Losses in Medium-Voltage Megawatt-Rated Direct AC/AC Power Electronics Converters

Direct ac/ac topologies for ac-to-ac power conversion benefit from the absence of dc-link capacitors, and therefore, are highly reliable and have low cost as compared to the traditional voltage-source inverter (VSI)-based topologies. This paper deals with one of the more important tradeoffs consider...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE transactions on power electronics 2015-07, Vol.30 (7), p.3553-3562
Hauptverfasser: Moghe, Rohit, Kandula, Rajendra P., Iyer, Amrit, Divan, Deepak
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Direct ac/ac topologies for ac-to-ac power conversion benefit from the absence of dc-link capacitors, and therefore, are highly reliable and have low cost as compared to the traditional voltage-source inverter (VSI)-based topologies. This paper deals with one of the more important tradeoffs considered in designing highly efficient converters: Losses. It is shown in this paper that the direct ac/ac converters have an inherently higher efficiency than their VSI-based back-to-back counterparts due to a dramatic reduction in switching losses (nearly 60%). Further, this paper compares the performance of three different device types (SiC MOSFETs, hybrid Si IGBT/SiC diode, and Si IGBTs) using wide-range device characterization that help to create detailed loss models. It is conjectured that traditional datasheets lack the level of detail needed for computing losses in direct ac/ac converters, and the availability of a multivalue voltage, current, and temperature-based loss profile is advocated. Using the obtained loss models, a comparison is drawn between the considered devices through simulations when operated in a 13-kV/1-MW direct ac/ac power flow controller, the controllable network transformer (CNT). The same loss-models are also used to compute losses in an experimental prototype of a 720-V, 10-kVA CNT and the results are compared with direct efficiency measurements. A similar computation is carried out for another experimental prototype at a 6.7-kV, 400-kVA, three-level, paralleled CNT. These experimental tests are used to confirm the validity of the analytical results presented in this paper.
ISSN:0885-8993
1941-0107
DOI:10.1109/TPEL.2014.2350003