Research Report
Fracture Data Analysis Technology
Fractured Reservoir Discrete Feature
Network Technologies
A Project of
Fundamental Geoscience
Research and Development
BDM-Oklahoma
U.S. Department of Energy
National Oil and Related Programs
Contract Number
#G4S51728
Prepared by:
William S. Dershowitz
Todd Foxford
Thomas Doe
Golder Associates Inc.
Redmond, Washington
March 9, 1998
963-1357.211
FracSys98.doc
TABLE OF CONTENTS
2.1 Discrete Feature Network Modeling
2.1.1 DFN Approach
2.1.2 Integration of Geological and Hydraulic Data3. NeurISIS FRACTURE SETS ANALYSIS
3.1 Algorithm
3.2 NeurISIS User Interface
3.3 Verification Case4. SPATIAL: SPATIAL LOCATION ANALYSIS
4.1 Algorithm
4.2 Spatial User Interface
4.3 Spatial Verification5. FracDim: FRACTIONAL DIMENSION TYPE CURVE ANALYSIS
6. FLARE: HYDRAULIC PARAMETER ANALYSIS
LIST OF TABLES
Table 2-1 Use of Field Data in the Discrete Feature Network Approach
Table 2-2 Input and Output Parameters for Fracture Conductivity Study
Table 3-1 Fracture Property Classes
Table 3-2 NeurISIS User Interface
Table 3-3 NeurISIS Data Format (.ISI)
Table 3-4 Borehole Data Format (.SAM)
Table 3-5 NeurISIS Verification Case
Table 4-1 Spatial Lineament Analysis Sequence
Table 5-1 Input file for FracDim Spreadsheet
Table 5-2 File Format for FracDim
LIST OF FIGURES
Figure 1-1 Fracture Data Analysis System
Figure 2-1 DFN Reservoir Model
Figure 2-2 Reservoir Modeling Approaches
Figure 2-4 DFN Implementation of Background Permeability
Figure 2-5 DFN Models at Scales from 10m to 100km
Figure 2-6 DFN Implementation of Discrete Fractures
Figure 2-7 Reservoir Heterogeneities Amenable to Discrete Feature Network Representation
Figure 2-8 Use of Geological Structure Databases in DFN Fracture Generation
Figure 2-9 DFN Approach for Integration of Characterization Data
Figure 2-10 Forward Modeling Approach
Figure 2-11 Neural Network Topology
Figure 2-12 Example Neural Network for Fracture Set Assignment
Figure 2-13 Hinton Diagram Using Continuous Variables
Figure 3-1 Probabilistic Neural Network Algorithm
Figure 3-2 NeurISIS User Interface
Figure 3-3 Object Oriented Data Model
Figure 3-4 View/Stereoplot and View/Histogram
Figure 3-6 Neural Net/Classify
Figure 3-7 NeurISIS Verification Case
Figure 3-8 Verification Case Statistics
Figure 4-1 Spatial Correlations
Figure 4-2 Spatial Analysis Algorithm
Figure 4-3 Examples of Scan Line, Borehole, and Trace Map Data
Figure 4-4 Stochastic/Probabilistic Set Definitions
Figure 4-5 Gridding Algorithms
Figure 4-6 Intensity Trend on Grid
Figure 4-7 Prior Set Correlation
Figure 4-8 Correlation Between Conjugate Shears
Figure 4-9 Spatial 1.0 Lineament Map Gridding
Figure 4-10 Spatial 1.0 Trend Analysis
Figure 4-11 Spatial 1.0 Dependency Analysis
Figure 4-12 Spatial 1.0 Correlation Analysis
Figure 4-13 Spatial Verification Case
Figure 4-14 Spatial Verification of Intensity P21, Length L, Orientation Cell Values
Figure 4-15 Spatial Verification of Projection Angle
Figure 5-1 Hydraulic Pathway Flow Dimension
Figure 5-2 Fractional Dimensional Flow
Figure 5-3 Fractional Dimension Type Curves
Figure 5-4 Fractional Dimension Type Curves
Figure 5-5 Production From Fractional Dimension Reservoirs
Figure 5-6 FracDim User Interface
Figure 5-7 FracDIM Verification
Figure 6-1 Distribution of Flow Dimension, Finnsjon and Äspö, Sweden and Kamaishi, Japan
Figure 6-2 Flare 1.0 Flow Chart
Figure 6-3 Flow Area vs. Radial Distance
Figure 6-4 Flow Width Channeling Factor Fi
Figure 6-5 Flare User Interface
This report presents theoretical development and implementation of a Windows 95 based data analysis system for fractured reservoir data. Quantitative procedures and software implementing those procedures are described for fracture orientation, spatial structure, flow dimension, and hydraulic parameters.
