Synchronous machine modeling precision and efficiency in electromagnetic transients

The main objective of this dissertation is the establishment of more efficient and more precise synchronous machine modeling approaches and solution algorithms for the computation of electromagnetic transients (EMT). Numerical integration time step size is a key factor in both aspects. The capability to use larger time steps in EMT-type simulation methods also contributes to the extension of such methods into the efficient simulation of electromechanical transients.;In this thesis, four new models are proposed in order to improve the precision of the classical dq0 model while maintaining its efficiency. Three of them use the classical dq0 model with increased accuracy. The most accurate models are: dq0 with internal intermediate time step usage, phase-domain and voltage behind reactance. Efficiency is maintained by restricting the accurate model usage to the transient intervals where the precision of the classical dq0 formulation decreases. This approach provides accuracy while maintaining classical dq0 computational speed. However, these three models are designed for typical transient stability cases and their performance deteriorates when the accurate model usage is needed for a large portion of the complete simulation interval. The last forth model proposed in this thesis is obtained by applying Park’s transformation to the discretized equations of the phase-domain model. This model maintains the precision of the phase-domain model and eliminates its computational inefficiencies through a constant admittance matrix. Unlike the first three models, its efficiency does not change with simulated system and phenomenon. Precision and efficiency assessment studies demonstrate that this model is superior in both aspects and should be chosen for the computation of both electromagnetic and electromechanical transients in the same computational framework.;The models proposed in this thesis are compared for practical cases and conditions. Precision analysis is performed within the accuracy constraints of the surrounding network and numerical efficiency assessment analysis accounts for the utilized sparse matrix solver and refactorization scheme. Hence, this work also contributes to better assessment of both numerical precision and efficiency for researched machine models in this thesis and in the recent literature.;This thesis also proposes two new synchronous machine representations for Modified-Augmented-Nodal-Analysis (MANA) formulation. In the first formulation, a machine Thevenin equivalent equation is inserted directly into the main network equations (MNE) using MANA. The second representation is proposed for phase-domain and voltage behind reactance models. In this representation all machine equations are inserted into the MNEs, thus eliminating the requirement of interfacing circuits. Although these formulations do not improve simulation speed, they demonstrate the modeling flexibility achievable through MANA and allow to verify performance hypothesis based on partial factorization.