Cover

Title Page

Copyright

Contents

Foreword

Preface

Acknowledgments

Chapter 1: Introduction to Data-Driven Concepts

Introduction

Current Approaches

Is There a Crisis in Geophysical and Petrophysical Analysis?

Applying an Analytical Approach

What Are Analytics and Data Science?

Meanwhile, Back in the Oil Industry

How Do I Do Analytics and Data Science?

What Are the Constituent Parts of an Upstream Data Science Team?

A Data-Driven Study Timeline

What Is Data Engineering?

A Workflow for Getting Started

Is It Induction or Deduction?

References

Chapter 2: Data-Driven Analytical Methods Used in E&P

Introduction

Spatial Datasets

Temporal Datasets

Soft Computing Techniques

Data Mining Nomenclature

Decision Trees

Rules-Based Methods

Regression

Classification Tasks

Ensemble Methodology

Partial Least Squares

Traditional Neural Networks: The Details

Simple Neural Networks

Random Forests

Gradient Boosting

Gradient Descent

Factorized Machine Learning

Evolutionary Computing and Genetic Algorithms

Artificial Intelligence: Machine and Deep Learning

References

Chapter 3: Advanced Geophysical and Petrophysical Methodologies

Introduction

Advanced Geophysical Methodologies

How Many Clusters?

Case Study: North Sea Mature Reservoir Synopsis

Case Study: Working with Passive Seismic Data

Advanced Petrophysical Methodologies

Well Logging and Petrophysical Data Types

Data Collection and Data Quality

What Does Well Logging Data Tell Us?

Stratigraphic Information

Integration with Stratigraphic Data

Extracting Useful Information from Well Reports

Integration with Other Well Information

Integration with Other Technical Domains at the Well Level

Fundamental Insights

Feature Engineering in Well Logs

Toward Machine Learning

Use Cases

Concluding Remarks

References

Chapter 4: Continuous Monitoring

Introduction

Continuous Monitoring in the Reservoir

Machine Learning Techniques for Temporal Data

Spatiotemporal Perspectives

Time Series Analysis

Advanced Time Series Prediction

Production Gap Analysis

Digital Signal Processing Theory

Hydraulic Fracture Monitoring and Mapping

Completions Evaluation

Reservoir Monitoring: Real-Time Data Quality

Distributed Acoustic Sensing

Distributed Temperature Sensing

Case Study: Time Series to Optimize Hydraulic Fracture Strategy

Reservoir Characterization and Tukey Diagrams

References

Chapter 5: Seismic Reservoir Characterization

Introduction

Seismic Reservoir Characterization: Key Parameters

Principal Component Analysis

Self-Organizing Maps

Modular Artificial Neural Networks

Wavelet Analysis

Wavelet Scalograms

Spectral Decomposition

First Arrivals

Noise Suppression

References

Chapter 6: Seismic Attribute Analysis

Introduction

Types of Seismic Attributes

Seismic Attribute Workflows

SEMMA Process

Seismic Facies Classification

Seismic Facies Dataset

Seismic Facies Study: Preprocessing

Hierarchical Clustering

k-means Clustering

Self-Organizing Maps (SOMs)

Normal Mixtures

Latent Class Analysis

Principal Component Analysis (PCA)

Statistical Assessment

References

Chapter 7: Geostatistics: Integrating Seismic and Petrophysical Data

Introduction

Data Description

Interpretation

Estimation

The Covariance and the Variogram

Case Study: Spatially Predicted Model of Anisotropic Permeability

What Is Anisotropy?

Analysis with Surface Trend Removal

Kriging and Co-kriging

Geostatistical Inversion

Geophysical Attribute: Acoustic Impedance

Petrophysical Properties: Density and Lithology

Knowledge Synthesis: Bayesian Maximum Entropy (BME)

References

Chapter 8: Artificial Intelligence: Machine and Deep Learning

Introduction

Data Management

Machine Learning Methodologies

Supervised Learning

Unsupervised Learning

Semi-Supervised Learning

Deep Learning Techniques

Semi-Supervised Learning

Supervised Learning

Unsupervised Learning

Deep Neural Network Architectures

Deep Forward Neural Network

Convolutional Deep Neural Network

Recurrent Deep Neural Network

Stacked Denoising Autoencoder

Seismic Feature Identification Workflow

Efficient Pattern Recognition Approach

Methods and Technologies: Decomposing Images into Patches

Representing Patches with a Dictionary

Stacked Autoencoder

References

Chapter 9: Case Studies: Deep Learning in E&P

Introduction

Reservoir Characterization

Case Study: Seismic Profile Analysis

Supervised and Unsupervised Experiments

Unsupervised Results

Case Study: Estimated Ultimate Recovery

Deep Learning for Time Series Modeling

Scaling Issues with Large Datasets

Conclusions

Case Study: Deep Learning Applied to Well Data

Introduction

Restricted Boltzmann Machines

Mathematics

Case Study: Geophysical Feature Extraction: Deep Neural Networks

CDNN Layer Development

Case Study: Well Log Data-Driven Evaluation for Petrophysical Insights

Case Study: Functional Data Analysis in Reservoir Management

References

Glossary

About the Authors

Index

EULA