Optimization of Low-Carbon Drilling Fluid Systems and Wellbore Stability Control for Shaximiao Formation in Sichuan Basin with a ‘Dual Carbon’ Background

Abstract

Driven by “Dual Carbon” goals, advancing the green development of oil and gas resources is imperative. The Shaximiao Formation tight gas reservoirs in the Sichuan Basin suffer from wellbore instability, impairing drilling efficiency and elevating energy use and emissions. This study integrates mineralogy, mechanics, drilling fluid optimization, and geostress modeling to address instability mechanisms and support low-carbon drilling. XRD shows that clay content decreases with depth (11–48%), while quartz and plagioclase dominate (45–80%). Synthetic-based drilling fluids fully inhibit clay swelling (0% expansion), outperforming calcium-based (2.4–3.1%) and water-based systems (5.4%). Synthetic and calcium-based fluids also reduce waste treatment difficulty and carbon intensity. Rolling recovery reaches 98.12% for synthetic-based vs. 78.18% for water-based. Strength tests reveal a 36.9% reduction after 14-day immersion in synthetic-based fluid, whereas water-based systems with nano-plugging agents show self-recovery, cutting energy use per foot by ~15%. Geostress modeling indicates a maximum horizontal stress of 90.08 MPa (NE114° ± 13°) and minimum of 67.2 MPa (NE24° ± 13°). Collapse pressure (48–60 MPa) varies azimuthally, requiring higher density (58–60 MPa) along the min. horizontal stress direction. A low-carbon mitigation strategy is proposed: prioritize synthetic or calcium-based drilling fluids, and optimize well trajectory using geostress models. This reduces fluid loss risk by >20%, limits methane emissions, shortens drilling cycles, and enhances efficiency while lowering carbon footprint. These insights support green and efficient natural gas development through intelligent drilling and eco-material applications.

1. Introduction

The Shaximiao Formation is a significant terrestrial sedimentary sequence within the Jurassic System of the Sichuan Basin, widely distributed in central, western, and northeastern Sichuan [1,2,3,4,5,6]. The formation exhibits typical fluvial–shallow lacustrine depositional characteristics, dominated by interbedded purplish-red mudstones and grayish-green-to-grayish-white sandstones [7,8,9,10]. Basal conglomerate layers are commonly present, indicative of a strongly oxidizing environment [11,12]. Regionally, the sedimentary facies show notable variation: in the Liangping area, northeastern Sichuan, the succession consists of braided river, meandering river, and littoral–shallow lacustrine facies, with micro-facies such as subaqueous distributary channels, mouth bars, and mud flats [13,14,15,16,17]. In contrast, the Wubaochang area, central Sichuan, is characterized by a shallow-water delta system, where delta-front subaqueous distributary channels and delta-plain distributary channels alternate, resulting in prominent lateral stacking of sand bodies [18,19,20]. Provenance analysis suggests that sediments were primarily derived from the NE–SW-trending Micang–Daba orogenic belt to the north, with secondary contributions from the eastern Xuefeng Mountains [21,22,23,24,25,26]. Reservoirs primarily consist of fine-to-medium-grained lithic arkose and feldspathic litharenite, characterized by low compositional maturity (quartz: 38–82.5%; feldspar: 7–31%; lithics: 4.5–40%) and moderate textural maturity [13,27,28]. Pore systems are dominated by residual intergranular pores, intergranular dissolution pores, and intragranular dissolution pores, with pore throats typically curved lamellar or necked types, exhibiting mesopore–mesothroat structures [29,30,31,32,33,34]. Petrophysical properties indicate low porosity and ultra-low permeability, with porosity generally ranging within 3.6–14% and permeability mostly below 1 mD (locally up to 3.84 mD), classifying it as a typical tight sandstone reservoir. Diagenesis is complex, involving intense compaction (average pore reduction: 27.91%), cementation (calcite, chlorite, silica), and dissolution (pore increase: 3.08%), placing it in middle diagenetic stages A–B. Chlorite coatings play a critical role in preserving primary pores. Formation water shows high salinity, dominated by CaCl₂-type brines enriched with trace elements (e.g., Sr, Ba) reflecting deep burial fluid–rock interactions, while hydrogen–oxygen isotopes indicate sedimentary origins [35]. Gas–water relationships are intricate, with free water commonly observed in structural lows and near faults. Gas well water production primarily consists of condensate water and capillary-bound water, exhibiting low water–gas ratios. The seal capacity of the Lower Sha-1 Member is superior to that of the Xujiahe Formation. Gas accumulation in the Shaximiao Formation follows a “tight before accumulation” model. The development of high-quality reservoirs is jointly controlled by sedimentary micro-facies (such as subaqueous distributary channels and mouth bars), dissolution intensity, and diagenetic mineral assemblages, making this formation a key target for tight gas exploration in the Sichuan Basin.

However, the development of Shaximiao Formation gas reservoirs in the Gaomiao area is plagued by persistent engineering challenges, with drilling complications including caving, collapse, and stuck pipe creating significant obstacles to well construction, wherein wellbore instability remains a core issue critically impacting drilling safety, economics, and efficiency [36]. Such instability-triggered incidents—stuck drill strings, borehole collapse, and formation damage—not only incur substantial financial losses (exceeding USD 600 million in global direct costs annually, per industry statistics, with indirect losses defying quantification), but also pose severe personnel safety risks. This complexity arises from the formation’s geological heterogeneity, the nonlinear mechanical behavior of downhole rocks, dynamic multiphysics coupling effects (thermo-hydro-mechano-chemical interactions), and multiscale mechanisms governing drilling fluid–formation interactions. In response, petroleum engineers worldwide have conducted extensive research, achieving significant progress in addressing wellbore stability challenges.

The mechanical essence of wellbore instability can be traced back to the disturbance and re-equilibration process of the original in situ stress field following wellbore formation. When the drilling fluid column pressure fails to effectively counteract the combined action of formation stress and pore pressure, the surrounding rock will undergo shear failure or tensile fracturing, leading to collapse or wellbore narrowing [37,38]. This process involves complex interactions between rock strength characteristics, in situ stress state, pore pressure distribution, drilling fluid performance parameters, and engineering operational conditions.

Early research primarily focused on static mechanical analysis based on linear elastic assumptions, employing classical failure criteria such as Mohr–Coulomb and Mogi–Coulomb. However, these models suffer from considerable limitations in capturing rock inelastic deformation, anisotropy, and damage evolution. The landmark study by Denney (2005) [5] highlighted the shortcomings of traditional methods, particularly their neglect of rock heterogeneity and insufficient calibration with logging data. He emphasized that probabilistic wellbore stability analysis requires quantifying the uncertainty of input parameters through Monte Carlo simulation. Nevertheless, his work also identified a major challenge: the difficulty in reasonably constructing probability distribution functions for geomechanical parameters—especially when field data are scarce—as subjective assumptions regarding distribution types may introduce systematic bias. This insight motivated subsequent researchers, such as Udegbunam et al. (2013) [22], to integrate uncertainty analysis within a deterministic fluid flow modeling framework. By using Monte Carlo simulation to generate statistical distributions of bottomhole pressure, they demonstrated that probabilistic approaches can broaden the predicted range of the safe mud weight window, thereby offering a more realistic basis for decision-making in underbalanced drilling design. The characterization of rock mechanical properties remains the cornerstone of wellbore stability research. Laboratory testing and log interpretation constitute dual pathways for obtaining rock strength parameters, yet both face inherent limitations. While core experiments provide precise mechanical parameter measurements for specific samples, they struggle to reflect heterogeneity at the formation scale; conversely, log data inversion offers spatial continuity, but the reliability of its correlation models is highly dependent on the similarity of regional geological characteristics. The probabilistic wellbore stability model developed by Al-Ajmi and Al-Harthy (2010) [2] identified maximum horizontal stress, cohesion, and internal friction angle as key sensitivity parameters via Monte Carlo simulation, a conclusion further validated by Yang Yugui et al. (2023) [27] in their study on fractured formations in the Bozhong Block. This research revealed the strength degradation effect caused by the development of microfractures (2~6 μm) in hard–brittle shales, emphasizing the critical importance of compound plugging materials (ultra-fine calcium carbonate, nano-silica, and white asphalt) in maintaining wellbore integrity.

Accurate characterization of the in-situ stress field is essential for wellbore stability analysis. Although conventional approaches such as stress polygon methods and poroelastic models are widely used, they exhibit limited adaptability to faulted structures, formation heterogeneity, and production-induced stress changes. Li et al. (2021) [13], through coupled seepage–stress numerical simulations, revealed how the orientation of faults relative to the loading direction reshapes the local stress field: faults perpendicular to the loading direction cause the most significant stress perturbations, while seepage effects further promote stress accumulation over time. This observation aligns with the findings of Li et al. (2021) [13] on complex fault systems, highlighting the importance of aligning numerical simulation boundary conditions with realistic geological prototypes. Lu et al. (2023) [17] developed a dynamics-enhanced poroelastic model for ultra-deep hard-rock formations, which goes beyond traditional static frameworks by incorporating inertial effects and viscous coupling into near-wellbore stress calculations. Their results indicate that formations with high eigenfrequencies are more susceptible to instability under pressure fluctuations, not only elucidating the underlying physics of dynamic wellbore failure, but also offering a theoretical foundation for real-time optimization of wellbore pressure control.

The analysis of multiphysics coupling effects represents the frontier of contemporary wellbore stability research. Song et al. (2023) [21] developed a multiphysics coupling model for mudstone formations in the Hutubi Anticline, quantitatively revealing the synergistic mechanism of seepage–hydration–mechanical deficiency: drilling fluid filtrate infiltration along fractures reduces effective stress, triggering a positive feedback loop of strength degradation and pore pressure elevation, demonstrating that merely increasing mud density may paradoxically exacerbate instability risks, overturning the conventional “density-for-stability” paradigm and extending the research focus to the synergistic optimization of sealing capacity and inhibition. Zhang et al. (2024) [28] further confirmed the nonlinear time-dependent behavior of collapse pressure—characterized by an initial decline followed by a subsequent escalation—in Minghuazhen Formation soft mudstones. They identified the expansive hydration of illite–montmorillonite mixed-layer minerals as the primary mechanism triggering wellbore instability. These findings emphasize the necessity of incorporating chemo-mechano-seepage coupling into stability analyses. Complementing this, Adham’s (2016) [1] stress perturbation equation system, which integrates membrane efficiency coefficients, thermal stresses, and flow-induced stress terms, offers a comprehensive mathematical framework for multiphysics modeling. Advances in probabilistic analysis and uncertainty quantification are progressively transforming risk assessment methodologies. Gangrade (2021) [39] highlighted this shift, which was further advanced by Chen et al. (2023) [4], who integrated the Nataf transformation with Monte Carlo simulation to characterize the non-independence of geomechanical parameters. Their study demonstrated that accounting for parameter correlations significantly narrows the safe mud weight window when reliability exceeds 50%. This approach not only quantifies risks during stress regime transitions—such as from strike–slip to normal faulting—but also reveals the influence of parameter sensitivity on well trajectory optimization. In a similar vein, Wang et al. (2018) [23] developed a reliability theory-based model that represents collapse and fracture pressures as probabilistic intervals rather than single-threshold values, substantially improving the robustness of engineering decisions. Meanwhile, Huang et al. (2023) [9] employed the Sobol’ method for the global sensitivity analysis of critical differential pressure in open-hole tests. Their results underscore the dominant role of the maximum horizontal stress and offer data-driven support for risk-tiered management strategies. Zhi et al. (2024) [36] demonstrated that hydration-induced strength degradation dominates instability processes in Qiongzhusi Formation shale horizontal wells. Long et al. (2023) [16] identified the inherent advantages of slim holes in deep shale gas development through comparative stability analysis across wellbore sizes, opening new pathways for well architecture optimization. Zhang’s (2023) [35] HPHT well study revealed nonlinear compressive strength reduction with temperature elevation and pressure-dependent bimodal collapse risks, challenging linear assumptions in conventional thermo-pressure coupling models.

For the Shaximiao Formation in the Sichuan Basin, addressing wellbore instability, reducing non-productive time (NPT), and shortening the drilling cycle necessitates enhancing the understanding of formation petrology and geomechanical characteristics, optimizing the appropriate drilling fluid system, and determining the optimal safe mud weight window [40]. To this end, this study characterized the mineral composition, microstructure, physicochemical properties, and mechanical strength of downhole cavings from the Shaximiao Formation. Furthermore, utilizing well logging and mud logging data, it investigated the in situ stress distribution within the formation, revealed variations in collapse pressure under fluid–rock interactions, and identified the dominant factors controlling wellbore instability. These findings provide critical support for the optimization of drilling fluid systems. The results offer significant practical value for enhancing the efficient development of gas reservoirs in the Shaximiao Formation and establish a scientific basis for safe, efficient, and intelligent drilling under complex geological conditions.

2. Methods and Research Procedure

This comprehensive study systematically investigates wellbore instability mechanisms in the Shaximiao Formation through an integrated experimental–modeling approach, commencing with the field sampling of downhole cavings from Well X across critical depths (2880–3287 m) for stratigraphic profiling and mineralogical characterization via X-ray diffraction analysis to quantify vertical heterogeneity in quartz, plagioclase, and clay content. Laboratory experimentation progresses through sequential fluid–rock interaction tests: artificial cores undergo linear swelling evaluation in five drilling fluid systems (freshwater, synthetic-based, near-oil-based, calcium-based, and high-performance water-based) at 80 °C to measure hydration inhibition capacity; 6–8 mesh cuttings are subjected to rolling recovery assessment through 16 h hot-rolling and post-sieving mass quantification; and standard cores experience controlled immersion for 2–14 days prior to uniaxial compressive strength testing with stress–strain evolution recording. Concurrently, geomechanical modeling employs the Composite Spring Model incorporating logging-derived parameters to compute continuous in situ stress profiles validated against caliper log interpretations of minimum horizontal stress direction, while 3D azimuth-inclination collapse pressure mapping applies the Mohr–Coulomb criterion through Newton–Raphson iteration across trajectory increments. The synthesized experimental datasets and model outputs ultimately inform the development of drilling fluid selection matrices and well trajectory optimization protocols, maintaining methodological integration through specimen traceability across XRD–mineralogical analysis, fluid interaction experiments, and strength degradation assessments, with experimental–model cross-validation ensuring comprehensive diagnosis of instability triggers across physicochemical and geomechanical domains. The research content and flowchart diagram are shown in Figure 1.

3. Study on Water–Rock Interaction Mechanisms

3.1. Mineral Composition Analysis

Each mineral possesses a distinct crystal structure, exhibiting “fingerprint” characteristics in its diffraction peak positions (2θ angles) and intensity distribution. By comparing these patterns with standard databases (e.g., the ICDD PDF database), mineral types can be identified. This study employed X-ray diffraction (XRD) technology, leveraging the diffraction effect of crystalline minerals on X-rays, to determine mineral types and relative abundances through analyzing diffraction patterns. The XRD pattern is a graphical output plotting diffracted X-ray intensity (y-axis) against diffraction angle 2θ (x-axis). Each mineral’s atomic lattice produces distinct peak positions and intensities, enabling mineral identification, quantification and crystallinity assessment. Whole-rock mineral analysis was conducted on cavings from Well X in the Shaximiao Formation (Figure 2). The research, as shown in

Table 1. Whole-rock mineral analysis.

Collapsed Depth (m)	Quartz (%)	K-Feldspar (%)	Plagioclase (%)	Calcite (%)	Dolomite (%)	Clay Minerals (%)
2881.66 m	42	0	40	0	0	18
2892.83 m	32	11	40	0	0	13
2941.76 m	33	0	56	0	0	11
3281.31 m	30	7	15	0	0	48
3285.46 m	36	4	13	4	0	43
3287.65 m	35	4	13	0	14	34

, reveals that within the Shaximiao Formation of the Sichuan Basin above 3000 m depth, quartz and plagioclase dominate the rock mineral composition, collectively accounting for up to 80% of the total content. Clay minerals constitute approximately 11–18%, and show a decreasing trend with increasing depth.

Below 3281.31 m, the rock composition is dominated by quartz, plagioclase, and clay minerals, with quartz and plagioclase contents reaching 45–49% and clay minerals 34–48%, exhibiting a decreasing trend in clay mineral content with increasing depth.

3.2. Linear Swelling Rate Test

Linear swelling rate testing is a standardized method for evaluating anti-swelling performance by simulating material water absorption or chemical reactions in realistic environments to measure dimensional changes over time. The principle relies on monitoring volumetric expansion caused by water absorption or physicochemical alterations in internal structures under controlled temperature, humidity, or fluid immersion conditions. High-precision transducers or optical instruments record displacement data converted to percentage swelling rate.

Cavings from the Shaximiao Formation in Well X were pulverized, sieved through a 200-mesh screen, dried to a constant weight, and pressed into thin artificial core discs using a hydraulic press. The linear expansion rate in five fluid types (fresh water, high-performance water-based drilling fluid (HPWBF), near-oil-based drilling fluid (NOBDF), calcium-based drilling fluid, and synthetic-based drilling fluid (SBM)) was measured using a linear expansion tester; the results are shown in Figure 3. Comparing the swelling behavior over a specific time span reveals differences in the ability of these fluids to inhibit formation mineral hydration and swelling. The freshwater curve (blue) displays rapid initial expansion, exceeding 20% within approximately 500 min and stabilizing near 24%, significantly higher than other fluids. This indicates the absence of inhibitive components allows for sustained clay hydration. The synthetic-based mud (SBM, light yellow) demonstrates optimal performance, maintaining 0% expansion throughout the test, confirming that its oil external phase effectively prevents water penetration. Both the calcium-based fluid (orange) and NOBDF (dark blue) provide secondary inhibition, stabilizing at 2.4% and 3.1% expansion, respectively. Their mechanisms involve calcium ions enhancing ionic lattice fixation and oil–wetting reducing hydration. The HPWBF (grey), though outperforming freshwater, still reached 5.4% expansion due to the presence of a residual active water phase. Notably, all drilling fluids except freshwater reached a stable expansion plateau after 500 min, reflecting the establishment of a dynamic inhibition equilibrium. Overall, this test provides clear engineering guidelines: SBM and calcium-based fluids are suitable for water-sensitive formations with a high clay content, while conventional water-based fluids are viable only for short-term application in non-swelling formations.

3.3. Roller Oven Recovery Test

Cavings from the Shaximiao Formation in Well X were dried to constant weight, crushed into 6–8 mesh particles, and placed in different drilling fluid types. Samples were hot-rolled at 80 °C for 16 h. Subsequently, they were sieved through a 40-mesh screen, carefully rinsed with fresh water, dried at 105 °C to constant weight, and the recovered cuttings percentage was calculated.

Based on rolling recovery rate tests, a bar chart compares the ability of five drilling fluids to inhibit cuttings dispersion (Figure 4). The results demonstrate that water-based drilling fluid (WBF) achieves only a 78.18% recovery rate, significantly lower than other types, indicating poor inhibition of cuttings dispersion that may trigger borehole enlargement or stuck pipe risks. The high-performance water-based drilling fluid (HPWBF) demonstrates improved performance through the use of polymeric inhibitors, though it remains less effective than oil-based systems. In contrast, near-oil-based (NOBF), calcium-based (CBF with 5% CaCl₂), and synthetic-based drilling fluids (SBFs) exhibit exceptional inhibition performance. Among these, SBF achieves a recovery rate of 98.12%, approaching complete cuttings inhibition. This high efficiency is attributed to its oil–continuous-phase characteristics and the physicochemical synergistic mechanism of sealants, which effectively protect micro-fractures on cutting surfaces. These results empirically demonstrate the technical superiority of oil-based fluids in complex fractured formations, while water-based systems require further additive enhancements to meet high-recovery requirements.

4. Strength Degradation Under Fluid–Rock Interactions

Uniaxial compressive strength (UCS) tests evaluated the wellbore stability impacts of deionized water, HPWBF, NOBF, CBF (5% CaCl₂), and SBF. Fifteen standard core samples (divided into three groups) were immersed in these five fluids at 80 °C for 2 d, 7 d, and 14 d before UCS measurement. To ensure experimental parallelism, cores with proximal serial numbers were selected for identical fluid exposures, maintaining consistent testing conditions except for fluid type and immersion duration variables.

The influence of immersion time on rock mechanical response is characterized by stress–strain curve variations after exposure to synthetic-based drilling fluid. As shown in Figure 5, positive strain represents axial deformation, while negative strain indicates radial deformation. The stress–strain curves can be divided into four distinct stages. ① Compaction Stage: At initial loading (low stress levels), rock deformation primarily manifests as closure of structural planes and compression of infill materials. The stress–strain curve exhibits a concave-upward nonlinear profile with low deformation modulus; total compression depends on structural plane geometry and infill properties. ② Elastic Stage: After sufficient compaction under increasing pressure, deformation shifts to collective elastic deformation of the rock matrix. Unloading at this stage fully recovers deformations, confirming reversible elastic behavior. ③ Plastic Stage: Once stress exceeds the elastic strength limit, irreversible plastic deformation initiates, governed by the intrinsic deformation characteristics of the rock matrix. ④ Failure Stage: At peak stress, compressive resistance deteriorates rapidly. Fracture planes propagate catastrophically, culminating in macroscopic core disintegration. Additionally, prolonged immersion reduces the slope of axial strain curves during the elastic phase, indicating the progressive degradation of the rock’s elastic modulus with extended fluid exposure.

The uniaxial compressive strength of rock samples after soaking in different drilling fluid systems for varying durations is shown in Figure 6. It can be observed that the UCS generally decreases with increasing soaking time across all tested fluid systems (deionized water, high-performance water-based (HPW), near-oil-based (NOB), calcium-based, and synthetic-based (SBM)). Both the type of soaking medium and the time effect exert significant control over UCS degradation. Within the 14-day period, water-based and SBM systems caused UCS reductions of 30.5% and 36.9%, respectively, exhibiting typical chemical degradation characteristics. The calcium-based system initially exhibited higher strength (35.5 MPa at 2 days), attributable to calcium ion-induced cementation. However, it experienced a 21.1% strength reduction over time, indicating a dynamic equilibrium between cementation and dissolution processes. In contrast, the high-performance water-based (HPW) system demonstrated a unique strength recovery capability, with unconfined compressive strength (UCS) rebounding to 24.8 MPa by 14 days. This behavior is hypothesized to result from nano-scale plugging agents facilitating micro-fracture self-healing through a dynamic adsorption–crystallization mechanism. Notably, the near-oil-based (NOB) system showed an anomalous UCS recovery, reaching 37.2 MPa at 14 days, exceeding its 7-day value of 35.0 MPa. This may be due to capillary pressure from the oil external phase effectively inhibiting fluid invasion. These findings offer critical criteria for the engineering selection of drilling fluid systems, underscoring the significant technical advantage of the HPW system’s strength recovery properties in long-term operations.

Based on systematic research of the Shaximiao Formation in the Sichuan Basin under the “dual-carbon” goals, the optimization of low-carbon drilling fluids and wellbore stability control technologies have significantly reduced carbon footprints through multi-dimensional collaborative innovations. For instance, synthetic-based drilling fluids (SBM) achieved zero clay expansion (0% linear expansion rate), completely inhibiting hydration compared to water-based systems (5.4%), while reducing filtration loss by over 20%. The calcium-based system exhibited only 2.4% to 3.1% expansion, simultaneously lowering waste treatment difficulty and carbon intensity. The high-performance water-based system (HPWBF), enhanced with nano-plugging agents, demonstrated a dynamic self-healing effect, with rock strength recovering to 24.8 MPa after 14 days of immersion. This prevented repeated wellbore collapse and the need for redrilling, reducing energy consumption per foot by approximately 15%. The SBM achieved a rolling recovery rate of 98.12% (compared to 78.18% for water-based systems), further minimizing resource consumption. A composite spring model revealed a maximum horizontal stress direction of NE114° ± 13° (90.08 MPa) and a minimum horizontal stress direction of NE24° ± 13° (67.2 MPa). Based on collapse pressure cloud analysis (48–60 MPa), designing the well trajectory along the minimum horizontal stress direction reduced the required drilling fluid density (58–60 MPa), thereby decreasing energy consumption. Improved wellbore stability reduced formation damage, and combined with intelligent drilling technology for precise trajectory control, effectively suppressed methane leakage during operations.

Compared with oil-based drilling fluids, water-based drilling fluids possess inherent advantages in biodegradability and low toxicity: their water continuous phase design avoids the risk of persistent organic pollution from mineral oils or synthetic hydrocarbons present in oil-based systems. Moreover, through the physical sealing mechanism of nano-plugging agents, they significantly reduce the reliance on heavy metal additives and high-molecular-weight polymers, thereby minimizing the threat of toxic substances (such as chromium and barium compounds, and sulfonated asphalt) to aquatic ecosystems from the source. Additionally, water-based drill cuttings can be directly treated via biodegradation, eliminating the need for thermal desorption or special landfill disposal required for oil-based cuttings. Combined with a 15% reduction in implicit carbon emissions during production and transportation, as well as the absence of risks associated with oil phase leakage contaminating highly mineralized formations, water-based systems establish a more sustainable technological pathway aligned with the “dual-carbon” goals.

5. Wellbore Stability Study

5.1. In Situ Stress Distribution

To analyze wellbore stability in the Shaximiao Formation of the Sichuan Basin, beyond understanding the rock strength response characteristics under drilling fluid–formation coupling, it is essential to obtain the original in situ stress distribution profile. While experimental methods and field fracturing tests can determine in situ stress parameters at discrete points, the limited measurement data only provide isolated values, failing to yield continuous stress profiles along a single wellbore or across multiple wells. Conversely, well logging data offer comprehensive records for individual wells and even entire regions, with high resolution, enabling the calculation and analysis of continuous in-situ stress profiles. Numerous in situ stress calculation models based on different geological assumptions have been proposed, including representative models such as the Mathews and Kelly model, Eaton model, Anderson model, Newberry model, Huang’s model, and the Composite Spring Model. Based on a comparative analysis of the advantages and disadvantages of various models, and considering the geological characteristics of the western Sichuan region and prior regional research insights, the Composite Spring Model was deemed most suitable for this study area. However, recognizing that in situ stress represents an internal stress state, a sedimentary structural geological model was established with a free upper boundary and horizontal stress constraints. The Composite Spring Model is a valuable “first-order approximation” tool. It can quickly and intuitively reveal the fundamental principles of stress distribution across rock layers of varying stiffness, particularly in relatively simple, layered, and laterally continuous sedimentary basins, where it provides satisfactory predictions of macroscopic trends. However, its applicability is highly limited in regions with significant heterogeneity, intense tectonic activity, or abundant faults and fractures. In such cases, more sophisticated numerical simulation methods (such as the Finite Element Method FEM or Discrete Element Method DEM) must be employed. These methods can explicitly capture heterogeneity, fault structures, and the nonlinear behavior of rocks, thereby providing more reliable predictions of in-situ stress. Therefore, when using the Spring Model, a clear understanding of its underlying assumptions and inherent limitations is a prerequisite for correctly interpreting and applying its results. Since the Shaximiao Formation in the Gaomiao area behaves as a porous elastic medium, the presence of formation pore pressure influences the overall stress magnitude. Consequently, the Composite Spring Model must also account for the pore pressure effect on in situ stress. The final in situ stress magnitude calculation model [34] is given by Equation (1):

$$\left\{\begin{array}{l}{\sigma }_{\mathrm{H}}={\displaystyle \frac{\mu }{1-\mu }}{\sigma }_{\mathrm{v}}+{\displaystyle \frac{E{\epsilon }_{\mathrm{H}}}{1-{\mu }^2}}+\left({\displaystyle \frac{1-2\mu }{1-\mu }}\right)\times \eta P_{\mathrm{p}}+{\displaystyle \frac{\mu E{\epsilon }_{\mathrm{h}}}{1-{\mu }^2}}\\ {\sigma }_{\mathrm{h}}={\displaystyle \frac{\mu }{1-\mu }}{\sigma }_{\mathrm{v}}+{\displaystyle \frac{E{\epsilon }_{\mathrm{h}}}{1-{\mu }^2}}+\left({\displaystyle \frac{1-2\mu }{1-\mu }}\right)\times \eta P_{\mathrm{p}}+{\displaystyle \frac{\mu E{\epsilon }_{\mathrm{H}}}{1-{\mu }^2}}\end{array}\right.$$

(1)

In the equation, σ_v represents vertical in situ stress in MPa; E and μ denote static elastic modulus and static Poisson’s ratio, respectively; P_p is formation pore pressure in MPa; η represents the poroelastic coefficient; εH and εh are tectonic strain coefficients in the directions of maximum and minimum horizontal stresses. σ_H signifies maximum horizontal stress in MPa, while σ_h indicates minimum horizontal stress in MPa.

As indicated by the in situ stress model, determination requires quantifying: vertical in situ stress, static elastic modulus, static Poisson’s ratio, formation pore pressure, poroelastic coefficient η, and tectonic strain coefficients εH, εh. Vertical stress is obtained by integrating density log measurements. Formation pressure can be directly measured in sandstone formations, while for low-permeability mud–shale formations, it is calculated using Eaton’s method. Static elastic modulus and Poisson’s ratio are derived from dynamic parameters (computed from acoustic log data) through dynamic-to-static conversion relationships established via laboratory triaxial tests. The poroelastic coefficient can be predicted from acoustic interval transit time, exhibiting a logarithmic correlation with compressional wave slowness [34], expressed in Equation (2):

$$\eta =0.562Ln\left(\Delta t\right)-2.6117$$

(2)

Regarding tectonic strain coefficients εH, εh, inversion can utilize acoustic and other log parameters. Alternatively, field fracturing test-derived horizontal principal stress values at discrete points are substituted into the in situ stress prediction model (Equation (1)). Combining known formation elastic parameters, pore pressure, and vertical stress values, point-specific εH, εh are calculated. Comprehensive regional tectonic strain coefficients are determined based on their vertical and horizontal distribution characteristics. This study utilized fracturing test parameters at 2880 m depth in Well X to derive εH = 3.66 × 10⁻³ and εh = 1.33 × 10⁻³.

Applying the density integration model, the vertical overburden stress at 2880 m in Well X is calculated as 69.52 MPa. Formation pressure determined via Eaton’s method is 38.88 MPa. By incorporating these parameters—including the derived poroelastic coefficient, tectonic strain coefficients, petromechanical properties, and formation pressure for the Shaximiao Formation—into the in situ stress prediction model, the minimum horizontal stress is calculated as 67.2 MPa and the maximum horizontal stress as approximately 90.08 MPa.

This study primarily employs the borehole breakout method to determine in situ stress magnitude and orientation based on regional data. After wellbore excavation, stress redistribution occurs around the borehole wall, where circumferential stress peaks along the minimum horizontal stress direction and reaches minima along the maximum horizontal stress direction. Consequently, shear failures primarily occur along the direction of Shmin, resulting in symmetric breakouts, while tensile failures produce drilling-induced fractures aligned with the S_Hmax orientation. The azimuth of Shmin was determined through inversion of caliper logging data, with Figure 7 presenting the caliper logs and interpreted Shmin orientation for Well X in the Zhongjiang area. The results indicate Shmin orientation at NE24° ± 13° and S_Hmax at NE114° ± 13°, revealing that the Shaximiao Formation exhibits higher stress magnitudes in near-EW directions and lower in near-NS directions, suggesting S_hmin aligns approximately N-S, while S_Hmax dominates E-W.

5.2. Collapse Pressure Distribution Characteristics

Borehole fluid column pressure critically governs wellbore stability. Constructing a 3D visualization model of formation collapse pressure enables intuitive identification of high-risk instability zones, providing scientific basis for optimizing mud weight, well architecture design, and drilling parameters to prevent wellbore collapse, reduce non-productive time (NPT), and mitigate unplanned costs. With acquired in situ stress parameters (magnitude/orientation) and rock–drilling fluid coupling responses, these inputs are incorporated into the near-wellbore stress equation (Equation (3)) combined with rock strength criteria to generate collapse pressure cloud maps. Treating the formation as a linear elastic medium, the Fairhurst stress model defines stress components on the borehole wall in polar coordinates as follows:

$$\left\{\begin{array}{l}{\sigma }_r=P_w-\alpha P_p\\ {\sigma }_{\theta }={\sigma }_x^b+{\sigma }_y^b-2\left({\sigma }_x^b-{\sigma }_y^b\right)\cos 2\theta +4{\tau }_{xy}^b\sin 2\theta -P_w-\alpha P_p\\ {\sigma }_z={\sigma }_z^b-2\nu \left({\sigma }_x^b-{\sigma }_y^b\right)\cos 2\theta -4\nu {\tau }_{xy}^b\sin 2\theta -\alpha P_p\\ {\tau }_{\theta z}=2{\tau }_{yz}^b\cos \theta -2{\tau }_{xz}^b\sin \theta \\ {\tau }_{rz}=0\\ {\tau }_{r\theta }=0\end{array}\right.$$

(3)

In the equation, $\left[{\sigma }_r,{\sigma }_{\theta },{\sigma }_z,{\tau }_{\theta z},{\tau }_{rz},{\tau }_{r\theta }\right]$ represents the stress component on the wellbore wall in polar coordinates, MPa; $\left[{\sigma }_x^b,{\sigma }_y^b,{\sigma }_z^b,{\tau }_{xy}^b,{\tau }_{yz}^b,{\tau }_{xz}^b\right]$ denotes the stress component of the in situ stress around the wellbore in the wellbore Cartesian coordinate system, MPa; θ is the azimuthal angle around the well circumference, °; Pp is the formation pore pressure, MPa; ν is the Poisson’s ratio of the formation rock, dimensionless; and Pw is the wellbore fluid (mud) pressure, MPa. The detailed solution process can be found in Reference [5] and is not reiterated here.

Coupled with the Mohr–Coulomb failure criterion, a wellbore collapse pressure prediction model was established. Using the Newton–Raphson iteration method, the wellbore azimuth angle was incremented from 0° to 360° in 5° steps, and the well inclination angle was incremented from 0° (vertical) to 90° (horizontal) in 5° steps. This yielded the collapse pressure contour map at 2880 m depth for Well X, as shown in Figure 8. In this figure, the color gradient from blue to purple indicates increasing collapse pressure magnitude. The circumferential direction of the contour map represents the wellbore azimuth angle, where 0° corresponds to the direction of the maximum horizontal principal stress and 90° corresponds to the direction of the minimum horizontal principal stress. The radial direction represents the well inclination angle, where 0° at the center denotes a vertical well and 90° at the periphery denotes a horizontal well. Additionally, the pink straight line in the figure indicates the direction of the maximum horizontal principal stress.

Figure 8 Analysis: Collapse pressure values exhibit significant azimuthal dependency on well trajectory. High-pressure zones (red areas, 58–60 MPa) concentrate within 0–60° and 180–240° azimuths, indicating requirements for elevated mud pressure to maintain wellbore stability. Conversely, low-pressure zones (blue areas, 48–52 MPa) distribute along 90–140° and 180–220° directions, reflecting enhanced shear failure resistance at these orientations. The cloud diagram displays symmetrical collapse pressure distribution (48–60 MPa range) relative to the maximum horizontal stress and minimum horizontal stress orientations. Notably, trajectories aligned with S_hmin experience higher collapse pressure. This symmetry pattern, governed by the orientation of S_Hmax and S_hmin, provides critical spatial guidance for targeted wellbore reinforcement and mud weight optimization in directional drilling operations.

A comparison was made with the actual drilling fluid density and downhole complications encountered in an exploration well in this block. At a depth of 2880 m in this vertical well, the actual bottomhole pressure measured 50.54 MPa, and complications such as blockages and stuck pipe, as well as difficulties during tripping, were observed. In contrast, a side-tracked horizontal well drilled along the direction of the minimum horizontal in situ stress at the same depth recorded an actual bottomhole pressure of 56.43 MPa. Logging data indicated a borehole enlargement rate within 5%, smooth tripping operations, and successful casing running. These observations are consistent with the collapse pressures predicted for vertical and horizontal wells in Figure 8, validating the rationality of the predictions presented in this study.

6. Conclusions

Wellbore instability in the Shaximiao Formation of the Sichuan Basin poses a major drilling challenge, frequently leading to cavings, collapse, and stuck pipe incidents that hinder operations in the Gaomiao area. This study combines experimental analysis and numerical simulations to systematically investigate rock mechanical degradation, drilling fluid compatibility, and collapse pressure behavior. The key findings include the following.

The mineral composition varies significantly with depth. Shallow sections (<3000 m) are dominated by quartz (30–42%) and plagioclase (40–56%), with clay content between 11 and 18%. Deeper zones (>3280 m) exhibit higher clay content (34–48%), although it paradoxically decreases with depth. Uniaxial compressive strength (UCS) is affected by both drilling fluid type and immersion time: after 14 days, freshwater immersion reduces UCS by 30.5%, while synthetic-based and calcium-based fluids lead to reductions of 36.9% and 21.1%, respectively. Notably, high-performance water-based drilling fluid (HPWDF) shows strength recovery, likely due to nano-plugging agents sealing micro-fractures. Synthetic-based fluid excels in inhibiting clay swelling and cuttings dispersion but is limited by cost and environmental concerns. Calcium-based fluid helps maintain initial strength via ion cementation, though prolonged exposure causes a 21.1% loss due to deteriorating cementation–dissolution equilibrium. Near-oil-based fluid exhibits unusual UCS recovery (37.2 MPa after 14 days), potentially from oleic-phase capillary effects limiting fluid invasion. For cost-effectiveness and engineering adaptability, HPWDF with nano-silica/asphalt plugging agents is recommended to balance short-term strength preservation and long-term inhibition.

In situ stress analysis using the Composite Spring Model indicates a maximum horizontal stress of 90.08 MPa (NE114° ± 13°) and a minimum of 67.2 MPa (NE24° ± 13%). Wellbore stability is highly trajectory-sensitive: boreholes near the minimum stress direction (approx. N-S) experience collapse pressures of 58–60 MPa, while those oriented between 90–140° (approx. E-W) reduce collapse pressure to 48–52 MPa. Well paths should be optimized using collapse pressure contour maps, favoring azimuths with lower stress perturbation, along with real-time mud weight adjustment (1.35–1.45 g/cm³) to dynamically balance wellbore pressure with in situ stress across trajectories. Through the synergistic integration of chemical inhibition in drilling fluids, restoration of rock mass strength, and in-situ stress modeling, the drilling cycle has been shortened by over 15%, while the carbon emission intensity per well has been reduced. This approach provides key technological support for the green and efficient development of natural gas, strongly aligning with the national “Dual Carbon” strategic goals.