The main complexity in multi-component, multi-phase simulation in unconventional reservoirs arises because of the thermodynamic phase behavior and transport mechanisms in such small pores. To model the fluid flow in unconventional reservoirs, a computationally cost-effective implicit model which adheres these issues is needed. This paper presents a new multi-component, multi-phase, dual-porosity numerical model including molecular diffusion for simulating fluid flow for the field applications in Eagle Ford formation.
The proposed model, which is based on the mass transport equations for each component, solves for the pressures and overall compositions simultaneously using a new discretization technique. This new model, which determines the effect of molecular diffusion on production, is computationally robust and efficient. The vapor-liquid equilibrium calculation is performed using Peng-Robinson equation of state including the phase shifts and the capillary pressure effects on phase behavior. The production data from Eagle Ford wells are used in conjunction with the simulation results to evaluate the accuracy and computational efficacy compared to the conventional volume balance technique.
The compositional mass transport equations can be solved using several variants of the fully implicit method in addition to a less stringent sequential volume balance technique. The latter is of interest because one can generate a simple compositional pressure diffusivity equation and sequentially solve for the composition. However, the volume balance technique is sensitive to the grid and time-step size. We compared our results with the volume balance technique that is implicit pressure and explicit composition and saturation (IMPECS) method to examine the computational times and accuracy for both solution techniques. The model presented in this paper solves for the crossing of phase boundaries more efficiently compared to other methods. Also, the degree of implicitness in our model not only provides better accuracy for the simulation but also eliminates the material balance errors occurring in explicit solutions. In particular, a decrease of thirty percent in cumulative oil production is achieved using the new model. Therefore, our method provides a better understanding of the physical phenomena of fluid flow processes in unconventional reservoirs.
Unconventional reservoirs consist of a matrix, macro- and micro-fractures, of which the difference in stress-dependent deformation affect the long-term production performance. Accurate representation and modeling dynamic behavior of this multiple continua system are needed to model fluid flow. Changes in reservoir pore pressure during production causes decrease in porosity, permeability, and reduction in the pore volume which, in turn, is reflected by an increase in pore compressibility. The combined effect of these variables both in the fracture and matrix determines the long-term performance of unconventional shale reservoirs. Most of the coupled geomechanics and fluid flow simulation models do not account for such details in pore compressibility. In our formulation which is based on linear poroelasticity theory, we calculate the varying Biot modulus with declining pressure. Moreover, this calculation provides the ever- changing pore compressibility rather than assuming constant compressibility. Therefore, this new model shows the effect of compaction drive on production in unconventional reservoirs. Field data from Bakken formation and a novel in-house fully-coupled multi-phase dual-porosity model are used to analyze the effect of pore compressibility alterations on reservoir performance.
We have observed that the energy provided by the reservoir compaction increases cumulative fluid production, and numerical models that has constant compressibility values in their transport equations underestimates the cumulative production. Specifically, the cumulative production is reduced by less than three percent if only the decrease in porosity and permeability is applied to fluid flow model with constant pore compressibilities. Whereas, the cumulative oil production is increased by eleven percent if the compaction drive including the changes in porosity, permeability, and pore compressibility is applied to the fluid flow model. Using the results of this work, we can demonstrate the relationship between elastic properties and stress-dependent rock parameters such as porosity and permeability on reservoir performance. Also, we could define the impact of factors such as pore compressibility to improve the accuracy of long-term production and forecast in unconventional shale reservoirs.
Our method presented in this work provides an accurate determination of the driving mechanisms in the unconventional reservoirs. The theory behind the proposed model and its applications will provide broader insights into the production-decline trends in unconventional reservoirs.
Pore diameters for shale reservoirs are on the order of few nanometers which become even smaller during production because of rock deformation. This dynamic interaction between pore fluid pressure and rock stress affects the phase behavior in unconventional reservoirs. In this paper, a new mathematical formulation of fully-coupled geomechanics and compositional dual-porosity model was used to determine the impact of rock deformation and confinement on the nanopore fluids as well as their effect on the production performance of Eagle Ford formation. The formulation presented was derived from our multiphase poroelasticity model which was an extension to the single-phase, single-porosity Biot's linear poroelasticity theory allowing to characterize the rock deformation and pore diameter reduction using the bulk modulus of the matrix-fracture system. Changes in reservoir pore pressure and rock deformation that cause the pore diameter to reduce increases the capillary pressure in the pores which affects the bubble-point pressure suppression and significant shift in the phase envelope, favoring longer period of single-phase production. It was observed that not taking rock deformation into account will lead to over estimation of production, whereas ignoring the effect of pore confinement would underestimate the production forecast. In an example field study based on Eagle Ford reservoir, an increase of around eight percent in cumulative oil production was achieved when the effect of rock deformation and confinement was included in the compositional model compared to the case where only the rock deformation was included. On the other hand, if only pore confinement effect was included in the simulation runs, four percent of increase was achieved.
Microseismic data acquired in a single observation well parallel to the axis of rotational symmetry of surrounding rocks - typically, in a vertical well drilled through a horizontally layered isotropic or vertically transversely isotropic formation - cannot be uniquely inverted for six independent components comprising the full seismic moment tensor. To constrain the inversion for such a survey geometry and medium symmetry, one might assume certain physical properties of seismic sources, the properties relating otherwise independent moment components to each other, regularizing moment tensor inversion, and helping reduce its ambiguity. Our paper examines one possibility of this kind: the assumption of a tensile fracture, rupturing the focal region along a plane of its greatest weakness. Mathematical formulation of inversion of single-well mi-croseismic records for the parameters of tensile fractures reveals that the true solution, always recoverable from properly acquired data, might be accompanied by two spurious solutions. The analysis of those solutions leads to a criterion that, although not perfect, makes it possible to select the correct solution in the majority of elastic models. We apply our methodology to a field data set recorded with multiple vertical downhole arrays and demonstrate that the results of dual-well moment tensor inversion can be replicated with single-well data.
Obtaining precise hypocenters of microseismic events is the primary objective of contemporary microseismic surveys. The sought precision usually hinges on two factors: an accurate velocity model, its inaccuracy biasing hypocenters of an entire event population, and an adequate data-recording aperture, its deficiency blurring the hypocenters of individual events and causing noisy appearance of event population computed even in a highly precise velocity model. Our paper explores the possibility of aperture enhancement through adding reflections to direct arrivals, conventionally used to locate microseismic events. We illustrate the performance of our multiphase (that is, direct reflected waves) event-location technique on a data set recorded in the Anadarko Basin Woodford play, Oklahoma, USA, and demonstrate that the use of reflected waves not only improves velocity model but also unambiguously places the recorded events in the Woodford formation, something that cannot be achieved with the direct arrivals alone.
Three algebraic surfaces – the slowness surface, the phase-velocity surface, and the group-velocity surface – play fundamental roles in the theory of seismic wave propagation in anisotropic media. While the slowness and phase-velocity surfaces are fairly simple and their main properties are well understood, the group-velocity surfaces or the elementary wavefronts are extremely complex; they are complex to the extent that even the algebraic degree,
This paper establishes the exact degree (
Presentation Date: Tuesday, September 26, 2017
Start Time: 10:10 AM
Presentation Type: ORAL
Hyperspectral imaging (HCI) is a non-destructive analytical technique that uses infrared light to produce a visual ‘map’ of the minerals in a core. A whole core from the Bakken formation (Mississippian-Devonian) in the Williston basin, North Dakota, was scanned using both low-resolution (1.5 mm) short-wave infrared energy (SWIR) and longwave infrared (LWIR) energy. Next, a new technology was employed that uses three cameras to simultaneously acquire high-resolution information over wider wavelengths of the electromagnetic spectrum. This new long-wave infrared (LWIR) spectrometer, the first in the United States, contains a specialized lens to obtain data at a high resolution of 300-500 μm pixels and measure responses from tectosilicates, carbonates and some clays, as well as hydroxides, sulfates and phosphates. The new SWIR, which also uses a specialized lens for a high resolution of 300-500 μm pixels, identifies carbonates, hydroxides, sulfates, hydrocarbons, other silicate minerals, and clays. The LWIR and SWIR data were co-registered with high resolution (160 μm) RGB core photographs taken under high-wattage white LED lights.
Mineral maps (both multi-mineral and single-mineral heat maps) of the core were obtained that display the textural relationships of the minerals in each core and distinguish subtle variations in mineral composition with depth, including silicate, carbonate and clay species. Curves of this mineralogical and textural data were then imported into petrophysical software to facilitate comparison of the older, low-resolution data with the data generated by the new high-resolution SWIR and LWIR technology. Furthermore, the SWIR and LWIR curves were overlaid with a variety of petrophysical curves, which allowed the visualization of the link between mineralogy and log measurements. Finally, quantification of continuous mineralogy from core provides calibration for petrophysical interpretation and integration of multiple mineralogy analysis from logs.
The use of hyperspectral image data to map minerals on the surface of the Earth was the result of decades of research conducted in both commercial and government sectors. Hyperspectral imaging of core using short-wave infrared light (SWIR) is a non-destructive analytical technique originally developed from airborne systems by the mining industry (Taranak and Aslett, 2009), and since the late 2000s has been widely applied in other industries, including food, forensics, pharma and art (e.g. Cucci et al., 2016; Sun, 2010; Edelman et al., 2012; Lu and Fei, 2014).
Uzun, Ilkay (Colorado School of Mines) | Eker, Erdinc (Colorado School of Mines) | Cho, Younki (Colorado School of Mines) | Kazemi, Hossein (Colorado School of Mines) | Rutledge, J. M. (Marathon Oil Company)
In reservoir evaluation of unconventional reservoirs, engineers use rate transient analysis (RTA) to assess the performance of hydraulically fractured horizontal wells. This technique consists of plotting rate-normalized pressure drop at the well sandface versus square root of time (i.e., linear flow) to calculate the reservoir effective permeability. In this paper, we report on the use of an in-house multiphase numerical simulator applied to production data from several Eagle Ford wells to assess both single-and multiphase production performance. Furthermore, a combination of numerical simulation and analytical solution techniques was used to validate the information obtained from the RTA in the Eagle Ford wells. This procedure further improved our insight into the flow behavior of unconventional reservoirs. The multiphase effective permeability calculated from the RTA was within a few percent of the input data to the numerical model. Furthermore, the multiphase RTA results produce more accurate than the conventional single-phase RTA approach; and the multiphase analysis results were more in line with the well performance observations than with when using the single-phase analysis approach. Furthermore, the multiphase results were more consistent with the success or failure of the hydraulic fracturing process.
The geometry of the hydraulic fractures affects well productivity in unconventional shale reservoirs. Modeling hydraulic fractures require simplifying assumptions because of the complexity of the rock deformation and the quality of rock physics data available. Nonetheless, modeling hydraulic fractures provides insight about the well productivity of unconventional reservoirs. In hydraulic fracture modeling, we generally assume the fracture geometry based on laboratory observations and invoke force balance on the conceptual physical model to arrive at governing equations. In this paper, we present a practical approach to modeling and estimating hydraulic fracture length and width using well stimulation treatment data in Eagle Ford formation. Our model relies on fracture propagation theory of Perkins, Kern, and Nordgren (PKN) and utilizes acoustic log rock mechanical properties and fracture treatment data for each stage. Sensitivity analysis was performed to assess the effect of fluid leak-off, rock mechanical and fracturing fluid properties on hydraulic fracture propagation. We applied the model to every fracture stage of several Eagle Ford wells. The application of our model to the analysis of Eagle Ford well data revealed that fracture length is significantly more sensitive to the Young modulus than the Poisson ratio, and, in turn, this affected well productivity. The predicted hydraulic fracture geometry is consistent with the rate transient analysis of production data.
A recently proposed paraxial ray-based technique for relative location of microseismicity is extended to accommodate several master events with respect to which other events, termed the slaves, are located. The multi-master extension addresses two issues inherent for the existing single-master algorithm: a gradual decrease of its accuracy with the distance from the master and less than satisfactory performance in the presence of strong velocity variations. Those deficiencies are handled by applying an improved paraxial traveltime formula, exact in homogeneous elliptically anisotropic media, and by distributing multiple masters in the subsurface to sample its heterogeneity. The contributions from different master events to the hypocenter of a given slave are automatically weighted to enhance the influence of adjacent masters, ensuring the precise slave location, and to suppress the influence of distant ones, tending to increase the slave-location errors. Tests of the multi-master relative event-location method on synthetic and field microseismic data demonstrate its precision and flexibility, as well as applicability to both surface and downhole microseismic geometries.
Presentation Date: Monday, October 17, 2016
Start Time: 3:45:00 PM
Presentation Type: ORAL