|Theme||Visible||Selectable||Appearance||Zoom Range (now: 0)|
Introduction This paper presents a practical analysis technique to determine actual fracture geometry and proppant profile using a three-dimensional (3D) hydraulic-fracturing simulator. The hydraulic-fracturing model used in this study considers the variation of in-situ stress, Young's modulus, Poisson's ratio, and net pay thickness in the productive interval. When our method is applied, the results from the fracture propagation model conform well with the results we obtain from pressure-buildup and production-data analyses. This study analyzed hydraulic-fracturing treatments from several wells in the Vicksburg formation of the McAllen Ranch area in south Texas. We have provided guidelines to properly describe the treatment interval, how to use this information in the analysis of such fracture treatments, and how to confirm the results using pressure-transient tests and production-data analyses. This paper presents examples illustrating that a detailed description of the reservoir layers is essential to properly evaluate hydraulic-fracture treatments. For the example wells presented in this paper, post-fracture-production and pressure-transient data were available. We have analyzed production and pressure-transient data to estimate permeability and fracture half-length. The values of fracture half-length used to analyze the production data matched closely with those predicted by the fracture model. Brief Field History The McAllen Ranch field was discovered in the 1950's. The Texaco portion of the McAllen Ranch field produces gas and condensate from the Vicksburg S, Vicksburg UV, Vicksburg Y, Vicksburg YE, and Guerra sands. These formations are highly geopressured, low-modulus, low-permeability gas reservoirs. The producing intervals are believed to be compartmentalized by faults and depositional features. The average depth of the Vicksburg formation is between 12,000 and 13,000 ft. The Guerra formation is located from 13,500 to 14,500 ft. The temperature is usually more than 320°F. A typical well drilled in any zone requires a fracture treatment to optimize gas recovery and economics. Methodology of Analyses Our method to analyze reservoirs that must be hydraulically fractured is summarized as follows.Analyze all available prefracture production and pressure-transient data to determine estimates of permeability, skin factor, and reservoir pressure. Analyze all available logs to divide the treatment interval into layers by category, based on lithology, fluid content, and porosity. Compute or estimate mechanical properties for all layers using logs. Use this layer description in a hydraulic-fracture-propagation model to compute created and propped-fracture properties. Use estimated values of reservoir permeability and fracture half-length to simulate post-fracture, single-well production performance. Match the production computed from the reservoir model with the actual well production data. Compare the estimates of propped-fracture half-length and formation permeability obtained from the pressure-transient analyses, the production-data analyses, and the fracture-propagation model. If these estimates do not agree, then alter the reservoir layer description, if possible, and repeat the analyses until reasonable agreement is achieved. Hypothetical Example This hypothetical example illustrates the importance of choosing a detailed in-situ-stress profile for the treatment interval to accurately design a fracture treatment. Fig. 1 shows the gamma ray and resistivity logs for a 720-ft interval. The reservoir consists of two 50-ft pay zones (Pay Zones A and B), separated by a 10-ft-thick shale interval. Using the logs, the entire interval can be divided into three major categories: the gas-bearing sandstones, the shale barriers, and the nonproducing, nonbarrier siltstones. Fig. 1 also shows the stress gradient in each layer. The two pay zones (A and B) along with the shale barrier in between are best represented by use of eight layers, which we call Case 2. However, some engineers may want to simplify the eight-layer system to minimize computing time. If one chooses to use a three-layer model, then the three layers would include an upper high-stress barrier, one thick productive zone (Pay Zones A and B combined), and a lower stress barrier. Case 1 in Fig. 1 illustrates the stress profile for the simple, three-layer case. Figs. 2 and 3 illustrate the created fracture dimensions and the proppant profiles for the same treatment design for Cases 1 and 2, respectively. For Case 1, which represents the treatment interval with only three layers, both Pay Zones A and B are effectively propped, and the propped-fracture half-length is about 600 ft. However, with the detailed eight-layer model to represent the treatment interval in a 3D layered hydraulic-fracturing simulator, we find a significant increase in fracture height at the wellbore, and most of the proppant settles into the siltstone below Pay Zone B. The average propped-fracture half-length for Case 2 is only 230 ft, less than one-half of the propped-fracture half-length obtained for Case 1. This significant difference in propped-fracture half-length for the two cases was caused by the way the engineer chose to represent the layers. Case 1 uses the average values for in-situ stresses and has oversimplified the treatment interval by using only a three-layer system. Case 2 uses a more detailed description of the treatment interval by using an eight-layer system. The results indicate the importance of using a detailed description of the reservoir. Field Examples We now present examples of actual wells where hydraulic-fracture treatments have been performed. We have analyzed more than 20 wells in our study. Four wells have been chosen to illustrate the necessity of developing a layered reservoir description to obtain reasonable and consistent estimates of propped-fracture dimensions. To assist our evaluation, we have analyzed pressure-transient test data before and/or after the hydraulic-fracture treatments. Pressure-Transient Analysis. Prestimulation Transient Tests. For wells where prestimulation pressure-transient data were available, test data were evaluated to compute values of formation permeability, skin factor, and average reservoir pressure. Table 1 presents reservoir data and results obtained from prestimulation pressure-transient tests.
The initiation and growth of a hydraulic fracture cause fundamental changes in the free-oscillations that occur in a well in response to an initial temporary excitation. These changes include a lengthening of the period of pressure oscillations and decreasing rates of attenuation. Modeling a fracture as a resistive-capacitive well termination simulates the behavior observed in field tests. Using elasticity to define the hydraulic resistance and capacitance in terms of fracture dimensions relates free oscillations to fracture geometry. The approach described herein is a specialized example of impedance analysis, a promising new tool for fracture diagnostics.
There are numerous reasons for mapping the geometry of man-made hydraulic fractures. In the petroleum industry, this capability can result in improved fracture design, lowering well stimulation costs and improving oil and gas recovery. Hydraulic-fracture orientations and dimensions are also of interest in tectonic studies, where fracturing is used to determine states of crustal stress (Zoback et al., 1978). Because it is rarely possible to directly determine geometry by mining or drilling into the fracture, indirect observation techniques are necessary. Two widely used approaches are based on measurements of nondynamic pressure changes in the well. The first considers changes resulting from porous diffusion of fluid into the fracture and surrounding medium using classical pressure transient analysis (e.g. Raghavan et al., 1980). The second interprets dimensions and growth on the basis of predicted changes of static pressure in an inflating fracture in an elastic medium (Nolte and Smith, 1981). In this paper we present a third approach. We examine how dynamic oscillatory pressures in a well are affected by the growth of a hydraulic fracture. In particular, we consider the propagation, reflection, and attenuation of pressure waves set up by an initial temporary excitation. The attendant pressure and flow fluctuations in the well are known as free oscillations.
Consider a pressurized fluid-filled well that is cased to the bottom, unperforated and not otherwise hydraulically connected to the rocks that it penetrates. There is no initial flow in the well. A valve is now abruptly opened at the wellhead, a small volume of fluid is released at the volumetric flow rate Q, and the valve is again shut. This procedure causes a low-pressure wave to propagate down the wellbore at the speed of sound a as fluid rushes upward to replace the volume that was removed. When the wave reaches the bottom of the well at depth L, no new fluid is available to move into the low-pressure zone. Fluid directly above the bottom must now flow downward in an attempt to equalize the pressure. The low-pressure wave is thus reflected and travels back toward the wellhead. Upon reaching the wellhead another reflection occurs and the process continues. A pressure transducer located at the wellhead would record a pressure oscillation with a period of 2L/a that would continue until damped out by the effects of fluid friction. If the boundary condition at the bottom of the well is one of constant pressure rather than zero flow, the period of the pressure oscillation is doubled.
The equilibrium acid fracturing technique has been developed to stimulate wells in the Wasson San Andres Denver Production Unit. This new treatment technique maximizes acid contact time with the fracture faces while allowing control of the created fracture dimensions. Maximum acid contact time is essential to create highly conductive etched pathways on the fracture faces of cool dolomite formations which react slowly with acid. Control of fracture dimensions is important in the Denver Unit San Andres because fractures tend to grow uncontained in at least one vertical direction and the oil column is bounded by permeable gas bearing intervals above and permeable water bearing intervals below.
Using this technique, a fracture of desired dimensions is first created by injecting acid at fracturing rates. The volume of acid required to create the desired fracture dimensions is determined by a two dimensional fracture geometry program using design parameters determined from mini-frac testing and laboratory testing. Injection is then continued at reduced rates which maintain equilibrium with the fluid leak-off rate from the created fracture faces. By maintaining equilibrium between injection and leak-off, the created fracture can be held open without significant further fracture extension. Equilibrium is achieved in the field by maintaining the injection pressure below the fracture extension pressure, but above the fracture closure pressure determined by mini-frac testing.
The background and theory of this technique will be presented in this paper along with design procedures, field examples, results, and conclusions. A comparison of the results of the equilibrium acid fracture treatments to the other acid stimulations performed in the Denver Unit is also shown.
IntroductionThe Denver Unit is one·of several production units in the Wasson San Andres Field located in the western Texas counties of Gaines and Yoakum. The location of the Denver Unit within the field is shown in Figure 1. The San Andres formation is a Permian dolomite. The target interval is at a depth of about 5000'. The Wasson San Andres Field was discovered in 1936. The Denver Unit was formed in 1964 and waterflooding began at that time. CO2 flooding in the Denver Unit was started in 1984 and expansion is ongoing today.
Summary When fluid injection is shut-off after a fracture stage has been pumped, the sudden change in injection rate leads to a pressure fluctuation called a water hammer. These pressure pulses are observed and available at no additional cost because the pressure and rate data are recorded for every shut-in during field treatments. This abundant field data is commonly ignored. In this paper, we show that this water hammer signature can provide diagnostic information on fracture geometry. We simulated the transient flow problem in a wellhead-wellbore-fracture system to match the water hammer signature, and the solution provides the fracture dimensions based on the resistance-capacitance-inertance (R-C-I) circuit analogy. The analysis of water hammer signatures has been applied to multi-stage hydraulic fracture treatments to show the effect of input parameters and stress interference between stimulation stages. Water hammer simulation also suggests an accurate method to estimate instantaneous bottom-hole shut-in pressure (ISIP). This ISIP estimation for multi-stage treatments clearly shows the impact of the inter-stage stress shadow effect when applied to multi-stage fracture diagnosis. Simulated results which include stress interference effects indicate variations in fracture dimensions. This analysis also shows that the net fracturing pressure, near-wellbore frictional pressure drop, and stress magnitudes are changed by the stress shadow in multi-stage fracture treatments. This work has demonstrated that water hammer simulations can provide valuable fracture diagnostic information which compliments other diagnostic methods such as microseismicity and long-term production. Intitletroduction A pressure pulse is created when the fluid flow in a pipe is suddenly shut-in. This fluctuation of pressure is called a water hammer signature. It is observed in many instances in the oilfield. When an offshore water injection well is shut down or pumping of fluid is shut in during hydraulic fracture treatment, a water hammer signature is almost always observed as shown in Fig. 1a. This pressure fluctuation originates from the momentum change of the fluid in the conduit when the fluid experiences a sudden change of flow rate in a confined system. This pressure pulse propagates through the wellbore up and down within a few seconds as shown in Fig. 1b, and attenuates over time (typically within a few seconds to nearly a minute, depending on the condition of wellbore, fluid, fracture, and reservoir).