This paper primarily focuses on simulating the variation of different impurity concentrations during the growth of polysilicon ingots. The simulation prototype employed was a directional solidification crystal growth furnace, and the governing equations of the Ansys Fluent module’s finite analysis software were utilized. The simulation models used in this study included the energy, flow field, phase change (solidification and melting), heat transfer, species transport, and finite volume models. The simulation outcomes were validated using process data from Eversol Technology Co., Ltd., Taiwan. This paper explores the concentration distributions of impurities, such as oxygen and carbon elements, in argon and molten silicon during the general crystal growth process. The simulation results indicate that during the initial 5 h of polysilicon growth, the input of a large amount of argon can be used to eliminate impurities. Moreover, the results suggest that reducing the argon flow rate at the inlet at about 30 h, in the late stage of polysilicon growth, can lead to a more uniform temperature distribution in the polysilicon ingot. To enhance the quality of the grown polysilicon ingot, the impurities were concentrated in the upper part of the ingot. Additionally, the volume of air extracted at the gas outlet was increased over time to decrease the concentrations of impurities in the gas phase in the growth furnace.