Machine learning-assisted log normalization and hydrocarbon correction
Challenge
- Automating hydrocarbon correction and log normalization with AI/ML for efficiency.
- Reducing subjectivity in logs' response for confident decision making.
- Implementing ML automation for faster processing and integrating human expertise for improved accuracy.
- Manual workload drain: Up to 50–70% of petrophysicists' time spent manually correcting logs.
- Decision confidence gap: subjectivity, overcorrection, and mistrust in data leads to lower confidence in decisions.
Solution
Our ML-assisted log quality control (QC) and reconstruction solution is a fully automated and assisted conditioning workflow that makes more data available for all geoscience workflows, reducing uncertainty and rejuvenating legacy data.
“The full cycle of ML training, logs edition and results review has been reduced from 15 days to two days, bringing true efficiency gain for the team.”
Middle Eastern NOC
Results
Automated log pre-processing.
Expert-driven ML.
ML-driven logs normalization and hydrocarbon.
5–10 x acceleration in log conditioning.
70%↑ leveraging petrophysical data.
2–3x resource efficiency.
