Applied AI for artificial lift and production
Machine learning is most useful in production operations when it stays close to the physics engineers already trust. These hands-on courses and consulting services apply data analytics and machine learning, with Python and generative-AI assistance, to real artificial lift problems.
Courses
- Data Analytics Workflows for Artificial Lift, Production and Facility Engineers (2 days / 16 hours): gas-lift injection optimization, choke flow-rate estimation, rod-pump dynamometer card classification, flow-pattern prediction and slug-catcher design, liquid loading, virtual flow meters, multiphase flow meter prediction.
- Surveillance, Optimization, Data Analytics and Machine Learning for ESP Engineers (1 day, extendable to 2): ESP state identification, virtual flow meter, choke flow-rate estimation, reservoir productivity assessment for unconventional wells, ESP failure analysis and prediction.
- Surveillance, Optimization, Data Analytics and Machine Learning for Gas Lift Engineers: gas-lift well state identification, virtual flow meters, single-point gas lift, reservoir productivity assessment for unconventional wells.
Taught in person or virtually, using Python notebooks in Google Colab. Client datasets can be worked into the class.
Consulting
Gap analysis of data-science models for artificial lift; virtual flow meters that combine ML with physics-based models; ESP failure prediction; and generative-AI applications for engineering knowledge capture and decision support. See Consulting.
Public notebooks
[CONFIRM whether to link github.com/SPE-PFAC01/ALCE or a separate public repo]