Discing Analysis
- Sam Johnson
- Yasin Dagasan
This document details Datarock’s Discing Analysis product.
Contents
Dependent Models
The outputs of the following models are used:
Model Name | Model Type |
|---|---|
Fracture Detection and Classification | Object Detection |
Drillers Break | Object Detection |
Fracture Mask | Instance Segmentation |
Data Processing
The following steps are taken to determine the Weighted Joint Density of a given interval:
Take the output of the fracture detection model, “simple” classes only
Run a segmentation model to extract the fracture profile
Measure the angle of each fracture with respect to the core axis
Identify potential discs by:
Measuring the distance between each fracture, referred to as the disc size
Checking the fractures on both sides of the disc have angles that are similar and near perpendicular to the core axis
Merge immediately adjust discs to create discing regions. A discing region is defined as an area containing at least three (default) disc-like objects (the number of discs can be adjusted as required, see Product Configuration Options section below).
Merge nearby discing regions if the depth between them is within the define threshold (see Product Configuration Options section below)
An example downhole plot can be seen below.
Product Configuration Options
The following parameters can be configured as part of the discing calculation:
Configuration | Options |
|---|---|
Maximum size of a disc | Threshold can be set to any value. Default value is 30mm |
Minimum number of discs required to define a discing region | Threshold can be set to any value. Default value is 3 discs |
Maximum distance between adjacent discing regions to merge into one region | Threshold can be set to any value. Default value is 300mm |
Output Data
The output data is based on identified intervals of discing.
User Data
User data is no necessary for this product.
Data Output
Results from this product is delivered in a batch nature.
Integration of the required technologies into Datarock production is ongoing.
The available CSV files include the following:
ProjectID_HoleID_discing.csv
The CSVs contains the following headers:
File Header | Description |
|---|---|
hole_id | Customer’s Hole ID |
depth_from | Start depth of discing discing region |
depth_to | End depth of identified discing region |
DISC_COUNT | Number of discs in region |
AVG_DISC_SIZE_m | Average length (m) of discs within the region |
AVG_ANGLES | Average angle of the of discs with respect to the core axis |
GOOD_ANGLE_COUNT | Count of angles within a discing region that are within threshold |
avg_abs_diff_angles | The average of the absolute difference between each discs two angles (LHS & RHS) within a region. This is a measure of the consistency of the fracture angles. |
C1 | Confidence metric based the fracture angles. Calculated using proportion of near perpendicular fractures within a discing region. |
C2 | Confidence metric based on the consistency of the fracture angles within a region. Calculated based on the average deviations of the fracture angles with a discing region. |
C3 | Confidence metric based the number of discs within a region. Calculated based on number of discs in a discing region. More discs indicate higher chances of the region being a truly discing region. |
Confidence | Overall confidence metric for a given region |
Product Limitations
Limitations | Comments |
|---|---|
Based on open joints only | Only open joints are measured, meaning the discing has to have developed enough |
Reliance on open joint detections | This product relies on assessment of measured angles of individual open joints/fractures. However, if a region contains intense discing that has broken/collapse as a result of the drilling or core handling, the fracture detection model will not detect each individual fracture. This is likely to result in a missed discing region (or portion of a region). |
Document Version
Version | Date | Author | Rationale |
|---|---|---|---|
1 | 5 April 2024 | S Johnson | Initial release |