查看更多>>摘要：Effective cuttings transport and accurate drilling hydraulics prediction remain issues of concern during drilling operations of horizontal, extended reach and multilateral wells. While several studies have adopted a two-or three-layered modelling approach to evaluate cuttings transport efficiency, they have neglected the effect of the gas-liquid fluid flow pattern within the annulus on cuttings transport. An experimental and theoretical study was carried out to evaluate the interplay between the two-phase gas-liquid flow patterns and the major drilling parameters and investigate its influence on the cuttings and fluid flow dynamics in a horizontal and inclined drilling wellbore. Several mathematical flow pattern dependent multi-layered models valid for any level of wellbore eccentricity were developed for the different cuttings transport mechanisms in the bubble, dispersed bubble, stratified and slug gas-liquid flow patterns, thereby providing a method to evaluate cuttings transport efficiency and perform wellbore hydraulics calculations for underbalanced drilling operations. Experimental results show that both fluid flow pattern and the drilling fluid flowrate are the most influential controllable parameters that affect the cuttings transport efficiency. Moreover, the hole cleaning requirements for an eccentric annulus is higher than that required for the concentric annulus of both single-phase and two-phase Newtonian or non-Newtonian fluids. Inclination angle was also found to influence hole cleaning and the degree of its effect is highly dependent on the fluid properties, the cutting transport mechanism and prevailing gas-liquid flow pattern. In the horizontal and inclined eccentric annuli, drillpipe rotation can improve cuttings transport for both single-phase and two-phase flows, but generally the effect of the drillpipe rotation on two-phase flow for cutting transport is much less than that of the single-phase flow. Overall, a good match was found between the mathematical flow pattern dependent multi-layered models and the experimental data. The findings of this study serve as a guide in the prediction of the wellbore dynamics for underbalanced drilling operations and provides a tool that can be applied for wellbore pressure management and the evaluation of hole cleaning based upon the specified flow conditions.
Rasoul MokhtariBenaiah U. AnabaraonyeArmin Afrough
14页
查看更多>>摘要：Chalk reservoirs, due to their high storage capacity and very low permeability, are one of the most interesting cases for reservoir engineering research on carbonates. They exhibit complex fluid-rock interactions because of their chemically active porous media. This study investigates the effect of brine composition, injection scenario, and temperature on oil recovery by low salinity water-flooding in chalk core samples from a Danish North Sea reservoir. The mechanisms governing oil-brine-rock interactions were also explored. Actual reservoir chalk core samples, without any open fractures, were selected using computed tomography analyses. These cores were saturated with representative fluids (crude oil and synthetic formation brine) and aged at reservoir conditions for approximately three weeks. The role of brine chemistry has been investigated through effluent analysis by ion chromatographyt and results indicate that low salinity diluted seawater promotes rock-surface reactions if left to incubate for at least 48 h. Rock dissolution, observed through the monitoring of effluent ions, increased both with increase in temperature and decrease in brine salinity. The recovery curves show that formation water and diluted seawater produce significantly more oil (of the order of 10 % more) at the secondary stage compared to seawater. Additionally, there is also some indication of an effect of low salinity brine at the tertiary stage. These experiments were performed on reservoir materials and corresponding crude oil samples and provide new data on the low salinity flooding potential for chalk, and provide further evidence for the applicability of the low salinity effect in carbonates.
查看更多>>摘要：The in-situ formation of foamy oil has been found to be a crucial mechanism accounting for the better-than-expected production performance in heavy oil reservoirs under solution gas drive. Not only have the physical laws dominating gas exsolution in foamy oil not yet been well understood, but also the different contributions of each component in a gas mixture to the generation of foamy oil associated with an extremely complicated dynamic process has not been quantified. In this study, a novel and pragmatic technique has been proposed and validated to quantify the gas exsolution and preferential diffusion for alkane solvent(s)-CO2-heavy oil systems under nonequilibrium conditions by taking gas bubble size distribution and preferential mass transfer of each gas component into account. Experimentally, constant-composition expansion (CCE) tests with various constant-pressure decline rates are utilized to describe gas exsolution behaviour of alkane solvent(s)-CO2-heavy oil systems under nonequilibrium conditions, during which not only pressure and volume are simultaneously monitored and measured, but also gas samples were respectively collected at the beginning and end of experiments to perform compositional analysis. Theoretically, a mathematical model has been formulated to quantify gas exsolution and preferential mass transfer between each gas component and the liquid phase in alkane solvent(s)-CO2-heavy oil systems under nonequilibrium conditions. More specifically, quasi-equilibrium boundary conditions, real gas equation, and Rayleigh distribution function are combined with the classical equation of motion, continuity equation, and mass transfer equation to form a novel equation matrix for quantifying gas bubble growth in foamy oil. Considering gas bubble size distribution and preferential diffusion of each component in a gas mixture, the total number of gas bubbles and individual diffusion coefficient of each gas component are determined by minimizing the discrepancy between the measured and calculated volume of alkane solvent(s)-CO2-heavy oil systems. More importantly, the dynamic composition of the gas phase and the amounts of both entrained gas and evolved gas can also be obtained simultaneously during the gas exsolution processes. Excellent agreements between the experimentally measured parameters (i.e., volume of foamy oil, composition of evolved gas, and volume of free gas) and the calculated ones have been respectively achieved. Compared with the individual diffusion coefficient for each component in a gas mixture determined under the traditional equilibrium conditions, a relatively large value has been found during mass transfer processes in a supersaturated oleic phase. Moreover, pseudo-bubblepoint pressure and gas exsolution rate are found to be two main mechanisms dominating the volume-growth rate of the evolved gas.
查看更多>>摘要：History matching can play a key role in improving geological characterization and reducing the uncertainty of reservoir model predictions. Application of reservoir history matching is restricted by the huge computational cost by amongst others the many runs of the full model. Surrogate models with a reduced complexity are therefore used to reduce the computational demands. This paper presents an efficient surrogate-assisted deterministic inversion framework to primarily explore the possibility of applying deep neural network (DNN) surrogate to approximate the gradient of large-scale history matching by using auto-differentiation (AD). In combination with the deep neural network model, the AD enables us to evaluate the gradients efficiently in a parallel manner. Furthermore, the benefits of using stochastic gradient optimizers in the deep learning practice, instead of foil gradient optimizers in conventional deterministic inversions, is investigated as well. Numerical experiments are conducted on a 3D benchmark reservoir model in the context of a water-flooding production scenario. The quantity of interest, e.g., dynamic saturation for an ensemble of test models, can be accurately predicted. The proposed surrogate-assisted inversion with stochastic gradient optimizer obtains a very quick convergence rate against the model and data noise for the high-dimensional history matching problem with a large number of data and parameters. In addition, we also conduct several comparisons and evaluations with our previously proposed projection-based subdomain POD-TPWL approach in terms of computational efficiency and accuracy. The subdomain POD-TPWL constructs a local surrogate model, which is repeatedly reconstructed a number of times for maintaining a satisfactory accuracy, while DNN constructs a global surrogate model based on the entire training data and generally does not require additional reconstructions. The subdomain POD-TPWL is very sensitive to how the domain is decomposed, increasing the training samples does not infinitely improve the history matching results by a fixed decomposition. Overall, these two kinds of surrogate models have demonstrated great potential in solving large-scale history matching problem. The DNN surrogate is particularly useful to generate multiple posteriors for model uncertainty quantification.
查看更多>>摘要：Amplitude versus offset (AVO) inversion is the process of transforming seismic reflection into elastic properties such as P-and S-impedance to estimate the interval properties and thickness of underlying geology using well log and post-or prestack seismic data. Recent applications of AVO inversion based on deep learning have shown excellent results and practical applicability. However, traditional deep learning methods yield only prediction results without any associated predictive uncertainty. Two types of predictive uncertainty should be considered;; aleatoric uncertainty, which occurs when noisy data are included;; and epistemic uncertainty, which is caused by a lack of data. To estimate the impedances and their uncertainties, Bayesian approximation using Monte Carlo dropout is applied, which simply approximates a Bayesian neural network. From the proposed method, we can not only predict impedances but also estimate their predictive uncertainties in the seismic survey area and determine whether prediction results are reliable.
查看更多>>摘要：Deep carbonate reservoirs with great resource potential are key areas for future hydrocarbon exploration with the continuous innovation of deep hydrocarbon accumulation theory and technology. Breakthrough pressure (BP) is the most important parameter used to indicate rock sealing. In this study, X-ray diffraction, CO2 adsorption, N2 adsorption, high-pressure mercury injection, thin section analysis, effective porosity-permeability and BP experiments are conducted to research pore-throat characteristic of carbonate rock and the influencing factor of BP. Pore volume increases with the specific surface area (SSA) of micropore, and has a weak negative and positive correlation with the SSA of mesopore and macropore, respectively. The BP increases with an increase in the micropore proportion and decreases with an increase in the macropore proportion. Compared with R50 (the pore-throat radii corresponding to these mercury saturation of 50%), R35 (the pore-throat radii corresponding to these mercury saturation of 35%) has a better correlation with porosity and permeability and can better characterize the influence of pore throats on the sealing performance of carbonate rock. A low clay content perhaps leads to an inapparent interrelation between clay content and BP. Because of high brittleness of quartz, an increase in quartz content will cause an increase of porosity and permeability, and a decrease in BP. The BP decreases with the increase in effective porosity and permeability, and linear correlations exist among the reciprocal of porosity (A), the-0.407 power of permeability (B), and the BP. The BP calculation formula is fitted based on A and B by the principal component regression (PCR) method of SAS software for carbonate rock in the Ordovician of the Gucheng area.
查看更多>>摘要：The Zaysan Basin in northeastern Kazakhstan is a petroliferous basin that closely correlates with the Central Asian Orogenic Belt (CAOB). The basin is characterized by multi-stage hydrocarbon migration and accumulation, while the details are poorly understood. Hydrocarbon accumulation processes were investigated within the constraints of the regional tectonic evolution and the timing of trap formation. Along with assessment of geochemical characteristics, hydrocarbon generation and accumulation models were established. Light blue-green fluorescence colours of intergranular oil were identified for the deep Permian and Jurassic reservoirs, but they were absent within the shallow Paleogene Severozagsan reservoirs. Geochemical results indicate that source rocks for the oils were deposited in anoxic brackish-saline environments, and mainly contain type I and type II kerogen. Oil and extracted bitumen from the Permian and Jurassic reservoirs originate from the same Permian source rocks and show low maturity. The biodegradation of oils within the Permian and Jurassic reservoirs, however, indicates different alteration processes. Paleogene gas is the biodegradation product of the oils. In the study area, two stages of oil charge and one stage of gas charge were identified. The first oil charge began about 207 Ma (Late Triassic) when the source rock entered the oil window. The oil migrated into the Permian reservoirs and subsequently underwent biodegradation. The second oil charge took place about 70~(-1)00 Ma and was controlled by Late Cretaceous cooling and exhumation events. Oil in the Jurassic reservoir originated by oil immigration from the Permian reservoir. The gas charge occurred during the Late Cenozoic. The gas was mainly the biodegradation of oils and accumulated in Paleogene reservoirs.
查看更多>>摘要：The article at hand focuses on the pressure drop at the hydraulic oscillator valve section. The flow path of the valve section is simplified through the hydraulic oscillator working principle and the valve structure. Further, based on the kinematic model of the positive displacement motor (PDM) and the moving valve, the flow area calculation model between valve sections was deduced. Next, using the simplified valve section flow path, said path is divided into four stages, and the analytical model of pressure drop was established by applying the fluid mechanic theory. Aiming to verify the analytical model, the transient finite element model, and the test bench were set up. Finally, three methods were used to calculate the pressure drop within the 1.3-inch and 1.4-inch stationary valves corresponding to the same input parameters. The results have shown that pressure drops obtained by all the methods are in good agreement;; the maximum error between the analytical and experimental results was 7.6 %. Additional comparisons were made between the analytical and finite element model (FEM) results at various flow rates. In that case, the maximum error between the two methods was 7.05 %. Thus, the model can be used to further study the hydraulic oscillator mechanisms and provide a useful reference for both the design and optimization.
查看更多>>摘要：The recognition of fluid-filled channels in cased holes is significant for the safety production of oil and gas wells. The fluid-filled channel at the cement-formation boundary is more difficult to be detected than that at the casing-cement boundary. To solve the key technical problem, a dual-frequency acoustic measurement method is proposed to obtain the cementation state of the two interfaces at the same time. The related theory is that the casing wave amplitude from the low-frequency source is sensitive to the channel at both two boundaries, but that from the high-frequency source is only sensitive to the channel at the casing-cement boundary. Therefore, high amplitudes of the casing waves from both two sources represent a channel on the casing-cement boundary, high amplitude from the low-frequency source but low amplitude from the high-frequency source represent a channel on the cement-formation boundary, and low amplitudes from both two sources represent no channel. Subsequently, numerical simulations are performed to examine the effects of several critical parameters on the dual-frequency acoustic measurement. The results show that the best combination of the measurement system is one low-frequency source at 10 kHz together with one high-frequency source at 150 kHz. For the low-frequency source, the casing wave amplitude increases obviously with the increasing channel thickness, regardless of the channel position. The increase of the cement thickness and the casing size damps out the casing wave. For the high-frequency source, the casing wave amplitude is little affected by these parameters except for the casing-cement channel. Moreover, the dual-frequency acoustic measurement method is suitable for common sandstone and shale formations. The findings of this study provide a theoretical foundation for next-generation cementing-quality-evaluation tool development.
查看更多>>摘要：The gas/fluid flowing behavior and fluid distribution under capillary forces in porous media are essential for evaluating the pore-fracture characteristics of coal. Capillary force is crucial for flow-back efficiency after hydraulic fracturing during enhanced coalbed methane (CBM) production. In this work, in situ dynamic X-ray micro-computed tomography (X-ray |i-CT), field emission scanning electron microscopy (FE-SEM) combined with mercury intrusion porosimetry (MIP) were used on two coal samples of different wettability to explore the behavior of spontaneous imbibition. The results show that both pore-fracture and mineral content could affect the imbibition behavior of coal, which the former has played a major role. The spontaneous imbibition process can be divided into 4 stages to reveal fluid flow characteristics by the fluid distribution at different time steps. The gas relative permeability of coal attenuates exponentially with imbibition time result for capillary force based on fractal theory (the gas relative permeability reduces its value from 0.72 mD to 0.01 mD for WD, 0.5 mD to 0.08 mD for WZX). A low flow-back rate of fracturing fluid would lead to severe permeability damage. The longer the return drainage time, the more severe the permeability damage.
NSTL