Testing for relationship between diagnostic tests and DC BDV of in-serviced aged stator bars of a hydro generator

Thesis / Dissertation

2024

Permanent link to this Item
Authors
Journal Title
Link to Journal
Journal ISSN
Volume Title
Publisher
Publisher

University of Cape Town

License
Series
Abstract
The effective management of assets currently presents a challenge to utilities. A cognisance and understanding of the actual condition of these assets will enable utilities to implement mitigation strategies and use these assets for the full duration of their lifespan. One of these assets could be hydro generators with the stator bars constituting one of the failing components. Integral to the satisfactory operation of a high voltage (HV) electrical rotating plant is the integrity of the electrical high voltage insulation of both the rotor and stator. During operation, high voltage generator components (for example stator bars), are continuously and simultaneously subjected to electrical, thermal, mechanical and ambient (environment) stresses. Subsequently, the complex interactions of these stresses degrade and deteriorate the insulation material gradually, resulting in a reduction of valuable operative life, potential forced outages and costly emergency repairs. Because both expected and unexpected stator bar in-service failures may lead to costly repairs, rotating machine users implement condition monitoring (CM), diagnostic tests and condition-based maintenance in order to mitigate risk. The contribution was executed in the form of removing aged stator bars from a generator and subjecting the bars to various diagnostic and direct current (DC) HV tests, testing for correlation. The diagnostic tests included insulation resistance (IR), polarization index (PI), partial discharge (PD), tan delta (TD) and Tennessee Valley Authority (TVA) tests. The HV breakdown tests included very low frequency (VLF) and DC tests. Despite using this wider range of tests than has been used before, the test results still confirm earlier reports that there are no clear correlations between the different diagnostic tests. There is also no clear relationship between the condition monitoring test results and physical breakdown of the insulation in the high voltage tests. Hence, it was concluded that these diagnostic tests on their own could not predict incipient insulation failure and that currently the most value derived from these tests would be to use the results as trending parameters rather than absolute values to predict insulation failure.
Description

Reference:

Collections