In generating a column grand composite curve (CGCC) the enthalpy values calculated by top-down and bottom-up approaches are different at the feed stage. An insight analysis of material and energy balances indicates that these differences are due to the feed stage treated as a stripping stage in the top-down approach and as a rectifying stage in the bottom-up approach. To consider the practical feed process and overcome the shortcomings in the existing CGCC generation approaches, a modified approach is proposed, dividing the column into three sections: the rectifying section (from the 1st to the f-2th stage, the condenser being the 1st stage), the feeding section (the f-1th and the fth stage) and the stripping section (from the f+1th to the nth stage, the reboiler being the nth stage). The modification is mainly focused on the feeding section stages. An indicator, the reflux energy-saving potential rate, is defined and used to evaluate these approaches. A benzene-toluene column is investigated to demonstrate the performance of the proposed approach, and a comparison is made with Aspen Plus Column Targeting. The results show that the modified approach is able to eliminate the ambiguity at the feed stage. Compared to Aspen Plus Column Targeting and the existing approaches, the modified approach could give better results.