A Simple Application for Executing Dynamic Simulations in Parallel on Cluster Machines

Dynamic simulations are a fundamental tool in the analysis of power networks. Typically spanning a time window of a few seconds, dynamic simulations make it possible to obtain the solution of the power system given an initial condition and perturbations applied at a specific time. Since the power system is a Differential-Algebraic System of Equations (DAE), at least two numerical algorithms are involved in this process: numerical integration and equation solving (Newton-Raphson). Therefore, the demand of computing resources is very high, and the simulation times tend to be tenths of minutes. Sometimes, it is necessary to execute multiple dynamic simulations. An example of this are the Transient Stability Studies for System Wide Analysis under high penetration of inverter based resources where different contingencies are tested to assess the stability impact of the interconnection of a new power generation in the system. Running the simulations with a High-Performance Computing framework could diminish the execution time and make much more efficient use of computing resources. This paper proposes a simple framework based on a task queue implementation for Python called Celery.