Dynamic risk modelling of a chemical reactor using Aspen Plus / Sharıfah Nur Hadzlin Wan Madehi, Ir. Mohd. Azahar Mohd Ariff, and Noor’aina Abdul Razak

• Main Authors: Wan Madehi, Sharıfah Nur Hadzlin, Mohd Ariff, Mohd. Azahar, Abdul Razak, Noor’aina
• Format: Conference or Workshop Item
• Language: English
• Published: 2020
• Subjects: Chemical elements
• Online Access: https://ir.uitm.edu.my/id/eprint/82514/1/82514.pdf; https://ir.uitm.edu.my/id/eprint/82514/
• Institution: Universiti Teknologi Mara

Risk assessment is an important step in predicting possible threats of hazards. The objectives of this research project are to identify potential risks that could possibly occur to a chemical reactor in a dynamic environment by using Fault Tree Analysis (FTA), and to simulate and evaluate the risks of a chemical reactor in a dynamic environment by using Aspen Plus and Bayesian Network. In a dynamic situation, a conventional risk assessment is lacking in providing the information needed. Aspen Plus is one of the most leading software used in the industry as it could be used to comprehend the condition of the reactor in a faulty condition. Sensitivity analysis is simulated in dynamic condition to observe the behavior of the methanol. The highest mole flow of methanol is produced when the model is deviated to 262 bars. Threats and severity of any faulty conditions such as leakage of reactor or thermal radiation from jet fire could be predicted by using Areal Location of Hazardous Atmosphere (ALOHA). From the leakage of the reactor, a small but dense flame pocket of concentration greater than 43080 ppm is detected. A threat of potentially lethal within 60 seconds is predicted as the thermal radiation from jet fire is simulated. From the point, the area of coverage for the threat is 1.38 yards off centerline and 1.72 yards downwind. A second-degree pain is possible at off centerline of 1.21 yards and 11.2 yards downwind. Fault tree analysis could be constructed based on the model simulations as the faults that caused the risk has been determined. A Bayesian Network could also be constructed as the threats from the risk also has been determined.