Here, an analytical model is proposed to solve in two dimensions the transport equations of the minority carriers, using the method of separation of variables. The present approach considers that the solar cell is composed, in addition to emitter and base regions, of a non-uniformly doped thin region at the back cell to improve the device output parameters. The model is used to investigate the influence of built-in electric field, grain size and recombination velocities (Sgb and Sb for the grain boundary and back surface respectively) on the distribution of excess carriers and the consequent photovoltaic characteristics. The results showed that, as compared to a typical n+p structure, the addition of a p+ rear surface field region enhances the solar cell's output characteristics under the AM1.5 spectrum. An optimum increase in conversion efficiency, open circuit voltage and photocurrent density were found to be 7.2% (from 14% to 15.02%), 6.4% and 5%, respectively. This demonstrates the potential of BSF cell designs to meaningfully improve commercial polycrystalline silicon solar cell's performance. Additional results indicate that higher performance parameters result from increasing grain size and decreasing grain boundary recombination velocity, and that only a modest electric field is sufficient to eliminate the impact of surface recombination velocity for values less or equal to approximately 5.10 3 cm.s -1 . Besides, to validate our approach, the values obtained for photovoltaic quantities were compared with other results reported in literature. A good agreement is found.