Landmark Graphics Corporation

Fan Jiang

Shengwen Jin

Phil Norlund

Youli Mao

Yiqing Ren

Alejandro Jaramillo

Steven Roy Angelovich

Philip Norlund

Zhili Wei

Meng Meng

Frequency-dependent machine-learning (ML) models can be used to interpret seismic data. A system can apply spectral decomposition to pre-processed training data to generate frequency-dependent training data of two or more frequencies. The system can train two or more ML models using the frequency-dependent training data. Subsequent to training the two or more ML models, the system can apply the two or more ML models to seismic data to generate two or more subterranean feature probability maps. The system can perform an analysis of aleatoric uncertainty on the two or more subterranean feature probability maps to create an uncertainty map for aleatoric uncertainty. Additionally, the system can generate a filtered subterranean feature probability map based on the uncertainty map for aleatoric uncertainty.

Frequency-dependent machine learning model in seismic interpretation

Seismic image data acquired for a subsurface formation from a data acquisition system is input into a deep neural network to generate fault detection data for the subsurface formation comprising probability values at a grid of locations in the subsurface formation. The fault detection data is preprocessed via downsampling and distributed weighted factors and inputted into a generative adversarial network (GAN) upscaling generator to create high resolution fault detection data with minimized distortion and artifacts. The GAN upscaling generator is pre trained on synthetic fault data in a GAN training system using adversarial training against a GAN upscaling discriminator, and both the GAN upscaling generator and the GAN upscaling discriminator learn to approximate the distribution of the synthetic fault data.

Geological feature detection using generative adversarial neural networks

A system can receive seismic data that can correlate to a subterranean formation. The system can derive a set of seismic attributes from the seismic data. The seismic attributes can include discontinuity-along-dip. The system can determine parameterized results by analyzing the seismic data and the seismic attributes using a deep learning neural network. The deep learning neural network can include a dilation module. The system can determine one or more fault probabilities of the subterranean formation using the parameterized results. The system can output the fault probabilities for use in a hydrocarbon exploration operation.

Deep learning model with dilation module for fault characterization

This disclosure presents a fault prediction system using a deep learning neural network, such as a convolutional neural network. The fault prediction system utilizes as input seismic data, and then derives various seismic attributes from the seismic data. In various aspects, the seismic attributes can be normalized and have importance coefficients determined. A sub-set of seismic attributes can be selected to reduce computing resources and processing time. The deep learning neural network can utilize the seismic data and seismic attributes to determine parameterized results representing fault probabilities. The fault prediction system can utilize the fault probabilities to determine fault predictions which can be represented as a predicted new seismic data, such as using a three-dimensional image.

Deep learning seismic attribute fault predictions

A system can receive fault likelihood data about a subterranean environment and apply a binary mask filter using a tuning parameter to convert the fault likelihood data to binary distribution data having a plurality of pixels arranged in a plurality of profiles in at least two directions. The system can perform, for each profile of the plurality of profiles, fault skeletonization on the binary distribution data to form fault skeletonization data with pixels connected that represent part of a fracture. The system can convert the fault skeletonization data to seismic volume data and combine and filter the seismic volume data in the at least two directions to form combined seismic volume data. The system can output the combined seismic volume data as an image for use in detecting objects to plan a wellbore operation.

Fault skeletonization for fault identification in a subterranean environment

Hydrocarbon exploration and extraction can be facilitated using machine-learning models. For example, a system described herein can receive seismic data indicating locations of geological bodies in a target area of a subterranean formation. The system can provide the seismic data as input to a trained machine-learning model for determining whether the target area of the subterranean formation includes one or more types of geological bodies. The system can receive an output from the trained machine-learning model indicating whether or not the target area of the subterranean formation includes the one or more types of geological bodies. The system can then execute one or more processing operations for facilitating hydrocarbon exploration or extraction based on the seismic data and the output from the trained machine-learning model.

Facilitating hydrocarbon exploration and extraction by applying a machine-learning model to seismic data

Systems and methods for simulating an effect of multiple seismic sources concurrently on a geologic formation are provided. Data is read from a seismic source file that describes at least two seismic source types. The concurrent propagation of acoustic energy from the at least two seismic sources types through the geologic formation is modeled. A seismic output file is then generated.

Methods and systems for seismic modeling using multiple seismic source types

Disclosed are systems and methods for numerically simulating seismic-wave propagation in tilted transversely isotropic (TTI) media, using an adaptive Lebedev staggered grid. In various embodiments, the adaptive grid includes multiple horizontal zones having different associated grid spacings, which may be determined based on a vertical wave-velocity model. The numerical simulation may involve iteratively solving a set of finite-difference equations including finite-difference coefficients that vary spatially depending on the grid spacing. Additional embodiments and features are described.

Seismic elastic wave simulation for tilted transversely isotropic media using adaptive lebedev staggered grid

Systems and methods for imaging seismic data using hybrid one-way wave-equation-migration in tilted transverse isotropic media and/or hybrid two-way reverse-time-migration in tilted transverse isotropic media.

Growing community of inventors

Sugar Land, TX, United States of America

Fan Jiang