Power system applications are computationally intense, and therefore there is a need to develop new methods for performing these calculations. This study proposes the use of numerical methods based on the Krylov subspace methodology on four areas of power systems: the power flow problem, the dynamic simulation, the trajectory sensitivity analysis and the model reduction. Krylov subspace techniques are tested and compared with traditional approaches. Additionally, a physical interpretation of this methodology is investigated, and links with well-established concepts in power systems are attempted. |