Simulating the Future Renewable-Based Power Network : High-Performance Computing for Power Systems Analysis

Martinelli, Samuele and Egea Alvarez, Agusti and Papadopoulos, Panagiotis (2021) Simulating the Future Renewable-Based Power Network : High-Performance Computing for Power Systems Analysis. University of Strathclyde, Glasgow.

[thumbnail of Martinelli-2021-Simulating-the-future-renewable-based-power-network]
Preview
 Text. Filename: Martinelli_2021_Simulating_the_future_renewable_based_power_network.pdf
Final Published Version
License: Strathprints license 1.0
Download (2MB)| Preview

Abstract

The standard simulation method for electric networks is Root Mean Squared (RMS) simulations, which models the grid lines at one fixed frequency, ignoring systems that work at different frequencies (e.g., power electronics). ElectroMagnetic Transient (EMT) simulations use time domain differential equations to model the components of the system. Allowing to capture with great accuracy the results of a grid simulation with components that work at different frequencies. However, EMT simulations take much longer time to simulate compared to RMS simulations. This project investigates different methods of simulating grid networks. MATLAB is used to simulate for RMS and EMT methods, investigating compilers ode23tb and ode45. PSCAD supports only EMT simulations, and two versions are used , PSCAD 5 with compiler GFortran 8.1 and educational license and PSCAD 4.6.3 with compiler GFortran 4.6.1 and professional license. Parallel core computing in PSCAD (PNI) is investigated as well. PNI is a technique that allows to split a big system (from one project) to multiple sub-systems (in more, interconnected, projects). Each sub-system is solved in a separated CPU core.

CORE (COnnecting REpositories)