Challenge

Deploying safer technologies for well control is a major focus of the offshore drilling industry. One of the key enabling technologies is managed pressure drilling (MPD), which introduces a closed-loop mud pump system to actively control pressure and detect kicks and losses during drilling.

MPD systems require control units for safety-critical hydraulic barriers that must remain functional under normal operation, degraded conditions and component failures. They must be continuously monitored during live drilling operations.

At the 2026 System Modeler Conference, Robert Prince-Wright and Konstantinos Antoniadis of Berkeley & Imperial demonstrated how they developed an operational digital-twin solution for MPD hydraulic control units, using System Modeler as the core physics-based simulation engine and integrating it with real equipment through industrial communication standards.

The challenges were to:

• Verify and validate complex hydraulic functions and failure behaviors
• Support systematic fault-injection and failure-mode analysis
• Visualize hydraulics and mechatronics as 2D animations
• Ensure the simulation could run in real time
• Synchronize digital-twin states with live sensor data from the rig
• Deploy multiple twin instances securely in operational environments
• Provide actionable diagnostics and alerts to operators in seconds
• Aggregate, analyze and visualize fleet-wide data

Solution

Berkeley & Imperial built a digital-twin technology stack centered around a System Modeler simulation of the MPD hydraulic control system.

A physics-based, computationally efficient hydraulic model was created in System Modeler and verified against system-integration test results. The model was designed to support real-time execution, robust initialization from steady-state conditions and stable behavior in both nominal and fault scenarios.

The digital-twin workflow combines:

• System Modeler for real-time simulation of the hydraulic and control behavior
• Wolfram Language as the orchestration, analytics and automation layer
• Structured fault-injection and reliability studies driven from notebooks
• Secure live data exchange using OPC UA

The team developed two operational modes.

In simulation mode, a standalone System Modeler simulation is used for design studies, optimization, sensitivity analysis, operator training and real-time diagnostics.

In sentry mode, a modified System Modeler simulation is synchronized with live equipment. Valve positions, pump speeds and pressures are streamed into the model via an embedded Wolfram OPC UA client, and the digital twin continuously computes the expected physical behavior of the system. Deviations between predicted and measured behavior are detected and alarmed in real time, both at the rig site and in a remote operations center.

To enable industrial deployment, the System Modeler simulation is exported as an executable and deployed in containerized environments, allowing multiple digital-twin instances to run in parallel for different wells or rigs.

Benefits

By using System Modeler as the core digital-twin engine, the Berkeley & Imperial team achieved safer operations, reduced risk and sustainable drilling, with:

• Real-time, physics-based digital twins for safety-critical hydraulic control systems
• Automated verification workflows, including fault-injection and failure-mode analysis
• Continuous operational validation by comparing live sensor data with digital-twin predictions
• Rapid anomaly detection and alerting for drilling operations
• Scalable deployment of multiple synchronized digital twins through containerized execution

This project moves System Modeler beyond offline simulation to become the foundation of operational digital twins for real-time monitoring and safety assurance in MPD systems.