Hydraulic fracture initiation dictates the communication path between the wellbore and fracture plane. Nonplanar fracture geometries such as multiple, T-Shaped, and reoriented fractures are not advantageous and they adversely affect the potential to achieve a desired stimulation treatment. Oriented perforation can be the solution to initiate a single wide fracture in vertical and deviated wells. Also oriented perforations may be used to create stable tunnels in poorly consolidated formations thus avoiding sand failure and consequently preventing sand production. This paper presents laboratory experimental results related to oriented perforations for hydraulic fracturing. It also discusses the use of oriented perforation for sand control. Experiments were designed to investigate the effect of perforation orientation in vertical and horizontal wells on hydraulic fracturing treatment. Introduction Oriented perforation has not been widely applied in the industry due to practical difficulties involved in this process. However, more applications are becoming candidates for oriented perforation, i.e., fracturing deviated and horizontal wells, controlling sand production, and solving wellbore instability problems. The first work on the effect of perforation on hydraulic fracturing was presented by Daneshy who showed that the direction of induced hydraulic fracture is not dictated by perforation orientation. He showed that in many cases, fluid traveled from the perforation through the area between the casing and formation to initiate a fracture in the direction of maximum horizontal stress. Several authors showed that the perforation orientation in horizontal wells should be in phase with the anticipated fracture direction. Field observations on the effect of oriented perforation were reported for vertical and deviated wells. Venditto, et al. reported a field observation on a vertical well. P. 411^

To date, the art of effective openhole horizontal well fracturing is not well defined. Difficulties in regional sealing hamper the fracturing task, and results are generally suspect. Without proper isolation methods, the use of openhole horizontal well fracturing is limited. During many fracturing processes, including fracture acidizing, fracture or acid placement often occurs where fluid first contacts the borehole, often at the heel of the well. A new method is now available that combines hydrajetting and fracturing techniques. By using this new method, operators can position a jetting tool at the exact point where the fracture is required without using sealing elements. Unlike other techniques, this new method allows operators to place multiple fractures in the same well; these fractures can be spaced evenly or unevenly as prescribed by the fracture design program. Large-sized fractures can be placed with this method. Because the method is simple, operators can economically bypass damage by placing hundreds of small fractures in a long horizontal section. To enhance the process even more, operators can use acid and/or propped sand techniques to place a combination of the two fracture types in the well. This paper discusses the basic principles of horizontal hydrajet fracturing and how Bernoulli's theorem was used to design a hydrajet fracturing technique. Laboratory test results for the new technique are provided on Page 4. P. 263

The rock mechanical poromechanics pressure property, or Biot's effective stress parameter, a, is an important rock matrix and grain characteristic. The Biot's parameter relates stress and pore pressure and weighs the effect of the pore pressure within the concept of the effective stress analysis, in the geomechanics disciplines, and in particular when applied to reservoir engineering and wellbore time-dependent drilling mechanics and stability. It measures the compressibility of the skeletal framework of the rock with respect to the solid material composing the rock. In addition, it also reflects the compressibility of the rock structure which is one of the most important parameters for predicting oil reserves. The poroelastic constant, a, is a complex function of the rock in-situ stress and porosity. The petroleum industry has historically calibrated empirical pore pressure relationships to the effective stress assuming a value of a to be unity or a constant value that may change as the reservoir is being depleted.A theoretical and experimental understanding of the measurements of the poroelastic constant, a, should improve the existing models in the areas lacking the data necessary for an accurate calibration. The paper discusses the various methods of measuring the rock poroelastic parameter, a. These methods include quasi-static and acoustic approaches. In the quasi-static approach, two experimental set-ups, known as the direct and indirect methods, were used in this study to determine a, compared simultaneously with the acoustic method which is based on the compressional and shear wave velocities measurements under hydrostatic loading. The direct method using the quasi-static approach utilizes the measurements of the change in pore volume and bulk volume of the sample for the calculation of a while the indirect method uses the bulk modulus of the fluid saturated rock sample and solid grains in the computation of a.The acoustic approach of measuring a utilizes the measurements of both compressional and shear wave velocities to compute bulk modulus which is then used to compute a; thus, very similar to the indirect technique. The measurements of a using these approaches were performed on fluid-saturated Berea cores, with mineral oil as the saturating fluid. The results obtained from these various techniques are discussed in this work. The indirect method reveals a higher magnitude of the poroelastic constant, a, compared with direct method measurements[1].This difference was found to be large for samples with high porosity, while for low porosity samples the magnitude for the poroelastic constant, a, measured using the two methods was comparable.This is due the fact that low porosity samples have high bulk modulus which tends to lower the magnitude of the poroelastic constant, a, when using the indirect method.The acoustic measurements showed the sensitivity of a to the stress level at which it was measured. The magnitude of the poroelastic constant, a, dropped by 20% for some samples when the pressure increased from 1000 to 9000 psi[2,3].The results of this work also indicate that the measurements of the poroelastic constant, a, using the direct method and acoustic method at early pressure are quite similar for the case of low porosity sample.For high porosity samples the magnitude of the poroelastic constant, a, measured from direct methods was found to fall in the range of acoustic measurements at high pressure. The quasi-static direct method showed a good prediction of a in the absence of the jacketing effect.

This paper presents experimental results related to the hydraulic fracturing of a deviated well, specifically, the impact that wellbore orientation with respect to the in-situ horizontal stresses can have on breakdown pressure and pressure decline after breakdown. Experiments were performed on 6 × 6 × 10-in. cast hydrostone samples. This paper discusses the pressure profile observed during hydraulic fracturing of openhole wellbores with various orientations with respect to the vertical and the maximum horizontal stress. The results of the experiments provide visual evidence of the complexity of induced fractures that can occur in deviated wellbores. Previous laboratory experiments have concentrated on stimulation of vertical or horizontal wells. This study was designed to bridge the gap between the extremes of wellbore orientations. Introduction A deviated well most frequently occurs where the surface location is far from the target position in the reservoir. Also, the angle buildup portion of a horizontal wellbore passes through varying inclinations. In many of the highly offset, deviated wells, the producing formations are encountered where the wellbore cannot be returned to a vertical orientation (as in the case of S-shaped wellbore trajectories). Because of the industry's inability to penetrate the formations vertically, these highly deviated wellbores must be stimulated. In many parts of the world, deviated wellbores are cased, cemented, and stimulated through perforations. Some formations have sufficient strength to be completed without the installation of casing. These deviated and horizontal wells are frequently fractured to improve productivity. Fracturing a deviated well sometimes results in premature screenouts and high treatment pressures. This work demonstrates the various pressure profiles that were observed during the fracturing of deviated, openhole intervals. This paper shows the complexity of fracture patterns that can be created from deviated wellbores. Several researchers have investigated critical mechanisms related to fracturing deviated wells. Daneshy showed that inclined hydraulic fractures exhibit tortuous surface "steps" as they propagate away from the wellbore toward the unaltered in-situ stress field. In his experimental work on fracture initiation from horizontal wells, El Rabaa showed that multiple fractures were created when the perforated interval was greater than four wellbore diameters and when the orientation angle was less than 750 (a deviation angle is referenced, in his work, from the minimum horizontal stress). Hallam et al. concluded that surface roughness of fractures initiated from deviated wellbores is caused by "starter fractures," which are individual fracture planes initiated from separate perforations. Abass et al. showed three types of fracture geometries: multiple parallel, reoriented, and T-shaped. They showed that the breakdown pressure is a function of wellbore deviation. However, the propagation pressure did not vary much with wellbore deviation. Experiments A series of laboratory experiments were conducted to promote research in the area of fracturing horizontal wells. The rock samples used in this study were rectangular blocks of hydrostone (gypsum cement) with dimensions of 6 × 6 × 10 in. These blocks were cast from mixing water and hydrostone with a weight ratio of 32:100, respectively. The physical and mechanical properties of this man-made rock are as follows: porosity = 26.5%; permeability (N2) = 3.9 md; grain density = 2.32 gm/cc; bulk density = 1.71 gm/cc; P. 269

This paper discusses multiple fracture creation, detection, and prevention. Multiple fracture creation will be discussed on the basis of rock mechanics theory, laboratory experiments, and field observations. In addition, several references from SPE papers will be outlined. The section on qualitative analysis will discuss several methods of detection, ranging from core and log analysis, production history matching, and pressure transient analysis to the use of a real-time, three-dimensional (3D) fracturing model that can possibly provide quantitative analysis. Mitigation methods with regard to perforation techniques will also be briefly discussed. Introduction Poor post-fracture well performance has long been attributed to such factors as proppant convecting out of zone or poor conductivity resulting from misapplied gel-break chemistry. Although evidence suggests that such effects could worsen poor fracturing treatment performance, many fracturing treatments that result in poor production may result primarily from the creation of multiple fractures. Many stimulation engineers have yet to accept this phenomenon fully because they believe that all multiple fractures result in screenouts. Screenouts may not always occur. This paper presents a successfully executed instance in which a 3D frac model revealed a surface treating pressure that indicated eight multiple fractures. In a recent work, Mahrer et al. cited 285 articles, reports, and other documents that provided qualified observations of multiple fractures. A vigorous research of the literature was performed to discover citings of single-wing, planar fractures. With the exception of a theoretical reference by Howard and Fast, Mahrer found no references to such fractures. At the presentation of his work, however, Mahrer was besieged by several single-wing fracture constituents who would not accept his research. Mahrer proposed that the paradigm of single planar fractures should be the exception, not the norm. This paper complements Mahrer's work by exhibiting other facets of multiple fractures. In rock mechanics theory, single-fracture planes are the given norm. This postulate is easy to address numerically and conceptually. However, many cases exist in which single planar fractures were not created during experimentation with hydrostone. Although this material is considered the most manageable rock available, it allows multiple fracture initiation in nonunique circumstances. Table 1 outlines several symptoms of multiple fractures. This paper will discuss each of these symptoms and their possible causes:sand production without the placement of more than six sand grains of proppant,cyclic production performance after the fracturing treatment,shallow mineback studies,pressure behavior during stimulation, and other lesser known illustrations. A detailed section is included that discusses how to use a 3D fracture simulator to qualify and possibly quantify multiple fractures. This section will describe the differences between single-plane tortuosity and a similar effect that occurs when multiple fractures are present. Types of Multiple Fractures and Their Environments This paper will focus on multiple fractures that coalesce, overlap, or compete for the same pore space (Fig. 1). Most experts consider noncompeting fractures (Fig. 2), such as those generated when long pay intervals are fractured with several perforated intervals, as no threat to the fracture treatment's success. An example of this type of noncompeting fracture occurs in the Hugoton wells, where several members of the Chase group are being stimulated simultaneously. Multiple fractures will most likely occur in (1) naturally fractured formations, (2) long intervals of perforations, with the perforating phase being 0 >to >180, and (3) strongly dipping planes and/or deviated wellbores to flat bedding planes. P. 163^