(624j) Real-Time Data Estimation Method for LNG Terminal Using Conventional Process Simulator | AIChE

(624j) Real-Time Data Estimation Method for LNG Terminal Using Conventional Process Simulator


Lee, S. - Presenter, Seoul National University
Lee, C. - Presenter, School of Chemical and Biological Engineering, Seoul National University
Cho, S. - Presenter, Seoul National University
Kim, J. - Presenter, Seoul National University
Han, C. - Presenter, Seoul National University
Lim, Y. - Presenter, Seoul National University

For analyzing the plant status exactly, it is necessary to get enough information from plant facilities by physical sensors such as flowmeters and thermocouples. But if every sensor were installed, the cost problem becomes serious therefore gathering sufficient data for analysis is unrealistic situation. Usuallly process simulation is an possible answer to solve this problem, thus many conventional simulators such as ASPEN PLUS, HYSYS and Pro/II offer various mathematical process model and with these models and sensored data, the unmeasured information can be calculated. Nevertheless the process simulation derives useful data, it is meaningless in field operation because the simulation couldn't be real-time.

LNG terminal which takes a part of storage and gasifying LNG have suffered from this issue. Especially on BOG(Boil-off Gas) estimation of storage tank inlet flow, real-time data about composition, flow rate, temperature and pressure are required to figure out exact amount of BOG flow rate though only temperature from a few thermocouple in the pipeline and pressure from the manometer inside the storage tank are available in real-time.

In this research, focusing on LNG terminal inlet pipeline, a simulation model that satisfies the steady state case's process data was built at first. After building a model, the model is validated in various cases - unloading, no-ship unloading and recirculation. Finally the model is solved during the sensored data was used as constraints of the dynamic simulation model and through this, BOG flow rate in the inlet pipeline which is the most important was estimated in real-time.