Studies to Utilize Loading Guides and ANN for Oil-Immersed Distribution Transformer Condition Monitoring

There are plenty of proper monitoring methods to evaluate the condition and possible incipient failures of a power transformer. For distribution transformer monitoring, the methods are usually too expensive and/or time consuming to use. However, cost-efficient methods for distribution transformer monitoring are needed and one possibility for this is to utilize loading and temperature information measured from the network. The monitoring methods presented in the paper are based on the existing IEC and IEEE standards and neural-network analysis. The methods are used to calculate the top-oil and hot-spot temperature as well as the loss of life of a transformer. The calculated results are verified with measured top-oil temperature values. The possibilities to utilize studied methods in a larger scale for distribution transformer monitoring are also discussed in this paper