On the establishment of effective condition-based maintenance for oil-filled electrical transformers to reduce failures due to copper corrosion problems and its role toward sustainable energy

Sammanfattning: Oil-filled electrical transformers are critical equipment for modern societies where a sustainable energy supply is required. Recently, many severe fire and explosion accidents of transformers have been reported due to copper corrosion problems in the form of copper sulfide deposits on the windings. The most negative impact of such accidents is safety threats to workers and people near the site, which can expose them to toxic chemicals, heat, sound pressure, etc. Another impact is the economic losses due to power outages and loss of profits. These accidents happen even though applying maintenance strategies for fault detection and diagnosis in power plants. Reviewing literature in the related topics revealed the inability of the applied maintenance strategies to detect copper corrosion problems due to shortcomings of an effective Condition Monitoring (CM) procedure that can be used to detect and track such problems in the initial stage, as well as a lack of a clear diagnosis model for defining a threshold that can early trigger an alarm before fault incidence. The main currently applied monitoring parameters are measuring corrosive sulfur compounds, evaluating the corrosivity level of the oil by corrosion tests, and measuring specific gases in the oil. Other parameters are also included, such as onsite electrical testing and online sensor device technology. All these parameters are insufficient for the early detection of copper corrosion problems. The main research problem addressed in this thesis is: how to establish an effective Condition-Based Maintenance (CBM) strategy for oil-filled electrical transformers to reduce copper corrosion problems and provide sustainable energy? The research is a mix of exploratory and explanatory approaches. For studying the current maintenance practices and identifying suitable ways for detecting and solving copper corrosion problems, literature reviews and experimental work were carried out, while artefact building and empirical studies were used for developing and testing solutions. One important result was identifying relevant parameters, such as hydrogen sulfide gas and toluene compound, which are by-products of the corrosion reaction. The finding was utilized in creating a procedure for early fault detection for the purpose of establishing an effective CBM strategy. In this strategy, the corrosion problems can be detected based on the relevant parameters and an Early Fault Diagnosis (EFD) model using a novel numerical method for defining an alarm threshold. The procedure was successfully verified and validated in two empirical studies in a power plant. The relationship between maintenance and sustainability was also explored to study the role of an effective maintenance strategy toward supplying sustainable energy. A framework for approaching sustainable maintenance was proposed based on three pillars, i.e., eco-efficiency indicator, asset management, and digitalization. In this context, a procedure called ABCD was used in an empirical study to assess the sustainability condition and find the prioritized actions toward sustainable energy. The main conclusion of this thesis is that the established CBM strategy, based on the proposed CM procedure, has the capability to reduce the probability of transformer accidents and their negative effects on social, environmental, and economic aspects, as well as provide sustainable energy to the society.