Product descriptions
      Back to home
      1. Datarock Knowledge Base
      2. Product descriptions

      Discing Analysis

      This document details Datarock’s Discing Analysis product.

      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

      We have two key methods for identifying discing regions:

      • Fracture angle analysis: This method assesses the angle of individual fractures to identify regions of medium to thick discs.
      • Crushed/thin discing detection: This method identifies regions of discing where individual discs cannot be assessed and are instead detected as a whole region.

      The diagram below provides an overview of the process.

      Fracture Angle Analysis

      1. Take the output of the fracture detection model, “simple” classes only

      2. Run a segmentation model to extract the fracture profile

      3. Measure the angle of each fracture with respect to the core axis

      4. Identify potential discs by:

        1. Measuring the distance between each fracture, referred to as the disc size

        2. Checking the fractures on both sides of the disc have angles that are similar and near perpendicular to the core axis

      Crushed/Thin Discing Detection

      1. Run an object detection model to identify regions of discing that won't be identified using the angle based method

      Generate Discing Regions

      1. Merge two methods and deal with overlapping detections
      2. Find discing regions by joining adjacent discs
      3. Merge ("lump") nearby discing regions if the depth between them is within the defined threshold (see Product Configuration Options section below)

      An example downhole plot can be seen below.

      downhole-1

      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 not required 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:

      • ProjectName_HoleID_discing.csv

      The CSVs contains the following headers: 

      File Header 

      Description 

      hole_id

      Customer’s Hole ID

      depth_from_m

      Start depth of discing discing region

      depth_to_m

      End depth of identified discing region
      disc_count

      Number of discs in region

      ave_disc_size_m

      Average length (m) of discs within the region

      ave_disc_angle

      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.
      broken_confidence_avg

      Average confidence score of the Crushed/Thin Discing Detection method

      merged_with_broken

      Yes or No response based on whether the two different methods of analysis are used for the given interval.

      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.

      C4

      Confidence of predictions based on the Crushed/Thin Discing Detection method.

      Confidence

      Overall confidence metric for a given region
       

      Product Limitations

       

      Limitations 

      Comments 

      Reliance on open joint detections (angle method) for disc counts

      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, we rely on detection of these regions as a whole, thus limiting the ability to provided detailed stats on the discs. 
       

      Document Version

       

      Version

      Date 

      Author 

      Rationale

      1

      5 April 2024

      S Johnson

      Initial release

      2

      18 October 2024

      S Johnson

      Updated with Crushed/Thin Discing Detection