Enhancement of carbon dioxcide (CO2) removal process in liquefied natural gas (LNG) production system

Removal of CO2 from natural gas is currently a global issue, apart from meeting the customer's contract specifications and for successful liquefaction process in any LNG project, it is also a measure for reducing the global CO 2 emission. The aims of this research are to present a comprehensive review for removal of CO 2 from natural gas to meet LNG production specifications and explore the capability of Aspen HYSYS process simulator to predict the CO 2 removal process. A base case of typical CO2 removal process is used to create a steady-state simulation using Aspen HYSYS 7.0 process simulator. Then, the simulation program is developed (Sulfinol process model) to modify the physical, thermodynamics and transport properties of the gas and the process units involved to improve process performance. Next, the constructed model was then validated against the existing plant data, which in turn provide information on potential problem areas within the current simulation process. Moreover, the model was then used to determine the CO 2 removal efficiency, maximize the heavier hydrocarbon recovery and reduce the power consumption at the optimum Sulfinol hybrid solution composition. The best optimum simulation result shows that increasing of CO 2 capturing capacity in the Sulfinol contactor to almost 84 percent. This process also met the LNG product specifications which is 1.69 mole percent of CO 2 in the LNG product stream and the reduction to about 11.14 percent of carbon dioxide slippage in sweet gas stream. In term of economics, this process can safe heat consumption at stripper reboiler up to 18.39 percent and power consumption at pump up to 6.68 percent. For the heavier hydrocarbons recovery, this process can recover to almost 8.89 kgmole per hour. As a conclusion, this research has achieved its objectives which are to improve the carbon dioxide removal process and also to model Sulfinol process model in Aspen HYSYS simulator. It is recommended to run a sensitivity analysis of this model when the feed to AGRU is increased in the case of "bottleneck" conditions