The objective of this paper is to optimize the thermal performance of a fractionation unit within a liquefied natural gas (LNG) facility. Typical fractionation units in an LNG facility consisting of three distillation columns, namely de-ethanizer, de-propanizer and de-butanizer were used in this study. A hierarchical approach is developed to optimizing the system. In this approach, increasing levels of model complexity are used and various thermal targets are set and implemented. The column targeting tool available within the simulation package of Aspen Plus software was used to optimize a fractionation unit in an LNG facility. First, integrated thermal analysis was used in identifying design targets for improvements in energy consumption and efficiency. The column targeting tool is used in the design of distillation columns by setting targets to reduce utility cost, improve energy efficiency, reduce capital investment and facilitate column debottlenecking. Starting from a short-cut distillation design calculation using the DSTWU method which is based on the well-known Fenske-Underwood-Gilliland correlations, the minimum and actual reflux ratios, minimum and actual number of stages, optimum feed location and condenser and reboiler duties were estimated. These estimates were used as starting points in the rigorous fractionation column design method RADFRAC available in Aspen Plus . The column Grand Composite Curve (CGCC) for each column was generated to give an insight of the actual operation and guide the optimization process. Starting with appropriate feed placement, the CGCC will show the scope for reflux ratio modification by increasing the number of stages. Feed would be either preheated or precooled due to the availability of sharp enthalpy change in the condenser or reboiler side. Finally, the scope for a side condensing or side reboiling can be identified from the area beneath or above the CGCC pinch point.