Emerson and Aramco Boost Refinery Efficiency with AI

[Ahmet Ö.]

Emerson has successfully implemented an AI-powered optimization solution for Aramco, one of the leading companies in the global energy and chemical sector. This collaboration began with the integration of Emerson's Aspen Hybrid Models product into Aramco's existing refinery planning system, thereby creating one of the world's largest multi-site and multi-period optimization models.

Aspen Hybrid Models combine first-principles models, deep industry expertise, and industrial artificial intelligence to capture non-linear relationships in refinery yield and quality responses. This significantly increased the accuracy of planning models, providing up to 98.5% yield and quality prediction in critical refinery units.

### Aspen Hybrid Models Use Cases

• Applied in Continuous Catalytic Regeneration (CCR) and Platformer Units.
• Ensures more precise feedstock blends.
• Minimizes gaps between planning and execution.
• Increases margin prediction accuracy across Aramco’s global refinery network.

Work is currently underway to extend this hybrid model approach to Aramco’s hydrocracker units. This expansion will further improve model accuracy and demonstrate the scalability and robustness of this AI-powered optimization strategy.

### Statements from Aramco and Emerson Officials

Ahmad Alkudmani, Director of Aramco's Global Optimizer Department, stated that this implementation marks a significant milestone for the company's AI strategy. He noted that improved model accuracy has led to better planning decisions, reduced manual interventions by engineers, and the discovery of new value across global assets.

Claudio Fayad, CTO of Emerson's Aspen Technology Business Unit, emphasized that they have demonstrated the tangible benefits of combining deep industry knowledge with advanced AI as Aramco progresses in its digital transformation goals.

### Key Advantages Achieved with Aspen Hybrid Models

• 98.5% yield and quality prediction accuracy
• Increased efficiency with optimized feedstock blends
• Reduced discrepancies between planning and execution
• Modeling of complex non-linear relationships in critical units like reactors
• Automated model updates and reduced manual adjustments
• Scalable and robust solution for global refinery operations

Thanks to these models, developed and implemented within Emerson's AspenTech Performance Engineering and Manufacturing and Supply Chain product suites, Aramco can perform highly accurate non-linear optimizations using thousands of simulation cases. This provides a robust and scalable tool for refinery planning worldwide.