Edge-based AI analytics for drilling optimization with 75-95% cost savings versus traditional EDR systems.
Industry: Oil & Gas (Upstream Drilling)
| Attribute | Value |
|---|---|
| Customer | Apache Corporation (NASDAQ: APA, Fortune 500) |
| Location | Permian Basin |
| Historical Wells | ~800 wells loaded for lessons learned |
| Data Rate | Up to 100 Hz |
| Cost Savings | 75-80% vs land-based EDR, 93-95% vs high-end solutions |
| Historian | OSIsoft PI (central) |
| Standards | OPC UA, MQTT, WITSML |
Improve well performance by leveraging multiple information sources and real-time data at Permian Basin operations, with remote access for faster data-driven decision-making.
Specific pain points:
Needed affordable solution to improve well performance
Required aggregation of multiple data sources (Rig PLC, Pason, MD Totco, service companies)
Wanted remote access control for better operational adjustments
Current land-based EDR systems too expensive
High-end solutions with real-time models priced out of reach
Legacy systems lacked interoperability
Impact: Without affordable real-time analytics, drilling optimization decisions relied on incomplete data and delayed insights.
| Location | Component | Capabilities |
|---|---|---|
| Rig | FrameworX Edge | Visualization, alarms, data acquisition, local store & forward, calculations |
| Rig | Local Data Log | RT/Archive interface |
| Rig | Event Detection Engine | Real-time algorithm processing |
| Rig | Models/Analytics | Python-based transformation |
| Office | FrameworX Central | Visualization, model building, real-time applications |
| Office | OSI PI Historian | Central data archive |
| Office | Relational DB | Structured data storage |
| Source | Data Type |
|---|---|
| Rig PLC | High-frequency sensor data |
| Pason | Drilling data acquisition |
| MD Totco | Drilling instrumentation |
| Service Companies | MWD, mud logging, cementing |
| Other Data Streams | Various operational inputs |
| Algorithm | Purpose |
|---|---|
| Rig State Determination | Automatic activity classification |
| Data Quality/Trustworthiness | Validate incoming data |
| Hydraulic Models | Wellbore pressure calculations |
| MSE Calculations | Mechanical Specific Energy analysis |
| Vibration Analysis | Drillstring dynamics |
| Time-to-Depth Transformations | Correlate time and depth domains |
FrameworX capabilities that made this solution possible:
| Capability | Application |
|---|---|
| Multi-Protocol Drivers | Connect real-time devices from multiple vendors |
| Local Store and Model | Store and process data locally at the rig |
| Python Analytics | Transform data with edge analytics |
| AI-Driven Events | Visualize and make decisions based on AI |
| OSIsoft PI Integration | Push key parameters to central historian |
| Open Standards | OPC UA, MQTT, WITSML for interoperability |
| Edge-to-Cloud Scalability | Plug & play connectivity from rig to enterprise |
75-80% Cost Savings — Direct cost per average well versus current land-based EDR systems
93-95% Cost Savings — Compared to other high-end solutions with real-time models and logging visualization
100 Hz Data Rates — High-frequency data capture for detailed analysis
~800 Wells Historical — Loaded for "lessons learned" analysis
Rig-Centric Approach — Proven successful with virtual, as-needed real-time control centers
Structured Insights — Unstructured data captured and transformed for better decision-making
Augmented Legacy Systems — Seamless integration optimizes operations within budget
This case demonstrates edge-based drilling optimization with AI analytics, achieving 75-95% cost savings versus traditional EDR and high-end solutions.