From an economic point of view Underground Coal Gasification (UCG) is a promising technology that can be used to reach coal resources that are difficult or expensive to by conventional mining methods. Furthermore, the process addresses safety concerns, by avoiding the presence of workers underground. An optimal UCG process requires the integration of various scientific fields (chemistry, geochemistry, geomechanics) and the demonstration of limited of environmental impacts. This paper focuses on the mechanical component of the UCG operation and its impact on the surrounding environment in terms of stability and land subsidence. The mechanical components are also considered. Underground mining by coal combustion UCG challenges include the mechanical behavior of the site and of stability of the overburden rock layers. By studying the underground reactor, its inlet and outlet, we confirm the key role played by mechanical damage and thermo-mechanical phenomena are identified. Deformation or collapse above the cavity may cause a collapse in the overlying layers or subsidence at the surface level. These phenomena are highly dependent on the thermoporomechanical behavior of the rock surrounding the cavity (the host rocks). Unlike conventional methods, the UCG technology introduces an additional variable into the physical problem: the high temperatures, which evolve with time and space. In this framework, we performed numerical analyses of the coal site that could be exploited using this method. The numerical results presented in this paper are derived from models based on different assumptions describing a raw geological background. Several 3D (3 dimensional) and 2D (2 dimensional, plane) nonlinear finite element modelings are performed based on two methods. The first assumes a rock medium as a perfect thermo-elastoplastic continuum. In the second, in order to simulate large space scale crack propagation explicitly, we develop a method based upon finite element deactivation. This method is built on a finite element mesh refinement and uses Mohr-Coulomb failure criterion. Based on the analysis of the numerical results, we can highlight two main factors influencing the behavior and the mechanical stability of the overburden, and consequently the UCG process evolution. The first is the size of the cavity. This geometrical parameter, which is common to all types of coal exploitation, is best controlled using the classic exploitation method. We show that in the case of UCG, the shape of the cavity and its evolution over time can be modified considerably by the thermomechanical behavior of the host rocks. The second is the presence of a heat source whose location and intensity evolve over time. Even if thermal diffusivity of the rock is low and only a small distance from the coal reactor is thermally affected, we show that the induced mechanical changes extend significantly in the overburden, and that subsidence can therefore be estimated at the surface. We conclude the integration of the mechanical analysis into a risk analysis process mechanical analysis can be integrated in a thorough risk analysis.