Results of the study on SO2 reduction in a utility boiler furnace by means of furnace sorbent injection are presented in this paper with analysis of major influential parameters. The Ca-based sorbent injection process in pulverized lignite fired boiler furnace with tangentially arranged burners is simulated. In simulations sorbent particles are distributed among the burner tiers, where they are injected together with coal, and also through sorbent injection ports located above the burners. The sorbent reactions model was adapted to be efficiently implemented in the code for CFD simulations of complex processes considering both the calculation time and the results accuracy. The sorbent particles reaction model was simplified with several assumptions to allow for faster calculations and significantly reduce simulation time without loss in calculation precision during the particle tracking in boiler furnace. Two phase gas-particle flow is modeled, with coal and sorbent particles reactions and interactions with gaseous phase. Test-cases based on fuels with different composition and combustion organization were simulated in details, and results showed that significant increase in reduction of SO2 at furnace exit could be achieved by proper sorbent injection. The sorbent injection locations were analyzed with special care to enable maximum SO2 capture in the case-study furnace under investigated conditions. Most of the test-cases with low SO2 capture had one or more of the following problems: intensive particle sintering, low local temperatures (leading to low calcination rates), or bad particles distribution. Significant SO2 retention was possible when the process was organized in such a way that particles were exposed to optimal temperature range, and injected in the furnace zones with high SO2 concentration simultaneously. It was shown that better results can be achieved by injection of sorbent through multiple burner tiers, with SO2 emission reduction efficiency around 60% at the furnace exit in several well optimized test-cases. (C) 2018 Elsevier Ltd. All rights reserved.