ABSTRACT: The role of faults and natural fractures on hydraulic fracture stimulation in the Vaca Muerta Play, of the Neuquén basin, Argentina, is investigated using a 3D, fully coupled, fracture and fluid flow simulator. During the development of the Vaca Muerta Shale, and due to the complex nature of the geology and geomechanical conditions of the play, well interference has been observed while pumping a vertical well. Microseismic events recorded during stimulation show a map-view “Z” pattern that has not been observed before and potentially identifies a trend of preexisting natural fractures/faults striking NE-SW and NW-SW, and steeply dipping that could hydraulically connect adjacent wells. We used a 3D coupled flow and geomechanical simulator to evaluate the geometric characteristics and material and fault properties that could produce well interference using the field parameters utilized during the stimulation operations in the study pad, and that would produce comparable synthetic microseismicity. Simulation results show that fluid pressure can communicate through sufficiently conductive faults for relatively long distances between 300 m and 700 m. In the case investigated here, hydraulic fractures can be arrested by faults and reinitiate from its tip. Fracture height growth exceeds the estimated propped height, consistent with the distribution of microseismicity. Results also indicate that the microseismic event cloud distribution follows large-scale structures, and therefore the presence of conductive faults could explain the well interference problems observed in the field.
The vast majority of the oil-and-gas-bearing reservoirs in the world contain rock discontinuities such as faults and fractures (Bratton et al., 2005) that affect how fluid is distributed in the reservoir. The properties of the faults and the rock surrounding them, such as fault gouge, fluid pressure, fluid types, lateral sedimentary facies variations, reservoir architecture, and stresses, dictate whether faults act as barriers or transmitters of fluid flow (Barr, 2007).
Izadi, Ghazal (Baker Hughes) | Moos, Daniel (Baker Hughes) | Cruz, Leonardo (Baker Hughes) | Gaither, Michael (Baker Hughes) | Chiaramonte, Laura (Baker Hughes) | Johnson, Scott (GeoNumerical Solutions)
Weak bedding planes create a unique mechanism for hydraulic fracture height containment, providing one possible explanation for unusual patterns of height growth in shale formations. This paper describes an investigation into how bedding planes modify the interactions between multiple, simultaneously propagating hydraulic fractures in a formation with weak horizontal interfaces with laterally varying properties. The investigation used a 3-D simulator that fully coupled geomechanics, fracture mechanics, and fluid behavior. Three equally spaced fractures were simulated along a horizontal trajectory. Fluid was injected simultaneously into all three locations, and partitioned according to maintain a specified total injection rate. Variations in perforation spacing, fluid viscosity and injection rate are modeled. The four designs investigated were: 1) 10-cp fluid, 20b-pm injection rate with 30-m cluster spacing. 2) 100-cp fluid and 20-bpm fluid injection rate with 30-m cluster spacing. 3) 10-cp fluid and 40-bpm injection rate with 30-m cluster spacing. 4) 10-cp fluid, 20-bpm and 45-m cluster spacing. Results showed how these changes affected fracture area and shape. The propped surface area for each scenario was also estimated. The results suggested that the presence of laterally varying weak interfaces can significantly affect fracture interference.
Cruz, Leonardo (Baker Hughes Inc.) | Fu, Pengcheng (Lawrence Livermore National Laboratory) | Izadi, Ghazal (Baker Hughes Inc.) | Moos, Daniel (Baker Hughes Inc.) | Sheridan, Judith (Baker Hughes Inc.) | Settgast, Randolph R. (Lawrence Livermore National Laboratory) | Ryerson, Frederick J. (Lawrence Livermore National Laboratory)
We use a 3D, fully coupled, fracture and fluid flow simulator to investigate the interactions between hydraulic fractures (HF) and pre-existing joints (J1 & J2), with stresses and joint characteristics based on typical Marcellus Shale properties. We consider two modeling scenarios in which we vary the orientation of the maximum horizontal stress (SHmax) with respect to the orientation of J1. Multiple stochastic realizations of the discrete fracture network (DFN) describing these joint sets, and two fluid viscosities end-members, were considered. Modeling results show that the fluid pressure perturbations extend to distances along J1 and J2 that are comparable to the microseismic (MS) event cloud, making joint shearing and “wet” events a probable consequence of the stimulation. Furthermore, the tendency of the main HF to be captured and offset by J1 and the extent of the offset depends in part on the fluid viscosity, the orientation of J1 with respect to SHmax, the length of J1, and the extent to which their height spans the zone of fracturable rock between the upper and lower fracture barriers. Comparing model predictions to MS interpretations suggests that the most probable geometric scenario for the pad site requires that SHmax and J1 are ~30 degrees apart.
The interactions between hydraulic fractures and geologic discontinuities, such as natural fractures, faults, and bedding that have been documented in sedimentary basins (e.g. the Appalachian Basin), where unconventional reservoirs are usually exploited, are very relevant to fracture stimulation design to enhance oil/gas recovery (Carlson and Mercer, 1991). An efficient fracture stimulation design, one that maximizes hydrocarbon recovery and reduces cost, should consider and possibly take advantage of these interactions, requiring a better understanding of the governing processes and the use of 3D fully-coupled reservoir simulators to effectively capture feedback mechanisms. Fracturing fluid can be diverted or “loss” into the discontinuities, which reduced the hydraulically fractured volume, but those preexisting discontinuities could also be stimulated, increasing the permeability pathways.
Settgast, Randolph R. (Lawrence Livermore National Laboratory) | Izadi, Ghazal (Baker Hughes Incorporated) | Hurt, Robert S. (Baker Hughes Incorporated) | Jo, Hyunil (Baker Hughes Incorporated, Applied Numerics LLC) | Johnson, Scott M. (Lawrence Livermore National Laboratory, Applied Numerics LLC) | Walsh, Stuart D. C. (Lawrence Livermore National Laboratory) | Moos, Daniel (Baker Hughes Incorporated) | Ryerson, Fredrick (Lawrence Livermore National Laboratory)
Design decisions for the layout and properties of perforation clusters in a hydraulic fracture stimulation job are typically based on idealizations that treat the fractures originating from each cluster identically. However, simulations of multi-clustered hydraulic fracturing stages have shown that some perforation clusters may be rendered ineffective due to an increase in confining stresses (i.e. stress shadow) induced by hydraulic fractures originating from neighboring clusters. Two methods to counteract the effects the inter-cluster hydraulic fracture interaction are using non-uniform cluster spacing, and varying the frictional properties of the perforation clusters themselves as investigated in [
In this work, the authors present a method for the evaluation of the effects that cluster spacing and frictional properties of perforation clusters have on the propagation of hydraulic fractures during a stimulation stage. This approach is done through the application of the hydraulic fracture simulation capabilities of the GEOS simulation framework, developed at Lawrence Livermore National Laboratory. GEOS provides a hybrid Finite Element Method/Finite Volume Method that fully couples the mechanics of rock deformation, the flow of fluid through the crack, and fluid flow through the rock matrix. This capability allows for the development of a method for the optimal design of hydraulic fracture stimulation staging that relies on basic engineering principles. For a given set of site properties, multiple simulations are performed with variations in cluster spacing, cluster configuration, fluid properties, and pumping pressure/rate.
Izadi, Ghazal (Baker Hughes Inc.) | Gaither, Michael (Baker Hughes Inc.) | Cruz, Leonardo (Baker Hughes Inc.) | Baba, Christine (Baker Hughes Inc.) | Moos, Daniel (Baker Hughes Inc.) | Fu, Pengcheng (Lawrence Livermore National Laboratory)
Hydraulic fracturing is a reservoir stimulation technique which in unconventional (e.g., shale gas or oil) plays is required to generate sufficient reservoir contact to enhance production. Frac design approaches use fracture spacing, sequence, and other parameters to enhance fracture complexity through modification of the local stresses within the stimulated area. Conventional frac designs consist of a series of stages within a well stimulated sequentially, from toe to heel. A newer promising design is a Zipper frac, where two or more parallel horizontal wells are stimulated stage by stage alternating between wells in a “Zipper” pattern. The effectiveness of Zipper fracs has been demonstrated by industry; however, the treatment optimization is still under development. In this paper a fully 3-D reservoir model is used to capture some of the fundamental effects which influence fracture growth during stimulation using three different stimulation approaches, for two different well spacings. The three designs are investigated are: Conventional, in which all stages in sequence are stimulated first in Well 1, followed by Well 2 and then Well 3; Zipper frac, in which Stage 1 is stimulated in Wells 1, 2, and 3 and then Stage 2 is stimulated in each of the wells, repeating until all stages are stimulated; and Modified Zipper frac, in which Stage 1 is stimulated in Well 1, then in Well 3, then in Well 2, and the same sequence is repeated. Following each stage the fractures are numerically propped so that their residual aperture is 50% of the total aperture, eliminating any surfaces with apertures smaller than 0.0025 meters as too small to receive proppant. In this way we can compare the relative fracture surface area which remains productive among the different approaches. The results reveal subtle but significant changes in the stress state due to the different stagings both as a function of time during the stimulation (which will affect subsequent propped aperture) as well as after the entire sequence is pumped. The greatest area is achieved for the modified Zipper sequence; the improvement is larger for the closer spacing of the wells. These changes in turn result in differences in total propped area which will lead to differences in IP. One interesting consequence of the analysis is that although the initial propped area is larger if wells are farther apart, the ratio of propped area to total area available to be drained is smaller suggesting that the initially higher IP for the larger well spacing is achieved at the cost that ultimate recovery for the field is likely to be lower.
One of the primary goals of hydraulic fracturing is to maximize the effective fracture surface area that connects the reservoir volume to the wells. The operational parameters that can be relatively easily manipulated include fracturing fluid properties, injection rate, proppant quantity and schedule, and the spacing and stimulation sequence of stages in a multi-well multistage system (pad). Recent studies (Roussel et al., 2011; Rafiee et al., 2012; Rios et al., 2013,) have suggested that it is possible to achieve greater “complexity” of the resulting fracture network through appropriate engineering of the staging and cross-well stimulation sequences. There is no clear definition on which the above papers agree for the term “complexity,” nor is it clear how they treat each fracture after injection into that fracture is complete. Rather than address this issue here, we focus in this paper on the differences among the different effective surface areas of the primary fractures. And, we clearly define the behavior of each fracture after it has been pumped and propped.
Gui, Feng (Baker Hughes) | Rahman, Khalil (Baker Hughes) | Moos, Daniel (Baker Hughes) | Vassilellis, George (Gaffney, Cline & Associates) | Li, Chao (Gaffney, Cline & Associates) | Liu, Qing (Baker Hughes) | Zhang, Fuxiang (PetroChina Tarim Oil Company) | Peng, Jianxin (PetroChina Tarim Oil Company) | Yuan, Xuefang (PetroChina Tarim Oil Company) | Zou, Guoqing (PetroChina Tarim Oil Company)
Shayegi, Sara (Shell) | Kabir, C. Shah (Hess Corporation) | If, Flemming (Hess Corporation) | Christensen, Soren (Hess Corporation) | Ken, Kosco (Hess Corporation) | Casasus-Bribian, Jaime (Hess Corporation) | Hasan, ABM K. (Hess Corporation) | Moos, Daniel (Dong E&P)
Underbalanced drilling (UBD) offers a unique opportunity to estimate undamaged, in-situ formation properties upon first contact with the formation while drilling. This paper compares well-testing techniques developed for UBD with conventional methods. The reservoir flow rates in combination with flowing bottomhole pressures (BHPs) acquired while drilling can be used to identify productive intervals and estimate dynamic reservoir properties.
Unlike typical UBD projects where reservoir benefits are the primary focus, the driver for this mature field was overcoming the drilling problems associated with the wide reservoir-pressure variability caused by nearby producers and injectors. UBD was piloted as a means to achieving the requisite lateral lengths for reserves capture and meeting production targets. Minimizing formation damage and characterizing the reservoir while drilling were added benefits.
Several reservoir-characterization methods based on rate-transient analysis (RTA) were used to perform well testing while drilling. Rate-integral-productivity-index (RIPI) analysis uses the rate and pressure data acquired during drilling to determine whether additional holes drilled contribute and to ascertain the relative quality of this rock. In the increasing-boundary method, real-time rate and pressure data during drilling, circulating, and tripping allowed assessment of formation properties through history matching. Pressure-buildup data were also available during trips because the concentric annuli allowed the pressure to be monitored below the downhole isolation valve. These data enabled the estimation of reservoir pressure and productivity index (PI) with a proxy vertical-well model for each productive interval drilled. These interpretation methods show close agreement in results and lend credence to the UBD-derived parameters.
The "Shale Engineering" approach and modeling addresses production forecasting in shale and tight formations. This new reservoir simulation methodology relies on modeling the propagation of the stimulated rock volume from the near-well vicinity to deep into the formation. Simulation models are built for individual fracturing stages and validated by matching treatment pressures and rates while conforming to geomechanical and microseismic observations. Stage models are then combined into a larger well model where individual stage contribution and early production performance are matched. This approach was applied on a project that was developed by EQT in an Upper Devonian shale formation in West Virginia. Data available for this project included fit-for-purpose formation evaluation description, production logs and downhole microseismic data with advanced processing and interpretation. The results provided a good match to early well performance, despite the complexity of having to match a combination of shales and partially depleted tight sandstones that had been stimulated by foam fracturing with proppant. This approach can be used not only to predict production, but also as a practical platform for field development design and optimization. Furthermore, the matched results validated the shear stimulation model developed by the authors for this type of application. The approach makes it possible to exploit microseismic observations in a more realistic way in order to describe the stimulated rock volume (SRV), and it explains early-life production logs that indicate uneven fracturing stage contribution. The model also can relate stimulation effectiveness to pre-existing formation rock and fluid properties, and thus can be used as a guide to identify optimal formation targets. The "Shale Engineering?? approach hinges on the premise that when unconventional "tight rocks?? containing hydrocarbons are modified by hydraulic fracture stimulation, the process converts them into "reservoir rocks??. In addition, interpretation of the newly created "artificial reservoirs?? is accomplished through multi-displinary expertise that is focused on providing a rate performance and predictive model to aid in reservoir development. Because unconventional resource/reservoir formations are unique and subject to a wide range of conditions, they require a production predictive method more suitable for this task than the commonly used "Type Curves". The advantage of the "Shale Engineering?? approach is that it allows validation with parameters that can be available at an early stage of the well life, which in turn are useful to constrain model solutions. It also offers the means to include geomechanics in a practical workflow that allows systematic workflow allows for step-by-step validation of the model. The suggested simulation process uses commerical software and it can be applied to either simple or complex cases.
Exploitation of unconventional shale gas reservoirs depends on successful hydraulic fracturing and horizontal drilling. Mineralogy, organic matter content, acoustic anisotropy, and in-situ stress all play an important role for well completion design. As part of a comprehensive site study of the Upper Devonian Huron shale, borehole acoustic and mineralogy logging data, in addition to conventional logs, were acquired in a vertical well prior to hydraulic fracturing and microseismic monitoring of a series of laterals drilled from the same location. The acoustic data was processed for compressional wave, cross-dipole shear, Stoneley-derived horizontal shear, radial velocity variations, and borehole Stoneley reflectivity indicators. The cross-dipole anisotropy and the near-well radial slowness variations provided information about intrinsic anisotropy and stress sensitivity to determine the source of dipole-mode anisotropy. Significant transverse acoustic anisotropy was detected and used to obtain vertical and horizontal dynamic elastic properties. The mineralogy and petrophysical analysis were used to generate a micromechanical constitutive model to reproduce numerically the laboratory stress-strain behavior of the rock, from which quasi-static mechanical properties were determined. These were calibrated against triaxial tests on core samples from an offset well, and the vertical and horizontal static elastic rock properties were used to estimate the vertical variation of the horizontal stress. The resulting stress profile, along with accurate mineralogy and petrophysical analysis, provides important information to select the best vertical locations of lateral wells and to identify natural fracture barriers.
Moos, Daniel (Baker Hughes) | Lacazette, Alfred (EQT Production) | Vassilellis, George D. (Gaffney, Cline & Associates Inc.) | Cade, Randall (Baker Hughes Inc) | Franquet, Javier Alejandro (Baker Hughes) | Bourtembourg, Eric (Baker Hughes) | Daniel, Guillaume (Baker Hughes)
A comprehensive site study carried out in an Upper Devonian shale gas reservoir at a site in southern West Virginia provided data to test a geomechanical model for stimulation of the Huron formation. Using a model in which natural fractures provide the primary conduits for production and are the major target for stimulation, and in which stimulation triggers shear slip on those pre-existing fractures, we were able to predict the shape of the reservoir volume stimulated by injection of high-quality foam and to match injection flow rates and pressures using a dual porosity dual permeability finite-difference flow simulator with anisotropic, pressure-sensitive reservoir properties. The resulting calibrated model matched both the relative contribu-tion of the individual stages measured by production logging and the early-life well production. This suggests that similar models may in the future provide earlier and better production predictions, guidance for completion and stimulation design, and recommendations to minimize production decline and maximize well value.