Li, Ningjun (Haimo Technologies Group Corp.) | Zheng, Ziqiong (Haimo Technologies Group Corp.) | Guo, Peihua (Haimo Technologies Group Corp.) | Hao, Xipeng (Haimo Technologies Group Corp.) | Chen, Bingwei (Haimo Technologies Group Corp.) | Ren, Yao (Haimo Technologies Group Corp.)
Ordos basin is known for its tight sandstone formations and fracturing has been the most effective approach to improve production[
To successfully treat and reuse flowback fluid in Ordos basin, two major obstacles have to be overcome: First, in the fracturing process, the local common practice is to add the entire designed amount of gel breaker at the end of propant pumping job, to avoid sand plugging and sanding out. This incorrect, but common practice results in incomplete breaking of gel of the frac fluid, which inevitably flows back leading to greatly increased difficulties in flowback fluid treatment. Secondly, organic boron crosslinking agent is widely used as crosslinking agent in the guar fluid system in this area. As boron compounds are extremely difficult to be removed during flowback fluid treatment, proven treatment methods alone cannot make the treated water reusable in making new frac fluids.
Technology and processes were developed to manage four key factors that affect the performance of guar frac fluid configured with treated flowback fluid: a) Metal ions, b) Bacteria, c) Breaking agent, d) Crosslinker. Mobile units developed in association with treatment processes and agents also help avoid secondary pollution from the transportation of fresh and flowback fluid. In 2017 and first quarter of 2018, more than 15,000 cubic meters of flowback fluid have been successfully treated and reused. One third of the treated water was guar frac fluid and was reused in making new frac fluid, reducing the need for fresh water significantly. Fracturing service company conducted tests on the treated water and found that the performance of the fluid configured with the treated water completely satisfy the requirements of the SY/T6376-2008 "General Technical Requirements for Fracture Fluid" and SY/T 5523-2016 "Oilfield Water Analysis Method" standard. Frac fluid configured with the treated water was successfully applied to the stimulation jobs of horizontal wells, resulting in double savings to the operators: purchase of fresh water and transportation of flowback fluid (to treatment centers) and fresh water, also avoided secondary environmental impacts such road safety hazard and fluid seepage.
With the treatment and reuse of flowback fluid, savings up to 8% of total frac costs per well were observed which could lead to 100+ million RMB within 2018 alone. Most importantly, the technology can effectively relieve environmental pressure and reduce the need of fresh water which is a scarce in this area.
The scope of applied geomechanics in the petroleum industry has been on the rise over the past decade. Geomechanics analysis has shown to increase the overall value to various projects and create positive technical synergies between various groups involved with project development. This paper highlights a comprehensive geomechanics study carried out on a deepwater Gulf of Mexico (GoM) field for optimizing field development with respect to critical drilling, completion and reservoir issues. A detailed and well-calibrated wellbore stability model applied to the Medusa field has shown to reduce the drilling operational costs by enhanced well planning. The log-based geomechanics models used to evaluate sand production and pressure dependent pore volume compressibility have aided in reducing project risks and increasing project life. The intent of the paper is to show the practical risk reduction and cost savings that are possible through geomechanics analyses in high-cost, high- risk projects in deepwater arenas such as the Gulf of Mexico.
Operations in deepwater are costly and time sensitive requiring optimum well construction design. Incorrect engineering analysis can affect the Net Present Value (NPV) of the project and also setback the future project objectives. In order to reduce operations and engineering uncertainties, a detailed geomechanics study was conducted for the Medusa field. The two main objectives of the study were:
Cost saving during operations
Risk estimations, project planning and optimization.
The deliverables for the geomechanics study were based on:
Drilling issues comprised of wellbore stability analysis for the development well program.
Completion issues relating to prediction of sand production and completion strategies.
Reservoir management issues consisting of pore volume compressibility determination for various pay sands.
Wider utilization of the study by mapping the mechanical characteristics of the reservoir, overburden and other relevant formations, to assist with frac and pack design and subsidence prediction.
An extensive array of drilling, logging, core and reservoir data from the few existing exploratory wells in the field provided the necessary framework and calibrations to characterize in-situ stresses, formation mechanical behavior and mechanisms of potential borehole instability.
The wellbore stability analysis, consisting of mud weight window prediction and optimal well trajectory analyses, was carried out for five of the development wells. The success of utilizing this stability analyses for two of the recently drilled wells are discussed, highlighting the value of conducting such studies on high impact wells.
Sand production prediction was carried out for all the main sand packages identified and encountered by the existing wells in the field. Formation compressive strengths and other mechanical parameters were computed every 0.5ft across the entire pay zones using a proprietary log-based prediction/simulation program. Critical drawdown pressures to predict sanding potentials were derived based on the static mechanical properties and expected flow conditions. Such detailed depth-based sand production prediction aided the project engineers in determining optimal production as well as completion strategies.
This paper presents an innovative log-based method of determining pore volume compressibility as a function of pore pressure depletion. The approach considers changes in reservoir stress associated with pore pressure change (stress path) and incorporates constraints that ensure deformations are within elastic bounds. The approach incorporates the effect of stress anisotropy by using elastic moduli derived from stress-strain curves under simulated triaxial loading conditions. The triaxial condition pore volume compressibility was then converted to that of unaxial strain equivalent, which best describes the existing reservoir characteristics. The proposed methodology is particularly useful for predicting pressure-dependent pore volume compressibility where core specimens are either not available or in situations where laboratory measurements are prohibitively laborious and time consuming.
For input, the method requires bulk modulus, compressive strength and other mechanical properties that characterize an elastic material, preferably predicted using a log-based mechanical property algorithm in order to generate a foot-by-foot profile of pore volume compressibility. A continuous profile of uniaxial strain pore volume compressibility with depth from log provides quick assessments of pore volume compressibility variations across the reservoir intervals. It is also useful and cost-effective for constraining pore volume compressibility of all the reservoirs penetrated by the well (and logged) but with only limited core data available for calibrations.
The example shallow oil well data illustrates that pore volume compressibility decreases with decreasing pore pressure (or increase effective stress). An inverse pore volume compressibility to strength relationship was also observed. It was also observed that pore volume compressibility decreases with increasing porosity until the effective porosity reaches a critical minimum value. At porosity higher than the critical value, the pore volume compressibility increases with increasing porosity. This may suggest that reservoirs with a porosity less than the critical value are more likely to be under pressure drive, while reservoirs with porosity higher than the critical value are more likely to be under compaction drive.
Pore volume compressibility, defined as the relative change in pore volume of a rock with respect to a change in pore pressure, is of fundamental importance in reservoir evaluation and management. It is an important parameter in material balance calculations and water/compaction drive performance studies. Its importance is becoming even more critical with recent fervent deepwater exploration and exploitation activities. These reservoirs, due to their depositional environments, tend to be weakly consolidated and oftentimes over-pressured. Compaction as a result of fluid withdrawal has major implications on reserve estimation, reservoir performance, casing integrity and seafloor subsidence. Due to the high risks and high level of uncertainties involved in deepwater projects, accurate estimations of pore volume compressibility, therefore, play a vital role in the project economics. In conventional depletion type reservoirs without strong pressure supports, reducing the fluid pressure causes changes in the reservoir effective stresses, which subsequently impact the volumetric changes of pore spaces. The engineering parameter quantifying these volumetric variations is pore volume compressibility.
Ong, Seehong (Baker Atlas) | May, Andy (Kerr-McGee Oil and Gas Corporation) | George, Ian (Kerr-McGee Oil and Gas Corporation) | Walker, Tim (Kerr-McGee Oil and Gas Corporation) | Zheng, Ziqiong (Baker Atlas)
This paper presents a case study where the creative use of wireline-derived sand strength in the offshore production test process improved operational efficiency. Sand strength was computed using wireline data only and subsequent testing of two wells verified the accuracy of the calculations.
Well logs provided the basis for evaluation of formation petrophysical and mechanical properties. These log-based properties enabled a foot-by-foot profile of the critical drawdown pressure to be generated. Analysis of critical drawdown pressures in the Bohai Wells A and B, located in the northeastern People's Republic of China, provided a cost-effective method to predict potential sand production problems. This information proved invaluable in reducing the risk and costs associated with the testing of viscous oils in the shallow, moderately unconsolidated sandstones of Bohai Bay.
The Bohai Bay basin is a fluvial environment located in northeastern People's Republic of China. The area has been home to numerous discoveries over the years. The shallow environment is one of heavy oil and unconsolidated fluvial sands. Kerr-McGee has been a partner with CNOOC for a number of years and is currently delineating a discovery in the Tertiary sandstones. Two wells, Bohai-A and Bohai-B, encountered the geological units of interest, which are the Upper and Lower Minghuazhen and Guantao formations.
Qualitative assessments of the compressional and shear wave slownesses suggested varying formation strengths ranging from weak but competent to unconsolidated. Because of the inherent characteristics of the formations, there was a concern about sand production during production testing. In unconsolidated sand, the decision to gravel pack is usually clear. The decision is harder for weak rock because the need for sand control often depends on the desired drawdown. The latter scenario has important implications in well test operations. The unnecessary application of sand-control techniques, as a precaution against anticipated sand production, can cause an increase in well testing cost and a possible reduction in formation productivity. However, if the operating conditions dictate the need for sand control, such techniques allow a formation, which otherwise might have to be abandoned, to be properly tested and evaluated. The ability to accurately predict the critical drawdown pressure (CDP) is, therefore, vital for sand-control decisions. This ability also allows intervals that have no sand control to be produced at optimal rates.
The CDP, which is the difference between the average reservoir pressure and the bottomhole flowing pressure above which mobilization of unconsolidated and/or disaggregated sand grains broken by perforation and concentrated stress around the borehole is expected to occur, provides useful information for offshore production tests and downhole completion designs. It allows sound engineering decisions to be made on the needs for 1) sand-controlled completion (e.g., gravel packing) and 2) the mobilization of gravel pack equipment for offshore well testing. The CDP profile, which is a continuous presentation of critical drawdown pressure with depth, also provides useful information for effective selective perforation designs. The latter provides an alternative to sand production mitigation without having to install any downhole sand filtration hardware.
This paper describes the approach for developing the critical drawdown pressure and illustrates how the information is being successfully applied in the design and implementation of offshore production tests of the Bohai Bay wells. The creative use of the information results in substantial savings for Kerr-McGee and partners in terms of the logistical costs associated with the mobilization of gravel pack equipment and personnel.
Knowledge of the stress path (ratio of the change of effective horizontal stress to the change of effective vertical stress due to the reservoir pore pressure drawdown) that reservoir rock will follow during production and its effect on reservoir properties are critical for reservoir management decisions to maximize productivity. However, in-situ stress measurements needed to determine reservoir stress path are difficult and expensive to conduct, and the measurement may take several years to collect. In situ stress measurements in carbonate and clastic reservoirs indicate that the reservoir stress path is not isotropic loading (equal to 1.0) and can range from 0.14 to 0.761-2. In the laboratory usually reservoir stress path is calculated using uniaxial strain condition. The calculated uniaxial strain stress path can be significantly different from the measured stress path (Table 1).
This paper presents a finite element numerical model to predict the reservoir stress path due to the reservoir drawdown. The model incorporates essential geologic and geomechanical factors that may control reservoir stress path during production. The effect of size and geometry of the reservoir, contrasts in elastic properties between the reservoir and the bounding formations rock, transverse isotropy in elastic properties, plasticity, far-field boundary conditions, anisotropy in the reservoir initial stress states and stratigraphy and structure of the reservoir was studied. For isotropic reservoir properties, stress path becomes lower as the aspect ratio of length to thickness increase. Lenticular sandstone reservoirs have a higher stress path than blanket sandstone reservoirs that are continuous across the basin. Transverse isotropy with lower vertical elasticity than horizontal of reservoir rock increases the stress path, compared to isotropic conditions. Plasticity of the reservoir rock decreases the stress path significantly at the lower aspect ratio of reservoir, but increases with the increase of aspect ratio. However, plasticity of the bounding formations reverses the process.
Validation of the simulated results was done by simulating and matching the stress path of an existing reservoir. Reservoir initial conditions were obtained from published literature2 and considering that initial conditions simulated reservoir stress path was matched with the actual field measurements by adjusting the magnitude and orientation of the geological and geomechanical parameters.
Matrix permeability and pore volume compressibility are fundamentally important characteristics of hydrocarbon reservoirs because they provide measures of reservoir volume and producibility. Laboratory studies have shown that these properties are stress-sensitive and are usually measured under hydrostatic (isotropic) stress conditions that do not truly reflect the anisotropic stress states in most reservoirs and do not adequately simulate the evolution of deviatoric stresses in a reservoir during production3,4. Recent laboratory studies have shown that permeability and compressibility are dependent on the deviatoric stress and change significantly with reservoir stress4 (Figure 1).
This paper presents a case example to illustrate how a consistent geomechanical approach and coherent data integration are used to provide the basis for the evaluation of formation mechanical properties, characterization of in-situ stresses, identification of wellbore failure mechanisms, and calibration of a wellbore stability model. The calibrated model is then used in conjunction with the proposed build-and-hold and catenary well profiles to generate the mud weight windows for an ERD well. To provide further insight into the wellbore stability with respect to wellbore trajectories and in-situ stresses, contour plots of mud weights are generated for a critical depth interval. These contour plots delineate potential instability areas and are critical for predrilled well trajectory designs. Additionally, an innovative approach to establish the lower limit of mud weight for wells that
Resak Field, located offshore of Peninsula Malaysia, is operated by PETRONAS Carigali Sdn. Bhd. (PCSB). The field is currently under development and 13 wells will be drilled radially from a production platform in a ‘bow-tie' pattern. In order to cost effectively drain the proven reserves without having to add a lightweight structure and pipelines connecting to the processing platform, some of the development wells will be drilled extended-reach. Resak LOC-11 is an extended-reach well designed to drain the southeastern portion of the field reserves. Two wellbore trajectories, build-and hold and catenary, were proposed. Fig. 1 and Fig. 2 show the schematics for the two well plans.
Extended-reach drilling (ERD) can be extremely costly if not optimally designed. Occurrences of borehole instability-related problems, such as stuck pipes, fishing and sidetracking operations, have been reported to increase with increasing wellbore inclinations1 and extended-reach drillings. In addition to escalating costs, wellbore instability can also result in poor hole conditions, which ultimately would affect the quality of reservoir characterization and the effectiveness of primary cementing. Poor primary cementing in high angle gas wells is of particular concern because of possible gas migration during cement setting. Faced with the challenges of delivering the well within AFE, the Resak project team decided to conduct the wellbore stability assessment for the proposed ERD well. The results of the geomechanical analysis will be used to: 1) decide which wellbore trajectory, and 2) establish the mud weight program that will mitigate wellbore instability and improve operational performance.
An extensive in-house database is available through comprehensive data acquisition programs during the exploration and appraisal phases. Due to the data quality of older wells, only the most recent appraisal well data (Resak 6F-18.4) was selected to provide the basis for rock mechanical properties and in-situ stress characterizations as well as wellbore stability model calibrations. The calibrated model was then used in conjunction with LOC-11's well plans (hole deviations and azimuths) to generate the stable mud weight profiles. The safe operating mud weight window can then be implemented together with good oilfield drilling practices to mitigate potential borehole instability-related problems. The purpose of this paper is to present a consistent geomechanical approach for developing the optimal mud weight program for LOC-11 ERD.
This paper presents an analytical model for the prediction of the on-set of sand production or critical drawdown pressure (CDP) in high rate gas wells. The model describes the perforation and open-hole cavity stability incorporating both rock and fluid mechanics fundamentals. The pore pressure gradient is calculated using the non-Darcy gas flow equation and coupled with the stress-state for a perfectly Mohr-Coulomb material. Sand production is assumed to initiate when the drawdown pressure condition (i.e. at CDP) induces tensile stresses across the cavity face. Both spherical and cylindrical models are presented. The spherical model is suitable for cased and perforated applications while the cylindrical model is used for a horizontal open-hole completion.
For input, the model requires cohesive strength and an internal friction angle that characterizes a Mohr-Coulomb material; preferably predicted using a log-based mechanical properties algorithm in order to generate a foot-by-foot profile of the maximum sand free drawdown for gas wells. The example GOM well illustrates a continuous profile of critical drawdown with depth, providing quick identification of potential sand producing zones. This allows a gravel pack decision to be made in the period between logging and completion. It also facilitates the design of selective perforation programs.
The model demonstrates that non-Darcy flow has a considerable effect on the sanding tendency of weak but competent rock, and completion decisions in high gas-rate wells that neglect the influence of non-Darcy flow could be overly optimistic. It also shows that the CDP of a horizontal well with slotted liner is less than that of the corresponding cased and perforated completion.
High-rate gas well completions are common practice in offshore developments and among some of the most prolific gas fields in the world. These fields typically have reservoirs that are highly porous and permeable but weakly consolidated or cemented, and sand production is a major concern. Because of the high gas velocity in the tubing, any sand production associated with this high velocity can be extremely detrimental to the integrity of surface and downhole equipment and pose extreme safety hazards. Prediction of a maximum sand free production rate is therefore critical, not only from a safety point of view but also economically. The unnecessary application of sand controlled techniques, as a precaution against anticipated sand production, can cause an increase in completion costs and a possible reduction in well productivity. However, if operating conditions dictate the need for sand exclusion, such techniques can make a well, which otherwise could have been abandoned or not developed, extremely profitable. The ability to accurately predict CDP is, therefore, critical to optimizing the completion strategy.
Two mechanisms responsible for sand production are compressive and tensile failures. Compressive failure refers to tangential stresses near the cavity wall exceeding the compressive strength of the formation. Both stress concentration and fluid withdrawal can trigger this condition. Tensile failure refers to tensile stress triggered exclusively by drawdown pressure exceeding the tensile failure criterion. Veeken et al.'s1 review noted that laboratory and production experiments support the existence of both types of failure mechanisms; with tensile failure predominating in unconsolidated sands and compressive failure in consolidated sandstone. The consensus is that near borehole stresses cause desegregation of the formation while the fluid drag forces provide the medium to remove the failed materials.
Applications of reliable compressibilities include reservoir pressure maintenance and subsidence evaluations and production forecasting. Due to their importance, compressibilities are routinely measured in laboratories. The most commonly used method is the hydrostatic compression test, in which a pre-saturated cylindrical rock sample is subjected to isotropic confining pressure. Often during a hydrostatic compression test, the hydrostatic confining stress is increased from zero to the "effective" in-situ vertical or horizontal stress. The pore pressure is maintained at atmospheric level and the volumetric strain and pore fluid volume expelled from the sample are measured and used for compressibility calculations. The compressibilities are then used either directly or after "correction" using elastic relationships in reservoir analysis [ 1-4].
Because the stress state during production is not isotropic (and, instead characterized by pore pressure drawdown with uniaxial strain due to horizontal crustal constraint and constant total axial stress due to overburden), laboratory experiments have been performed and compressibilities have been measured under hydrostaticompression and uniaxial strain conditions [ 1, 4 and 6]. These research work demonstrated that the compressibilities measured under uniaxial strain behave differently from those measured under hydrostaticompression, for sandstone and coal. They concluded that compressibilities during production can not be derived from hydrostatic compression tests, using existing "corrections" . However, results shown by both Lachance and Andersen  and Rhett and Teufel [ 1] were from experiments conducted in the absence of pore pressure, and the effect of pore pressure on compressibilities is not considered. In addition, the pore volume compressibilities were derived by assuming the grain compressibility be negligible and porosity remains constant during uniaxial strain. Results of by Zheng et al.  under simulated production conditions were on coal, which differs significantly from conventional oil/gas reservoir rock. Experimental work presented in this paper verifies the previous research findings, determines the effect of pore pressure on the compressibilities of oil/gas reservoir rocks under hydrostatic compression and simulated production boundary conditions.
Zheng, Ziqiong (Department of Materials Science & Mineral Engineering, University of California) | Cook, Neville G.W. (Department of Materials Science & Mineral Engineering, University of California) | Myer, Larry R. (Earth Sciences Division, Lawrence Berkeley Laboratory)
ABSTRACT: Cylindrical specimens of Indiana limestone have been tested in uniaxial, confined and partially confined compression. A low melting point metal alloy was used as a pore fluid. At the stress of interest this alloy was solidified in place to preserve the stress- induced microcracks. Optical and electron micrographs of cross sections from the specimens were used to study the density, orientation, thickness and mechanism of these microcracks as a function of confinement.
Rock specimens tested in compression display initial consolidation, linear elastic and then strain hardening deformation up to the peak stress, followed by strain softening in the post failure regime. The non-linear modes of deformation and dilatation are probably due to the formation of extensile microcracks under differential compression (Brace et al. 1966, Scholz, 1968 and Rao and Ramana, 1974). Many theories of extensile microcrack generation and propagation in compression have been proposed, including the sliding crack mechanism (Hock and Bieniawski, 1965, Brace et al., 1966, Horri and Nemat-Nasser, 1985, Ashby and Hallam, 1986 and Kemeny and Cook, 1987; the generation of microcracks by local tensile stresses resulting from mismatched elastic properties between grains, or indentation of one grain by another (Dey and Wang, 1981); tensile stresses inside a grain due to Brazilian type loading (Gallagher et al., 1974), and microcracks from tensile stresses adjacent to pores (Sammis and Ashby, 1986). Detailed measurements of microcrack populations in laboratory specimens have been made (Peng and Johnson, 1972, Hallbauer et al., 1973, Gallagher et al., 1974, Olsson, 1974, Hadley, 1976 and Tapponier and Brace, 1976, etc.), but mostly on crystalline rocks. The observation of microcracks in rocks as they exist under load is very difficult because existing microcracks may close and new microcracks may be generated by unloading and the preparation of sections. To distinguish microcracks produced during loading from those produced by unloading, we have used a low melting point metal alloy -- Cerrosafe ® -- which is liquid at 87°C, as a pore fluid. The metal alloy can be solidified at any stage of an experiment to preserve the microcracks as they exist under loading and to distinguish these microcracks from those initiated later. At a pore pressure of 10MPa the liquid metal penetrates openings as small as 0.08µm. Specimens of Indiana limestone were tested in uniaxial, partially conf'med and confined compression, and complete stress-strain curves were recorded. Observations of microcracks were made on polished sections of these specimens using optical and scanning electron microscopy at different magnification levels. Statistical data on the distributions of microcrack density, crack lengths, thickness and orientation are presented. The results show that the microcracks are mostly aligned in the direction of the maximum compressive stress. The distributions of crack length, density and thickness depend strongly on the confining stresses. The volumetric strain and the final porosity of the rock specimens were measured and found to depend on the confining stresses.
Table 1. Summary of test conditions and results for specimens of Indiana limestone confined with stainless steel wire.(available in full paper)
Zheng, Ziqiong (Department of Material Sciences and Mineral Engineering, University of California) | Cook, Neville G.W. (Department of Material Sciences and Mineral Engineering, University of California) | Myer, Larry R. (Earth Sciences Division, Lawrence Berkeley Laboratory)
Dedicated to Charles Fairhurst for his pioneering contributions to stress measurement in rock by hydraulic fracturing.
ABSTRACT: Elongated breakout cross sections for different rock stresses and strengths have been produced using a numerical simulation based on boundary element methods and a micromechanics model of extensile splitting. Although these breakout cross sections are completely stable for an elastic-brittle rock they are not related uniquely to the stresses and strengths, and depend upon the sequence in which holes are drilled and the stresses applied. The depth of any breakout depends mainly on the initial breakout angle. The same initial breakout angle can result from many different combinations of stress. We conclude that breakout shapes cannot be used to infer the magnitudes of the stresses orthogonal to a borehole. The interpretation of hydraulic fracturing measurements based on the stress around a circular hole does not result in large errors if the hole has moderate breakouts. Large breakouts result in tensile fractures that preclude measurements of breakdown pressures.
The phenomenon of rock fracturing by spalling from the walls of boreholes, is referred to as "borehole breakout". The breakout problem has been the subject of intensive field studies and observations (Leeman 1960; Cox 1970; Bell and Gough 1979, 1982; Hickman et al. 1985; Plumb and Hickman 1985; Kim et al. 1986), laboratory experiments (Gay 1976; Mastin 1984; Haimson and Herrick 1985) and theoretical analysis (Gough and Bell 1982; Zoback et al. 1985; Zheng and Cook 1985, Ewy et al. 1987, Guenot, 1987 and Maury, 1987). All the studies agree that breakouts take place in regions of high stress concentration and that the elongation of the diameter of a borehole is parallel to the direction of the minimum principal stress orthogonal to the borehole axis. In this paper, we wish to report recent theoretical results concerning the effect of borehole breakouts on in situ stress measurements. We investigate the possibility of finding the magnitudes of the in situ stresses from borehole breakouts and the effect of breakouts on the results of stress measurement by hydraulic fracturing. Results from the numerical simulations show that borehole breakouts are stress history dependent, that is, the shape and size of a breakout depends on the sequence in which the hole is drilled and the stresses are applied. Results also show that the initial breakout angle is the main factor that controls the breakout depth and the same initial breakout angle can be obtained from different stress- strength combinations so that there is a non-unique relationship between the in situ stresses and the breakout shape and size. The effect of borehole breakout on the hydraulic fracture technique, however, is not important for moderate breakouts but for larger breakouts, in which tensile fractures have already been generated, only the "shut in" pressure can be obtained.
2 NUMERICALLY GENERATED BREAKOUTS AND THEIR NONUNIQUE RELATIONSHIP TO IN situ STRESSES
When a borehole is drilled, the original stress field changes and results in concentration of tangential stress around the borehole boundary. The stress distribution around a circular borehole can be obtained from the Kirsch's solution (Jaeger and Cook, 1979).