Petrophysics and rock physics consulting services
Work Examples: All images are created from public or released data.
Log Editing is a time consuming but critical first step in petrophysical interpretation. This Log on the right used Outlier Detection, Multi-Linear Regression, and Neural Networks to automatically generate a corrected Bulk Density curve. Original log data is preserved when data is high quality.
It is desirable to use multiple standard interpretation methods when analyzing conventional reservoirs. In this example we compare the results from quick look, shaly sand, and deterministic models for porosity and water saturation determination. Multiple equations were involved in estimating the permeabilities of the porous sand reservoirs.
Unconventional reservoirs require advanced techniques for accurate interpretations. In this case we used statistical and machine learning methods to generate a quality model of a complex formation. At FEJ Petrophysical we have over 30 years of experience with unconventional resources.
Examples: Cross Sections, Machine Learning, and Rock Physics
Structural Cross Section showing productive Wilcox sands in the Fandango field. The water saturation and porosity curves were generated using Shaly Sand analysis.
Random Forest clustering is used to compare a 4 facies model to a 6 facies model with the associated probabilities. Facies classification is just one of the machine learning tools commonly used in petrophysical interpretation today. Synthetic curve generation is the other popular machine learning package used in log analysis.
Rock Physics modeling is an interactive process between the petrophysics and the rock physics. The petrophysical interpretation is essential to building an accurate rock physics model. Volume of dry clay is often the critical component that needs to be adjusted to result in a quality final product.