A methodology that combines the Eulerian/Lagrangian droplet flow model and the droplet impingement erosion model is proposed in this paper to qualitatively predict the distributions of wall thinning locations on the shell wall of feedwater heaters (FWHs). This hybrid computational fluid dynamics model can simulate the three-dimensional distribution of steam flow and the rebound characteristics of droplets within the FWH shell. Coupled with the droplet flow characteristics, an appropriate indicator derived from the droplet impingement model is used to calculate the possible locations of severe wall thinning. The predicted wear sites on the shell wall correspond well with the plant-measured ones. Based on this good agreement, the methodology proposed herein might be used to help construct the monitoring project of wall thickness measurement for FWHs in the power plant, especially in the measuring areas on the shell wall.