Abstract
Transformers that are grid connected are routinely subjected to severe electrical, chemical, and mechanical stresses, and the operational weather conditions can be extreme. Furthermore, as the operating time increases, the dielectric and mechanical strength of transformer’s insulation materials deteriorates, thereby increasing the chance of failures and decreasing residual life span. The inner insulation condition is influenced by various operating settings, such as electromagnetic fields and temperature profiles. Water content, contaminants, defects, winding thermal transients, and other factors all affect transformer’s internal insulation and safety. Transformers have complex configurations and many other accessories. The relationship between failure reasons and transformer components is not always evident. Certain failures have inductivity and compliance characteristics. One fault may trigger another fault, and the latter may be induced by it. Transformer condition monitoring and diagnosis are primarily based on data collected from in-service transformers. As a result, the approaches used for data collection, transmission, processing, and interpretation have a significant impact on determining the transformer's status.
In the age of interconnected technologies, power transformers, the backbone of electricity distribution networks, are subjected to a drastic change in terms of how they are monitored, assessed and maintained. The emergence of new technologies like the Internet of Things (IoT) and the introduction of online monitoring technologies has changed the landscape of transformer condition monitoring by allowing engineers and operators to gather real-time data that has a substantial impact on decision-making processes. This research work intends to modernize power transformer condition monitoring by creating an IoT-based system for condition monitoring of power transformers. The proposed IoT system will further automates routine testing outlined in the IEC 60076-1 standard. The goal is to eliminate human error, reduce the need for on-site technicians, and enhance cost-efficiency. Through this innovation, the study seeks to enable accurate fault detection and comparison of cost savings with traditional methods. This dissertation introduces the IoT framework that will replace the need for manual routine testing. This IoT system continuously measures the relevant electrical parameters and perform routine tests with the transformer in-service.
iii
A 2MVA, 6.3/0.42kV three-phase dry-type transformer Dyn1 from Avon Peaking Power Plant was used for this research work. This transformer is located in unit 13 transformer room substation, it is being used for distributing power to the 400VAC Motor Control Centre (MCC) board at Avon Power Plant. The system was tested at Avon Peaking Power, Durban, South Africa, with the IoT prototype designed based on local power requirements and data. The system was designed using EasyEDA. The tests performed during this study assumed a three-phase balanced load system. The Internet of Things (IoT) system for transformer condition monitoring was set up by considering the four IoT elements (i.e., Sensors, Communication channels, IoT cloud and data processing). The four sensors utilized for this research work are voltage, current, temperature and oil level sensors. Sensors were strategically placed around the transformer to gather real-time data for the following parameters of the transformer (Voltage, Current, Oil Level and Temperature). These sensors are constantly monitoring the transformer’s parameters and feeding the data to Adafruit IO cloud platform through the communication device ESP8266 using a Wi-Fi communication protocol. The IoT-based transformer condition monitoring system was further programmed to perform three routine tests adopted from IEC60076 standard. These tests are the transformer turns ratio, no load loss and load loss. The Arduino Nano board was selected for this research work to control these sensors as well as all the other peripheral devices like power sources, auxiliary communication devices and ports (i.e., USB).