Lu, Mingjing (China University of Petroleum, Colorado School of Mines) | Su, Yuliang (China University of Petroleum) | Wang, Wendong (China University of Petroleum) | Zhang, Ge (Xianhe Oil producing Plant, Shengli Oilfield, Sinopec)
Refracturing treatment are performed since stimulation effect won't last for entire life. Screening wells for refracturing needs a systematic analysis of huge amounts of data. With literature review, it is obviously that there are many factors controlling the success of refracturing and factors may vary in different oilfields. Proper factors and data processing are the primary principle in candidate selection. The Integrated Multiple Parameters (IMP) method is presented to provide assists in selecting candidate wells.
After deeply researching over 200 restimulated wells, all factors thought to be related with success of refracturing are listed and analyzed, results show that single factor may have great influence on restimulation but no significant patterns can be obtained since too many factors making things complicated. The IMP method proposes five parameters which are all integrated by those single factors. It is emphasized that all parameters have physical or engineering meanings which makes it easier to quantify their correlation in refracturing. Besides, all the parameters are dimensionless which makes it easier for using in mathematical models and statistical analysis.
The five dimensionless parameters are developed considering the most important aspects of candidate wells selection which are showed as followed: fracture reorientation, well completion, reservoir depletion, production decline, oil-water well connectivity. Parameters are calculated for all the restimulated wells to dig into their correlation with the outcomes of refracturing. A simple decision model is built to help with screening wells for refracturing. Results shows that it is more executable to evaluate and predict the success of refracturing with these dimensionless parameters. Fracture reorientation parameter is the primary one to be considered since it leads to fracture reorientation which brings significant production increment. Then two types of potential wells are picked: (a) wells with dissatisfied initial well completion, low production decline rate and high oil-water connectivity parameter; (b) wells with satisfied initial well completion, high well completion parameter, low production decline parameter, reservoir depletion parameter and low oil-water connectivity parameter for wells that are not easy for fracture reorientation. Wells selected are proved to be refracturing potential which verify the reliability and accuracy of IMP method.
The IMP method is an improved approach integrating most of the important factors which makes candidate selection much more predictable and it succeeds in screening out more than 80% of the potential wells in field test. Also, it can be applied widely in different oilfields since all the parameters are dimensionless. By combining with some mathematical methods such as neural networks, it can even predict increment of the restimulation treatment.
Shale reservoirs contain predominantly micro and mesopores (<50 nm), within which gas is stored as free or adsorbed gas. Due to the ultra-small pore size, multiple transport mechanisms coexist in shale reservoirs, including gas slippage, Knudsen diffusion of free gas and surface diffusion of adsorbed gas. In this work, we propose a new transport model, valid for all ranges of Knudsen number, which combines all transport mechanisms with different weighting coefficients. To quantify the effects of influence factors, we introduce the compressibility factor for real gas effect and effective pore radius for gas adsorption and stress dependence. The model is proven to be more accurate than existing models since the deviation of the analytical solution of our model (3%) from published molecular simulation data is lower than that of existing models (10~20%). Based on this model, we compare (1) the contribution of each transport mechanism to gas transport in pores of different radii, (2) shale permeability measured in laboratory and at reservoir conditions, and (3) permeability of nanopores and natural fractures. It is found that gas transport is dominated by gas slippage and surface diffusion when the pore radius is over 10 nm and below 5 nm, respectively. Knudsen diffusion only becomes significant when the pore radius is between 2 and 25 nm and pore pressure is below 1000 psi. Furthermore, laboratory measurements usually over-estimate shale permeability. We also propose a promising enhanced gas recovery method, which is to open and prop up closed natural fractures using micro size proppants.
Lascelles, Peter (EP Energy) | Wan, Jichun (EP Energy) | Robinson, Lauren (EP Energy) | Allmon, Randy (EP Energy) | Evans, Grant (EP Energy) | Ursell, Luke (Biota Technology) | Scott, Nicole M. (Biota Technology) | Chase, John (Biota Technology) | Jablanovic, Jelena (Biota Technology) | Karimi, Moji (Biota Technology) | Rao, Vik (Biota Technology)
DNA diagnostics is a new reservoir characterization tool with potential to maximize reservoir production in tight rock formations. DNA extracted from rock layers provides high resolution fingerprints that define a "DNA stratigraphy" for organic intervals like the Wolfcamp. DNA sequences originate from microbes feeding on organic matter or minerals within the formation. A DNA stratigraphic profile, or type section, was assembled from a vertical pilot well's cuttings and core. The DNA signature from produced oil from offset laterals was subsequently compared against the DNA type section to provide estimated effective drainage height. Cuttings from a lateral well were compared with DNA from its produced oil to construct a production profile comparable to a traditional production log. In addition, when oil samples are collected over time, the method provides insight on interference, completion effectiveness, and SRV (Stimulated Reservoir Volume) changes with time.
An optimized development plan in unconventional reservoirs requires operators to understand parameters such as effective drainage height, hydraulic fracture half-length and individual stage contributions resulting from their completions. Wolfcamp reservoirs consist of highly laminated mudrocks interbedded with limestones that have quite different mechanical properties. These contrasting lithologies make it difficult to estimate resultant completion geometries, SRV, and well-to- well interactions. Also, using costly production logs, individual stage contributions are difficult to obtain in lower pressure reservoirs like the Wolfcamp. However, these reservoir performance parameters are required to set benchmarks and continuously uplift the EUR by taking advantage of insightful diagnostics.
Production logs, micro-seismic, chemical or radioactive tracers are all useful in understanding the subsurface, but can be expensive and can pose operational challenges. Subsurface DNA sequencing is a relatively low cost new data source that can be used to gain subsurface insights in complicated reservoirs. DNA stratigraphy can help assess critical geometric parameters resulting from stimulation by employing non-invasive sampling that enables lifetime well monitoring to track the flow of oil and provide engineers the basis to optimize completions and development plans.
An 8 well "subsurface" lab was selected for the experiment. The project included one vertical pilot hole with cuttings, and 8 horizontal wells landed in two Wolfcamp pay zones (one of the laterals was extended from the same vertical pilot). Three horizontals had been on production for 11 months before the pilot well and 6 additional laterals were drilled. The pilot well and its sidetracked lateral had cuttings extracted for DNA sequencing. DNA signatures from the pilot well and lateral well were compiled to produce vertical and lateral DNA stratigraphic profiles. The DNA stratigraphic profiles were then compared to DNA from oil produced in the 7 offset laterals. DNA profiles were also compared to standard geologic parameters using pilot well e-logs, particularly mechanical stratigraphy. Lateral wells were sampled at various times after initial production to assess changes with time. Blind tests were designed to check the method as a reasonable estimator for effective drainage height and communication.
DNA stratigraphy provides a more informed view of well spacing, completion design and well performance to help increase efficiency and asset value.
Summary In this study, we investigate the reported ultrasonic measurements of core samples of organic shale and our lab measurements. We focus on Vp/Vs ratios, P-impedance, and velocity anisotropies for different types of organic shales. A layered model is constructed to explain why the silica-rich shales tend to have higher S-wave anisotropy and calcareous shales tend to have higher P-wave anisotropy. Hence, ratio may be potentially treated as an indicator for mineralogical composition in shale reservoirs. We also observe linear trends between Thomsen's parameters, / So rock physics models developed for clay-dominated mudstones are inappropriate and cannot be applied to organic shale.
Wu, Keliu (University of Calgary and China University of Petroleum) | Li, Xiangfang (China University of Petroleum (Beijing)) | Guo, Chaohua (Missouri University of Science and Technology) | Wang, Chenchen (University of Calgary) | Chen, Zhangxin (University of Calgary)
A model for gas transfer in nanopores is the basis for accurate numerical simulation, which has important implications for economic development of shale-gas reservoirs (SGRs). The gas-transfer mechanism in SGRs is significantly different from that of conventional gas reservoirs, which is mainly caused by the nanoscale phenomena and organic matter as a medium of gas sourcing and storage. The gas-transfer mechanism includes bulk-gas transfer and adsorption-gas surface diffusion in nanopores of SGRs, where the bulk-gas-transfer mechanism includes continuous flow, slip flow, and Knudsen diffusion. First, a model for bulk-gas transfer in nanopores was established, which was dependent on slip flow and Knudsen diffusion. The total gas flux in the bulk phase is not a simple sum of slip-flow flux and Knudsen-diffusion flux but a weighted sum on the basis of corresponding contributions. The weighted factors are primarily controlled by the mutual interaction between slip flow and Knudsen diffusion, which is determined by probabilities between gas molecules colliding with each other and colliding with nanopore surface in this newly proposed model. Second, a model for adsorbed-gas surface diffusion in nanopores was established, which was modeled after the Hwang and Kammermeyer (1966) model and considered the effect of gas coverage under a high-pressure condition. Finally, with the combination of these two models, a unified model for gas transport in nanopores of SGRs was established, and this model was validated through molecular simulation and experimental data. Results show that:
Production rapid decline is the major problem for the tight sandstone reservoirs in Jilin oilfield. For the particular reservoir investigated in this study, production is not only subjected to the reservoir properties, but also the well completion designs especially fracturing. A comprehensive study has been conducted for multi-stage fractured horizontal wells. New fracturing improvement strategies are presented in this paper for future operations in the studied field and also those who may have similar tight sandstone reservoirs to share.
Through the integrated studies of the petrophysical characteristics, geomechanical properties and fracturing data from the fractured wells of the tight oil reservoirs in Jilin Field, numerous fracturing modeling scenarios were compared with actual fracturing monitoring data. A fully three dimension finite element simulation, associated with the analytical result from earlier production data, and the theory of interaction between fracture clusters, were built in this study. We conducted the inversing design parameters from the multi-stage hydraulic fracture with some monitoring data to improve the understanding of the reservoir properties. Additionally, a calibrated geomechanical stress model for a completed well in this field was built. At the end, the production model was presented. Data was provided to facilitate later comparison with the actual multi-stage hydraulic fracture production and valuable lessons have learned through those iteration studies.
With thoroughly trained and well calibrated model, a new fracturing strategy has been developed for the studied tight oil field. The best NPV can be achieved with the optimal fracture conductivity, fracture geometry and well performance. But first of all, the most valuable lesson we learned is that, the Effective Propped Volume (EPV) is the dominating factor for the fractured well performance, instead of the so-called Stimulated Reservoir Volume (SRV). SRV is a misinterpreted concept yet un-calculable. By adopting a numerical simulator and a proficient technology, we developed the most suitable design (perforation, fracture spacing etc.) and the fluid system (slick water, linear gel etc.) for this reservoir so that the optimal fracture geometry and fracture conductivity can be achieved. Besides that, the fracture geometry and proppant distribution were simulated. The simulated oil production data from the finite element fracture and production software is highly matched with the recorded oil production data.
An adaptability evaluation was conducted along with this study. To ensure the relevance and the authenticity of design, we analyzed the effective factors of treating material from both the laboratory and the field data in this field. A novel fracturing fluid system was applied. The fluids are more effective and leave less damage to the formation.
Hydrocarbon-bearing ultra-tight formations generally exhibit heterogeneous, anisotropic, and pressure-dependent petrophysical properties. Consequently, various laboratory measurements on separate core plugs and crushed rock samples from tight formations tend to generate inconsistent petrophysical estimates. These inconsistencies are further escalated by the existence of varied pressure- and pore-size-dependent fluid flow mechanisms in the nanopores of ultra-tight formations. We circumvent such discrepancies in petrophysical estimates by simultaneously estimating six petrophysical parameters from laboratory-based pressure-step-decay measurement on a single ultra-tight rock sample. The proposed method involves nitrogen gas injection into an ultra-tight rock sample at multiple stepwise pressure increments, high-resolution pressure-decay measurement at the outlet, followed by a deterministic inversion of the measured downstream pressure data based on numerical finite-difference modeling of nitrogen gas flow in the ultra-tight rock sample.
This work is performed with an aim to improve the petrophysical estimates previously obtained from pressure-step-decay measurements using only a Klinkenberg-type gas slippage model. We implement a transitional transport model that can handle both slip and diffusion. The proposed method was applied to nine 2-cm-long, 2.5-cm-diameter core plugs extracted from a 1-ft3 ultra-tight pyrophyllite block. We estimated the intrinsic permeability, effective porosity, pore-volume compressibility, pore throat diameter, and two slippage-Knudsen diffusion weight factors parameters. Accuracy of the estimates depends on the physical models incorporated in the forward model and on the error minimization algorithm implemented in the inversion scheme. The estimation results are independent of initial guess of intrinsic permeability, effective porosity pore-volume compressibility, and pore throat diameter in the range of 3 nd to 300 nd, 0.01 to 0.15, 10-2 to 10-6 psi-1, and 60 nm to 500 nm, respectively. The average estimated values of intrinsic permeability, effective porosity, pore-volume compressibility, and pore throat diameter of the nine ultra-tight samples are 86 nd, 0.036, 2.6E-03 psi-1, and 195 nm, respectively. Notably, the two inverted slippage-Knudsen diffusion weight factors indicate that the gas transport mechanism in the nine ultra-tight pyrophyllite samples is completely dominated by slip flow without any Knudsen diffusion or transitional flow even though the Knudsen numbers across the samples during the entire duration of the pressure-step-decay measurements are in the range of 0.01 to 1.
Reservoir depletion results in changes in effective stresses, which may lead to significant changes in reservoir permeability. These changes are associated with matrix compaction, fracture closure and potential slip. A depletion-induced increase in effective stresses often leads to a decrease in permeability. However, the opposite is observed to happen in some fractured gas reservoirs with an organic rock matrix that exhibits strong sorption-mechanical coupling. During depletion, an adsorbed portion of the gas desorbs from micropores resulting in shrinkage of the organic components in the rock matrix, effective stress relaxation and a potential increase in fracture permeability. The objective of this study is to develop a reservoir simulator with a full mechanical coupling accounting for sorption-induced change of stresses. This paper aims to estimate the influence of the parameters affecting reservoir permeability and to predict its evolution during reservoir depletion. We compare two natural gas fields with strong (San Juan coal basin) and weak (Barnett shale formation) sorption-mechanical coupling. The results of the study highlight the interplay between mechanical moduli, swelling isotherm parameters, and fracture compressibility in determining the impact of desorption on fracture permeability evolution during depletion.
Natural gas consumption currently constitutes a fifth of the total energy sources . About a half of nonassociated gas accrues to non-conventional gas reservoirs, mainly organic shales and coal seams . Non-conventional tight reservoirs have an extremely low permeability, a fair portion of which pertains to fractures as main fluid conduits. The openings of these fractures are dictated by lithology and the reservoir stresses, which may alter during reservoir development [2-5]. Two competitive geomechanical processes are known to affect stresses during depletion in organic-rich rocks: pressure drawdown and desorption-induced shrinkage. The latter is of significant importance in coals because sorbed gas constitutes more than 50% of total gas in place and desorption induces a substantial amount of rock shrinkage [6-8]. Sorbed gas in hydrocarbon-bearing shales constitutes 5-15% of the total gas in place. Sorption capacity is usually proportional to total organic carbon (TOC) in shales . Decreases in pore pressure associated with reservoir depletion cause increases in effective stresses, which often leads to fracture closure and a decrease in permeability. In contrast, desorption and matrix shrinkage result in a drop in effective stresses and an increase in permeability [8, 10, 11].
In 1898, Kirsch published equations describing the elastic stresses around a circular hole that are still used today in wellbore pressure breakdown calculations. These equations are standard instruments used in multiple areas of petroleum engineering, however, the original equations were developed strictly for vertical well settings. In today's common directional or horizontal well situations, the equations need adjusted for both deviation from the vertical plane and orientation to the maximum and minimum horizontal in-situ stress anisotropy. This paper provides the mathematical development of these modified breakdown equations, along with examples of the implications in varying strike-slip and pore pressure settings. These examples show conditions where it is not unusual for breakdown pressure gradients to exceed 1.0 psi/ft and describes why certain stages in “porpoising” horizontal wells experience extreme breakdown issues during hydraulic fracturing treatments. The paper also discusses how, in most directional situations, the wellbore will almost always fail initially in a longitudinal direction at the borehole wall, after which the far-field stresses will take over and transverse components can be developed. Tortuosity and near wellbore friction pressure can actually add to forcing the initiation of such longitudinal fractures, which can then have cascading effects on other growth parameters such as cluster-to-cluster and stage-to-stage stress shadowing. Special considerations for highly laminated anisotropic formations, where shear failure of the wellbore may precede or preclude tensile failure, are also introduced. Such failure behaviors have significant implications on near wellbore conductivity requirements and can also greatly impact well production and recovery efforts.
In vertical wells it is relatively easy to accept that a tensile fracture will initiate along the axis of the well and propagate in the direction of maximum horizontal stress, opening perpendicular to the minimum stress. As the industry has increasingly moved to horizontal well completion and stimulation, the paradigm that fractures will orient in the plane of the maximum and intermediate stress of the earth tensor, and continue to open against the minimum horizontal stress, has been retained. Fieldscale development plans are based on the expectation of nearly parallel vertical fractures that are orthogonal to the planned borehole (transverse), or set at a predictable angle to the well axis. Azimuth or strike of the fracture is expected to be given by the azimuth of maximum horizontal stress. When the horizontal stress anisotropy is low, a more complex system of possibly orthogonal fractures, or some complex joint network, is expected.
Over the past decade, multiple-fracture horizontal wells (MFHWs) have proved successful in recovering oil and particularly gas from very-low-permeability reservoirs. For MFHW planning and design, it is important to be able to make decisions on several variables, such as number of wells, number of fractures per well, amount and type of proppant per fracture, fracture dimensions (length, width, and height), and possibly others. Standard practice for making such decisions is mostly empirical. However, empiricism may not be as successful for new cases that are significantly different from old ones. In such cases, what-if analysis is used by combining intuition with numerical simulators to assess the outcomes of possible decisions. Even though this approach may produce feasible results, it is unlikely that it leads to optimal decisions because of the complexity of the problem. Therefore, a systematic methodology is needed for MFHW planning and design that produces optimal solutions efficiently and effectively.
In this work, we develop such a methodology that optimizes net present value (NPV). The methodology poses the design problem as nested optimization in which the outer-optimization shell involves decisions on the number of wells, number of fractures per well, and amount of proppant, whereas the inner optimization maximizes the productivity index (PI) by selecting optimal fracture dimensions. The proposed methodology integrates in a systematic way a fracture-design module, a production-estimation module, and an economics module. Because the methodology is automated and numerically efficient, one can use it effectively for field development.
A case study is used to illustrate the applicability of the proposed methodology, and suggestions are made for further improvements.