Study of Predictive Control Model for Cooling Process of Mark III LNG Bunker

Abstract When loading liquefied natural gas (LNG) onto a dual-fuel LNG container ship fuelled by LNG, there is a considerable temperature difference between LNG and the fuel tank at room temperature. The current solution is to pre-cool the tank with LNG through a spray line but the cooling process, if not correctly handled, can result in excessive cooling rates and Boil-Off Gas (BOG), which can expose the tank to increased temperature stress and gas pressure. Therefore, this paper takes the Mark III fuel tank of a specific type of LNG container ship as the object and realises a real-time predictive control system by writing a UDF (User Defined Function) to simulate and analyse the influence of LNG spray rate on the change of cooling effect, cooling time and cooling cost under the unidirectional LNG spray cooling mode. Compared with the results of the fuel tank gas experiment, the deviation of numerical model simulation results is less than 5%. Under the same cooling rate, the real-time control scheme can achieve a more uniform cooling rate and reduce the total LNG consumption by 25%. With the increase in LNG cooling rate, the cooling time, LNG usage, and the total BOG exhaust volume all decrease; however, the decreased range gradually decreases as well. The results of this paper provide parameters and suggestions for optimising and improving the LNG fuel tank cooling monitoring and control system.