Search

Unconsolidated sand reservoirs containing heavy oil play a significant role in hydrocarbon resources, characterized by high porosity and permeability alongside abundant movable fines. During production, these fines can detach and migrate with the reservoir fluids, causing pore plugging and reduced productivity. Visualizing and quantitatively evaluating the evolution of pore structure caused by fines migration under various influencing factors at the microscale is fundamental for devising effective prevention and mitigation measures. This study employs on-line NMR experiments and CFD-DEM simulations to investigate fines migration processes and their effects on physical properties and pore structure at the pore scale. Results indicate that fines migration initiates the formation of a preferential network of migration pathways. The evolution of pore structure demonstrates zonal characteristics along the flow direction, with fines plugging and residual accumulation primarily occurring in the middle/rear section of the core. As the core’s skeleton porosity decreases, fines plugging intensifies; however, at high injection velocity, new dominant flow channels may emerge, leading to a transition from a single-peak to a double-peak T2 spectrum. Below the critical velocity (0.5–1 mL/min), an increase in flow velocity exacerbates severe fines plugging. Conversely, above the critical velocity, an increase in flow velocity results in a more pronounced enhancement of permeability.

Conventional galaxy mass estimation methods suffer from model assumptions and degeneracies. Machine learning (ML), which reduces the reliance on such assumptions, can be used to determine how well present-day observations can yield predictions for the distributions of stellar and dark matter. In this work, we use a general sample of galaxies from the TNG100 simulation to investigate the ability of multibranch convolutional neural network (CNN) based ML methods to predict the central (i.e. within 1−2 effective radii) stellar and total masses, and the stellar mass-to-light ratio (M */ L). These models take galaxy images and spatially resolved mean velocity and velocity dispersion maps as inputs. Such CNN-based models can, in general, break the degeneracy between baryonic and dark matter in the sense that the model can make reliable predictions on the individual contributions of each component. For example, with r -band images and two galaxy kinematic maps as inputs, our model predicting M */ L has a prediction uncertainty of 0.04 dex. Moreover, to investigate which (global) features significantly contribute to the correct predictions of the properties above, we utilize a gradient-boosting machine. We find that galaxy luminosity dominates the prediction of all masses in the central regions, with stellar velocity dispersion coming next. We also investigate the main contributing features when predicting stellar and dark matter mass fractions (f *, f DM) and the dark matter mass M DM, and discuss the underlying astrophysics. [ABSTRACT FROM AUTHOR]

Particle migration in oil and gas reservoirs is a common phenomenon in the process of oil and gas development, and is considered to be an important reason for the damage of reservoir permeability and the reduction of oil and gas productivity. The mechanism of this phenomenon includes the desorption, migration, and precipitation of particles, which eventually clogs the throat and causes reservoir damage. Therefore, it is necessary to accurately characterize the complex mechanism of particle migration and identify the main controlling factors of particle migration, which is very important for efficient oilfield development and plugging solution. First, the reservoir types are divided into three types and the pore structure models of different types of reservoirs are established. Then, computational fluid dynamics and discrete element coupling method numerical simulation and microscopic visualization of pore throat structure model were combined, to characterize the rules of particles and migration, and analyze the main controlling factors. Finally, a typical model of particle migration and clogging is established. The results show that particle size/throat and particle concentration are the key factors affecting particle plugging, and particle migration has the least effect on the permeability of Type I reservoir and the greatest damage to Type III reservoir. According to the mechanical and hydrodynamic behavior of particles in porous media, three mechanisms and six modes of particle plugging are proposed. [ABSTRACT FROM AUTHOR]