An open-source approach to accelerating power system dynamic simulation

Summary form only given. Power systems are inherently dynamic machines, but the majority of real-time operational tools are based on a static estimation of inherently dynamic states. Such a discrepancy is primarily due to the fact that the software tools for simulating power system dynamics are not able to keep up with real-time requirements for large-scale systems. Advanced computing holds the promise for significantly accelerating dynamic simulation. There are various options in utilizing advanced computing hardware and programming models for the purpose of parallelizing dynamic simulation. But the barrier in moving from today´s sequential software tools to parallel platforms is the lack of a readily accessible and usable parallel numerical library that would expose parallelism in dynamic simulation problems but hide parallel computing complexity. Several generic numerical libraries for parallel computing exist, but using them for developing power grid simulation software tools is time consuming and thus hinders the application and adoption of parallel computing for power system simulation. This presentation explores the development of fast dynamic simulation using an open-source numerical library - GridPACKTM - that is specifically tailored for power system applications. The goal of GridPACKTM is to provide developers with a comprehensive set of modules that can substantially reduce the complexity of writing software for parallel computers while still providing efficient and scalable software solutions. GridPACK^TM leverages existing numerical libraries and provides access to a wide range of mathematical solvers in those libraries. The approach of dynamic simulation powered by GridPACKTM is presented and contrasted with other approaches to demonstrate the efficiency in software development and the speedup of performance while maintaining the accuracy of simulation. Such an open-source approach facilitates integration of mathematics and models and enab- es interoperability among software modules as opposed to the traditional approach of vertically integrated software infrastructure. It also has the benefit of sharing achievements and lessons learned and thus accelerating innovations and reducing software validation and verification costs.