Identifying Transformer Faults with DGA

Imagine trying to explain power transformer fault detection to your child’s second grade class or to a CFO who does not have a technology background.  The conversation might start like this: “like a doctor takes a blood sample from your body to detect an illness, an engineer may take an oil sample from a transformer to detect a problem or fault inside the transformer.” Okay, that was the easy part! Now how do you explain what to do with that oil sample and why it is important to the utility?

DGA and Fault-Related Gases
Dissolved gas analysis (DGA) is used to identify faulty transformers and to monitor transformers that are under stress or behaving abnormally. The key to trouble detection is understanding that abnormal heat and electrical discharges cause the production of fault-related gases from the affected liquid and paper insulation. When DGA detects a transformer fault, then DGA can also be used to diagnose the fault.

How to use DGA data to identify faults
We want to know the state of a transformer based on recent conditions, not a mushy “average” of all the past and present contributors to the gas concentrations found in the oil. When increasing fault gas concentrations raise suspicion of a fault, the gas increments (i.e. the changes in gas concentrations) generated by the gassing event of interest must be determined. These gas increments, not the total gas concentrations reported by the laboratory (or DGA monitor), are the proper basis for fault type identification, which may be determined with diagnostic tools such as the Duval Triangle.

In Summary
A sudden or prolonged rise in concentrations of fault gases, produced from liquid (e.g. oil) and solid (e.g. paper) insulation exposed to abnormal heat or electrical energy, indicates the presence of a fault inside a transformer. Gas increments calculated over a specified interval of active gas production can be used for identifying the type of fault responsible for producing the gas.

Why is DGA important?
Many transformer failures have been avoided by the use of DGA, but if you talk with any experienced high-voltage apparatus engineer, s/he will tell you that applying DGA is a complex activity and is not fool-proof. It is important to arm your maintenance teams with the best tools available that go above-and-beyond conventional methods and allow users to apply rules specific to their operating conditions and organizational policies. Considering that transformer fleets are worth millions of dollars, marginal improvements in the analysis can result in substantial savings from optimizing maintenance, increasing reliability and avoiding catastrophic failures.

For a fuller discussion about transformer fault identification using DGA, please download the technical paper here.