An Enhanced CAPE-OPEN-Based Digital Twin Platform Architecture for Chemical Processes

Digital Twin (DT) technology enables real-time monitoring, optimization, and decision support for complex chemical processes. Yet, industrial deployment is often limited by heterogeneous interfaces, insufficient computational efficiency, and the closed architecture of conventional commercial simulators. This study proposes an enhanced CAPE-OPEN-based DT platform architecture for chemical processes. By incorporating a global interface management mechanism into the COM-based CAPE-OPEN framework, the communication pathway for high-frequency thermodynamic calculations is streamlined, improving real-time computational performance while maintaining openness and interoperability. The proposed framework is benchmarked against the original CAPE-OPEN framework using two distillation column models and two refinery tower models, with performance evaluated by ideal-gas enthalpy calculation time and overall model convergence time. The enhanced framework achieves a maximum CPU time reduction of 94.79% for ideal-gas enthalpy evaluation and 94.85% for module convergence, demonstrating systematic gains from optimized scheduling and data access. In addition, the proposed platform integrates a five-dimensional DT model with real-time data-driven methodologies to establish a closed-loop workflow encompassing data interaction, high-fidelity model simulation, and application services. Industrial validation is conducted on an S Zorb gasoline adsorption desulfurization unit, where the DT system provides accurate predictions of product properties and key operating parameters and supports stable long-term online operation. Moreover, real-time optimization based on the proposed platform reduces octane number loss under compliant sulfur specifications, confirming the engineering feasibility and industrial applicability of the proposed DT architecture.