Underground coal gasification (UCG) is the in-situ gasification process carried on in un-mineable coal seams using injection and production wells drilled from the surface, which enables the coal to be converted into combustible gas. UCG process may provide a secure energy supply and reduce green house gas emissions and is capable of producing commercial quantities of gas to be used as a chemical feedstock or as fuel for power generation. As gasification proceeds, an underground cavity is formed. The volume of the cavity increases progressively with coal spalling from the roof and getting consumed. As underground coal gasification cavities are of irregular three-dimensional shapes, computational fluid dynamics studies (CFD) are essential in order to understand the complex flow patterns involved. The characterization of the non-ideal flow patterns in UCG is an important aspect as it is likely to influence the process performance significantly. The main objective of this work is to understand the velocity distribution and perform residence time distribution (RTD) studies in the UCG cavity. Based on the RTD studies, the actual UCG cavity at different times has been modeled as a simplified network of ideal reactors, which may offer a computationally less expensive and easier option to determine UCG process performance at any given time.