H2scan TM Catalog 2025 - Flipbook - Page 29
One common theory often applied to transformer failure is the bathtub
curve, and after the transformer has been in service for about 昀椀ve years,
owners expect to see a reliable life of up to about 30 years
In Fig. 1, a chart from a 2014 US Department of Energy study
on large power transformers shows that—as the buildout of
electrical infrastructure in North America was planned for
expected future capacity--very few transformers over 100
MVA were added to the grid in the 1990s, which is why many
factories making large power transformers in North America
closed during that time or soon thereafter. Consequently,
many new, large power transformers in North America need
to be imported from countries that still make these expensive,
critical assets.
Besides data available at IEEE and CIGRE, a few entities
accumulate data on in-service transformers and on transformer failures. One of those entities is Doble Engineering. In
addition to accumulating data on various failures, they work
with transformer owners to understand how to extend the life
of these critical assets. As the diagram in Fig. 2 depicts, transformers are much more complex than most people realize.
Many things contribute to the reduction in a transformer’s life.
Most of the time, people try to simplify transformer failure
rates into concepts that are easy to understand based on
statistical theory. One common theory often applied to
transformer failure is the bathtub curve (Fig. 3). This concept
seems to make sense because there are failures when
transformers are new due to design, manufacturing, or
transportation issues.After the transformer has been in
service for about 昀椀ve years, owners expect to see a reliable
life of up to about 30 years. After 30 years, the failure rate
starts increasing rapidly.
Unfortunately, this model has not proven to be accurate since it
does not account for real-world design, construction, operation,
and maintenance. If transformers are subjected to repeated
through-faults or cooling systems are not maintained, these
types of operational issues will likely cause premature failure.
The diagram in Fig. 4 illustrates actual data from over 25,000
transformers in the Doble database. Each line represents the
failure rate of transformers produced in a particular decade.
For instance, transformers manufactured in the 1950s had
very low failure rates until the transformers were about 40
years old. Contrasting that with transformers made in the
1980s, those had high failure rates in the 11-30- year range.
Figure 2. There are many potential modes of transformer failure per ISO 18095 [5]
90000258
pg.29
27215CA
Turnberry
Lane, Unit
Valencia, CA 91355,
USA | +1 661-522-8775
| sales@h2scan.com
www.h2scan.com
27215 Turnberry Lane, Unit A, Valencia,
91355, USA
| +1A,661-775-9575
| hello@h2scan.com
| www.h2scan.com
|| 90000209
R1, ECO
24-229
©2025|H2scan
Corporation
