Computers & Geosciences最新文献

In this study, we propose an improved lattice Boltzmann flux solver (LBFS) to simulate the thermal-hydraulic (TH) processes within fractured porous media. In LBFS, the flux at cell interfaces is calculated using a locally reconstructed lattice Boltzmann model (LBM). Unlike conventional methods that use direct mathematical approximations, LBFS can suppress the oscillation of solutions and has better accuracy. However, when simulating two-dimensional fractured porous media problems, the rock matrix is divided into surface cells, while fractures are usually divided into line cells. This increases the complexity of implementing the LBFS, as the reconstruction of interface flux in different dimensions requires the use of discrete velocity models (DmQn) in different dimensions. To address this challenge, we introduce an innovative interpolation scheme based on the improved D1Q3 model, thereby establishing a dimensionally independent approach for the reconstruction of the interface flux. This approach greatly reduces the complexity of applying the LBFS to hybrid dimensional problems and simplifies the computational process. The present method is validated by simulating three typical cases and the results show good agreement with the reference solutions. Finally, the improved LBFS is applied to analyze the TH coupling behavior in fractured porous media with a single fracture and a more complex scenario involving two intersected fractures.

Well log data provide key subsurface information, which is crucial for lithology evaluation and reservoir characterization. However, due to technical issues, well log data may contain missing values at certain depth intervals, which can be detrimental for data analysis. The best method is to reacquire the missing data by relogging, but this increases operational costs. Thus, a cost-efficient method for restoring the lost data is needed to overcome this issue. We propose an imputation method for missing well log data using collaborative filtering, a widely used algorithm for making new item recommendations to users. Although collaborative filtering is mainly used in recommendation systems, its fundamental principle allows us to utilize it to help make predictions for missing log data. The method is applied to a well log dataset obtained from the North Sea near Norway. The results show that the collaborative filtering algorithm has the potential to be a powerful imputation method for missing well log data, but there are some limitations that need to be addressed.

Most companies struggle to find and extract relevant information from their technical documents. In particular, the Oil and Gas (O&G) industry faces the challenge of dealing with large amounts of data hidden within old and new geoscientific reports collected over decades of operation. Making this information available in a structured format can unlock valuable information among these mountains of data, which is crucial to support a wide range of industrial and academic applications. However, most natural language processing resources were built from general domain corpora extracted from the Internet and primarily written in English. This paper presents Petro NLP, a comprehensive set of natural language processing and information extraction resources for the oil and gas industry in Portuguese.

We connected an interdisciplinary team of geoscientists, linguists, computer scientists, petroleum engineers, librarians, and ontologists to build a knowledge graph and several annotated corpora. The Petro NLP resources comprise: (i) Petro KGraph– a knowledge graph populated with entities and relations commonly found on technical reports; and (ii) Petrolês, PetroGold, PetroNER, and PetroRE– sets of corpora containing raw text and documents annotated with morphosyntactic labels, named entities, and relations. These resources are fundamental infrastructure for future research in natural language processing and information extraction in the oil industry. Our ongoing research uses these datasets to train and enhance pre-trained machine learning models that automatically extract information from geoscientific technical documents.

THEPORE (THErmo-POro-Elastic solutions) is an open source software to perform forward and inverse modeling of ground displacements induced by thermo-poro-elastic sources. The software, implemented in MATLAB, offers a library of analytical and semi-analytical solutions to compute ground displacements induced by thermo-poro-elastic deformation sources of different geometries, embedded in an elastic, homogeneous and isotropic half-space. The solutions have been verified against finite-element simulations. THEPORE includes also an inversion procedure of the deformation data to constrain the source parameters that better fit the observed signals.

The software's functionality is showcased by inverting the GPS deformation data recorded on Vulcano Island at the onset of the 2021 unrest, in order to estimate the position and volume change of the source responsible for the observed deformations. The results encourage to consider THEPORE as a practical tool suitable for a fast preliminary estimation of the deformation source during a volcanic crisis.

Geophysical logging is a geo-scientific instrument that detects information such as electric, acoustic, and radioactive properties of a well. Its data plays a vital role in interpreting subsurface geology. However, since logging data is an indirect reflection of rocks, it requires the construction of a logging interpretation model in combination with core samples. Obtaining and analysing all core samples in a well is not practical due to their enormous cost, leading to the problem of scarce core sample labels. This problem can be addressed using semi-supervised learning. Existing studies on lithology identification using logging data mostly utilize graph-based semi-supervised learning, which requires known features to establish a graph Laplacian matrix. Therefore, these methods often use logging values at certain depths to construct feature vectors and cannot learn the shape information of logging curves. In this paper, we propose a semi-supervised learning method with feature learning capability based on semi-supervised generative adversarial network (SSGAN) to learn the shape information of logging curves while utilizing unlabelled logging curves. Additionally, considering the problem of insufficient use of labels when dividing a validation set in extremely scarce-label situations, we propose a strategy of weighted averaging of three sub-models, which effectively improves model performance. We verify the effectiveness of our proposed method on five wells and demonstrate the mechanism of semi-supervised learning using adversarial learning through extensive visualization methods.

In this study, we proposed a general workflow that aims to enhance the ML-based geothermobarometer modelling. Our workflow focuses on three key areas. Firstly, we developed a robust pre-processing pipeline that addresses data imbalance, feature engineering, and data augmentation. Secondly, we assessed modelling errors using a Monte Carlo approach to quantify the impact of analytical uncertainties on the final pressure and temperature estimates. Thirdly, we implemented a robust strategy to validate and test the ML models to avoid over- and under-fitting issues while correcting biases associated with the application of specific ML models (i.e., tree-based ensembles).

To facilitate the use of our workflow, we have developed a web app (https://bit.ly/ml-pt-web) and a Python module (https://bit.ly/ml-pt-py). The robustness of this strategy has been tested on two calibrations: clinopyroxene (cpx) and clinopyroxene-liquid (cpx-liq). Our results show a significant reduction in errors compared to the baseline model, as well as good generalization ability on an independent external dataset. The Root Mean Squared Errors are 57 °C and 2.5 kbar for the cpx calibration, and 36 °C and 2.1 kbar for the cpx-liq calibration. Finally, our models show improved outcomes on the external dataset compared to existing ML and classical cpx and cpx-liq thermobarometers.

The present study proposes a double-branch classification network, DPNet (Double Path Net), for the classification and identification of microseismic and blasting signals based on multimodal feature extraction. The vibration signals’ one-dimensional spectrogram and two-dimensional wavelet time–frequency graph are inputted into the double branch network. Subsequently, convolutional neural networks and ResNet are employed to extract the one-dimensional frequency features and two-dimensional time–frequency features of the vibration signals, respectively. Experimental results demonstrate that our proposed method achieves outstanding classification performance with an accuracy of 97.34% for microseismic signals and blasting signals. This research not only provides innovative solutions to practical problems but also introduces a novel idea of multimodal feature extraction at a theoretical level. By successfully applying it to efficiently classify complex signals in mining engineering, we offer a feasible solution that holds promising prospects for practical applications in this field.

Numerical modeling of electromagnetic (EM) fields in a conductive marine environment is crucial for marine EM data interpretation. During marine controlled-source electromagnetic (MCSEM) surveys, a variety of transmitter locations are used to introduce electric currents. The resulting electric and magnetic fields are then concurrently logged by a network of receivers. The forward simulation of MCSEM data for a subsea structure whose electrical properties vary in all three dimensions is computationally intensive. We demonstrate how such computations may be substantially accelerated by adapting algorithms to operate efficiently on modern GPUs with many core architectures. The algorithm we present features a hybrid MPI-CUDA programming model suitable for multi-GPU computers and consists of three levels of parallelism. We design the optimal kernels for different components to minimize redundant memory accesses. We have tested the algorithm on NVIDIA Kepler architecture and achieved up to 105 × speedup compared with the serial code version. We further showcased the algorithm's performance advantages through its application to a realistic marine model featuring complex geological structures. Our algorithm's significant efficiency increase opens the possibility of 3D MCSEM data interpretation based on probabilistic or machine learning approaches, which require tens of thousands of forward simulations for every survey.

Mine induced seismic events are a major safety concern in mining and require careful monitoring and management to reduce their effects. Therefore, an essential step in assessing seismic and rock burst hazards is the analysis of mine seismicity. Recently, deep neural networks have been used to automatically determine seismic wave arrival times, surpassing human performance and allowing their use in seismic data analysis such as seismic event location and seismic energy calculation. In order to properly automate the rockburst and seismic hazard assessment deep neural network phase picker and a spatio-temporal clustering method were utilized. Seismic and rockburst hazards were statistically quantified using two-way contingency tables for two categorical variables: seismic energy level of mine tremors and number of clusters. Correlations between several spatio-temporal clusters and a statistical association between two categorical variables: seismic energy level and cluster number indicate an increase of seismic hazard in the Marcel hard coal mine in Poland. A new automated tool has been elaborated to automatically identify high-stress areas in mines in the form of spatio-temporal clusters.

Some target seismic signals in the earthquake data can be very weak compared with interfering phases, and are thus difficult to detect, which further hinders the effective usage of these weak phases for subsequent high-resolution imaging of earth interiors. The strong ambient noise makes this situation even more troublesome since the weak signals can be mostly buried in the noise. Here, we present an open-source package for uncovering the weak phases from global seismograms. We adopt a two-step scheme to reconstruct and denoise array data. The first step is weighted average interpolation which puts the data into irregular grids. The second step adopts the weighted projection-onto-convex sets based on damped rank-reduction to further interpolate and denoise for the binned data. Taking the complexity of the weak signal into consideration, we adopt the automatic strategy to select an appropriate rank in different localized windows. We conduct several synthetic tests to carefully investigate the performance regarding effectiveness, robustness, and efficiency, and compare the algorithm with the frequency–wavenumber-domain projection onto convex sets method that is already used in the global seismology literature. Finally, the proposed framework is validated via a recorded array data set of the 1995 May 5 Philippines earthquake.