Runhai Feng

Runhai Feng

Auto-generated profile, based on publicly available information

Runhai Feng

Niels Bohr Institute, University of Copenhagen, Copenhagen, Denmark

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 November 2020 in Geophysics
Verified
Journal Article
Closed access
RFRunhai Feng
Runhai Feng
THThomas Mejer Hansen
Thomas Mejer Hansen
DGDario Grana
Dario Grana
NBNiels Balling
Niels Balling
We propose to invert reservoir porosity from poststack seismic data using an innovative approach based on deep-learning methods. We develop an unsupervised approach to circumvent the requirement of large volumes of labeled data sets for a conventional learning process. We apply convolutional neural networks (CNN) on seismic data to predict the relative porosity that is to be added to a low-frequency prior component. We then apply a forward model to synthesize seismic data based on a source wavelet and an acoustic impedance converted from the network-determined porosity. The parameters in the CNN are iteratively updated to minimize the error between recorded and simulated seismic data. We test the capability of our deep-learning approach to estimate reservoir porosity using a synthetic rock-physics model with two different signal-to-noise ratios. We also apply the proposed method to a real case study of seismic data acquired for hydrocarbon exploration of clastic reservoirs in the Vienna Basin. Instead of randomly assigning neural parameters, we use pretrained weights and biases at a previous location as initialization values for the next location, to preserve the geologically lateral continuity of the layers’ physical properties. As shown by these analyses, the unsupervised CNN-based scheme provides more or equally accurate results than standard methods for porosity estimation from seismically inverted acoustic impedance, which makes it a promising tool in seismic reservoir characterization with less user intervention.

28 January 2021 in IEEE Transactions on Geoscience and Remote Sensing
Verified
Journal Article
Closed access
RFRunhai Feng
Runhai Feng
NBNiels Balling
Niels Balling
DGDario Grana
Dario Grana
JDJesper Soren Dramsch
Jesper Soren Dramsch
THThomas Mejer Hansen
Thomas Mejer Hansen
The seismic response of geological reservoirs is a function of the elastic properties of porous rocks, which depends on rock types, petrophysical features, and geological environments. Such rock characteristics are generally classified into geological facies. We propose to use the convolutional neural networks in a Bayesian framework to predict facies based on seismic data and quantify the uncertainty in the classification. A variational approach is adopted to approximate the posterior distribution of neural parameters that is mathematically intractable. The network is trained on labeled examples. The mean and the standard deviation of the distribution of neural parameters are randomly drawn from predefined Gaussian functions for the initialization, and are updated by minimizing the negative evidence lower bound. The facies classification is applied to seismic sections not included in the training data set. We draw multiple random samples from the trained variational posterior distribution to simulate an ensemble predictor and classify the most probable seismic facies. We implement the proposed network in the open-source library of TensorFlow Probability, for its convenience and flexibility. The applications show that the internal regions of the seismic sections are generally classified with higher confidence than their boundaries, as measured by the predictive entropy that is calculated based on a multiclass probability across the possible facies. A plain neural network is also applied for comparison, by assigning fixed values to the neural parameters using a classical backpropagation technique. The comparison shows consistent results; however, the proposed approach is able to assess the uncertainty in the predictions.

19 January 2021 in Geophysics
Verified
Journal Article
Closed access
RFRunhai Feng
Runhai Feng
DGDario Grana
Dario Grana
NBNiels Balling
Niels Balling
Segmentation of faults based on seismic images is an important step in reservoir characterization. With the recent developments of deep-learning methods and the availability of massive computing power, automatic interpretation of seismic faults has become possible. The likelihood of occurrence for a fault can be quantified using a sigmoid function. Our goal is to quantify the fault model uncertainty that is generally not captured by deep-learning tools. We have used the dropout approach, a regularization technique to prevent overfitting and coadaptation in hidden units, to approximate the Bayesian inference and estimate the principled uncertainty over functions. Particularly, the variance of the learned model has been decomposed into aleatoric and epistemic parts. Our method is applied to a real data set from the Netherlands F3 block with two different dropout ratios in convolutional neural networks. The aleatoric uncertainty is irreducible because it relates to the stochastic dependency within the input observations. As the number of Monte Carlo realizations increases, the epistemic uncertainty asymptotically converges and the model standard deviation decreases because the variability of the model parameters is better simulated or explained with a larger sample size. This analysis can quantify the confidence to use fault predictions with less uncertainty. In addition, the analysis suggests where more training data are needed to reduce the uncertainty in low-confidence regions.

1 May 2020 in Journal of Natural Gas Science and Engineering
Verified
Journal Article
Closed access
RFRunhai Feng
Runhai Feng

21 April 2021 in Geophysics
Verified
Journal Article
Closed access
RFRunhai Feng
Runhai Feng
DGDario Grana
Dario Grana
NBNiels Balling
Niels Balling
We have introduced a Bayesian neural network in quantitative log prediction studies with the goal of improving the petrophysical characterization and quantifying the uncertainty of model predictions. Neural network (NN) methods are gaining popularity in the petrophysics and geophysics communities; however, uncertainty quantification in model predictions is often neglected in the available literature, where the prediction is frequently performed in a deterministic setting. Determination of the uncertainty of the petrophysical model requires the estimation of the posterior distribution of the neural parameters that is generally mathematically intractable; for this reason, we adopt a variational approach to approximate the posterior model of the Bayesian network. To represent the uncertainty, we randomly draw samples from the posterior distribution of neural parameters to predict the model variables given the input data, leading to a learned-ensemble predictor. The proposed approach combines the ability of the NN of extracting hidden relations within the data set that physical relations cannot describe and the probabilistic framework for uncertainty quantification. We apply the proposed method to a well-log data set from the Volve Field, offshore Norway, to predict well logs in intervals where the data are incomplete or missing due to operational issues in the drilling procedure. The proposed approach is validated in intervals where the true data are available but not included in the training process. In the proposed application, the correlation coefficient between predictions and true data is greater than 0.9. In terms of accuracy, the results are comparable to those obtained using a traditional NN approach; however, the proposed method also provides a quantification of the uncertainty in the results, which offers additional information on the confidence in the predictions.

16 March 2020 in Geophysical Journal International
Verified
Journal Article
Closed access
PDF file
RFRunhai Feng
Runhai Feng
SUMMARY: Lithofacies is one of the most important reservoir parameters, which could provide a qualitative description for hydrocarbon and geothermal reservoirs. Various techniques, such as artificial neural networks (ANN) and hidden Markov models (HMM), have been applied to extract this information, with the well log suites as inputs. However, both of these methods have their own limitations, such as no geological priors in ANN, since log samples along the depth direction are treated independently. While in HMM, a simple Gaussian assumption is usually adopted, which may not be sufficient for complex data distributions. In order to address these problems, a new method is proposed, which combines ANN and HMM into a hybrid system. This new technique allows for a more flexible approach to the probability distributions of rock properties, without any Gaussian assumption being made. At the same time, the geological dependence between adjacent samples is introduced by a representative transition matrix of lithofacies. The output probability from ANN must be reformulated to an emission function before it can be fed into the HMM, which is achieved via the Bayes’ rule. Then the Viterbi algorithm in HMM is applied for the decoding of sequential patterns in the subsurface. In this fashion, the classification process can be proceeded statistically and geologically. Better performance of the proposed approach, compared with other classification methods, is demonstrated in two case studies.