首页期刊导航|Journal of Petroleum Science & Engineering
期刊信息/Journal information
Journal of Petroleum Science & Engineering
Elsevier Science B.V.
Journal of Petroleum Science & Engineering

Elsevier Science B.V.

0920-4105

Journal of Petroleum Science & Engineering/Journal Journal of Petroleum Science & Engineering
正式出版
收录年代

    A comparison of machine learning surrogate models for net present value prediction from well placement binary data

    Joao Roberto Bertini JuniorSergio Ferreira Batista FilhoMei Abe Funcia
    11页
    查看更多>>摘要：Net Present Value (NPV) is an important indicator to guide investment decisions. In oil production planning, NPV is employed to evaluate and select among different production strategies. However, NPV estimation requires computational costly numerical simulations. So, evaluating as many production strategies as is desirable may be prohibitive. Therefore, one can only evaluate a small part of the search space, decreasing the chance of finding a near-optimal production strategy. To speed up the searching process, a much faster, but error-prone, surrogate model is used to approximate the simulator output. Data-driven surrogate modeling depends on both;; 1) building a simple model to reproduce the quality of a high-fidelity model, while 2) considering a large volume of data to build it. In this work, we address the well placement optimization task by considering a binary data representation, indicating the presence or absence of a given well in a production strategy. We show the possibility of predicting the NPV from binary data, thus reducing data dimension and model complexity. Specifically, we compare six machine learning regression algorithms to predict the NPV. The simulations conducted in a benchmark case, based on a real field, showed that some regression algorithms can be used as a surrogate model to the simulator to efficiently perform well placement optimization considering binary data. The best results were obtained with Multi-Layer Percegtron, whose estimations covered a wide range of NPV with a small and constant error.
      原文链接:
    • NSTL

    The flow of water-in-oil emulsion in heterogeneous parallel model

    Jierui LiWeidong LiuSunan Cong
    10页
    查看更多>>摘要：Emulsification has the potential to profoundly change enhanced oil recovery (EOR) and to improve the mechanism of recovery, and the emulsification performance of the chemical is an important basis for the selection of the injection agent for the surfactant-polymer flooding. The objective of this research is the study of the transportation of the oil-rich zone of surfactant-polymer flooding, its ability to adjust liquid production profile in heterogeneous parallel conglomerate cores at different permeability. A stable water-in-oil emulsion was screened by ultrasonic emulsification experiment with the condition of oil-water ratio at 7:3, the ultrasonic power at 900 W. When the emulsion flowed through the conglomerate core, the blockage of pores by emulsion droplets included the blockage by single droplet and multiple droplets accumulation. A one-dimensional flow characterization model of the water-in-oil emulsion flow was established based on the generalized Darcy's law and corroborated with the single-core emulsion flow experiment. The water-in-oil emulsion had a conformance control when flowed in heterogeneous parallel conglomerate cores. When the pore size of the hyperpermeable layer of the heterogeneous layer is much larger than the size of the emulsion, the fluidity of the emulsion in the hyperpermeable layer would mainly be adjusted with high viscosity, so its adjustment ability is poor. When considering the influence of emulsification on the enhanced oil recovery, it is necessary to fully consider the compatibility of emulsion properties and reservoirs.
      原文链接:
    • NSTL

    Nonlinear viscoelastic rheological behavior of bentonite and sepiolite drilling fluids under large amplitude oscillatory shear

    Ali EttehadiCeylan UlkerGursat Altun
    21页
    查看更多>>摘要：Common models used to describe rheological behavior of drilling fluids are built on pure viscous flow assumption. Full rheological characterization of drilling fluid under this assumption is not feasible since drilling fluids are viscoelastic materials and exhibit both elastic and viscous features. Gel strength development, yield stress value at near zero shear rates, and sag tendency in drilling fluids are the strong function of viscoelastic responses. Linear and nonlinear viscoelastic responses should be measured to provide complete viscoelastic analysis. Two water-based drilling fluids, sepiolite and bentonite drilling muds, each in four states, were subjected to testing using rheometer. Nonlinear viscoelastic parameters and their indications for both fluid systems were characterized for the first time. Large amplitude oscillation sweep tests were conducted as a function of strain and strain rate at four temperatures and frequencies. Stress response was decomposed through Fourier transform into elastic and viscous stress components. Lissajous-Bowditch loops were assessed as rheological fingerprints to detect the initiation of nonlinear region and the nature of nonlinearity. Results revealed that bentonite fluid systems provided stronger gel strength, a more stable network, and high mechanical stability compared to sepiolite fluid samples up to 50 °C. However, the sepiolite fluid systems outperformed at high temperatures (100 and 150 °C). The elastic nonlinear response was determined to be strain/strain rate softening, and the nonlinear viscous response was shear rate thinning at large strain rates for both fluid systems. Contrary to common knowledge found in literature (shear rate thinning), it was revealed that both fluid systems demonstrated shear rate thickening behavior at low to moderate strain rates. This finding is of importance in understanding gel strength development and evaluating sag tendency in drilling operations.
      原文链接:
    • NSTL

    A review of experimental studies on the proppant settling in hydraulic fractures

    Ke HaiChunli ChangShanshan Yao
    24页
    查看更多>>摘要：The hike of hydraulic fracturing in North America and beyond leads to significantly enhanced hydrocarbon production especially in low-permeability geological formations. Proppant settling is one of the most important particle motions during a hydraulic fracturing process, which to a large extent determines the created fracture's conductivity. This paper provides a critical review of the proppant settling in hydraulic fractures, with an emphasis on the experimental studies published in the past several decades. Six factors are identified out of a proppant/fracturing-fluid/fracture system;; wall retardation, fluid rheology, proppant non-uniformity, proppant surface wettability, proppant concentration, and fracture complexity. Influences of these factors on the proppant settling are summarized based on the analysis of published experimental data, images, and correlations. Notwithstanding the extensive experimental studies, more experiments are in demand to systematically investigate proppant settling behavior in various kinds of fracturing fluids confined by hydraulic fractures. In addition, a gap is found to exist between the experimental studies of proppant transport and the application of the experimental findings to numerical simulations of hydraulic fracturing.
      原文链接:
    • NSTL

    Three-dimensional physical simulation of water huff-n-puffin a tight oil reservoir with stimulated reservoir volume

    Ke SunHuiqing LiuJing Wang
    16页
    查看更多>>摘要：Water huff-n-puff has been considered as a cost-effective method to enhance oil recovery in tight reservoirs. However, the physical simulation method for water huff-n-puff is still lacking up to now, so that the production behaviors of actual tight reservoirs cannot be reproduced and the EOR potentials cannot be evaluated directly by scaled models in the laboratory. In this study, mathematical models for water huff-n-puff processes in tight oil reservoirs with stimulated reservoir volume (SRV) were established, then the corresponding scaling criteria were proposed and verified. Taking the tight reservoir ZMY in the Ordos Basin as a prototype, three groups of three-dimensional physical simulations of water huff-n-puff under different conditions (PSM~(-1), PSM~(-2) and PSM-3) were designed based on the scaled parameters respectively, then performed favorably and compared comprehensively. The experimental results show that oil recovery of PSM~(-1) can be increased by 20.16% within 8 cycles, which shows great EOR potentials for the tight reservoir ZMY. Compared with PSM~(-1), oil recovery of PSM~(-2) reduces by 8.23%, which indicates that forced imbibition is quite a helpful and promoting mechanism, and oil recovery enhancement will be greatly limited if no imbibition effect is given play between the matrix and fractures. Comparatively, oil recovery reduces by 3.39% for PSM-3, which can be inferred that if fracture density relatively increases and all fractures are well connected, the matrix imbibition effect can be further strengthened, the high pressure can propagate faster, the elastic energy can supplement into the matrix deeper, and the effect of gravitational differentiation can be better utilized as well. It is found that cumulative oil production of the three physical simulations in the first several cycles is always relatively high, which implies that sufficient oil supply in physical models can be maintained. As oil-water interface rises, less oil and more water will be produced and different dynamic characteristics of cumulative oil and water production are shown. It is also found that all three groups of water cut curves have the same upward trend and the Sigmoidal-Logistic model can be employed to accurately depict the dynamic characteristics of water cut at different cycles. This study aims to establish a physical simulation method for water huff-n-puffin tight oil reservoirs with SRV based on the three-dimensional scaled model, which will be greatly helpful to understand the EOR mechanisms and production characteristics of water huff-n-puff
      原文链接:
    • NSTL

    Production decline analysis for a fractured vertical well with reorientated fractures in an anisotropic formation with an arbitrary shape using the boundary element method

    Liwu JiangJinju LiuTongjing Liu
    16页
    查看更多>>摘要：In this work, theoretical models have been formulated, validated, and applied to evaluate the production decline performance of a fractured vertical well with reorientated fractures in an anisotropic formation with an arbitrary shape by using the boundary element method. More specifically, the coordinate transformation method was applied to convert the anisotropic system to an equivalent isotropic system, while a coupled matrix-fracture flow model is proposed in the Laplace domain. Then, the boundary element method is applied to solve the fluid flow problem in the matrix, while the Laplace-transform finite difference method is used to numerically obtain the pressure solutions for each fracture segment. Furthermore, the effect of orientated fractures can be examined by generating the Agarwal-Gardner type decline curves, during which two flow regimes including transient flow regime and boundary dominant flow regime can be found. The orientated fractures only affect the transient flow regimes and the boundary dominant flow can be affected by the area and shape of the reservoir. The fracture conductivity is found to be the most important parameter dominating the early production rate. The larger the anisotropic factor, the larger production rate during the early times. Also, the primary fracture angle is more sensitive than the orientated fracture angle. Subsequently, this newly proposed method has been validated and then extended to a field application, demonstrating that the production decline curves for a fractured well with reorientated fractures can be analyzed in a reasonable and accurate manner.
      原文链接:
    • NSTL

    Analytical and numerical study of thermal and solvent-based gravity drainage for heavy oil recovery

    Min YangZhangxin ChenMaojie Chai
    15页
    查看更多>>摘要：With the increasing demand for energy, an initial screening for heavy oil recovery options is important. However, the calculation of the gravity drainage recovery of heavy oil is complex and involves coupled fields and changes in the properties of heavy oil or its mixtures. There is currently no general analytical treatment for heavy oil gravity drainage processes, such as steam-assisted gravity drainage (SAGD), vapor extraction (VAPEX) and warm VAPEX, with the simplest possible assumptions but with the necessary mechanisms. In this study, a novel discrete hypothetical model has been proposed to simplify the calculation of gravity drainage flow in porous media. Then, a validated equation with a reasonable mixing rule is used to describe the property changes of heavy oil or its mixtures under decoupled temperature and concentration variables. Finally, internal velocity distributions for different gravity drainage processes are validated with numerical modeling results. Our results agree with previous work on the importance of mass transfer coefficients. Moreover, different from the previous work, the effect of slightly water-soluble solvents (dimethyl ether) has been investigated on diffusion in both the oil phase and the water phase. For solvent-based heavy oil recovery, a slightly water-soluble solvent shows great advantages with promising application potentials. The results of this study can guide the initial screening for a gravity drainage process and the analysis of its key mechanisms.
      原文链接:
    • NSTL

    Circulating preheating model of full-lengi horizontal wellbore in heavy oil reservoirs with multiple thermal fluid injection

    Bin Nie
    11页
    查看更多>>摘要：Heavy oil reservoirs are rich in resources and have broad prospects for exploration and development. Wellbore heat transfer is one of the important areas of heavy oil research. However, the current research on the preheating of multi-element thermal fluid circulation in the wellbore is still insufficient. Based on thermodynamic equations and fluid dynamics equations, this paper establishes a cyclic preheating model of multi-element thermal fluid in the wellbore. The model considers the mixing characteristics of superheated steam and non-condensed gas, and is applicable to both vertical and horizontal wellbores. Results show that the temperature curve can be divided into four parts;; the temperature of the inner tube in the vertical section of the wellbore, the temperature of the inner tube in the horizontal section of the wellbore, the temperature of the annular space in the horizontal section of the wellbore, and the temperature of the annular space in the vertical section of the wellbore;; In the horizontal section of the wellbore, the multi-element thermal fluid in the inner tube releases heat energy into the annular space, and there is a large temperature difference at the heel end and a small temperature difference at the toe end.
      原文链接:
    • NSTL

    Application of statistical machine learning clustering algorithms to improve EUR predictions using decline curve analysis in shale-gas reservoirs

    Zijuan ChenWei YuJenn-Tai Liang
    18页
    查看更多>>摘要：A robust machine learning workflow using two well-established statistical clustering algorithms, K-Means Clustering (K-MC) and K-Nearest Neighbors (K-NN) was developed to improve ultimate recovery (EUR) predictions of new wells in shale-gas reservoirs using a decline curve model developed by Duong (2011). These two clustering algorithms can handle big datasets with multiple well parameters with high computational efficiency. Out of a total of 55,623 dry-gas wells from seven shale gas formations, 7631 wells that fit the criteria for Duong's decline curve model were selected in this study for further analysis. K-MC and K-NN were applied to group wells with similar well characteristics using well location, well depth, well length, and production starting year as parameters. Locations of the grouped wells show that the clusters of wells with similar well characteristics identified by both clustering methods scatter over a big area. These findings clearly demonstrate that simply grouping wells by the proximity of physical locations is not a good way of identifying wells with similar well characteristics. The results from this study also suggest that the selection of optimal clustering methods is highly dependent on the shale formation and the method used for EUR prediction. Even with the limited number of well parameters available in the datasets used in this study, the machine learning clustering algorithms managed to improve the EUR prediction accuracy by around 20%. More available well information will allow the machine learning algorithms to capture more of the well characteristics, thereby further increasing the accuracy of EUR predictions.
      原文链接:
    • NSTL

    Controls of interlayers on the development and distribution of natural fractures in lacustrine shale reservoirs;; A case study of the Da'anzhai member in the Fuling area in the eastern Sichuan Basin

    Xiaoju ZhangJianhua HeHucheng Deng
    21页
    查看更多>>摘要：Lacustrine shale reservoirs with interbedded organic-rich shales and thinly layered tight shell limestones are mainly developed in the Da'anzhai Member of the Lower Jurassic Ziliujing Formation in the Fuling area. These reservoirs generally have low porosities and permeabilities. Natural fractures provide effective spaces for these reservoirs and significantly improve fluid flow capability. However, little research has been conducted on the genesis and distribution of fractures in the Da'anzhai Shale Member. This study aims to understand the features, distribution, and factors that influence natural fractures of these shale reservoirs in the study area. Based on field outcrop observations, core descriptions, and thin section analysis, four types of natural fractures are developed in the study area;; tensile fractures, shear fractures, interlayer bedding fractures, and diagenetic shrinkage fractures. Most low-angle and horizontal natural fractures have lengths of 5 cm~(-1)0 cm, and the effective fracture density is 0.90 pieces/m. The lithology (Y), thickness of the rock layer (If), regional tectonic stress (F), and fault scale (S) have a significant impact on the development of natural fractures. Most natural fractures develop when there is a large number of laminations in thin rock layers. The large regional tectonic stress in the Late Himalayan, which had a large fracture density index, leads to the formation of natural fractures in numerical simulations. Large faults would have co-developed with more fractures than small faults due to the high energy of large faults. The relationship between the effective permeability and the four parameters (i.e., Y, H, F, and S) was established. Based on this relationship, it is concluded that His the most critical factor influencing the effective permeability of natural fractures in the study area. As a result, the middle to upper section in the 2nd subsection of the Da'anzhai Member, in which the lithofacies is thinly laminated shales with shells, can be considered to contain favourable shale gas layers.
      原文链接:
    • NSTL