Multiattribute reservoir parameter estimation based on a machine learning technique

Yuan, Zhenyu; Handong, Huang; Jiang, Yuxin; Tang, Jinbiao

doi:10.1190/segam2018-2992296.1

Cited by 3 publications

(1 citation statement)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In recent years, the rapid development of big data, high-performance computing, and artificial intelligence has promoted the application of data-based deep learning methods in HR seismic processing. Yuan et al [17] constructed a regression model by means of support vector machine that can be used for HR seismic processing when the target curve is an HR seismic trace. However, such method treats seismic signals discretely so that the consequence of spatial correlation and structural features of the target strata is ignored.…”

Section: Introductionmentioning

confidence: 99%

Untitled

View full text Add to dashboard Cite

Section: Introductionmentioning

confidence: 99%

Untitled

View full text Add to dashboard Cite

Seismic resolution improving by a sequential convolutional neural network

Yuan,

Jiang,

et al. 2024

PLoS ONE

View full text Add to dashboard Cite

Thin-bed soft rock is one of the main factors causing large deformations of tunnels. In addition to relying on some innovative construction techniques, detecting thin beds early during surface geological exploration and advanced geological prediction can provide a basis for planning and implementing effective coping measures. The commonly used seismic methods cannot meet the requirement for thin beds detection accuracy. A high-resolution (HR) seismic signal processing method is proposed by introducing a sequential convolutional neural network (SCNN). The deep learning dataset including low-resolution (LR) and HR seismic is firstly prepared through forward modeling. Then, a one-dimension (1D) SCNN architecture is proposed to establish the mapping relationship between LR and HR sequences. Training on the prepared dataset, the HR seismic processing model with high accuracy is achieved and applied to some practical seismic data. The applications on both poststack and prestack seismic data demonstrate that the trained HR processing model can effectively improve the seismic resolution and restore the high-frequency seismic energy so that to recognize the thin-bed rocks.

show abstract

Seismic Inversion of Reservoir Porosity with Neural Network Technology-A Case Study in Central Iraq

Chen

Deng

et al. 2022

Day 2 Tue, March 22, 2022

View full text Add to dashboard Cite

Understanding the interwell distribution of the reservoir porosity is of great importance for well deployment to improve EOR. Seismic inversion with seismic and logging data is an efficient method to obtain the reservoir porosity. This study aims to demonstrate the spatial distribution of the reservoir porosity in an important formation in Central Iraq. 3D seismic attributes and logging data were combined to invert the reservoir porosity through neural network technology. The migrated 3D seismic volume, inverted P-wave impedance volume, seismic attributes and the logging data of 10 wells, were employed for training neural networks. Based on the training network, we generated the 3D porosity volume. To verify the accuracy of the inverted result, the inverted porosity were compared with those from the logging data of other 5 wells. Data slices were extracted with seismic horizons to show the lateral distribution of the reservoir porosity. The validation error shows the best multi-attribute pair is the pair of square root of P-wave impedance, quadrature trace, and instantaneous frequency. Neural network was trained with the three attribute pair. Analysis of the correlations between the predicted and the logging porosity showed the correlations from neural network training were higher than those achieved with multi-attribute regression. The porosity from the logging data of the 5 wells, which were not evolved in neural network training, coincided well with those from the inverted 3D porosity volume. That verified the accuracy of the inverted porosity volume from neural network inversion. Vertical sections and lateral slices of the inverted porosity volume were extracted to demonstrate the vertical and lateral distributions of the porosity, respectively. Data slices showed that the porosity were higher in the north and south area, and lower in the middle area. The study shows the porosity inverted from neural network technology is more reliable than that from muti-attribute regression. In addition, through this study, we demonstrate the porosity distribution in the project area. The new knowledge of the spatial distribution of reservoir porosity provides important guidance to the well deployment in the oilfield.

show abstract

Multiattribute reservoir parameter estimation based on a machine learning technique

Cited by 3 publications

References 3 publications

Untitled

Untitled

Seismic resolution improving by a sequential convolutional neural network

Seismic Inversion of Reservoir Porosity with Neural Network Technology-A Case Study in Central Iraq

Contact Info

Product

Resources

About