The spPillar Stability Experiment (APSE) was conducted to study the rock mass response in a heated rock pillar between two large boreholes. This paper summarizes the back calculations of the APSE using a two-dimensional (2D) fracture propagation code FRACOD. To be able to model all the loading phases of the APSE, including the thermal loading, the code was improved in several ways. A sequential excavation function was developed to model promptly the stepwise changing loading geometry. Prior to the mod-elling, short-term compressive strength test models were set up aiming to reproduce the stress-strain behaviour observed for the sp diorite in laboratory. These models simulate both the axial and lateral strains of radial-controlled laboratory tests. The volumetric strain was calculated from the simulations and compared with the laboratory results. The pillar models include vertical and horizontal 2D models from where the stress in the pillar wall was investigated. The vertical model assesses the stability of the experimental rock volume and suggests the resultant stress below the tunnel floor in the pillar area. The horizontal model considers cross-sections of the pillar between the two large boreholes. The horizontal model is used to simulate the evolution of the stress in the rock mass during the excavation of the boreholes and during and the heating phase to give an estimation of the spalling strength. The modelling results suggest that the excavation-induced stresses will cause slight fracturing in the pillar walls, if the strength of the APSE pillar is set to about 123 MPa. Fracture propagation driven by thermal loading leads to minor spalling. The thermal evolution, elastic behaviour and brittle failure observed in the experiment are well reflected by the models. The spPillar Stability Experiment (APSE) was conducted to study the rock mass response in a heated pillar between two large boreholes. This paper summarizes the back calculations of the APSE using a two-dimensional (2D) fracture propagation code FRACOD. To be able to model all the loading phases of the APSE, including the thermal loading, the code was improved in several ways. A sequential excavation function was developed to model promptly the stepwise changing loading geometry. Prior to the mod- elling, short- These models simulate both the axial and lateral strains of radial-controlled laboratory tests. The volumetric strain was calculated from the simulations and compared with the laboratory results. The pillar models include vertical and horizontal 2D models from where the stress in the pillar wall was investigated. The vertical mode l assesses the stability of the experimental rock volume and suggests the resultant stress below the tunnel floor in the pillar area. The horizontal model considers cross-sections of the pillar between the two large boreholes. The horizontal model is used to simulate the evolution of the stress in the rock mass during the excavation of the boreholes and during and the heating phase to give an estimation of the spalling strength. The modelling results suggest that the excavation-induced stresses will cause slight fracturing in the pillar walls, if the strength of the The thermal evolution, elastic behavior and brittle failure observed in the experiment are well reflected by the models.