Vous êtes sur la page 1sur 8

Distinguished Author Series

Off-Balance Growth: A New Concept in Hydraulic Fracturing


A. Ali Daneshy

Abstract This paper offers a new fracturing concept that explains how and why actual hydraulic fractures deviate from the commonly accepted single-fracture growth models. It demonstrates that the fracture follows the local path of least resistance, not the global path, and this leads to substantial branching, presence of extensive shear fractures, and a growth pattern that is dominated by conditions at the tip of the propagating fracture, thus making its growth haphazard and off balance. Within each fracture, branching and shear fracturing create a complex path for the fluid flow and proppant transport, as well as production of reservoir fluid. As a result, most hydraulic fractures are substantially shorter and narrower than the intended design and yield a suboptimal production increase because of inadequate fracture length and conductivity. Often, common well-completion designs also lead to creation of multiple fractures at the wellbore, a situation further aggravated by prevailing fracture designs. Off balance growth can obstruct proppant flow, which may lead to screenout. Introduction The oil and gas industry has long recognized the inadequacy of existing theories to predict the behavior and outcome of hydraulic fracturing treatments. Treatments require higher pressures than predicted by theory. After-fracture pressure buildup tests often behave more like those of radial flow from wells with negative skin than of fractured wells.1 Actual production increases from fractured wells are lower than predicted from fracture design computations. Several field observations show very complex fracture paths, presence of multiple fractures, and random-appearing proppant distribution. Warpinski et al.2-5 provide a comprehensive review of some of the major experiments conducted by Sandia Natl. Laboratories at the Nevada Test Site and by the Gas Technology Inst. and U.S. Dept. of Energy at M-Site. These reports detail important features of these fractures, together with an explanation of their cause. These results are used extensively as background and support for the new off-balance fracture growth concept. The concept of fracture tortuosity was introduced to account for a nonplanar near-wellbore fracture path. Although this term is not scientifically defined, its use is very widespread in fracturing literature and generally means any wellbore effect that complicates the near-wellCopyright 2003 Society of Petroleum Engineers This is paper SPE 80992. Distinguished Author Series articles are general, descriptive representations that summarize the state of the art in an area of technology by describing recent developments for readers who are not specialists in the topics discussed. Written by individuals recognized as experts in the area, these articles provide key references to more definitive work and present specific details only to illustrate the technology. Purpose: to inform the general readership of recent advances in various areas of petroleum engineering.

bore creation and extension of the fracture. The focus of most existing papers is more on recognition of the problem than on the underlying cause. The general approach has been to relate tortuosity to near-wellbore events and conditions. Most literature refers to two specific causes for tortuosity: near-wellbore change in fracture direction and initiation of multiple fractures by use of perforations. Mahrer et al.6 give a comprehensive account of events and publications leading to the current industry understanding and history of tortuosity. Aud et al.7 attribute most screenouts to near-wellbore events most likely caused by tortuosity. Cleary et al.8 provide a description of some of the events that can cause tortuosity. Weijers et al.9 consider natural fracturing as the main cause of multiple fracturing. Shlyapobersky and Chudnovsky10 attributes the complex fracture behavior to the presence of a near-tip process zone that increases the dissipation of energy during hydraulic fracturing. In this paper, the concept of off-balance fracture growth is introduced and used to show how it dominates the growth pattern of a large majority of hydraulic fractures, both near the wellbore and away from it. Two distinct fracture characteristics are defined and described. These are multiple fracturing (a near-wellbore effect caused mainly by well-completion details) and branching and shear fracturing (which exist within the main body of the fracture and control its global growth pattern). Off-balance growth occurs everywhere along the fracture and is not restricted to the near-wellbore region. Its consequences are very narrow fracture widths, short fracture lengths (created and propped), severe branching, high pressure drops along the fracture, and preferential fluid and proppant movement dominated by shear fractures. Fundamental Fracturing Concepts Fractures in rocks are created by any one or a combination of three mechanisms: opening, sliding-mode, and tearingmode fractures.11 Tensile Fractures. In this mode, the stress creating the fracture is tensile and separates the two faces of the fracture (Fig. 1a). Because rocks are particularly weak in tension, this type of fracture usually requires the least amount of force and energy for its creation. Tensile fracturing is the dominant mechanism in hydraulic fracturing. The main resistance to fracturing comes from the least insitu principal stress, min. Fracture opening (width) is the result of compression of the formation adjacent to it. Thus, higher incremental pressures (fluid pressure inside the fracture minus the least in-situ principal stress) create wider fractures. Fracture width is also directly proportional to its length and height and related to formation

78

APRIL 2003

mechanical properties. Most fracturing models are based on tensile fracturing. The force requirement for creation of the fracture is: Apf.pf=Af.min, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .(1) in which Af =fracture area perpendicular to the least insitu principal stress, Apf =pressurized area of the fracture, pf =constant pressure equivalent of pressure in the fracture, and min=least in-situ principal stress. Fracture opening (width), , is related to formation mechanical properties E and , the difference between fluid pressure pf in the fracture and min, and fracture lateral dimension l (length, height, or radius) through the equation: = (12).(pf min)l E in which is a constant and depends on fracture shape (e.g., circular, rectangular, or elliptical). Without proppant, tensile fractures close when pf <min. Sliding-Mode Fractures. Fracturing occurs when the induced shear stress along a given plane exceeds the shear strength of the rock in that plane (Fig. 1b). Failure (fracture) occurs when the two faces of the fracture slide over each other in opposite directions but without separating the faces by forming a gap between them. Thus, it has no volume. In general, because the shear strength of rock is greater than its tensile strength, shear fracturing requires higher pressure. The more common occurrences of sliding shear fractures are along existing natural fractures and planes of weakness in the formation (i.e., shear strength is very low), during the fracture reorientation process, or as a connecting mechanism to join two adjacent fracture branches. The latter two situations will be discussed in more detail later. Tearing-Mode Fractures. In this type of fracture, the failure mode is a tearing action, in which the two faces of the fracture twist away from each other (Fig. 1c). The basic . . . . . . . . . . . . . . . . . . . . . .(2)

tearing action does not create a gap between the two faces of the fracture and does not create additional volume. Again, this type of failure is more likely to occur along natural fractures or existing planes of weakness during the fracture reorientation process, or as a mechanism to join two fracture branches. Both types of shear fractures do not close as easily as tensile fractures. Closure involves reverse sliding of the two fracture faces. Although the shear strength of the formation along the shear fracture plane is zero (already fractured), resistance to closure comes from frictional forces along the fracture faces. Mixed-Mode Fractures. These fractures are created by combined tensile and shear forces. Figs. 1d and 1e show the two opposite faces of the fracture separated from each other with a gap, as well as laterally shifted from each other. Fracture Branching. Branching occurs when the fracture propagates along two or more different but connected paths (same as the term strand used by Warpinski2). A branch of a fracture may, and usually does, have its own branches. As discussed later, branching is caused mainly by formation characteristics and treatment parameters. Multiple Fracturing. In this paper, this term refers to separate fractures created at the wellbore. Each of the multiple fractures has its own off-balance growth pattern. Some publications have used branching and multiple fracturing interchangeably. However, significant differences exist between the two. For example, the path of multiple fractures can be divergent, such that they gradually grow out of each others influence zone while branches of a fracture are always within its influence zone. Fluid distribution into multiple fractures is highly influenced by near-wellbore characteristics, including the number of perforations connected to each fracture, whereas branches compete for fluid within the same fracture and are somewhat independent of the near-wellbore behavior. Multiple fracturing is the result of well-completion design (e.g., borehole inclination; number, size, and distribution of perforations; etc.) and details of the treatment (e.g., injection rate and fluid viscosity), whereas branching is controlled mostly by formation properties. From a practical point of view, it is easier to influence multiple fractures than branching. Off-Balance Fracture Growth Almost all theoretical developments of hydraulic fracturing are based on an idealized assumption of a single fracture in a plane perpendicular to the least in-situ principal stress with a simple periphery (mostly rectangular, circular, or occasionally elliptical) that extends progressively away from the wellbore. As more fluid is pumped into the fracture, its dimensions grow, and the fracture becomes larger in its original plane. The fracture is created by simple tensile stresses, and it exists in a single plane and grows in an orderly and predictable manner. Such a growth pattern is defined as balanced. Consider a hydraulic fracture with the dominant extent, l, and fluid pressure, pf, where pf and l are at equilibrium (Fig. 2). As more fluid is injected into the fracture, its pressure will rise until it causes instability and extension of the fracture. The extension path of the fracture is along its

Fig. 1Different fracturing modes.

APRIL 2003

79

min min

d = cos()

Pf

Fig. 3Effect of fracture orientation on width.

Fig. 2Fracture growth along local path of least resistance.

local path of least resistance and depends on the state of stress, as well as the strength of the formation at its tip. In a homogeneous and isotropic material, the most likely path is toward Point A, Fig. 2a. But material weaknesses such as natural fractures can change this path: for example, Paths B or C in Fig. 2a. Each of these paths will have a new and different pressure for onset of its next instability, as shown by Points A, B, and C in Fig. 2b. Point A in this example is perpendicular to the least in-situ principal stress and in purely opening mode. It will require the least incremental pressure for its next extension and is, therefore, the least stable of these paths. Point C requires the highest incremental pressure for its next extension and is, therefore, the most stable. Whether the fracture will undergo another extension at Point C will depend on whether fluid pressure reaches the required value for onset of its next instability. The opening of the fracture is also a function of its path, widest along Path A and narrowest

along Path C, as shown in Fig. 3. Failure along any path other than Path A will involve shear. Furthermore, the next growth step of the fracture from any point other than Point A will also involve shear failure, and in a different path than the original path. Thus, one shear failure leads to a progression of other shear failures, which result in a rugged and discontinuous fracture plane. The tip of a hydraulic fracture spreads over a large area (from tens to several thousand feet). Because of material anisotropy and heterogeneity, the resistance to fracturing varies along the tip. A fracture grows wherever and whenever the induced stresses exceed the corresponding strength of the formation. Fracture growth occurs at small, discrete local steps and randomly moves along its tip. At each local extension increment, the fracture follows its local path of least resistance, which can lead to creation of many interdependent branches along its tip, as shown graphically in Fig. 4. These branches can be very close to each other, and some of them may link through local small shear fractures. An example of this situation is shown in Fig. 5.3 In this experiment, the front face of the fracture has been removed to show details on the other face. The dyed fracturing fluid helps to identify extension sequence

Fig. 4Fracture branching.

Fig. 5Branching caused by discrete fracture growth. (Courtesy of N.R. Warpinski)

80

APRIL 2003

pattern. Larger fluid volumes do not guarantee a longer fracture length. More proppant does not mean longer propped length. Once a fracture begins to grow off balance, its extension becomes progressively more off balance and creates a network of randomly distributed and propped branches. Multiple Fracturing. Usually, multiple fractures are the result of an excessive supply of fluid and occur when the initial fracture is not able to accept the total injected rate. The excess supply of fluid causes the wellbore pressure to rise until it exceeds the formation strength at the next weakest perforations, at which point another fracture initiates and begins to accept fluid. The process continues until enough fractures are created to balance the volume injected into and out of the wellbore. As noted by Cleary et al.,8 the breakdown pressure of each perforation is as related to its own characteristics as it is to its alignment with the preferred fracture direction. Therefore, it is possible for some perforations to stop taking fluid after other more preferably oriented perforations break down and provide an easier link to their associated fracture. Causes of Off-Balance Growth The following offers some of the more common causes of off-balance growth. Borehole Inclination With Respect to Principal Stresses. Fig. 6 shows a laboratory sample with an inclined hydraulic fracture created from an open hole.12 It shows extensive shear fracturing and branching. Inclined boreholes also are likely to create multiple fractures, each with its own complex behavior and growth pattern. The number of perforations linked with each fracture will control the distribution of the injected fluid between them. Abass et al.13 recommend substantially reducing the length of the perforated interval to avoid creating multiple fractures. Reducing or eliminating multiple fractures increases the effective length and productivity of the fracture, as well as increases the probability of successful completion of the treatment. Perforation Pattern. The capacity of the initial fracture to accept fluid is limited by the number of perforations linked with it. If the injection rate is higher than this capacity, an increase in wellbore pressure will cause breakdown of other perforations, leading to initiation of multiple isolated fractures from the same wellbore. Multiple fractures may manifest themselves by multiple breakdowns, which can be detected on the treatment pressure record. This situation is quite common in the industry. Fracture initiation from a perforated interval can be quite complex. For example, a fracture may initiate from only one side of the borehole, the two sides of the fracture on the opposite sides of the borehole may not be in the same plane, the fracture plane may deviate vertically to link with adjacent perforations, or combinations of these. Furthermore, in cases of multiple fractures, creation of width in one fracture may disrupt the alignment between adjacent perforations and their respective fractures.

Fig. 6Branching and shear fracturing in included holes.12

through differences in color intensity. Each of the branches on the left tip of the fracture corresponds to a different spurt in fracture growth and clearly demonstrates its discrete nature. While these have the same general direction and are linked to the main fracture, a very short distance also separates them from each other. Closer examination of the fracture shows it to have at least three very close, larger branches. The darker colors correspond to older fractures and show that, even though adjacent to each other, these branches were created at different times, and that the main branch connected to the fracture tip is the youngest. Because of this closeness, all branches are within each others influence zone, increasing the effective value of min. Furthermore, the limited continuity of each branch will result in a small value of l (lateral branch extent). Thus, referring to Eq. 2, it is clear that the width of the fracture at each branch will be quite small, substantially smaller than computed from any of the existing models. Measurements of fracture width reported by Warpinski et al.3 confirm this conclusion. The fracture in Fig. 5 contains many smaller shear fractures, but they are not easily visible in the photograph. In the above example, if int is horizontal, then branching will more likely be vertical or subvertical. If int is vertical, then branches will tend toward horizontal. From the above discussion, once a hydraulic fracture deviates from its natural plane (normal to the min direction), its growth becomes associated with branching and shear fracturing. The nonplanar geometry and presence of shear fracturing and branching exert shear stresses at the fracture tips such that growth is not in simple opening mode and not strictly in the plane perpendicular to min. The outcome of this process is a fracture created in mixed mode (in varying degrees at different locations) that contains numerous branches. The extensive tip of the fracture has many areas that lie within the fractured zone and not at its leading edge farthest away from the wellbore. Fluid and proppant travel along a very complex path very different from the piston-like motion assumed in many fracturing models. The fracture growth staggers from one part to another in a random and unpredictable manner. Thus, the term off balance is used to describe the fracture growth

APRIL 2003

81

Fig. 8Proppant blockage by shear fractures parallel with borehole.

extreme vertical dimension of the branches. Width will vary greatly along the fracture and may restrict the proppant entry into very narrow segments of the fracture.
Fig. 7Narrow width of shear fracture.

Natural Fractures and Planes Of Weakness. Most formations contain natural fractures and planes of weakness. If a propagating hydraulic fracture intersects these, depending on the relative magnitude of the formation tensile strength along the original fracture plane compared with the shear strength along the plane of weakness, the fracture may extend along the plane of weakness, thus temporarily deviating from its path. Although the fracture width () measured perpendicular to its original direction may stay the same, its width for fluid and proppant flow is reduced to cos, where is the angle between the old and new planes of fracture (Fig. 3). As increases, the real fracture opening decreases and may, in fact, be insufficient to allow proppant entry into this portion of the fracture. Fig. 7 shows a view of tensile fractures joined together by shear fractures in the mine-back experiments at Nevada Test Site.* It clearly shows that shear fractures are narrower than tensile fractures, and that the width has decreased with increase in the angle . The sharp turns in the fracture path are effective for trapping the proppant and preventing its forward movement, thus causing a pressure increase inside the fracture. Even when the deviation in a fracture plane is very local, the reorientation process extends to a larger area than the natural fracture and requires additional shear fracturing, some of it in intact rock with larger shear strength than the natural fractures. If fracture pressure and dimensions do not create enough leverage to overcome the formation shear strength, then part of the fracture will become dormant unless increases in fluid pressure and fracture growth elsewhere provides the necessary leverage to overcome formation shear strength. These pockets of unbroken rock pinch the fracture width and obstruct movement of the slurry. Thus, even small natural fractures can create off-balance growth that extends beyond their immediate area. This effect also shows that the extending edge of the fracture is not necessarily at its farthest points from the wellbore. In fact, two points geometrically very near each other in the fracture may have been created at very different times. Consequences of Off-Balance Growth Fracture Dimensions. Conventional fracture dimensions: length, width, and height take different meanings in off-balance growth. Each branch of the fracture will have a different set of dimensions, all of them different from those of the overall fracture. Fracture length will refer to the farthest point reached by any of the branches, and height means the
* Personal communication with N.R. Warpinski

Fluid Leakoff. Off-balance fractures have very high ratios of fracture surface area to volume. Thus, fluid leakoff constitutes a large percentage of the injected fluid. The complex fracture geometry makes application of existing leakoff computations made on the basis of simple fracturing theory (Nolte14 and Lee15) questionable. Small branches with narrow width may, in fact, get plugged off with gel residue left behind after fluid leakoff into the formation. Preferential Fluid and Proppant Movement. The narrow opening of shear fractures makes fluid movement between adjacent branches more difficult than along each of them, resulting in a preferential direction for fluid movement and, consequently, fracture growth. This situation is exacerbated with proppant. As the slurry tries to move through the shear fractures, proppant particles can get trapped in the narrow width and cause further decrease in the area open to flow. The net effect is development of a dominant flow path along the length of the branches and restriction perpendicular to it, thus creating a de facto barrier in the direction normal to their major growth dimension. The restriction in fluid flow also creates a pressure difference between branches that may cause further shear fracturing and associated effects on growth pattern. Because all of the fluid and proppant has to enter the fracture through perforations, their shape and distribution affect treatment execution. Narrow fracture widths (more likely caused by shear than by branching near the wellbore) can trap the proppant and form a nucleus that can spread and block an entire part of the fracture. Shear fractures parallel with the borehole will have a bigger effect on proppant distribution than those perpendicular to it. The shear fracture in Fig. 8 imposes a minor obstruction to fluid flow. However, when proppant enters the fracture, the flow area can be substantially reduced if proppant particles get trapped behind it. The slurry will have to find an alternate path for flow, which will also be at a higher pressure with the possibility of a similar blockage. Screenout results when these alternate paths are insufficient to transmit the injected fluid under allowable fracturing pressures. Under similar conditions, shear fractures perpendicular to the borehole will have much less effect on the fracture. Fracturing Pressure. The complex nature of fracture growth and slurry path result in larger frictional pressure drops along the fracture. This is not limited to the nearwellbore region and spreads across the entire fracture area. The pressure increases associated with shear fracturing and branching are more prominent early in the fracturing treat-

82

APRIL 2003

ment. With short fractures, the main mechanism for overcoming shear resistance of the formation is fluid pressure. When fracture dimensions become larger, they can exert sufficient leverage to break the formation under shear with lower pressures. This explanation is much more plausible for the early pressure increase seen in some fracturing treatments than the commonly used reasoning of PerkinsKerns-Nordgren type treatment with limited height.16 In fact, as discussed earlier, in off-balance growth, the terms fracture length and height have different meaning, from those in classical theory of balanced fracturing. Production Increase. The complex connection between different parts of the fracture creates an undesirable path for reservoir fluid flow. Because proppant is distributed inside many small branches, the local flow capacity is also much smaller than computed. The result is large production-pressure drops along the fracture. This pressure drop is why hydraulic fractures very seldom provide the expected theoretical production increase and after-fracture well testing that often indicates the behavior of a well with a small negative skin rather than a fractured well. Fracture Closure. Closure of shear fractures is more difficult than tensile fractures and requires a larger force. Therefore, some shear fractures will not return to their original position, which results in keeping some of the branches open even without proppant. Whether this will provide the fracture with some intrinsic flow capacity depends on the interconnection of these branches with the dominant path of reservoir fluid flow inside the fracture. The other aspect of this phenomenon is that closure in its classical form does not take place in off-balance fracturing. This aspect casts some doubt on the validity of the techniques currently used by the industry to define closure stress. Screenout. Screenouts are caused by obstruction to proppant movement to such an extent that the fracturing pressure exceeds the allowable maximum pressure of the treatment. The narrow width of shear fractures and branches can block or hinder proppant movement. Away from the wellbore, the disruption increases the fluid pressure, which in turn increases the local fracture width. As pressure increases, some of the dormant tips of the fracture can reactivate and extend, creating new avenues for slurry movement. Increase in fluid pressure means these obstructions have moved closer to the wellbore where the fracture has fewer tips for reactivation, and the increase in fluid pressure acts on a smaller area. Thus, the probability of any particular combination of shear fracturing or branching causing screenout is inversely related to its distance from the wellbore. This distance also affects the time and speed of screenout; the closer the obstruction is to the wellbore, the earlier and faster the screenout. Beneficial Applications of Off-Balance Growth. One application in which off-balance growth may be superior to conventional fracturing is stimulating high-permeability reservoirs. In these situations, multiple fractures and off-balance growth create a short, high-permeability zone around the wellbore that will bypass the damage and provide sufficient conductivity to promote efficient radial fluid flow. This situation can help delay or reduce sand production.

Hindering Off-Balance Growth There are two key issueshow to decrease the number of multiple fractures and how to decrease the amount of offbalance growth in each fracture. Two main contributors to creation of multiple fractures are fluid pressure inside the wellbore (which is controlled by rate, viscosity, proppant concentration, etc.) and number and type of perforations along with details of the well completion, especially well inclination. Longer perforated intervals increase the probability of multiple fracturing. The most desirable scheme for perforating is hydrojetting a short slot. Another option is to use larger-diameter concentrated perforations, as recommended by Stadulis,1 Aud et al.,7 Cleary et al.,8 Weijers et al.,9 and others. Even when the intent is to create several fractures in different sections of the producing interval, it is better to keep each perforated interval as short as possible. Some authors recommend perforating normal to the min direction (Weijers et al.9). The dilemma is trying to determine the orientation before performing hydraulic fracturing. The probability of multiple fracturing also increases with injection rate. If the initial number of fractured perforations is insufficient to transmit all the fluid, pressure will increase and cause breakdown of other perforations until a balance is reached between pressure, rate, and the number of perforations linked to multiple fractures. Some authors recommend the use of high breakdown rates with the intent of reducing near-wellbore tortuosity, Weijers et al.9 This technique will work if all the perforations are connected to one fracture. Perforation geometry has a bigger effect on control of fluid supply for each of multiple fractures than the injection rate. Some authors recommend use of high-viscosity fluids to reduce tortuosity (Aud et al.7 and Weijers et al.9). Use of high-viscosity fluids increases the wellbore pressure and the possibility of multiple fractures. Also, it increases the frictional pressure caused by slurry flow in the fracture and probability of internal branching. The drag force on the proppant caused by high viscosity is useful only if proppant is able to move forward inside the fracture. Borehole inclination is another source of multiple fracturing. Even in vertical wells, doglegs and deviations from vertical within the pay zone can result in misalignment of the borehole with the principal stresses and cause multiple fractures. The severest situation occurs during fracturing horizontal holes, especially if fracture orientation is perpendicular to the borehole direction. In the latter case, several authors have pointed out that creation of multiple fractures is a near certainty.13,19 Cleary et al.8 recommend injection of a small volume proppant slug during the minifracture treatment or the pad stage to clear the path for the main treatment. McDaniel et al.17 also cite case histories in support of this approach. Stadulis1 extends this approach to include several high-concentration proppant slugs (5 to 12 lbm/gal) separated by small volumes of clean fluid. It is difficult to explain why the proppant slug will have a different effect on the fracture than the proppant that is pumped during the main treatment. Because of the many unknowns at the start of a fracturing treatment, it is best to design the job for maximum flexibility, namely the ability to change rate, fluid viscosity, volume of each stage, and proppant concentration. A mini-

APRIL 2003

83

Case History Example 1. The treatment consisted of C two back-to-back minifractures followed 8,000 B 50 by the main treatment. The minifracture D 7,000 E data is presented in Fig. 9. During the A I first minifracture, the data show at least 6,000 40 three distinct pressure peaks, shown as 5,000 Points A, B, and C, indicating multiple 30 F fracturing, with the first and the third tak4,000 ing a larger percentage of the fluid than 3,000 20 the second. The large pressure drop after Point C indicates initiation of a dominant 2,000 Tubing Pressure, psi fracture at this point. The rise and fall in 10 Slurry Rate, bbl/min 1,000 fluid pressure around Points D and E could indicate removal of large-scale 0 0 120 130 140 150 160 170 180 190 200 110 obstructions to fracture growth, possibly Time, min by shear fracturing, to create better linkage within the fractures. Although there Fig. 9Minifracture chart for Example 1. are several sharp pressure drops after Point E, the associated changes in rate do fracture treatment provides the first glimpse into the frac- not allow confident interpretation of their cause. ture. The first step in the recommended treatment consists The rapid pressure drop after Point F confirms the earliof determining the fracture reopening pressure. This deter- er diagnostics. The frictional resistance of shear fractures is mination can be made by injecting at the lowest possible supporting part of the closure force, thus relieving the fluid rate and observing fluid-pressure variations. The rate can from providing full support. be increased gradually to observe its effect on computed The initial pressure decrease during the second mini(or preferably measured) bottomhole treatment pressure. fracture (Point I) indicates easing of fluid movement due The maximum treatment rate can be established by this to better linkage between more branches. At this stage, approach. If the job data indicates off-balance growth, then shear fracturing can take place at lower pressures because it is wise to adjust the treatment accordingly. Large pad of larger fracture extent. volumes do not provide a safeguard and may reduce the The surface pressure and rate, together with computed propped length of the fracture (and may increase cost). In bottomhole pressure and proppant concentration data for this case, one should be prepared for possible screenout. the main fracture treatment, are presented in Fig. 10. The At this stage of the treatment, it is too late to modify com- relatively constant pressure between Points A and B indipletion details for this treatment, and it is recommended cates steady fracture growth (which includes consistent that notes be made of the results for use in designing the branching and shear fracturing) during this period. The completion of the next well in the same formation. computed bottomhole pressure shows an instantaneous Higher fluid viscosity increases the drag force on the increase as soon as the proppant enters the fracture. This proppant and helps move it deeper into the fracture. This is consistent with the earlier diagnostic. The surface and property is more essential for off-balance fractures because bottomhole pressure increase indicates steady growth of proppant has to move inside individual branches with very the blockage to proppant flow. At Point C, there is a narrow widths. The increase in drag force also increases marked increase in the rate of pressure increase. Although the fluid pressure inside the fracture and causes more this pressure change appears to be related to proppant conbranching, which will result in a shorter fracture length. It centration, the two events may not be connected to each is difficult to determine the net effect of viscosity on prop- other. The more likely cause of the increase in pressure is pant transport and, in particular, on effective fracture con- that one of the multiple fractures may have screened out. ductivity. It seems reasonable to expect that moderate-vis- Shortly after this point, the job screened out, and attempts cosity fluids (10 to 50 cp) will yield a more conductive, to re-establish injection failed to produce a positive result. The very high pad volume in this job did not help overalbeit shorter, effective fracture. come the obstructions created by shear fracturing. Closure Stress. Reasonable determination of closure stress is very critical for on-site analysis of job data and establish- Conclusions ing the occurrence and extent of off-balance growth. Exist- Off-balance growth creates fractures that are short, naring techniques of closure stress determination on the basis row, and contain many branches and shear fractures. The of after-fracture pressure falloff data do not help in this sit- proppant distribution inside the fracture is suboptimal uation. The simplest way to determine min is to estimate it and not suited for generation of sustained high well profrom fracture reopening pressure. The recommended prac- ductivities. Therefore, industrial hydraulic fractures do tice is to break down the formation at the lowest rate allow- not achieve the production increases estimated from theable with the equipment on location, pump fluid for a very oretical computations. JPT short time (2 to 5 minutes), shut the well for a few minutes, and reinject at the same low rate. The pressure at which the Nomenclature fracture opens is a reasonable estimate of min and is suitAf = fracture area perpendicular to the least in-situ able for future diagnostic work. principal stress
9,000 60

84

Surface Pressure, psi

Slurry Rate, bbl/min

APRIL 2003

18,000

16,000

14,000

C 12,000 A B

10,000

8,000

6,000

BHP, psi
4,000

Surface Pressure Slurry Rate, bbl/min BH Prop Conc, lbm/gal slurry


3.5 lbm/gal

2,000

0 0 10 20 30 40 50 60 70 80

Time, min

Fig. 10Main fracture chart for Example 1.

Apf = E= l= pf = = = = max= int= min= =

pressurized area of the fracture Youngs modulus fracture lateral extent (height, length) constant pressure equivalent of pressure in the fracture angle between hydraulic and natural fracture planes constant Poissons ratio Maximum in-situ principal stress Intermediate in-situ principal stress least in-situ principal stress fracture width

Acknowledgments The author would like to thank Norm Warpinski, Robert Langedijk, and PDO Oman, and Buddy McDaniel, Halliburton Energy Services. References
1. Stadulis, J.M.: Development of a Completion Design to Control Screenouts Caused by Multiple Near-Wellbore Fractures, paper SPE 29549 presented at the 1995 SPE Rocky Mountain Regional/Low Permeability Reservoirs Symposium, Denver, 2022 March. 2. Warpinski, N.R. et al.: Direct Observation of a Sand-Propped Hydraulic Fracture, Sandia Report SAND-81-0225 (May 1981). 3. Warpinski, N.R.: In-Situ Measurement of Hydraulic Fracture Behavior, PTE-3 Final Report, Sandia Report SAND-83-1826 (July 1985). 4. Warpinski, N.R. et al.: The Interface Test Series: An In-Situ Study of Factors Affecting the Containment of Hydraulic Fractures, Sandia Report SAND81-2408 (February 1982). 5. Warpinski, N.R. et al.: An Interpretation of M-Site Hydraulic Fracturing Diagnostic Results, paper SPE 39950 presented at the 1998 SPE Rocky Mountain Regional/Low Permeability Reservoirs Symposium, Denver, 58 April. 6. Mahrer, K.D., Aud, W.W., and Hansen, J.T.: Far-Field Hydraulic Fracture Geometry: A Changing Paradigm, paper SPE 36441 presented at the 1996 SPE Annual Technical Conference and Exhibition, Denver, 69 October. 7. Aud, W.W. et al.: The Effect of Viscosity on Near-Wellbore Tortuosity and Premature Screenouts, paper SPE 28492 presented at the 1994 SPE Annual Technical Conference and Exhibition, New Orleans, 2528 September.

18. Cleary, M.P . et al.: Field Implementation of Proppant Slugs to Avoid Premature Screenout of Hydraulic Fractures With Adequate Proppant Con50 centration, paper SPE 25892 presented at the 1993 SPE Rocky Mountain Regional/Low Permeability 40 Reservoirs Symposium, Denver, 1214 April. 19. Weijers, L. et al.: Simultaneous Propagation of 30 Multiple Hydraulic FracturesEvidence, Impact, and Modeling Implications, paper SPE 64772 pre20 sented at the 2000 International Oil and Gas Conference and Exhibition in China, Beijing, 710 10 November. 10. Shlyapobersky, J. and Chudnovsky, A.: Fracture 0 Mechanics in Hydraulic Fracturing presented at 90 the 33rd U.S. Symposium on Rock Mechanics, Santa Fe, New Mexico, 35 June 1992. 11. Daneshy, A.A.: Hydraulic Fracture Propagation in the Presence of Planes of Weakness, paper SPE 4852 presented at the 1974 SPE European Spring Meeting, Amsterdam, 2930 May. 12. Daneshy, A.A. A Study of Inclined Hydraulic Fractures, SPEJ, (August 1973) 61. 13. Abass, H.H., Hedayati, S., and Meadows, D.L.: Nonplanar Fracture Propagation From a Horizontal Wellbore: Experimental Study, SPEPF (August, 1996) 133. 14. Nolte, K.G.: A General Analysis of Fracturing Pressure Decline With Application to Three Models, SPEFE (December 1986) 571; Trans., AIME, 281. 15. Lee, W.S.: New Minifrac Procedure for Simultaneous Determination of Early- and Late-Time Fluid Loss During Fracturing, paper SPE 21463 presented at the 1990 SPE Eastern Regional Meeting, Columbus, Ohio, 31 October2 November. 16. Nolte, K.G. and Smith, B.S.: Interpretation of Fracturing Pressures, paper SPE 8297 presented at the 1979 SPE Annual Technical Conference and Exhibition, Las Vegas, Nevada, 2326 September. 17. McDaniel, B.W., MeMechan, D.E., and Stegent, N.A.: Proper Use of Proppant Slugs and Viscous Gel Slugs Can Improve Proppant Placement During Hydraulic Fracturing Applications, paper SPE 71661 presented at the 2001 SPE Annual Technical Conference and Exhibition, New Orleans, 30 September3 October. 18. Daneshy, A.A.: Three-Dimensional Propagation of Hydraulic Fractures Extending From Open Holes 15th U.S. National Committee for Rock Mechanics, Application of Rock Mechanics Symposium Proc. 157179. 19. Abou-Sayed, I.S. et al.: Multiple Hydraulic Fracture Stimulation in a Deep Horizontal Tight Gas Well, paper SPE 30532 presented at the 1995 SPE Annual Technical Conference and Exhibition, Dallas, 2225 October.
60

Ali Daneshy, SPE, is President of Daneshy Consultants Intl. He holds a PhD degree in rock mechanics from the U. of Missouri-Rolla. Daneshy has been involved in hydraulic fracturing for more than 30 years including laboratory and theoretical research, numerical simulation, fracture design, and field execution of treatments. His publications include many technical papers as well as chapters in several books on hydraulic fracturing. Daneshy's other areas of technical interest include well completion, control of excess water produced with oil and gas, and intelligent wells. He is a recipient of SPE Distinguished Member and Distinguished Service awards.

Rate, bbl/min, or, Prop Conc., lbm/gal

Bottomhole Pressure, psi

APRIL 2003

85

Vous aimerez peut-être aussi