Once again about the Steinmetz transformer model and the ongoing subdivision of its leakage inductance

Abstract

Purpose

Despite its well-founded criticism and lack of proper justification under core saturation conditions, the T-equivalent transformer model (Steinmetz scheme) is obviously championing in the literature. This educational paper aims to explain in a simple manner the limitations of the T-model of a low-frequency transformer and critically analyses some attempts to improve it.

Design/methodology/approach

Using a simplified examination of magnetic fluxes in the core and windings and using the modeling in ATPDraw, it is shown that transient transformer models with the indivisible leakage inductance allow circumventing the drawbacks of the T-model.

Findings

The authors show the absence of valid grounds for subdividing the leakage inductance of a transformer between its primary and secondary windings. The connection between the use of individual leakage inductances and inaccurate prediction of inrush current peaks is outlined as an important example.

Practical implications

The presented models can be used either as independent tools or serve as a reference for subsequent developments.

Social implications

Over generations, the habitual transformer T-equivalent is widely used by engineers and Electromagnetic Transients Program experts with no attention to its inadequacy under core saturation conditions. Having studied typical winding configurations, the authors have shown that neither of them has any relation to the T-equivalent.

Originality/value

This educational paper will contribute to the correct understanding of the transients occurring in a transformer under abnormal conditions such as inrush current or ferroresonance events, as well as during an out-of-phase synchronization of step-up generator transformers.

Citation

Zirka, S.E., Moroz, Y.I. and Arturi, C.M. (2022), "Once again about the Steinmetz transformer model and the ongoing subdivision of its leakage inductance", COMPEL - The international journal for computation and mathematics in electrical and electronic engineering, Vol. 41 No. 1, pp. 81-95. https://doi.org/10.1108/COMPEL-06-2021-0190