Monitoring mining-induced ground subsidence is critical for geotechnical safety, yet conventional borehole-installed sensors in aeolian sand areas are prone to burial and damage. This study employs an integrated monitoring system combining Brillouin optical time domain reflectometry (BOTDR), global navigation satellite system (GNSS), and interferometric synthetic aperture radar (InSAR) to evaluate mining-induced subsidence. A 1-km horizontally deployed fiber optic sensing nerve (FOSN) was implemented to capture the spatiotemporal evolution of subsidence troughs. To enhance cable-soil coupling in the field, specialized T-shaped anchors were designed and deployed. The distributed strain results were combined with GNSS and InSAR measurements to provide a coordinated point-line-plane characterization of mining-induced ground subsidence and its impact on critical surface structures such as high voltage towers (HVTs). Furthermore, laboratory experiments were conducted to analyze deformation patterns under uneven subsidence using fiber optic strain data. Based on the experimentally obtained datasets, a transformation model was developed to quantitatively estimate ground subsidence from distributed strain measurements. This work highlights the effectiveness of horizontal FOSN for ground subsidence monitoring and provides practical guidance for optimizing anchor layouts in aeolian sand regions.

This work is the first computational study of proppant leak-off through coal cleats that accounts for proppant retention in cleats, occlusion formation at cleat entrances, the resulting control of fluid leak-off, and the influence of realistic cleat roughness on these factors. Suspensions are simulated with a coupled lattice Boltzmann method-discrete element method, which explicitly models all physics, including shear-thinning fluid rheology. Firstly, using a simplified computational geometry, it is demonstrated that leak-off and mounding are both minimised when proppant invades and is retained in the cleat. This occurs most effectively with wide proppant size distributions, such as 100/635 mesh. However, when the proppant is larger than the cleat aperture, occlusions form at the cleat entrance, which can lead to significant mounding; this is observed for 100 mesh and 40/70 mesh. These findings are commensurate with an existing benchmark experiment. The present study additionally demonstrates that mounding is significantly reduced for shear-thinning fluids compared to Newtonian fluids. Simulations are then conducted with rough cleats in a realistic fracture channel. Leak-off is smallest for high cleat roughness and when cleats are narrower than a critical width, while proppant retention is largest at the same critical width but only above a certain roughness. Mounding is primarily dependent on the width, as opposed to the roughness. These results are presented as high-fidelity maps which can be directly incorporated into hydraulic fracturing simulations for improved predictions of fluid leak-off and propped reservoir volumes, and which can be tailored for different treatment and reservoir conditions.

The transmission of viruses through the air plays a crucial role in the spread of viral diseases in enclosed environments. The mobility of individuals is a potential factor that contributes to the increment of the propagation of respiratory infections through the air. This research comprehensively focuses on transient modelling of the spread of solid-containing droplets during a cough from a moving person inside a ventilated room through CFD approach. This study investigates a range of moving speeds, from 0 to 1.5 m/s, to illustrate differences in patterns and concentration of droplets during both mobile and stationary conditions of an individual considering the interactions among gas, liquid and solid phases. Interactions between phases are considered through a coupled Eulerian-Lagrangian approach, and discrete phase model (DPM), turbulence model, species transport model, evaporation model and dynamic mesh technique are integrated. Moreover, the influences of effective forces such as buoyancy, Brownian motion, drag, lift, and gravitational forces are included. Regarding the results, motion of individuals significantly affects the airflow pattern and dispersion of droplets, particularly for walking speeds of more than 1 m/s. The results also elaborately indicate that person's movement (from 0 to 1.5 m/s) considerably enhances the turbulent intensity (about 40 %), average air velocity and oscillations in pressure distribution, especially, pressure gradient before and after the moving person (about 1.5 Pa). Additionally, when the person walks at speeds exceeding 1 m/s, most of the particles cannot attach to the person's body due to insufficient time for settling, resulting in an increment in the total number of particles that remain suspended in the air.

The movement of mining equipment complicates predicting and managing methane dispersion, with the dynamic interaction between machinery and gas flows significantly impacting methane concentrations and safety. This computational study investigates the impact of shearer operation on airflow, methane and oxygen dispersions in longwall mining using advanced transient computational fluid dynamics (CFD) simulations. For this purpose, dynamic and overset meshing techniques are innovatively integrated to model both translational (0.5 m/s) and rotational (60RPM) shearer movements, providing an advanced approach to improving accuracy while reducing computational costs in dynamic scenarios. To capture the intricate interactions between the moving shearer and methane dispersion, the computational domain is also divided into five sub-zones. The dynamic model's performance is validated using existing experimental data. Additionally, unlike earlier studies that simplified methane flow from mining face as uniform, our research presents a more realistic scenario where methane concentration peaks in front of the moving shearer and dynamically shifts along the mining face as the shearer progresses. The results indicate that the shearer's movement and counter-rotating cutting drums create significant turbulence, altering airflow that can substantially affect the dispersion of gas around the cutting face and tailgate. The average gas flow velocity at the tailgate after 80 s increases from 3.59 (non-operating condition) to approximately 5.78 m/s as the shearer advances, while average methane concentration at the tailgate reaches 2.4 %. The findings of this research contribute significantly to improving mining safety and ventilation design, particularly through the improvement of real-time methane measurement under dynamic conditions in underground mining.

The ventilation airflow in the longwall section of underground coal mines is significantly affected by the motion of mine equipment, particularly shearer operation. To develop an effective and efficient longwall ventilation design, it is essential to examine the effects of equipment operations on the airflow, dust, and gas distribution patterns. This study aims to numerically investigate the distribution of airflow, methane, and moisture content in the longwall section of an underground coal mine. The longwall shearer is considered to have a translational speed of 0.5 m/s and a cutting drum rotational speed of 55 rpm, traveling from the maingate to the tailgate. The ventilation air is considered to enter the longwall section with a velocity of 2 m/s, a dry bulb temperature of 15°C, and 70 % humidity. The numerical modelling is based on advanced hybrid dynamic and overset meshing techniques. The performance of the hybrid meshing approach is validated against available experimental data. The shearer operation proved to impact the gas flow distribution, especially near the cutting drum picks. The shearer motion, however, showed minimal impact on the moisture distribution along the longwall face. It is evident that as the air moisture content increases, the methane gas flow distribution varies along the longwall.

Horizontal boreholes have been widely used to extract natural gas from coal seams. However, these boreholes can encounter severe instability issues leading to production interruption. Optimizing drilling azimuth is a potential solution for enhancing borehole stability while considering gas production. In this work, we improved and implemented a dual-porosity, fully coupled geomechanical-hydraulic numerical model into COMSOL Multiphysics to investigate into this factor. The sophisticated numerical model incorporates various critical factors, including desorption-induced matrix shrinkage, stress-dependent anisotropic fracture permeability, and the interactions of gas flow and reservoir deformation in matrices and fractures.A suite of simulation scenarios (e.g., varying coal strength) was carried out to quantify the impact of drilling azimuth on coal permeability evolution, cumulative gas production, and the borehole break-out width for Goonyella Middle Seam of Bowen Basin, Australia. The model was calibrated against both theoretical permeability values and field gas production data. Due to the lack of directly measured matrix permeability data, the actual gas production was used to back calculate the best-matched matrix permeability, which is 0.65 μD for this particular work. Moreover, based on the breakout shape and induced volumetric strains around the borehole, drilling along the maximum horizontal stress does not necessarily lead to the best stability of the borehole, as generally believed. A drilling azimuth between 0° and 60° results in similar breakout width, whereas a drilling azimuth between 60° and 90° achieves the most efficient gas production. By considering both gas production efficiency and borehole stability, for this particular reservoir condition, the optimum drilling azimuth is determined to be between 45° and 60°.This study presents a practical approach for determining the optimum drilling azimuth in coal seam gas extraction through in seam boreholes.

Solar energy, particularly solar thermal technology, has gained popularity as a possible long-term replacement to fossil fuels. The application of concentrated photovoltaic-solar thermal (CPV/T) collectors, which are improved by spectral filter fluids (SFF) and nanotechnology, has the potential to provide both higher thermal power for heating and cooling as well as improved electrical power generation. This work contributes new insight by quantifying the influence of nanoparticle agglomeration on collector performance and highlighting the challenges associated with the heterogeneous distribution of nanoparticles in CPV/T systems. The study employed coupled Eulerian multiphase modeling and discrete ordinate (DO) radiation modeling to examine slip velocity, nanoparticle diameter (including agglomeration), and suspension concentration. Population balance modeling (PBM) was utilized to determine the nanoparticle size distribution, and the obtained results were validated through comparison with experimental and numerical studies. When neglecting the effect of agglomeration and breakage of the non-solar participating media, the maximum error for this configuration was found to be 3.58% when compared to experimental work. From the solar participating study, in terms of electrical energy production, the best performance obtained was 16.64% with a volume fraction of 0.005% when considering agglomeration and breakage. It was also found that the Sauter diameter increases with volume fraction as the tendency for nanoparticle agglomeration increases. This study provides a broader view of the application of multiphase modeling in solar participating and non-solar participating media and, additionally, provides insight on the effect of various boundary conditions on the key system performance indicators. To extend this work, the flow Reynolds number can be increased in addition to varying the type of working fluid.

The behaviour of non-Newtonian fluids, and their interaction with other fluid phases and components, is of interest in a diverse range of scientific and engineering problems. In the context of the lattice Boltzmann method (LBM), both non-Newtonian rheology and multiphase flows have received significant attention in the literature. This study builds on that work by presenting the development and validation of a phase-field LBM which combines these features in three-dimensional flows. Specifically, the model presented herein combines the simulation of Herschel-Bulkley fluids, which exhibit both a yield stress and power-law dependence on shear rate, interacting with a Newtonian fluid. The developed model is verified and validated using a diverse set of rheological properties and flow conditions, which in their totality represent an additional contribution of this work. Comparison with steady-state layered Poiseuille flow, where one fluid is Newtonian and the other is non-Newtonian, showed excellent correlation with the corresponding analytic solution. Validation against analytic solutions for the rise of a power-law fluid in a capillary tube also showed good correlation, but highlighted some sensitivity to initial conditions and high velocities occurring early in the simulation. A demonstration of the model in a microfluidic junction highlighted how non-Newtonian rheology can alter behaviour from cases where only Newtonian fluids are present. It also showed that significant changes in behaviour can occur when making small and smooth changes in non-Newtonian parameters. To summarise, this work broadens the range of physical phenomena that can be captured in computational analysis of complex fluid flows using the LBM.

The application of spectral filter fluids (SFF) in concentrated photovoltaic thermal (CPV/T) collectors is becoming increasingly popular due to their ability to improve energy yield by reducing the photovoltaic (PV) cell temperature. Such an integrated system is also known as spectral filter CPV/T (SF-CPV/T). The material used to construct a PV cell in a SF-CPV/T possesses a specific wavelength window, known as the ideal spectral response (ISR), within which the transmittance is higher and PV power generation is maximised. ISR could be varied depending on the context of PV spectral response and solar spectrum. In this work, a coupled numerical approach that combines modelling of radiation, electrical, and thermal power has been developed and applied to study the behaviour of a SF-CPV/T with SFF. Five commercial solar fluids, namely Therminol®VP1, 5g/L CoSO4, Therminol®66, Duratherm S, and Dowtherm A, were investigated due to their good optical transmittance and stability at high operating temperature. A merit function was defined and calculated to assess each fluid’s performance and economic feasibility. It was found that the viscosity of the fluid has the greatest effect on increasing the bulk fluid temperature, followed by other thermophysical properties. Non-dimensional analysis of the collector indicates that, when using 5g/L CoSO4, the collector length must be longer to achieve the same power output as when using Therminol®66 under the same operating conditions. However, Therminol®VP1 was found to result in a higher merit function, especially when the worth factor was higher. By including the ideal spectrum criteria for the PV cell and non-dimensional analysis of the candidate solar fluids and band-dependent optical parameters in the numerical modelling, the performance of the SF-CPV/T collector can be more accurately predicted. To extend this work, future investigation will look into the influence of the heat loss coefficients and the local climate conditions on the performance of the collector. Additionally, methodological development to include band-based solar radiation and multiphase modelling of the nanofluid will be conducted to further improve model accuracy.

Given the concerns surrounding the possibility of crosscontamination caused by the airborne transmission of respiratory aerosols (> 5 mu m in diameter) and droplets (> 5 mu m in diameter) containing infectious viruses, there is a great need for simulations that reliably characterize the behaviour of these particles in real-world scenarios. This study performs a comprehensive transient CFD analysis to investigate the transmission of virus-carrying aerosols and droplets released through coughing by a mobile patient within a typical room equipped with a ventilation system. This computational study elaborately examines how particle size and relative humidity impact the dispersion of aerosols and droplets carrying virus in both mobile and stationary conditions of patients. To enhance the accuracy of this study, effective factors such as evaporation of liquid content within aerosols and droplets and random distribution of the particles, along with considerations for buoyancy, drag, lift, Brownian motion, and gravitational forces, are taken into account. To investigate the influence of aerosol and droplet size, this study considers uniform size distributions of 1, 10, and 100 mu m in diameter, comprising 98.2% liquid water and 1.8% solid content. Additionally, different relative humidity levels, 0%, 50%, and 90%, are incorporated to indicate their impact on the dispersion pattern and residence time of the particles in both stationary and dynamic scenarios. According to the results, high levels of relative humidity and individuals' movement significantly affect the turbulence intensity, airflow pattern, travelling distance, residence time and trajectory of particles, air pressure, and density distributions in such environments.

A growing number of hydraulic fracturing stimulation treatments rely solely on the deployment of 100-mesh (i.e. 150μm) proppant. Further, these may be preceded by post-pad injection of microparticles (i.e. 5-75μm) with the intent of activating natural fractures. The objective of this work is to describe fundamental new insights on the behaviour of polydisperse microparticle suspensions in hydraulic fractures in terms of screen-out and leak-off. This study was undertaken using a high-fidelity computational model of suspension transport based on the lattice Boltzmann method (for fluid mechanics) and discrete element method (for particle mechanics). The approach has been previously validated against analytical solutions and experimental data for suspension flows in channels, and applied to study fundamental aspects of size segregation and clogging (i.e. screen-out) of microproppant. This fully-resolved modelling approach captures two-way hydrodynamic coupling, electrostatic interactions, fracture roughness and tortuosity, and non-Newtonian fluid rheology. The sole assumption of significance is that particles are assumed to be perfectly spherical. One of the challenges of deploying micron-scale proppant in hydraulic fracturing stimulation treatments is poor control of the particle size distribution. This inevitably results in a proportion of the injected proppant being small enough to be susceptible to electrostatic (rather than just mechanical and hydrodynamic) forces. This study clearly demonstrates the increased probability of screen-out that results as a consequence of electrostatic particle-particle and particle-wall interactions, and shows how this effect reduces as the minimum particle size increases. Analysis of particle leak-off off into transverse cleats also demonstrated how the combination of cleat width and particle size result in the formation of occlusions at cleat intersections, the reduction of leak-off rates in cleats, and the retardation of proppant transport in the primary fracture. These findings are significant because tight size control is not economically feasible for microproppant, and so hydraulic fracture engineering must accommodate their characteristic behaviours. This will result in more effective stimulation of coalbeds, but is applicable to all jobs where small proppant is proposed for injection into natural fracture systems.

The depletion of fossil fuels and the accumulation of plastic waste are both serious issues facing the world today. This study investigated the potential use of plastic waste as an alternative fuel by utilizing diesel (D) to depolymerize low-density polyethylene (LDPE) in the presence of co-solvents. Several co-solvents were examined: a-pinene, toluene, and o-xylene, with ratios of 10% weight in the LDPE/diesel blend. The blends were theoretically tested in a diesel engine to investigate the impact on the combustion characteristics and emissions. According to tested data, the in-cylinder peak pressure and temperature decreased within the blend of D/LDPE in comparison to diesel at all engine loads. However, when co-solvents were added to the D/LDPE blend, in-cylinder pressures increased during the combustion at all loads, with improvements of 0.48%, 0.5%, 0.6%, and 0.4% for the blends of D/LDPE (10% toluene and 1 bar), D/LDPE (10% a-pinene and 1 bar), D/LDPE (10% xylene and 1 bar), and D/ LDPE (10% xylene and 10 bar), respectively, in comparison with diesel at 100% load. Co-solvent additions such as a-pinene and o-xylene to the plastic/diesel blend resulted in a substantial and gradual increase in the heat release rate at high loads. In respect to emissions, carbon monoxide, unburnt hydrocarbon, and soot levels were raised by adding LDPE plastic to D and reduced by adding co-solvents. The most effective co-solvents for decreasing carbon monoxide, unburnt hydrocarbon, and soot emissions were pinene, followed by xylene.

Hydrodynamic clogging in planar channels is studied via direct numerical simulation for the first time, utilising a novel numerical test cell and stochastic methodology with special focus on the influence of electrostatic forces. Electrostatic physics is incorporated into an existing coupled lattice Boltzmann-discrete element method framework, which is verified rigorously. First, the dynamics of the problem is governed by the Stokes number, $St$ . At low $St$ , the clogging probability, $P$ , increases with $St$ due to increasing collision frequency. At high $St$ , however, $P$ decreases with $St$ due to quadratic scaling of hydrodynamic force acting on arches. Under electrostatic forces, clogging is well represented by the wall adhesion number, $Ad_w$ . For $Ad_w \lesssim 4$ , the mechanical dependence on $St$ is exhibited, while for $4 < Ad_w < 20$ , there is a transition to high $P$ as sliding along, and attachment to, the channel surface occurs increasingly. For $Ad_w \gtrsim 20$ , clogging occurs with $P > 0.95$ . Particle agglomeration, however, can also decrease $P$ due to diminished interaction with channel walls. Distinct parametric regions of clogging are also observed in relation to the channel width, while a critical width $w/d^*=2.6$ is reported, which increases to $w/d^*=4$ with strong electrostatic surface attachment. The number of particles that form stable arches across a planar channel is determined to be $n=\left \lceil {w/d}\right \rceil + 1$ . Finally, sensitivity to the Coulomb friction coefficient is determined in favour of calibrating numerical parameters to bulk system behaviour. The greatest sensitivities occur in situations where the arch stability is lowest, while clogging becomes independent of friction for strong wall adhesion.

Shale anisotropy can lead to deviated patterns of hydraulic fractures in multi-stage hydraulic fracturing (MSHF) operations, significantly hindering hydrocarbon production. Such deviation usually occurs in formations where the induced fractures interfere with natural bedding planes or where the stress shadow is created. In this work, the elastic behavior of a transversely isotropic shale rock during MSHF is investigated using the extended finite element method (XFEM) in conjunction with the cohesive zone model (CZM) in ABAQUS. In comparison to the linear elastic fracture mechanics approach, the CZM considers the plasticity and softening effects at the fracture tip, and therefore, leads to a more accurate prediction of fracture geometry. The XFEM model is capable of simulating fractures with an arbitrary path without requiring re-meshing. The 2D numerical model is first verified against the Kristianovich-Geertsma-de Klerk (KGD) analytical solutions. Five injection clusters in a single fracturing stage are simulated to analyze the effects of the material isotropy, the dip angle of shale bedding, and the fracture spacing on the geometry of multiple hydraulic fractures. The transversely isotropic poroelasticity model predicts narrower and longer hydraulic fractures than the isotropic poroelasticity model. The bedding dip angle plays a crucial role in the deviation and the shape of multiple hydraulic fractures. Increasing the dip angle from zero to 90 degrees results in an increase in the average fracture width and an increase-peak-decrease trend in the fracture length. A strong correlation is observed between the fracture spacing and the fracture geometry, and a spacing of 75 m is considered optimal because it results in no discernible stress shadow at all dip angle cases.

Plastic pollution is increasing as the management of plastic waste has become a global challenge. The possibility of turning plastic waste into fuel at low temperatures has been successfully reported in our recent research. However, the low recycling ratio and higher fuel viscosity were the main limitations. The focus of this research was to improve the depolymerization process of low-density polyethylene (LDPE) in diesel by introducing various amounts of co-solvents. The resultant product can be directly used as transportation fuel without further refining. Several co-solvents were examined in the LDPE/diesel blend: di-butyl ether, a-pinene, toluene, and o-xylene under various pressures (1, 5, 10, and 20 bar). The quality of fuel products was evaluated using differential scanning calorimetry (DSC), an Anton Paar MCR302 rheometer, and a bomb calorimetric thermometer. The degradation behavior of LDPE in diesel is affected by the co-solvents, and the highest dissolution occurred with 10 wt% co-solvent in diesel. Among the tested reactor pressures, the most effective pressure was 10 bar resulting in 92.4% LDPE dissolution. The addition of co-solvents improved the quality of the fuel blend in terms of the kinematic and dynamic viscosities compared with the PD/LDPE blend. The higher heating value (HHV) of the resultant fuel blend was higher compared to non-solvent, pure diesel-derived LDPE fuel blends. The values of HHV for PD, PD/LDPE (10% a-pinene and 1 bar), PD/LDPE (10% xylene and 1 bar), and PD/LDPE (10% xylene and 10 bar) were 45.24 MJ/kg, 45.60 MJ/kg, 46.05 MJ/kg, and 46.14 MJ/kg, respectively.

Magnetorheological plastomer (MRP) dampers are controllable semi-active smart devices, which are commonly used in the field of vehicle suspension systems, structural engineering, aerospace engineering and military equipment. This study was conducted to numerically investigate the performance of MRP dampers with different annular shear gap sizes. The MRP used in the analysis consisted of 60% Carbonyl Iron Powder (CIP) and 40% Polyurethane Matrix (MRP60). The focus of the analysis was on the MRP60 flow behaviour in the annular shear gap region, where the induced magnetic field generated a range of dynamic damping forces depending on the size of the annular shear gap. The numerical simulation involved an integration of the finite element analysis (FEA) to examine the magnetic field in the annular shear gap region as well as the computational fluid dynamics (CFD) analysis to simulate the damping characteristics of the MRP damper. The relationship between the yield stress and the magnetic flux density for MRP60 was used for the analysis. The results of the FEA showed that as the size of the annular shear gap increased, the magnetic flux density and the yield stress decreased. From the CFD results, it was observed that the damping force decreased as a result of increasing the size of annular shear gap.

Carbon Capture and Storage (CCS) is considered an important means to reducing CO2 emissions. One of the key technical risks associated with large scale CCS implementation is the potentially induced fault reactivation. To assess this risk, a coupled numerical approach was established by integrating three different software packages over four steps. Firstly, a field-scale structural geological modelling was conducted using the JewelSuite based on available geological, geophysical, and logging data. Secondly, dynamic CO2 injection was simulated using the reservoir simulator IMEX to predict reservoir pressure build-up. Thirdly, the pressure build-up and the associated reservoir uplift were simulated in the Finite Element software Abaqus. Finally, the Finite Element output database was re-imported into the JewelSuite to delineate the risk areas of fault reactivation. The feasibility of this new approach was tested on a real geological site featuring two faults. The results suggest that, for a given fault setting, the horizontal well placement and bottom hole injection pressure (BHIP) are two critical factors in determining the risk of fault reactivation. To minimise the risk, horizontal wells should not be placed within the intermediate region between faults. Where the borehole placement cannot be optimised, the critical BHIP should be estimated and controlled during operation.

This study utilises computational modelling to investigate the application of thermal nanofluids in direct absorption solar collectors (DASC), including the effect of optical and thermal properties on the prediction of thermal energy production. A single-phase, computational fluid dynamics (CFD) model has been used to understand the hydrothermal behaviour of nanofluids operated at high temperature. A non-gray discrete ordinate radiation model (DORM) was coupled to the CFD to capture the dependence of optical parameters on the bulk temperature variation. In order to increase the robustness of the numerical modelling approach, the extinction coefficient of the solar participating media, including the reflective and transmissive behaviour of the semitransparent wall, were taken into consideration. Results show that the exergy efficiency is lower when using fully diffusive conditions at the semitransparent wall as opposed to fully specular conditions. Overall, exergy efficiency decreases with the Nusselt number, Nu, (until Nu=10∼12), although it increases with non-dimensional irradiation. When modelling with temperature-dependent as opposed to temperature-independent properties of the working fluid, the Carnot efficiency was found to vary by 0.67% to 3.30%. Additionally, a correlation between the hydraulic entrance length and Reynolds number, Re, (in the range Re<1,000) has been established and this can be applied for the further hydrothermal analysis. This study serves as a basis for future numerical modelling of any volumetric solar receiver that takes into account the various key system performance indicators of the solar participating media.

Horizontal boreholes have been routinely applied to coal seams as a cost-effective way to maximize coal seam gas production. However, these wells can encounter severe instability issues during field development due to significant horizontal stress loss and change in deviatoric stresses acting on the borehole. In this work, a general dual-porosity dual-permeability model is established and assigned to a coupled gas flow and coal deformation numerical model to investigate permeability change and borehole break-out regarding different in-situ stress regimes around a horizontal borehole. Mohr-Coulomb failure criterion is used in this model. The results show that drilling parallel to the maximum horizontal stress direction neither achieves the best stability of the borehole nor maximizes the permeability ratio. Drilling along the minimum horizontal stress direction would maximize the permeability ratio, but it has the worst stability. The optimal drilling direction window considering both permeability ratio and borehole stability is recommended to be between 45– 60°.