, which is interpreted in available seismic as a regional compressional event. The whole Ara group in the south Oman salt basin is divided into six sequences, “A0 to A6.” In the study area, it is difficult to distinguish between these sequences in real seismic. The main reason for that is the significant effect of the salt tectonic and thrusting process, which resulted in a randomly distributed stringers as patchy within the salt. Figure 1 shows the location of the study area and the Ara group’s stratigraphy.

Figure 1
figure 1

A map showing the study area’s location and associated stratigraphic chart. The chart is from Grotzinger and Al Rawahi12.

Full size image

The 3D seismic acquired in the area is 3-D Wide Azimuth (WAZ), oriented to the north direction. Below are some of the seismic acquisition parameters:

  • - Source interval: 50 m.

  • - Receiver interval: 50 m.

  • - Source line interval: 150 m.

  • - Receiver line interval: 150 m.

The data was sampled at 4 ms and processed using a long list of advanced processing tools with the main objective in mind of achieving a high-quality seismic imaging for top salt, base salt, and intra salt carbonate stringers. The list of processing includes but not limited to surface amplitude corrections, structure-oriented filtering, spectral balancing, footprint noise attenuation, multiple attenuation, zero-phasing, Prestack Time Migration (PSTM) velocity analysis, Full volume PSTM, velocity model building, Prestack Depth Migration.

Analog fields

In this section we analyze two examples of carbonate stringers from analogue fields from two world salt basins. The first field is from Oman Salt Basin, while the second is from North Sea. As shown in Fig. 2, the carbonate stringers in Birba field distinguish themselves from their encasing salt by their higher Gamma Ray, low sonic transit time, and significantly higher density. Their high velocity and density are the cause of high acoustic impedance contrast between them and the hosting salt. The Ara group in this area is found to contain at least six tectono-eustatic evaporate—carbonate cycles and the cyclicity represent the change in water depth and water salinity. All identified carbonates stringers in this group are found to be marine in origin8. The lower unit of each cycle is formed by evaporite that is precipitated during basin restriction condition and commonly contain characteristics of anhydrite-halite-anhydrite. The carbonate stringers in Ara group can be divided to “carbonate-evaporite transition facies association, deep ramp and slop facies association, subtidal microbialites facies association, clastic textured carbonate facies and restricted pretidal carbonate facies association” (Grotzinger and Al Rawahi12). The similarity between the study area and Birba field analog in Oman lies in the fact that both fields exhibit similar rock properties and seismic responses. However, in the area of our study, the stringers are found to be heavily affected by the Western Deformation Front which resulted in distributing stringers as patchy within salt3.

Figure 2
figure 2

An example of carbonate stringers from Birba field, from SOSB (Al-Syiabi 2005).

Full size image

The other analog is from North Sea. The carbonate stringers in this area show high similarity with the stringers from Oman. They show higher natural radioactivity, high density, and sonic transit time as shown in Fig. 3. However, they are so thin and sometimes fall below the seismic resolution. In other words, the stringers from North Sea have lower Net (carbonate) to Gross (encasing salt) compared to the ones from South Oman Salt Basins. However, seismically, data from South Oman (land data) are considerably lower in resolution and signal to noise ratio compared to the seismic data from North Sea (marine data). Also, another different is that in Oman Ara group contain six sequence of carbonates and evaporites whereas Zechstein in North Sea contain three cycles of carbonates and evaporites2.

Figure 3
figure 3

An example of North Sea penetrated carbonate stringers.

Full size image

Methods

The first task in the study is to use the newly acquired seismic data to perform a careful seismic attribute analysis followed by a detailed structural seismic interpretation. After tying the well data with seismic at the well locations a number of key horizons were picked such as top Mahwis, base Mahwis, and top Salt, top of carbonate stringers and base salt which represents top Nafun surfaces. The interpreted horizons are used to build geophysical models of P-wave and S-wave velocity models, and also a density model. The values of Vp, Vs, and density are extracted from nearby drilled wells within the area (Table 1). The velocity and density models are used to generate synthetic gathers and stacked seismic sections. The created gathers and the stacked seismic section are utilized to understand the seismic response of carbonate stringers. The next task was accomplished by inverting the created stack seismic section to impedance using post-stack inversion. Two pseudo wells were created to help conduct the inversion. Next, the inversion was performed on the real seismic data using post-stack seismic inversion. For this purpose, the main interpreted horizons, two pseudo-wells along with an available well were used to create the low-frequency model needed to perform the post-stack model-based inversion. Finally, we integrate both the results from seismic attributes, post-stack inverted data from the modeled seismic data and from real seismic data inversion to characterize the carbonate stringers. Figure 4 shows the project workflow.

Table 1 Velocity and density information from nearby wells.
Full size table
Figure 4
figure 4

The workflow used to accomplish the tasks of the study.

Full size image

Results and discussion

Figure 5 displays seismic data from old data published by Al-Barawani and McClay3 and the newly acquired data. In Fig. 6, a seismic cross-section from the new seismic data is depicted. The newly processed data exhibit better quality (resolution and signal to noise ratio) compared to the old legacy one. It is worth noting that the use of some postprocessing filters, namely the dip steering median filter, has significantly contributed to the enhancement of the data. The latter filter is structure-oriented filter that becomes widely used in data cleaning and pre-conditioning5,13, Farfour and Russell.

Figure 5
figure 5

A seismic section from the old seismic data showing present-day structure (Al-Barawani and McClay3).

Full size image
Figure 6
figure 6

A seismic section from the new seismic data showing the same structure.

Full size image

This appears clear especially in terms of multiples attenuation and imaging quality of key seismic reflectors. The data also show a better continuity of some markers such as Amin and the Base of Ara Salt. Figure 7 displays the results of the detailed and careful structural interpretation conducted on the data to extract the maximum of its structural features.

Figure 7
figure 7

An interpreted seismic section close to the well location showing the key horizons identified and mapped on the new 3D seismic.

Full size image

Figure 8 depicts some of the attributes extracted from the data. The pseudo-relief and energy are seismic attributes that are commonly used to detect abnormal changes in seismic amplitude. The latter changes can be associated with hydrocarbon-charged reservoirs, igneous rock intrusions, salt piercing, carbonate stringers14,15. The analysis and interpretation of the extracted post-stack attributes show that “top salt” is characterized by strong amplitude which is attributed to the high acoustic impedance contrast between the salt in the area and Nimr clastic. In addition, the interpretation infers that “Pre-Mahwis” sequence boundary and “top Mahwis” have continuous reflectors with strong amplitude response and high frequencies. On the other hand, the base salt “top Nafun” exhibits a discontinuous reflector with high amplitude response which makes it challenging to interpret.

Figure 8
figure 8

Pseudo relief attribute and Energy attribute extracted from seismic data.

Full size image

Figure 9 shows the resulting time map of Top of Ara salt formation. The horizon displays the highly variable topography of the salt formation, salt ridges, salt pillow, and salt weld.

Figure 9
figure 9

Seismic structural map (two-way time) of top Ara Salt in study area, South Oman Salt Basin picked using new seismic data.

Full size image

The extracted structural features from the interpretation are used to build velocity and density models from which a stack seismic section is produced. The created P-wave velocity model is shown in Fig. 10. Figure 11 illustrates the seismic stack section computed from the velocity and density models using convolutional theory and wave theory at normal incidence. The section mimics the real seismic data but with primary wave only.

Figure 10
figure 10

Compressional velocity model of the area created from the new seismic.

Full size image
Figure 11
figure 11

Stack seismic section created from elastic models.

Full size image

We observe that the stringers top and base exhibit strong seismic amplitude responses similar to the real seismic. The large amplitude is caused by their large acoustic impedance contrast with the encasing salt. After integrating all interpreted horizons and the pseudo-wells, we conduct a model-based post-stack inversion to convert the modeled seismic data to impedance. Figure 12 and Fig. 13 depict the results of inversion conducted at the well locations and over the entire modeled stack seismic section. The inversion allowed the stringer bodies to stand out of their background with their remarkable high impedance. The results of the inversion are much closer to the velocity models we built from the seismic.

Figure 12
figure 12

Inversion conducted at the pseudo-well location.

Full size image
Figure 13
figure 13

Impedance model obtained from Post-stack inversion applied on synthetic seismic data showing carbonate stringers with high impedance compared to their encasing salt.

Full size image

With this in mind, we carried out the same model-based inversion on the real seismic data. As we have only limited well available for the study, we have created Pseudo-wells from the initial inverted section. The wells have been integrated to improve the final inversion results. Figure 14 displays one of the wells created from the inverted impedance model.

Figure 14
figure 14

Pseudo wells (right) generated from initial impedance data (left) and used for final post-stack inversion of real seismic data.

Full size image

In Fig. 15, the final impedance model from the inversion is presented. The carbonate stringers exhibit large acoustic impedance that differentiate them from their surrounding salt.

Figure 15
figure 15

Seismic section with Pseudo well and inverted section from it. Stringers are showing high impedance compared to other surrounding rocks.

Full size image

Figure 16 shows a final interpreted P-impedance model where all results are incorporated.

Figure 16
figure 16

An interpreted impedance section where stringers are identified.

Full size image

The impedance model indicated that even thin carbonate stringers can be detected. This is because post-stack inversion enhances seismic resolution and removes the effect of the wavelet (interference and side lobs). The model also demonstrated that at the middle part there is clear gradual increase in impedance as we go deeper to the top of the salt. The increase is attributed to the compaction effect caused by the load of overlying sediments. Moving from the center, we observe different trends of impedance increasing with depth. This variation in compaction from area to another is caused by the salt geometry which controls the accommodation space for the overlaying sediment. In fact, the salt has unique physical properties such as their low density, impermeability and a ductile nature. Interacting with sediment loads creates a different form of salt tectonic related features such as diapirs, ridges, withdrawal synclinal basins, overhangs, canopies, etc. During the formation of these features, the force and orientation of the stress fields and compaction change16.

Results obtained from inverting the real data are found in a good agreement with the modeling and inversion experiment carried out on synthetic data. In fact, the forward and reverse modeling (inversion) that we carried out allowed us to improve our understanding of the carbonate and helped us interpret the real seismic and impedance data.

The above example of integration of geophysical modeling and inversion coupled with attribute analysis, structural interpretation to understand salt carbonate reservoirs can be applied on other similar fields within Oman basins or other analogue world basins.

It is worth noting that the lack of well data presents a significant challenge for such exploration studies and efforts, potentially increases the uncertainty regarding any findings derived from the available data. Therefore, future drilling remains the only approach to reduce this uncertainty, validate interpretations, and enhance understanding of targeted formations.

Conclusion

A detailed seismic interpretation study was carried out on newly acquired and processed seismic data from South Oman Salt Basin. The study aimed to detect potential carbonate stringers that can be a good drilling target. A thorough seismic attribute analysis was followed by carful structural interpretation. The interpreted horizons are utilized to build elastic and seismic models of the area. The integration of attribute analysis, seismic modeling and inversion helped us understand carbonate stringers’ seismic response and identify them with a good confidence in the real seismic data. It also allowed us to detect some thin carbonate stringers that were not visible in seismic data. The present study demonstrates how the integration of attribute analysis, modeling and inversion can help improve the interpreter’s understanding of subsurface formations, hence reduce the risk and uncertainty related to geology.