Digital Twins for Oil and Gas Production Systems

Agenda

• Challenges in oil and gas production and process plants.
• How digital twins address these challenges.
• Success stories of digital twins in the industry.
• Introduction to Simscape.
• Demonstration on creating a digital twin.

Introduction to MathWorks

• MathWorks software utilized across various industries:
  • Aerospace and Defense
  • Automotive
  • Energy and Industrial
  • Financial Services
  • Medical and Health
  • Electronics and Semiconductor
  • Communications (e.g., 5G)
• Focus on reducing carbon emissions and optimizing operations.
• Importance of innovation in engineering and science to derive business value.

Challenges in Oil and Gas Industry

• High-level challenges:
  • Regulatory and competitive pressures.
  • Achieving growth in both topline and bottomline.
• Operational challenges:
  • Monitoring vs. optimizing operations.
  • Adding sensors does not directly correlate with process KPIs (e.g., quality, yield).
  • Lag in understanding process impacts due to lab testing.
  • Complex physical behaviors and models that are difficult to interpret.

Digital Twins Overview

• Digital twins allow for smarter and connected operational systems.
• Benefits include:
  • Visibility into operations.
  • Predictive maintenance capabilities.
  • Simulation of various scenarios and optimization of operations.
• Use of statistical approaches (e.g., PCA) and AI to analyze system-level KPIs.
• Importance of combining first principles models and data-driven techniques.

Building Digital Twins

• Utilize operational data from system operators.
• Apply physical domains and conservation laws to enhance models.
• Combine physics-based techniques with data-driven approaches for hybrid digital twins.
• Aim to achieve:
  • Operations optimization.
  • Predictive maintenance.
  • Fault detection and diagnostics.

Case Study: RAG Austria

• RAG Austria is a European gas storage operator.
• They created a digital twin of their adsorption dehydration units using Simscape.
• Simulated performance against plant data to optimize operations.
• Key features of Simscape used:
  • Custom unit operations and libraries.
  • Solver capabilities for handling significant changes in input parameters.

Introduction to Simscape

• Simscape is a 1D physical modeling tool integrated with Simulink.
• Provides:
  • Causal modeling style.
  • Foundation library and custom components.
  • Support for various physical domains (e.g., electrical, thermal, hydraulic).
• Integration with MATLAB for data processing and control design.

Demonstration Overview

• The demonstration focuses on modeling the adsorption process using Simscape.
• Key features demonstrated include:
  • Model calibration using measurement data.
  • System-level optimization within the modeling environment.
  • Use of custom components for specific processes.

Custom Components in Simscape

• Creating custom components using Simscape language (SSC).
• Process to develop and integrate custom models with existing Simscape libraries.
• Example: Modeling the absorption process to dry wet gas using silica gel.

Modeling Approach

1. Define physical nodes and parameters for the custom component.
2. Incorporate internal variables for the absorption process.
3. Create a physical network to represent the complete absorption system.
4. Automate element creation for scalability.

Simulation and Data Analysis

• Use of Simscape logging for tracking variables during simulation.
• Options for performing parameter estimation and calibration.
• Fast Restart mode for efficient simulation cycles.

Conclusion

• Simscape allows for detailed modeling of complex processes in the oil and gas industry.
• Provides capabilities for custom modeling, optimization, and integration with existing systems.
• Resources for further learning and development are available through MathWorks.

Additional Resources

• Training programs and on-ramps for new users.
• Consulting services for project development and optimization assessment.