Recent industry polls show winding hot-spot temperature is the most critical parameter utilities mention when considering a transformer-monitoring program. Six myths exist today concerning the use of fiber-optic sensors for hot-spot temperature measurements.

Myth No. 1: Fiber optics are too fragile for transformers

During the initial usage of fiber optics in transformers, the industry experienced a high percentage of fiber failures due to breakage. Fortunately, we have seen drastic improvements in the design and construction of the fiber-optic probes used within transformers. Transformer manufacturers have gained invaluable experience handling and installing fiber optics, significantly reducing fiber-optic breakage. LumaSense's (formerly Luxtron) redesigned probes now show a 95% installation and transportation survival rate.

Myth No. 2: Fiber optics are too expensive

Any new technology has higher initial costs. Now, with a product that has been on the market for 15 years, we have seen the costs of fiber optics tumble by as much as 75%. At the same time, we've seen an increase in functionality. As transformers can cost in excess of US$2 million, fiber-optic temperature-monitoring systems can be a fraction of 1% of the costs associated with large power transformers.

Myth No. 3: Fiber-optic temperature measurements are not reliable

Winding temperature indicators (WTIs) consist of gages that have been used on transformers for top oil indication and winding hot-spot simulation. This simulation is based on heating an oil bath by a current transformer that is connected to one of the phases of the transformer. LumaSense has documented instances where the actual temperature measured using fiber optics located directly at the winding hot-spot region were up to 20°C (68°F) different than indicated with a WTI. Additionally, a WTI reading can have up to four hours of delay. Also, older instruments do not predict the elevated temperature that would occur with a loss of cooling fans and oil-circulation pumps, since predictions are based solely on the load of the transformer.

Myth No. 4: Fiber-optic temperature-monitoring systems require more maintenance

Compared to existing top oil and simulated hot-spot sensing systems, maintenance on some fiber-optic monitoring systems is minimal to nonexistent. Some fiber-optic monitoring systems are now available that use a long-life LED as the light source. These LEDs can last 50 years without replacement. Conversely, WTI instruments must be calibrated, typically on an annual cycle.

Myth No. 5: Fiber optics cannot last for the lifetime of a transformer

The material used in fiber optics is long lasting in an oil-submersed (or SF₆-filled/dry) environment. Transformers now have been in operation for more than 25 years with properly functioning fiber-optic sensors.

Myth No. 6: I don't overload my transformer, so I don't need fiber optics

While initial new transformers put into service are rarely used at 100% capacity, there is no guarantee that you would be able to keep your transformer at low load through its entire life. Also, transformers do fail, which means other transformers have to pick up the load. Additional reasons to use fiber optics for direct temperature measurement include design verification, dynamic loading, transformer life calculations, transformer cooling performance and condition-based maintenance.

With two decades of experience measuring winding hot-spot temperatures accurately and in real time, LumaSense has a robust system now operating in more than 1000 transformers globally. For more information, go to www.lumasenseinc.com or contact Bret Sargent at B.Sargent@LumasenseInc.com.