Permeability Prediction of Nanoscale Porous Materials Using Discrete Cosine Transform-Based Artificial Neural Networks.

The permeability of porous materials determines the fluid flow rate and aids in the prediction of their mechanical properties. This study developed a novel approach that combines the discrete cosine transform (DCT) and artificial neural networks (ANN) for permeability analysis and prediction in digital rock images, focusing on nanoscale porous materials in shale formations. The DCT effectively captured the morphology and spatial distribution of material structure at the nanoscale and enhanced the computational efficiency, which was crucial for handling the complexity and high dimensionality of the digital rock images. The ANN model, trained using the Levenberg-Marquardt algorithm, preserved essential features and demonstrated exceptional accuracy for permeability prediction from the DCT-processed rock images. Our approach offers versatility and efficiency in handling diverse rock samples, from nanoscale shale to microscale sandstone. This work contributes to the comprehension and exploitation of unconventional resources, especially those preserved in nanoscale pore structures.

An Analytical Model for Hysteretic Pressure-Sensitive Permeability of Nanoporous Media.

Hysteretic pressure-sensitive permeability of nanohybrids composed of substantial nanopores is critical to characterizing fluid flow through nanoporous media. Due to the nanoscale effect (gas slippage), complex and heterogeneous pore structures of nanoporous media, the essential controls on permeability hysteresis of nanohybrids are not determined. In this study, a hysteretic pressure sensitive permeability model for nitrogen flow through dry nanoporous media is proposed. The derived model takes into account the nanoscale effect and pore deformation due to effective stress. The model is validated by comparing it with the experimental data. The results show that the calculated permeability and porosity are consistent with the measured results with the maximum relative error of 6.08% and 0.5%, respectively. Moreover, the hysteretic pressure-sensitive permeability of nanohybrids is related to effective stress, gas slippage, pore microstructure parameters, grain quadrilateral angle, and the loss rate of grain quadrilateral angle. The nanoscale effect is crucial to the permeability of nanoporous media. In addition, as impacted by the comprehensive impact of multiple relevant influential parameters, permeability during the pressure unloading process is not a monotonous function but presents complicated shapes. The proposed model can explain, quantify, and predict the permeability hysteresis effect of nanoporous media reasonably well.

Hyperparameter Tuning of Artificial Neural Networks for Well Production Estimation Considering the Uncertainty in Initialized Parameters.

A well production rate is an essential parameter in oil and gas field development. Traditional models have limitations for the well production rate estimation, e.g., numerical simulations are computation-expensive, and empirical models are based on oversimplified assumptions. An artificial neural network (ANN) is an artificial intelligence method commonly used in regression problems. This work aims to apply an ANN model to estimate the oil production rate (OPR), water oil ratio (WOR), and gas oil ratio (GOR). Specifically, data analysis was first performed to select the appropriate well operation parameters for OPR, WOR, and GOR. Different ANN hyperparameters (network, training function, and transfer function) were then evaluated to determine the optimal ANN setting. Transfer function groups were further analyzed to determine the best combination of transfer functions in the hidden layers. In addition, this study adopted the relative root mean square error with the statistical parameters from a stochastic point of view to select the optimal transfer functions. The optimal ANN model's average relative root mean square error reached 6.8% for OPR, 18.0% for WOR, and 1.98% for GOR, which indicated the effectiveness of the optimized ANN model for well production estimation. Furthermore, comparison with the empirical model and the inputs effect through a Monte Carlo simulation illustrated the strength and limitation of the ANN model.

A Novel Analytical Model for Pore Volume Compressibility of Fractal Porous Media.

Over the past decades, many scholars have been studying the pore volume compressibility (PVC) of porous media. However, the fundamental controls on PVC of porous media are not yet definitive. Some scholars suggest a negative correlation between PVC and initial porosity, while others suggest a positive correlation. Motivated by this discrepancy, this paper presents a new analytical model to study the PVC of fractal porous media. The predictions are compared with test results and thereby validated to be accurate. In our attempt not only to complement but also to extend the capability beyond available models, the derived model accounts for multiple fundamental variables, such as the microstructural parameters and rock lithology of porous media. Results suggest that, there is a negative correlation between PVC and initial porosity, if all other parameters are fixed, the relationship between initial porosity and PVC is not monotonic. In addition, PVC decreases with rougher pore surfaces and smaller initial minimum pore radius. Besides providing theoretical foundations for quantifying PVC of porous media, this analytical model could be applied to estimate pore structure parameters of porous media using inverse modeling.

Physically-based urban stormwater quality modelling: An efficient approach for calibration and sensitivity analysis.

Physically-based urban stormwater quality modelling is helpful for increasing the understanding of spatial-temporal dynamics of urban pollution, and for designing innovative management technologies. However, because of the high computational cost, calibration and validation of physically-based models is still challenging. In this context, this study aims to develop a new meta-model based framework for efficient calibration and sensitivity analysis of complex and computationally intensive physically-based models. The proposed approach is applied to the FullSWOF-HR model. According to the average rainfall intensity, 21 rainfall events are categorized into three groups, such as 9 light rains, 6 moderate rains and 6 heavy rains. After upscaling the original high-resolution model, 77 parameter nodes are selected by using the adaptive stochastic collocation method with sparse grids algorithm on the lower-resolution surrogate. 77 simulation runs are then performed with the original model for three representative rainfall events, respectively. The interpolating polynomials of the original models are hence generated. Once the meta-model is constructed, we performed the sensitivity analysis with the variance-based Sobol's method, the results of which are consistent with our previous studies. Calibration process of the meta-model is based on the Markov chain Monte Carlo method. The optimized parameters are verified with the original model and then validated for different rainfall events. These promising results show that the proposed meta-model based approach can efficiently perform sensitivity analysis and parameter optimization for complex physical stormwater quality models, and hence will be very helpful for spreading the detailed water quantity and quality modelling for urban water management issues.

A new analytical model for flow in acidized fractured-vuggy porous media.

Acidizing is one of the widely used technologies that makes the development of naturally fractured-vuggy reservoirs effective. During the process of acidizing, carbonate minerals are dissolved by hydrochloric acid, which can create high conductivity channels and wormholes to connect fractures and pores. In this work, a new analytical model, incorporating the heterogeneity of the pore networks into acidizing region, is proposed to study the flow characteristics in acidized fractured-vuggy reservoirs. The model is coupled by an acidized inner region and a conceptualized outer region of common triple medium. The porosity and permeability of inner region, which are rather heterogeneous and disordered when observed at different length scales, can be well addressed by fractal theory. The properties of the outer region can be described with three basic parameters: the matrix block size LM, the space interval of fracture LF and the radius of the vug Lv. Results show that the flow characteristic curves can be characterized by six flow stages (i.e. wellbore storage stage, radial flow stage in the interior region, fracture-vug inter-porosity flow stage, transition flow stage, fracture-matrix inter-porosity flow stage and external boundary response stage). It can be applied to estimate reservoir parameters for uncertainty reduction using inverse modeling.