Power transformers, though very small in number, play a critical role in the transmission of power in an electric grid. They are the vital nodes that carry huge volume of electricity. During its operation, power transformers are subjected to various stresses including thermal and electrical, resulting in liberation of gases from the hydrocarbon mineral oil which is used as insulant and coolant. If not monitored, these abnormalities will deteriorate transformer windings and produce incipient faults that go unnoticed.
The concentration and evolution rate of different gases released in the transformer chamber will have a bunch of information hidden. A proper analysis will help identify the real cause of the gas generation and thereby locate the criticality for maintenance or replacement. Dissolved Gas Analysis (DGA) is one of the most successful and used technique to analyze the condition of the power transformer.
DGA results can be interpreted using multiple techniques that are formulated by various standard organizations. To accurately locate the underlying problem and take necessary action, a combination of these standards must be effectively clubbed with intelligent analytics. Key Gas Method is one of the standard methodologies used in DGA. This will provide an insight into inherent problem by calculating the percentage of key gases that are specifically linked directly to certain faults.
For instance, the percentage of hydrogen and acetylene in the total dissolved gas is directly related to arcing and similarly carbon monoxide indicates over heating of cellulose. The accumulated data of these key gas emissions over a time frame could be used to predict the faults which these gases could possibly bring to a transformer. Clubbing Ratio Methods (Rogers and IEC) along with the Key Gas Method make the analysis more accurate delivering better information.
Having above said condition based monitoring system on cloud helps utilities and industries to perform effective asset management. Our demonstration system implements this and will help create, co-create and support on similar environments.
Written by: Inbu Raja EK