It's not always clear how AI and tech fit into Oil & Gas. This living list of real tangible use cases is here to provide both clarity and inspiration.
| # | Use Case | Segment |
|---|---|---|
LLM | ||
| 1 | LLMs to process JIB files of any format and transform them into structured, usable outputs. | Upstream |
| 2 | LLMs for well knowledge management, capturing notes and history data, then enabling users to query and learn from that information in a ChatGPT-like interface. | Upstream & Midstream |
| 3 | LLMs to categorize unstructured notes/text, such as downtime codes, equipment IDs, or event tags. | Upstream & Midstream |
| 4 | LLMs to analyze accounting data, automatically generating PDF insight reports and distributing them to teams. | Upstream & Midstream |
| 5 | Automated RRC filings, using AI to parse data from PDFs, databases, and other sources to generate filings automatically. | Upstream & Midstream |
Machine Learning (ML) | ||
| 6 | ML models to approximate physics simulations, enabling thousands of scenarios to run significantly faster. | Upstream & Midstream |
| 7 | ML for well design optimization, improving completion design based on historical and simulated data. | Upstream |
| 8 | ML to predict well liquid loading 30+ days in advance, allowing proactive artificial lift installations and reducing downtime. | Upstream |
| 9 | Time-series anomaly detection for spotting outliers in daily production, meter readings, and real-time SCADA data. | Upstream & Midstream |
| 10 | ML models to predict failure event types, notifying maintenance teams in advance. | Upstream & Midstream |
| 11 | Edge-based ML for leak detection, continuously evaluating sensor data and automatically notifying the control room when anomalies are detected. | Upstream & Midstream |
Constraint Optimization | ||
| 12 | Constraint optimization of offshore gas lift, reallocating gas during downtime to minimize deferred production. | Upstream |
| 13 | Constraint optimization for midstream fluid flow, maximizing margins by accounting for royalty costs, contractual obligations, SWD capacity constraints, and operating expenses. | Midstream |
Automation | ||
| 14 | Automated time-series segmentation to identify unique well/meter/field behaviors, alerting teams to flow regime changes or potential issues. | Upstream & Midstream |
| 15 | Automated FBHP calculations at scale, enabling rapid analysis across 1,000+ wells while reducing manual effort and generating robust datasets for history matching. | Upstream |
| 16 | Automated production allocation, reducing multi-team effort to a single click while seamlessly handling edge cases and exceptions. | Midstream |
| 17 | Automatically generate workover procedures that align with internal standards and templates, incorporating well history. | Upstream & Midstream |
| 18 | Automated Texas RRC regulatory filings by extracting PDFs, validating well data, and generating compliant submissions. | Upstream |
AI Agent | ||
| 19 | Agents to capture project updates from field personnel, automatically populating Excel sheets and other systems, running continuously in the background. | Upstream & Midstream |
| 20 | Agents to dynamically edit BI reports (Power BI / Spotfire) based on user requests. | Upstream & Midstream |
| 21 | Agent to auto generate A&D decks that tend to be very manual and repetitive. | Upstream & Midstream |
| 22 | Internal engineering agent that surfaces best practices from senior engineers to guide design decisions. | Upstream & Midstream |