Digital twin and control of an industrial-scale bitumen extraction process

• Published in: Computers & chemical engineering 2024-06, Vol.185, p.108671, Article 108671
• Main Authors: Soesanto, Jansen Fajar, Maciszewski, Bart, Mirmontazeri, Leyli, Romero, Sabrina, Michonski, Mike, Milne, Andrew, Huang, Biao
• Format: Article
• Language: English
• Subjects: Bitumen extraction; Digital twin; Flotation cells; Model predictive control; Oil sands; Primary separation vessel
• ISSN: 0098-1354
• DOI: 10.1016/j.compchemeng.2024.108671

This article presents a digital twin solution for bitumen extraction process required to study process dynamics, control, and optimization of this critical oil sands process. It captures the actual process interconnectivity and scale, enabling the bidirectional data connection between the virtual and physical entities. The proposed model was validated using data from an oil sands extraction facility. Results demonstrate the model well mimics real process behavior in both steady-state and dynamic conditions, with tunable parameters tailored for any further adaption. The model was used to study extraction performance under various production scenarios. It enables analyzing process behavior and testing strategies to optimize bitumen recovery. •The digital twin captures plant-wide interconnectivity of the oil extraction.•Optimization and sensitivity analysis systematically identify a vast set of plant-wide model parameters.•Validating steady-state and dynamic behavior with actual process data ensures high fidelity.•The digital twin showcases its potential to study the process dynamics and innovate in control and optimization.