Steering Capacitor Film Development with Methods for Correct and Adequate Dielectric Performance Assessment

Abstract

The transition of electric power systems towards renewable generation has created an increasing market for power electronics using film capacitors as one of their key components. Size, weight, and cost reduction can be achieved with better capacitors – an objective achievable with advanced dielectric films. The current state-of-the-art biaxially oriented polypropylene (BOPP) films are already operated close to their fundamental limits, causing a growing demand for next-generation technologies. To perform well when used in a capacitor, a film needs to have a wide range of fundamental and applied properties, all of which should be evaluated during film development to ensure there are no unwanted trade-offs. Power capacitors are used in applications with high downtime costs, e.g. HVDC, thus especially the reliability aspects must be given scrutiny. This thesis work was inspired by the lack of knowledge of the long-term performance of next generation dielectrics, e.g. polymer nanocomposites. Equally important was to fill the gaps in published knowledge of measurement methods to evaluate long-term properties, voltage endurance, and surprisingly, also the dielectric permittivity of thin (≈10 μm) low-loss films. In this thesis, a suitable measurement for each three is presented along with examples of their capability and an approach to applying them to steer film development.

The large-area multi-breakdown method developed in our research group is extended to measurements at realistic operating temperatures, and industrial BOPP films are shown to exhibit an 11–20 % decrease in the DC breakdown strength between room temperature and 100 ◦C. The results align with literature, which supports the validity of the approach. BOPP films made of base materials varying in terms of molecular weight are measured: these films exhibit similar short-term breakdown performance at room temperature, yet at 100 ◦C differences emerge. The difference did not correlate with the reduction of breakdown performance after DC electro-thermal aging, demonstrating the necessity of long-term tests.

Electron beam evaporation in high vacuum (P<10−6 mbar) is established as a repeatable and suitable method to metallize electrodes on ultra low-loss BOPP films, solving earlier issues of abnormally high dielectric losses or unrealistically low real permittivity. Metallization process is identified as the crucial factor: no pre- or post-treatments are required, and valid results are obtainable with various electrode metals. The method was demonstrated by measuring true “literature value” dielectric permittivity of commercial BOPP films: E≈2.25 and tanδ≈10−4. The importance of successful metallization process for measuring the intrinsic losses is demonstrated: samples with sputter deposited electrodes exhibited abnormally high dielectric losses, as also did samples metallized using another e-beam evaporator.

The multi-breakdown approach is also extended to times-to-breakdown tests, and accelerated ageing tests are conducted on an industrial BOPP film. High-field degradation and drastically reduced insulation life are observed. Analysis of the Weibull failure rate supports the notion that at current design stresses, BOPP is already operated close to the fundamental material limits, and also that the life in operating conditions cannot be determined by simple inverse power law extrapolation of accelerated rapid ageing data. Again, long-term ageing testing is advocated. Space charge measurements on “classic” BOPP films reveal charge accumulation at high fields, as expected. Interestingly, no space charge accumulation is detected in a novel nanostructured material under similar conditions, demonstrating the potentiality of nanofilled DC insulation.

A DC electro-thermal ageing test method is presented to investigate long-term phenomena in realistic operating conditions. Two 1000 h DC electro-thermal ageing tests associate ageing with the formation of electrically weak points. Large-area breakdown behavior, being sensitive to local changes, is established as a recommended ageing indicator. Material characterization does not reveal ageing-induced changes in bulk properties, supporting the literature-backed conclusion that early ageing progresses by localized degradation. A trial with eight pilot-scale materials demonstrate that weak point formation may be inhibited in nanostructured materials, but also that material-specific optimization of film processing is required to reach optimal dielectric performance.

Ultimately, the methods developed are fused into one resource-efficient approach to capacitor film development, in which the short-, mid-, and long-term properties are evaluated in three overlapping phases. Reliance on individual performance metrics to steer film development is discouraged: all properties need to be at an appropriate level for a film to perform in application, and there are trade-offs to be managed.