Power transformers represent the largest portion of capital investment in transmission and distribution substations. The financial consequences of losing a single unit can have a multimillion-dollar impact. By contrast, a failing transformer removed from service in time can usually be economically reconditioned.

Fault conditions in a power transformer are detected in several ways. One method is based on detection of the degradation products of the insulating oil – usually dissolved gases – which are produced as the result of an abnormal dissipation of energy within the transformer. However, this energy, released through fault processes such as overheating, partial discharge (pd) and arcing, is often sufficient to generate the fault gases initially in the form of bubbles. Also, high moisture conditions and sudden overloads can cause the inception of moisture vapor bubbles released from conductor insulation.

When it has been established that a transformer is gassing – from abnormal dissolved gas analysis (DGA) results - most transformers owners implement a program of more frequent testing – weekly or even daily – to try to understand what is happening inside the transformer and to prevent a catastrophic failure.

Acoustic Emission techniques have been used throughout the years to detect and locate partial discharge/arcing. However, there are some cases where acoustic emissions have been detected in the absence of PD. MISTRAS was the first company in the world to document, through laboratory and on-site tests, that Acoustic Emissions can be produced by other fault mechanisms such as overheating (see EPRI web site). It has also been known for several years that AE can be caused by mechanical problems.

The mechanisms that produce signals are directly related to a fault in the transformer. However, there are other mechanisms that will generate AE activity that are not directly related to a problem in the unit. Some of these are: environmental sources, such as rain, snow, ice, or dust impacting the transformer, areas of turbulent oil flow within the unit (pump operation), load tap changer operations, magnetostrictive noise, etc.

Over the years, we have developed data filtering techniques that allow us to differentiate between relevant and irrelevant data. There are several different factors that determine what filtering scheme will be used for a given situation. Adequate data filtering is critical to obtain a true indication of the condition of the transformer.

MISTRAS has been participating in an EPRI Tailored Collaboration Project along with several US and international utilities. This ongoing project started in September 2001 and at this time is in its second phase. The objective of this project is to improve the AE technique and develop a better understanding of the correlation between acoustic activity and gassing. Some of the results obtained in Phase I include: improvement of the test procedure, refinement of the algorithm used for source location, performance of a more detailed data analysis, development of a database containing 61 different tests, creation of a grading system to assess the severity of the acoustic emissions detected and development of a classifier to distinguish between different acoustic sources (electrical, thermal, noise, etc). More detailed information can be found on the EPRI web site.

The objectives of Phase III are to continue the development of the test database, develop a classifier system and advance the evaluation of load tap changers using a combination of acoustic and vibration techniques. The development of an on-line monitoring device is also a goal during this phase.