On a new method of estimating shear wave velocity from conventional well logs

Wang, Pan

doi:10.1016/j.petrol.2019.05.033

Cited by 48 publications

(10 citation statements)

References 52 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The results show that the prediction accuracy of HFEN is higher than that of random forest. In addition, the prediction results of the proposed method also show advantages compared with similar studies (Wang and Peng, 2019;Wang et al, 2021;Mehrad et al, 2022). Compared to a previous study regarding the JSO field (Song, et al, 2021), HFEN provides results that are closer to actual measurements with lower errors which prove that the network is effective.…”

Section: Performance Evaluation Of Hybrid Neural Networkmentioning

confidence: 57%

Acoustic Prediction of a Multilateral-Well Unconventional Reservoir Based on a Hybrid Feature-Enhancement Long Short-Term Memory Neural Network

Song

et al. 2022

Front. Energy Res.

View full text Add to dashboard Cite

Due to the complexity of unconventional reservoir measurement, log data acquired are often incomplete, which results in inaccurate formation evaluation and higher operational risks. Common solutions, such as coring, are typically high cost related while not being sufficiently representative. In recent years, neural network has received increasing attention given its strong ability in data prediction. Nevertheless, most neural networks only focus on one specific feature of the selected data, thus prohibiting their prediction accuracy for reservoir logs where data are often dominated by more than one key feature. To address this challenge, a novel multi-channel hybrid Long Short-Term Memory (LSTM) neural network for effective acoustic log prediction is proposed. The network combines Convolutional Neural Network (CNN) and LSTM, where CNN is used to extract spatial features of the logs and LSTM network extracts temporal features with the assistance of an adaptive attention mechanism implemented for key feature recognition. In addition, the strong heterogeneity of unconventional reservoirs also increases the difficulty of prediction. Therefore, according to the characteristics of the unconventional reservoir, we designed three feature enhancement methods to mine the hidden information of logs. To prove the performance of the proposed network, a case study is presented with data acquired from Jimusar Oilfield, one of the largest unconventional reservoirs in China. Four groups of experiments are conducted, and the proposed network is employed for acoustic log prediction. The predicted results are validated against measurement (R2: 92.27%, 91.42%, 93.31%, and 92.03%; RMSE: 3.32%, 3.92%, 3.06%, and 3.53%). The performance of the proposed network is compared to other networks such as CNN, LSTM, CNN-LSTM, and random forest (RF). The comparisons show that the proposed network has the highest accuracy level of prediction, which means it provides an effective approach to complement missing data during complicated unconventional reservoir measurement and, therefore, could be of significant potential in energy exploration.

show abstract

Section: Performance Evaluation Of Hybrid Neural Networkmentioning

confidence: 57%

Acoustic Prediction of a Multilateral-Well Unconventional Reservoir Based on a Hybrid Feature-Enhancement Long Short-Term Memory Neural Network

Song

et al. 2022

Front. Energy Res.

View full text Add to dashboard Cite

show abstract

“…The prevalent deep learning algorithms in geophysics are recurrent neural networks (RNNs) for sequence data and convolutional neural networks (CNNs) for image recognition, which are commonly used for micro-seismic monitoring, seismic first break pick, and oil flow rates prediction (Yuan et al, 2018;Abad et al, 2021;, etc. Logging curves are typical of depth sequence data, so many scholars have introduced RNNs and their variants Long short-term memory (LSTM) networks and Gated Recurrent Unit (GRU) networks into Swave velocity prediction (Sun and Liu, 2020;Zhang et al, 2020;Wang and Cao, 2021;You et al, 2021;Wang et al, 2022) feature extraction, so S-wave velocity prediction based on CNN has been widely applied in recent years (Wang and Peng, 2019;Zhang et al, 2021), but it is necessary to mention that CNNs can only extract local features and appear to be powerless for features of long sequences of logging curves. RNNs can only indirectly obtain the information of the previous sequence using hidden states, and since the dimension of the hidden states must be much smaller than the connected dimension of all samples of the previous sequence, this approach will inevitably lose some information, leading to the problem of exploding or vanishing gradients (Bengio et al, 1994;Chen et al, 2020).…”

Section: Introductionmentioning

confidence: 99%

Shear wave velocity prediction based on 1DCNN-BiLSTM network with attention mechanism

Feng,

Liu,

Yang

et al. 2024

Front. Earth Sci.

View full text Add to dashboard Cite

The Shear wave (S-wave) velocity is an essential parameter in reservoir characterization and evaluation, fluid identification, and prestack inversion. However, the cost of obtaining S-wave velocities directly from dipole acoustic logging is relatively high. At the same time, conventional data-driven S-wave velocity prediction methods exhibit several limitations, such as poor accuracy and generalization of empirical formulas, inadequate exploration of logging curve patterns of traditional fully connected neural networks, and gradient explosion and gradient vanishing problems of recurrent neural networks (RNNs). In this study, we present a reliable and low-cost deep learning (DL) approach for S-wave velocity prediction from real logging data to facilitate the solution of these problems. We designed a new network sensitive to depth sequence logging data using conventional neural networks. The new network is composed of one-dimensional (1D) convolutional, bidirectional long short-term memory (BiLSTM), attention, and fully connected layers. First, the network extracts the local features of the logging curves using a 1D convolutional layer, and then extracts the long-term sequence features of the logging curves using the BiLSTM layer, while adding an attention layer behind the BiLSTM network to further highlight the features that are more significant for S-wave velocity prediction and minimize the influence of other features to improve the accuracy of S-wave velocity prediction. Afterward, the nonlinear mapping relationship between logging data and S-wave velocity is established using several fully connected layers. We applied the new network to real field data and compared its performance with three traditional methods, including a long short-term memory (LSTM) network, a back-propagation neural network (BPNN), and an empirical formula. The performance of the four methods was quantified in terms of their coefficient of determination (R2), root mean square error (RMSE), and mean absolute error (MAE). The new network exhibited better performance and generalization ability, with R2 greater than 0.95 (0.9546, 0.9752, and 0.9680, respectively), RMSE less than 57 m/s (56.29, 23.18, and 30.17 m/s, respectively), and MAE less than 35 m/s (34.68, 16.49, and 21.47 m/s, respectively) for the three wells. The test results demonstrate the efficacy of the proposed approach, which has the potential to be widely applied in real areas where S-wave velocity logging data are not available. Furthermore, the findings of this study can help for a better understanding of the superiority of deep learning schemes and attention mechanisms for logging parameter prediction.

show abstract

“…These models excel in complex function approximation and layer-wise feature transformations, making them a compelling choice in the field. Convolutional Neural Networks (CNNs), renowned for their ability to capture spatial features, have demonstrated remarkable performance across diverse geophysical domains, including well-logging data interpretation [13], seismic interpretation [14], and seismic inversion [15,16]. Because Wang et al (2020) leveraged the long-term correlation patterns observed in well-logging data, the Long Short-Term Memory (LSTM) network emerged as a viable choice for shear wave velocity prediction and the identification of complex reservoir geophysical parameters [17,18].…”

Section: Introductionmentioning

confidence: 99%

Predicting Shear Wave Velocity Using a Convolutional Neural Network and Dual-Constraint Calculation for Anisotropic Parameters Incorporating Compressional and Shear Wave Velocities

Liu,

Gui,

Gao

et al. 2023

Processes

View full text Add to dashboard Cite

As the exploration of unconventional reservoirs progresses, characterizing challenging formations like tight sandstone becomes increasingly complex. Anisotropic parameters play a vital role in accurately characterizing these unconventional reservoirs. In light of this, this paper introduces an approach that uses a dual-constraint anisotropic rock physics model based on compressional and shear wave velocities. The proposed method aims to enhance the precision of anisotropic parameter calculations, thus improving the overall accuracy of reservoir characterization. The paper begins by applying a convolutional neural network (CNN) to predict shear wave velocity, effectively resolving the issue of incomplete shear wave logging data. Subsequently, an anisotropic rock physics model is developed specifically for tight sandstone. A comprehensive analysis is conducted to examine the influence of quartz, clay porosity aspect ratio, and fracture density on compressional and shear wave velocities. Trial calculations using the anisotropic model data demonstrated that the accuracy of calculating anisotropic parameters significantly improved when both compressional and transverse wave velocity constraints were taken into account, as opposed to relying solely on the compressional wave velocity constraint. Furthermore, the rationality of predicting anisotropic parameters using both the shear wave velocity predicted by the convolutional neural network and the measured compressional wave velocity was confirmed using the example of deep tight sandstone in the Junggar Basin.

show abstract

On a new method of estimating shear wave velocity from conventional well logs

Cited by 48 publications

References 52 publications

Acoustic Prediction of a Multilateral-Well Unconventional Reservoir Based on a Hybrid Feature-Enhancement Long Short-Term Memory Neural Network

Acoustic Prediction of a Multilateral-Well Unconventional Reservoir Based on a Hybrid Feature-Enhancement Long Short-Term Memory Neural Network

Shear wave velocity prediction based on 1DCNN-BiLSTM network with attention mechanism

Predicting Shear Wave Velocity Using a Convolutional Neural Network and Dual-Constraint Calculation for Anisotropic Parameters Incorporating Compressional and Shear Wave Velocities

Contact Info

Product

Resources

About