Aerosol agglomerates are generated artificially for industrial use and unintentionally as air pollutants. Aerosol models are needed in efforts to control and/or minimize particle formation. The sectional method has been the most widely used method for solving the aerosol population balance equation. The fixed-sectional method suffers from numerical diffusion when dealing with surface growth. The moving-sectional method introduces much less numerical diffusion, but is not easily applied to spatially inhomogeneous phenomena. In this paper, we present a novel fixed-sectional method that can significantly reduce numerical diffusion. It is compared to the moving-sectional method and a good agreement is obtained. The error caused by numerical diffusion was found to increase with the ratio of characteristic coagulation time over the characteristic surface growth time. Incorporating the new methodology for solving surface growth, we develop an accurate and time-efficient sectional model for the formation and growth of aerosol agglomerates. Simultaneous nucleation, surface growth, coagulation, sintering, and primary particle obliteration by surface growth are incorporated in the model. The "representative size" method is used, and the numbers of agglomerates and primary particles in each section are selected as the two variables representing size distribution and structure of particles. Our model is used to simulate the titania particle synthesis in a premixed flame. The effect of primary particle obliteration due to surface growth is shown to be significant when surface growth is the main particle growth mechanism. (C) 2004 Published by Elsevier Ltd.