Supervisor:
Claus Leth Bak

Co-Supervisor:
Huai Wang

Assessment Committee:
Sanjay Chaudhary(Chair)
Professor Joachim Holbøll, DTU
Professor Yurly Serdyuk, Chalmers

Moderator:
Zhe Chen

Abstract:

Modern power systems are undergoing a major transformation by integrating renewable energy sources and power electronics technology. As the core components in power conversion and transmission, high-voltage power assets and power electronics converters are now required to achieve maximum power density, facing more severe electrothermal stresses than ever. These challenges are intensified by power electronics switching impulses with fast rise times and higher frequencies, which impose stringent demands on system performance. Such operating conditions pose serious risks to electrical insulation systems, often the most fragile components within power assets. The cumulative stress from these conditions can significantly reduce the endurance and lifetime of insulation systems. As driven by the actual situations, it is essential to investigate the frequency-related dielectric/impedance properties and the relevant basic theory of electrical insulation systems, aiming to enhance insulation system design and support prognostics and health management (PHM) of power assets under multifrequency conditions. Therefore, this Ph.D. project mainly focuses on the underlying dielectric physics mechanism analysis and modeling characterization of broadband impedance properties, as well as the insulation resilience response to stresses, with a specific focus on high-voltage polymeric insulation systems in multifrequency power electronics applications.

Firstly, this work comprehensively investigates the impacts of various transient temperature stresses (below/above glass transition temperature, Tg) on frequency-domain dielectric properties and their corresponding underlying multi-timescale dynamics response mechanisms for epoxy resin insulation materials in multifrequency power electronics applications. Four different dielectric analysis techniques, including complex permittivity, dielectric loss tangent, complex conductivity, and electric modulus, are employed to reveal the significant influence behaviors brought by a glass transition process. Additionally, the nature of conductance and relaxation polarization behaviors is analyzed based on the circuit modelling tool. As a takeaway, the potential application directions motivated by the analysis performed by this work are discussed, supporting the insulation system design and condition estimation in multifrequency power assets.

Secondly, this work rethinks the principle, modelling, and application of the conventional Equivalent Debye Circuit model (EDCM), with a discussion on its limitations and shortcomings. Its uniqueness lies in that this work is the first to discuss in detail the performance and problems faced by widely used EDCMs without circumventing any issues. Different modelling strategies of EDCM with different RC relaxation branches are applied to reconstruct the measured broadband impedance properties of polymeric insulation systems under different temperature conditions (seven cases). Additionally, the Nyquist plots of seven cases are analyzed to reveal the underlying physical behaviors of polymeric insulation systems. As an enhancement supplement of physical mechanisms, the combination use of EDCM and Nyquist plot is recommended.

Thirdly, an innovative grey-box modelling approach, the Fractional-Order Equivalent Circuit Model (FO-ECM), for characterizing the frequency-dependent impedance properties of polymeric insulation systems are proposed, which can effectively capture the underlying nonideal and nonlinear dynamics behaviors in multi-frequency and high-voltage applications. The relevant theory basis and modeling processes of the novel FO-ECM are analyzed. The first four cases, epoxy resin under four different temperatures, are employed to conduct the illustrative validation to fit the measured broadband impedance. The strong reconstruction abilities demonstrate the effectiveness and feasibility of the proposed FO-ECM. In addition, to further validate the generality of the FO-ECM, another four cases involving different polymer materials with complex chemical structures and physical properties (polyester-imide resin, crosslinked polyethylene, and polyethylene terephthalate), measurement conditions (temperature, voltage, frequency range), aging states, and setup differences, are applied to reconstruct the measured impedance, also exhibiting better performance to capture the dynamics mechanisms.

Finally, a novel conceptual methodology, insulation resilience response, is proposed to characterize the ability of electrical insulation systems to preserve their intrinsic operational properties across varying environmental conditions, particularly under extreme circumstances. The basic concept and modelling processes of insulation resilience response are analysed. In addition, three study cases, where the stress of each case is temperature, are used to illustrate the insulation resilience behaviors. Multidimensional sensitivity indicators to temperature stresses are extracted to characterize the insulation resilience response. The Radar plot tool is also used to visualize the insulation resilience response of electrical insulation systems. Moreover, a composite integration metric, the resilience index, is developed to quantify the insulation resilience response of electrical insulation systems. To support the insulation health monitoring and estimation, an application framework is provided based on a time-varied resilience index.

Overall, the above investigation contents have been carried out through the comprehensive analysis of the frequency-dependent dielectric dynamics mechanisms and linear and nonlinear grey-box equivalent circuit modelling, as well as insulation resilience response for electrical insulation systems in multifrequency and high-voltage power assets and power electronics converters. The findings and outcomes of this Ph.D. project will support the insulation system design and insulation health management of high-voltage power assets and multifrequency power electronics applications.