Dynamic modeling and predictive control of boil-off gas generation during LNG loading

• Published in: Computers & chemical engineering 2022-04, Vol.160, p.107698, Article 107698
• Main Authors: Shin, Kyeongseok, Son, Sanghwan, Moon, Jiyoung, Jo, Yeonpyeong, Kwon, Joseph Sang-II, Hwang, Sungwon
• Format: Article
• Language: English
• Subjects: Boil-off gas (BOG); Dynamic modeling; LNG tank cool-down process; model predictive control; Pressure of LNG tank
• ISSN: 0098-1354; 1873-4375
• DOI: 10.1016/j.compchemeng.2022.107698

•A high-fidelity model was developed to describe BOG generation during LNG loading.•A computationally efficient model was derived for efficient use of the high-fidelity model.•An MPC system was designed to regulate the pressure and temperature of an LNG tank.•In the MPC, optimal vapor outlet flow and amount of LNG spray were manipulated.•The designed MPC was able to achieve a safe operation of the cool down process. During liquefied natural gas (LNG) loading, excessive boil-off gas (BOG) is generated due to the temperature difference between the LNG and a tank at room temperature. While one way to deal with this challenge is to gradually reduce the tank temperature via an LNG, this cool-down process may also produce excessive BOG, if not properly handled, leading to a sudden pressure change inside the tank. To this end, a model predictive control (MPC) system was formulated in this work to simultaneously regulate the pressure and temperature of the tank by manipulating the vapor outlet flow rate and the amount of LNG spray injected during the cool-down process. Specifically, a high-fidelity model of cool-down process was developed and simplified to a reduced-order model (ROM). Then, the ROM was incorporated into the developed MPC scheme which successfully computed an input profile to drive the pressure and tank temperature to target values.