Advanced, process-based geomechanical modeling reduces risk in layered salt basins

Successful hydrocarbon exploration in the world’s salt basins has highlighted the geological significance of salt rock formations in the creation and preservation of hydrocarbon traps. A key aspect of this success is the development of fit-for-purpose kinematic techniques, most notably in the Gulf of Mexico (GOM), which provided valuable insights into the timing and structural evolution of reservoirs near salt bodies, and thereby reduced the exploration risk. The kinematic techniques assume an end-member scenario, where the “salt” formation is homogenous, a reasonable assumption for the allochthonous salt in the GOM. 

However, most salt formations worldwide, on the contrary, are comprised of heterogeneous layered salt, commonly referred to as layered evaporite sequences (LES’s) or “bedded” salt (Warren 2016). Some examples of LES can be found in the Oman salt basin (Li 2012), the Gulf of Suez, and the Levant basin in the Mediterranean Sea (Cartwright 2012). The heterogeneity in LES’s could be caused by variations in salt composition (halite, tachyhydrite, carnallite) or be due to inter-layering of salt with sedimentary layers (SLs) (carbonates and siliciclastics). Traditional kinematic techniques, developed for an end-member scenario, such as the GOM, cannot be applied directly to constrain the structural evolution in LES’s. The author will demonstrate that process-based geomechanical modeling is a powerful technique to understand the evolution of geological structures in LES’s and complement traditional kinematic techniques (Thigpen 2019). 

KINEMATIC METHOD 

At the core of kinematic techniques are the assumptions that rocks within a layered sequence deform in a similar manner and that some geometric rules (line-length balance) may be used for restoring geological deformation. Such assumptions can be invalid in LES settings, in which the geological deformation in the viscous salt intervals and frictional-plastic SLs can be drastically different. Recent physical modeling studies (Cartwright 2012) showed that deformation in the SLs and salt can be drastically different. Although physical models provide valuable insights into the 3D evolution of structures in LES’s, their applicability to natural examples is not straightforward for the following reasons. Firstly, scalability of mechanical properties of analogue materials in physical models to Km-scale natural examples is debatable. 

Secondly, no robust techniques exist for scaling the thermal properties. The latter are particularly critical for natural rock salt, which exhibits a T-dependent non-Newtonian rheology. As a result, physical models cannot be used to constrain the ambient conditions (stresses, pore-pressures) in real-world subsurface conditions. Conversely, process-based modeling can fill this technology gap and support investigation of ambient subsurface conditions (pore-pressures, stresses) in structures in LES’s.

Fig. 1. Schematic listing challenges posed by LES for structural modeling and drilling.

Understanding the mechanical evolution of structures in LES’s has economic significance. SLs in LES’s could be valuable for resource extraction (hydrocarbons, minerals) or could be excellent sites for storage of nuclear waste, as they are surrounded by impermeable salt formations. From an operational viewpoint, LES’s pose significant drilling challenges, due to either fluids trapped within the SLs (gas kicks, over-pressures), intensely deformed SLs (wellbore stability issues near rubble zones) or highly mobile salt intervals (wellbore closure), Fig. 1. Process-based geomechanical models can provide distinct insights that can be used for assessing the likelihood of encountering such challenges pre-drill.

Modeling the geomechanical evolution of LES structures. For a comprehensive insight into the evolution of structures in LES’s, some key processes associated with salt tectonics (Fig. 2a and Fig. 2b) need to be simulated in an internally consistent manner. In this study, we specifically focus on the evolution of the strains and stresses in the SLs during salt diapirism, and structural controls on the integrity of the SLs, and briefly discuss the implications of the results for exploration risking and well-planning in LES structures. The model below is for an active salt diapir initiated by the deposition of an LES on pre-existing salt and sediment layers. The general approach can be tailored easily for simulating the large array of structural styles (canopies, mini-basins) distinct to salt basins (Hudec 2007).

Fig. 2. Primary controls (a) on halokinesis in salt basins. Models of salt—sediment interactions (b) should use material properties that support compaction and dilation in sediments, in addition to viscous flow in salt and design of the salt diapir model, as discussed (c).

The initial geometry comprises a horizontal basal salt layer overlain by sediment layer, Fig. 2c. Perturbation of the salt-sediment interface in the center of the model domain was used to seed the diapir. The simulation comprises of two stages—a “settling stage” in which geostatic equilibrium was established for the initial layers, followed by a “deposition” stage, in which the LES was deposited. The LES in the models comprises alternating salt and SLs, which are deposited during the simulation. Deposition of SLs was achieved by draping the model surface with a uniformly thick (150-m) layer. Salt, on the other hand, is deposited in the region between the current model surface and a prescribed horizontal surface at a given time.

The lateral boundaries were constrained in the horizontal direction, and model base was constrained in the vertical direction. No tectonic boundary conditions were imposed on the model. A lateral earth coefficient of 0.7 was assigned to the sediments in the settling stage. A constant geothermal gradient of 40.5°C/km was assigned, and surface temperature dynamically adjusts to 21°C during deposition. Acceleration, due to gravity (9.8 m/s²), was assigned in both stages of the simulation.

We assigned an advanced elastic–plastic rheology (SR3 material, Crook 2006), which also supports modeling sediment compaction and attendant property changes in all the SLs. Depending on the degree of mechanical compaction, the Young’s modulus varies between 7–18 GPa (shale), 15–40 GPa (sandstone), and the Poisson’s ratio between 0.2–0.25. Salt was modeled as a non-Newtonian incompressible material (Fredrich 2007), with an effective viscosity that varies between 10¹⁸–10²³ Pa’s. We refer to Goteti (2017) for a detailed description of the model design and material properties. For the simulation, we used the finite element software ELFEN, due to its robust, large strain modeling capabilities involving geological deformation.

RESULTS

The left panel in Fig. 3 (a–e) shows the evolution of the model geometry and the plastic strains in the SLs during the syn- and post-depositional deformation associated with diapirism. Due to model symmetry, we only show the strains and stresses in the region highlighted by the dashed rectangle, Fig. 2c.

Fig. 3. Evolution of plastic strains in the SLs during diapirism (a–e). Distribution of shear stress, (f), effective stress ratio (g) and dip of minimum principal stress in SLs (h). Stress state in salt remains isotropic. Ma—geological duration in million years represented by the model.

During the early stages of LES deposition, the diapirism is minimal to non-existent. In these stages, the difference in pressure between points A and B in the base salt layer (P_A-P_B <0, Fig. 3b) is not large enough to initiate a density-driven instability and the growth of the diapir. Once P_A exceeds P_B and overcomes the resistance from LES overburden, the incipient diapir grows rapidly, both in the vertical and lateral directions. Since the LES predominantly comprises many weak salt layers, which offer minimal lateral resistance, widening of the diapir occurs at the same time as its vertical growth, Fig. 3c–e. The widening of the diapir results in significant shortening in the adjacent SLs in the LES. If the side-burden were comprised of only sediments of finite strength, the diapir widening and, hence, shortening in adjacent sediments could be lower. 

The relative proportion of compaction and dilatational strains in the SLs is dependent on the timing of their deposition vis-à-vis growth of the salt diapir. Sediments, deposited early (Fig. 3e) in model evolution, undergo compaction, and their elasticity and strength are established before the onset of diapirism. As a result, deeper SLs exhibit relatively rigid behavior, with minimal hinge thickening or thinning during folding. Due to the increased rigidity and strength of deeper sediment layers, they develop folds with larger wavelengths than those in shallower SLs. Shallower SLs deposited later during diapirism, on the other hand, are less compacted and remain relatively weak during diapirism. As a result, they develop low-amplitude folds. In other words, deformation in a single LES column could result in disharmonic folding, and assumptions of harmonic folding for structural modeling purposes can lead to erroneous interpolation of geometry of traps and seals in SLs.

In-situ stresses in the sediments. The stress magnitudes and orientations evolve in a complex manner, depending on the structural position. High shear stresses develop in the tightly folded SLs, Fig. 3f. The elevated shear stresses are particularly noticeable in overturned/dipping SLs, and in regions where the surrounding salt layer has thinned (expelled) significantly.  Such zones of high shear stresses may indicate structural positions, where one can expect significant faulting/fracturing and/or wellbore instability issues when drilling through the SLs.

Stress orientations: Away from the salt diapir, the minimum compressive stresses are sub-horizontal (Fig. 3g) and are parallel to the salt-sediment interfaces. The maximum principal stress is vertical. However closer to the diapir, the complex folding of the SLs results in drastic variations in the orientation of in-situ stresses over short distances. The high shear stresses and the unfavorable orientations of the principal stresses in multiple locations, suggests that significant geomechanical hazards may be encountered, when the SLs are intensely deformed.

Application of process-based modeling in salt basins with LES’s: (title). Salt basins with LES’s pose distinct challenges associated with the seismic visibility of the SLs. Sub-seismic SLs, and the low signal-to-noise ratio near some salt structures, can preclude robust interpretation and structural modeling in an LES.

Our model results show that in LES’s, salt at multiple structural levels decouples the deformation between adjacent SLs.

As a result, two successive SLs may evolve along very contrasting strain paths, depending on the thickness and the effective viscosity of intervening salt layer(s), Figs. 3e–h. Disharmonic folding may, therefore, be a common feature in an LES. Process-based geomechanical modeling can be used to test multiple geological scenarios of salt—sediment distribution in an LES, and help constrain the likely structural evolution of SLs at different positions.

Assessing the continuity of SLs can be critical for mapping the extent of a resource. Even when the SL is seismically resolvable, seismic continuity does not necessarily translate to physical continuity. Sub-seismic faulting and boudinage, for example, can adversely impact access to resources or pose severe drilling challenges. Geomechanical models can be used to understand the extent of deformation in SLs and place constraints on their continuity as a function of structural position.

For example, in our model, the overturned portions of the deeper SLs undergo significantly higher amount of deformation than shallower SLs, Fig. 3. Therefore, SLs in such locations may be prone to reduced continuity, due to faulting, fracturing or rubble zones. Occurrence of moderate-high plastic strains in the overturned and dipping SLs, and relatively low strains in sub-horizontal SLs, suggests that the intensity of deformation is controlled primarily by two factors viz., the proximity to the diapir and dip of the SLs. Curvature, a commonly used proxy for predicting intensity of faulting or fracturing in traditional kinematic techniques, may not be a reliable indicator of deformation in the SLs in LES’s.

Finally, geometry of sediment onlaps against salt diapirs is often used to place constraints on timing of salt-related structural traps. However, due to the presence of interlayered salt, seismic images of LES’s are not always amenable for characterizing the detailed geometry of sediment onlaps. Since all the SLs are deposited at uniform thickness in our model, we do not reproduce the depositional thinning (sediment onlaps) commonly observed near salt bodies. The results, nevertheless, show that one can expect significant structural thinning of SLs in an LES. The amount of thinning depends on the structural position, lithology and the degree of compaction in the SL. Since it may not be possible to differentiate depositional and structural thinning from seismic data in LES’s, caution should be exercised when attributing the thickness changes in seismically resolvable SLs to the timing of LES salt structures.

Implications for drill hazard assessment and well planning in LES’s: The key advantage of process-based modeling is the ability to place realistic constraints on the ambient subsurface conditions (stresses, pore-pressures) and assess geomechanical hazards, pre-drill. Such constraints cannot be derived from physical modeling or simplistic analytical approaches of LES structures.

Two important parameters in pre-drill risk assessment in LES’s are the orientation and magnitude of the principal stresses in the SLs. Our model results suggest that, in regions of low–moderate strains (i.e., distal regions away from the diapir), the salt-sediment interfaces track the orientation of one of the principal stresses. The dip of the SL, where resolvable in seismic images, is a good indication of the in-situ stress orientations in distal regions. However, near the diapir, intense folding can reorient the stress field within the SLs. For example, in the region highlighted in Fig. 3g, both the maximum and intermediate (out-of-plane) principal compressive stresses are sub-horizontal. In addition, high plastic strains in sediments in this region also suggest the continuity of SLs may be reduced. Therefore, wellbore instability, due to unfavorable stress orientations (casing shear) or abrupt changes in pore-pressures, due to unanticipated transition between sediments and salt, may be encountered when drilling in these locations.

The perturbation of the regional stress field around salt structures is well-documented (Dussealt 2004). These perturbations can be more widespread in an LES, due to repeated occurrence of salt layers in a heterolithic sequence. Salt can be assumed to be pseudo-fluidic over geological timescales. Therefore, K, defined as the ratio of the effective horizontal and vertical stress magnitudes, is equal to 1 in salt. Estimating the K value in the intervening sediments is not straightforward. Traditional approaches assume a constant K (Zoback, 1984) or a depth-dependent K (Matthews, 1967). This, in combination with vertical stress (integrated density logs) is used to estimate horizontal stresses in the SLs.

Our model results show that K varies in a complex manner (Fig. 3h), due to stress perturbations from multiple salt layers in an LES. Such variations in K cannot be predicted by the analytical approaches mentioned above. K values transition from that characteristic of an extensional tectonic setting (K < 0.7) to that of a contractional setting (K >1.1), closer to the diapir. Partially compacted, weak, shallow SLs in the intervening regions have a high value of K, due to their inability to sustain high differential stresses. In addition, the rapid variations in K with depth, owing to transitions between salt (K = 1) and sediments, can introduce significant challenges in designing wells in an LES. For example, stuck pipe incidents in salt could be mitigated by increasing the mud weight. However, this increase needs to take into account the reduced S_min in the adjacent SL.

Fig. 4. Profiles of total vertical stress, minimum stresses, and pore-fluid pressures (filled red rectangles) along wells AA’ (above) and BB’ (below). HYD = hydrostatic pressure profile. Well locations are shown in Fig. 3e.

In the absence of offset well data, idealized geomechanical models also can be used to gain useful insights into in-situ stresses along various well paths. Consider two hypothetical well trajectories at different structural positions in the model (AA’ and BB’), Fig. 3e. Wells AA’ and BB’ are respectively distal and proximal vis-à-vis the diapir. For both the trajectories (Fig. 4), S_min and S_v are identical in all the salt layers and in shallow SLs, both of which cannot sustain large differential stresses. In compacted, deeper SLs, S_min is less than S_v. S_min values, in general, are smaller for well path AA’ (Fig. 4a) and this is consistent with lower K in this location (Fig. 3h).

Closer to the diapir (Fig. 4b), S_min increases in magnitude, due to the increased compression from the salt diapir. Pore-fluid pressures (P_f) in both the locations remain close to hydrostatic in the sandstone layers and are elevated in the shale layers in the LES. Lower S_min and elevated P_f for well AA’ could indicate narrow mud-weight windows (MWW) in such locations, thereby increasing the risk of gas kicks and lost circulation. Higher S_min in SLs, in well BB’ (closer to the diapir), can support wider MWW. However, the cumulative plastic strains in the sediments (Fig. 3e) are higher near BB’. Therefore, in spite of the wider MWW, there could be issues related to continuity of the SLs along BB’. Geomechanical forward modeling studies can be used to investigate various subsurface scenarios and high-grade favorable well paths.

VALUE ADDED 

Advanced geomechanical modeling is a powerful technique to assess the geological- and drilling-related risk factors in basins with LES’s. Idealized models can: 1) yield useful insights into the deformation of reservoir intervals; 2) provide constraints for seismic interpretation in areas of poor imaging; and 3) complement physical or kinematic modeling approaches. Process-based models support investigation of the impact of sedimentation, compaction, salt mobility and pore-fluid pressures on LES tectonics in an internally consistent manner. We demonstrated how the insights from the simulations can provide insights for exploration well planning and drill hazard assessment in salt basins containing LES formations. 

Historically, due to the short time-cycles involved in business decision-making, turnkey analytical solutions or kinematic techniques were routinely employed to arrive at the “80% solution” for deciphering structural evolution in salt basins. Representative process-based modeling was not routinely employed, due to the amount of time it took in designing and running numerical simulations. For example, a decade ago, the model presented above required run time ranging from days to weeks. Emerging convergence between high-end computational power and efficient numerical formulations now supports running these advanced simulations in a matter of hours, or even minutes, on a single workstation!

This technical advance supports usage of process-based models for rapid geological scenario testing and placing realistic constraints on ambient conditions in the subsurface for drilling purposes. In addition, results from such models also can be used as synthetic training data sets for various applications (improve seismic interpretations) in the emerging areas of machine learning and data analytics. 

Structural geologists and operations engineers should leverage such process-based geomechanical simulations and incorporate them in their routine toolkit for risk assessment in salt basins as occurrence of heterogeneous salt formations (LES’s), which are not amenable to traditional structural analysis techniques, is the norm and not the exception. 

ACKNOWLEDGMENT

The author would like to thank the American Rock Mechanics Association for permission to publish portions of Goteti (2017) cited in the text.

REFERENCES

1. Cartwright, J., M. Jackson, T. Dooley and S. Higgins, “Strain partitioning in gravity-driven shortening of a thick, multilayered evaporite sequence” Geological Society, London, Special Publications, 363(1), pp. 449-470, 2012.
2. Crook, A.J.L., S.M. Wilson, J.G. Yu and D.R.J Owen, “Predictive modelling of structure evolution in sandbox experiments, Journal of Structural Geology, 28(5), pp. 729-744, 2016.
3. Dusseault, M.B., V. Maury, F. Sanfilippo and F.J. Santarelli, “Drilling through salt: Constitutive behavior and drilling strategies in Gulf Rocks 2004,” the 6th North America Rock Mechanics Symposium (NARMS), American Rock Mechanics Association, January 2014.
4. Fredrich, J.T., B.P. Engler, J.A. Smith, E.C. Onyia and D. Tolman, “Pre-drill estimation of sub-salt fracture gradient: Analysis of the Spa prospect to validate non-linear finite element stress analyses, SPE/IADC Drilling Conference, January 2007.
5. Goteti, R., S.M. Agar, J.P. Brown, H.J. Sibon and R. Zuhlke “Deformation of siliciclastic stringers in a layered evaporite sequence (LES): Insights from geomechanical forward modeling,” American Rock Mechanics Association. 51st Symposium, August 2017.
6. Hudec, M.R., and M.P. Jackson, “Terra infirma: Understanding salt tectonics.” Earth-Science Reviews, 82(1-2), pp. 1-28, 2007.
7. Li, S., S. Abe, L. Reuning, S. Becker, J.L Urai and P.A. Kukla, “Numerical modelling of the displacement and deformation of embedded rock bodies during salt tectonics: A case study from the South Oman salt basin,” Geological Society, London, Special Publications, 363(1), pp. 503-520, 2012.
8. Matthews, W.R., and J. Kelly, “How to predict formation pressure and fracture gradient,” Oil and Gas Journal, 65(8), pp. 92-106, 1967.
9. Thigpen, J.R., D. Roberts, J.K. Snow, C.D. Walker and A. Bere, “Integrating kinematic restoration and forward finite element simulations to constrain the evolution of salt diapirism and overburden deformation in evaporite basins,” Journal of Structural Geology, 118, pp.68-86, 2019.
10. Warren, J.K., “Evaporites: A geological compendium, Springer, 2016.
11. Zoback, M.D., and J.H. Healy “Friction, faulting and in situ stress,” In Annales geophysicae (1983), Vol. 2, No. 6, pp. 689-698, 1984.