Reservoir Characterization of the Horn River Shale in British Columbia, Canada : Integration of Core, Well log and Seismic data

다국어 초록 (Multilingual Abstract)

The geoscience-based reservoir characterization essentially identifies lithofacies in a different scale based on diverse data such as core, well logs and seismic profiles. The Horn River shale represents about 25.7% (11.1Tcf) of the remaining recoverable raw gas reserves in British Columbia, Canada. The Horn River shale consists of the Muskwa, Otter Park and Evie shales that are underlain by the Keg River carbonates and overlain by the Fort Simpson shales. This study focuses on the stratigraphic and sedimentological features of 162m long borehole in the Horn River shale. Six lithofacies are identified: faintly laminated siliceous mudstone (FLSM), homogeneous siliceous mudstone (HSM), laminated siliceous mudstone (LSM), laminated mixed mudstone (LMM), argillaceous mudstone (AM), and calcareous mudstone (CM). The FLSM and HSM are the most dominant facies of the Evie and Muskwa shales. These facies have high silica and organic contents.
In order to establish a quantitative relationship between the core lithology and wireline logs, electrofacies are derived from wireline log data through the pattern recognition method. Prediction of electrofacies in shale succession with core is based on the comparison of results from various supervised methods (MRGC, ANN, and SVM). SVM and ANN methods are important to better understand the complex relationships between core lithofacies and well logs. However, these algorithms have limitations when they are applied to non-cored wells, so the MRGC algorithm was selected to propagate the electrofacies.
To build up the seismic-scale facies models, the stochastic methods are appropriate, particularly in case of insufficient electrofacies data. In order to enhance the accuracy and applicability of these methods, this study applies trend data for sequential indicator simulation modeling. The trend data is four equally weighted seismic attributes: density and gamma ray inversion, envelope, and spectral decomposition volume. The 3D model of lithofacies provides lateral facies distribution, and it is useful to visualize porosity, permeability and brittleness of shale succession.
The physical, biological and chemical observations in various scales can be integrated within the sequence stratigraphic framework. It is the main procedure of sequence stratigraphic interpretation to recognize key stratal surfaces and stratal stacking patterns. The Th / U ratios against the spectral uranium curve represent a proxy record of sea level change that influences the depositional regime and the accumulation of organic matter. In the shale succession, the T-R sequence bounding surfaces can be recognized from analyses of vertical trends such as lithofacies, spectral gamma-ray, and TOC. The sequence stratigraphic analysis of a vertical succession of variable lithofacies provides geological framework to figure out the sediment type, redox condition, and possible reservoir potential.
This study demonstrates that the integration of lithofacies, TOC, spectral gamma ray ratio, and geochemical data is useful for estimating the variation of lithology, detrital flux, redox conditions, and organic matter accumulation in the Horn River shale. The Horn River shales can be divided into six 4th order T-R sequence that are grouped into two 3rd order T-R sequences of a few millions of years.

목차 (Table of Contents)

• Chapter 1. Introduction 12
• 1.1 Purpose of this Study 12
• 1.2 Study workflow 14
• 1.3 Thesis layout 16
• References 17
• Chapter 2. Characteristics of Horn River shale lithofacies: sedimentary description combining the ECS log and petrographic analysis 19
• 2.1 Introduction 19
• 2.2 Regional stratigraphy 22
• 2.3 Data 27
• 2.4 Methodology 30
• 2.5 Lithofacies 33
• 2.5.1 Faintly Laminated Siliceous Mudstone (FLSM) 33
• 2.5.2 Homogeneous Siliceous Mudstone (HSM) 34
• 2.5.3 Laminated Siliceous Mudstone (LSM) 35
• 2.5.4 Laminated Mixed Mudstone (LMM) 35
• 2.5.5 Argillaceous Mudstone (AM) 36
• 2.5.6 Calcareous Mudstone (CM) 37
• 2.6 Discussion 45
• 2.7 Conclusion 51
• References 53
• Chapter 3. Application of supervised and unsupervised methods to building electrofacies of Horn River shale 60
• 3.1 Introduction 60
• 3.2 Method 64
• 3.2.1 Multi Resolution Graph based Clustering (MRGC) method 64
• 3.2.2 Artificial Neural Networks (ANN) 66
• 3.2.3 Support Vector Machine (SVM) 68
• 3.2.4 Distance parameter 71
• 3.3 Geological setting 72
• 3.4 Horn River shale Lithofacies 74
• 3.5 Data 77
• 3.6 Procedure 78
• 3.6.1 Preparation of well log data 78
• 3.6.2 Clustering 82
• 3.7 Results and Discussion 85
• 3.8 Conclusion 95
• References 97
• Chapter 4. Integration of spectral gamma ray, TOC, and lithofacies and its application to the recognition of T-R stratigraphic surfaces in the Horn River shales 103
• 4.1 Introduction 103
• 4.2 Regional stratigraphy 105
• 4.3 Data and methodology 109
• 4.4 Sequence stratigraphy 110
• 4.5 Result and discussion 115
• 4.5.1 T-R sequence stratigraphy 115
• 4.5.2 4th order Sequence of the Horn River Formation 123
• 4.5.3 Correlation of systems tract and comparison of previous study 124
• 4.6 Conclusion 132
• References 135
• Chapter 5 Conclusion 142
• Abstract(in Korean) 145
• APPENDICES 148