The Rock Physics license has received several improvements for 2024.3
A few highlights:
With this software release, we're focused on making it easier to start RokDoc projects and execute multi-well workflows with fewer clicks. This is done through multi-well operations and by providing automated QCs so that users don't need to manually build plots to assess results. Overall, the time spent in the software is more effective and users can reach and incorporate deeper understanding of their data faster than ever.Project Startup Efficiency
- New options in the Multi-Well TVDss to TWT Conversion Creator
Rapidly import existing time-depth relationships for all wells with TWT logs or use checkshot sets. This will allows users to more easily transfer data into RokDoc without re-calculating and re-verifying already-established time-depth relationships
- Removed requirement for ML checkshot for offshore wells
- Checkshot set editing by interval velocity
Updated Rock Physics Model Workflows
- Parameters collapsible by default
With the recent refresh to the RPM interfaces, it's easier than ever to navigate rock physics workflows. Calibration parameters can be quickly documented with screed capture or by highlighting the text in the summary window and the RPM description can be positioned to the side while you work, for quick reference. Now, the rock physics models have certain parameters collapsed by default, so that the most important inputs are adjusted first. This should make working with complex models more straightforward. This functionality is available for external interface users as well, so custom RPMs can be similarly structured for simplicity and accessibility.
- Multi-well Forward and Reverse Rock Physics Modeling
To save time on multi-well applications with faster modeling turnaround and testing, users can now apply rock physics models in forward (predict elastic properties) and reverse (predict petrophysical properties) across multiple wells simultaneously. In addition, reverse modeling can be run on up to 3 properties simultaneously (e.g. porosity, saturation, and mineralogy).
- Deep QI: RPML Enhancements
- Addition of Xu-Payne Model
- Bulk and Shear Modulus Custom QC Plots
The RPML library now covers a larger family of theoretical models (Inclusion/DEM) and provides more QC tools for assessing the predicted classification and fundamental effectiveness of the model fits.
- Deep QI: ML Training Randomized Search Algorithms
A new training function algorithm has been added to produce faster training for machine learning applications. Users can track the progress and accuracy of the training with automated QCs for the following algorithms:
-
- Random Forest (RF) Classification
- RF Regression
- eXtreme Gradient Boost (XGB) Regression
- XGB Multi Regression
We look forward to continuing to hear your feedback for future RokDoc releases!
