Avalanches pose significant threats to human life, infrastructure, and the environment, making accurate prediction and mitigation strategies essential. Avalanches are modelled using the shallow-water equations and the friction law for granular materials flowing down an inclined plane. The resulting equations are known as depth-averaged avalanche equations. This work implements the depth-averaged avalanche equations in ExaHyPE 2, a high-performance computing framework for simulating hyperbolic PDEs. A preprocessing routine is also implemented to facilitate the calculation of complex derivative terms in the equations. The validation of the solver application is mainly done by simulating the flow of a circular patch of masonry sand down a rectangular plane inclined at an angle of 35 degrees. The finite volume method with Rusanov flux and adaptive time stepping is used for the simulations. A comparison between the flow of masonry sand and carborundum particles is done to qualitatively validate the dependence of friction on material parameters. The results of the validation are found to be in agreement with the physics of the governing equations. The availability of the static mesh refinement feature in ExaHyPE 2 is demonstrated by refining one-half of the domain while coarsening the other half. The solver application is a high-performance computing solution that provides detailed insights into avalanche dynamics.