Heavy-duty centrifugal fans are employed for the transport of dust-laden flows in industrial plants, for cement or steel production, or in energy production of plants and mine ventilation systems. These applications require the disposal of huge amounts of suspended particles, which can lead to a gradual erosion of the machine parts where impacts take place. The wear of the fan components can lead to premature failure of the machine, threatening human safety and reliability of the whole plant. The assessment of the wear severity of the machine, according to the process parameters, can aid the plant owner in scheduling overhaul operations along with the operative life of the machine. Moreover, through a reliable estimation of the wear severity, fan manufacturers can optimize the whole machine design process, from the material selection to the warranty time assessment. This work aims to develop a semi-empirical method capable of estimating the erosion severity of centrifugal fans employed for heavy-duty operations. A mechanistic model which accounts for the erosion on the blade leading edge is proposed. The model is derived by means of an analytical approach and accounts for a number of operating parameters (i.e.,, fan geometry, fan operating point, particle concentration in the flow, fluid properties, and material erosion resistance). A comparison of the theoretical model to the computational fluid dynamic (CFD) simulation results obtained through the use of multiphase particle tracking is also provided to assess the reliability of the present method.