Transformer Hot Spot Monitoring Comes of Age

Surinder Sandhu and Ed Oh, Luxtron Corp.

For years now, there has been much discussion in the industry about applying new technologies toward improving T&D grid reliability. One such technology-direct hot spot monitoring of transformer windings using fiber optics-is actually not so new having been field-tested and refined over the course of the last 20 years in hundreds of installations worldwide.

During this time, industry standards bodies like the IEEE and the IEC, as well as the industry in general have extensively studied, identified, and tested fiber optic hot spot monitoring, the benefits of which have been enumerated elsewhere with several key ones being:

“- protecting transformers from exceeding maximum safe hot spot temperatures during transient overloading;

“- enabling lifetime prediction of transformer assets; and

Click here to enlarge image

“- in conjunction with modern control schemes, enabling dynamic loading of transformers for most efficient asset utilization.

Improved fiber optic feed through tank wall plate. The traditional NPT threaded connections have been replaced by an all welded design, eliminating the possibility of oil leakage.
Click here to enlarge image

So well-established are these benefits, that industry motivation to adopt fiber optics is no longer an issue. Rather, what is hindering adoption of this potentially beneficial and field-proven technology are two lingering industry misconceptions about fiber optics, namely fragility of the equipment-especially probes-and high cost. While both were issues previously, equipment manufacturers have aggressively tackled both ruggedness and cost. Today, neither are real barriers to adoption.

Probe Ruggedness

In the early days of fiber optic temperature measurement, probes were quite fragile, requiring careful handling incompatible with the rigors of the installation process. It was not uncommon to have a 50 percent failure rate necessitating the installation of two probes at every measurement point in order to ensure one working.

Click here to enlarge image

But in 1997, in close cooperation with leading transformer original equipment manufacturers (OEMs), ruggedized probes were introduced to the industry with greater flexibility, allowing a tighter bend radius, and tough, protective jacketing providing impact immunity and improved tensile strength, all of which reduced the care required during the installation process.

These ruggedized probes also use specialized fiber designs to ensure complete penetration of oil and exclusion of air that could lead to failure. One manufacturer performs dielectric testing to ensure the fiber probes are PD (potential discharge)-free to 6.5 kV/mm.

Coupled with improved installation methods and subsequent OEM personnel training, today installation success rates greater than 95 percent are typical, and at least one manufacturer now offers a warranty on its probes equal to the lifetime of the transformer.

This improvement in installation success rate coupled with a concurrent 40 percent decline in probe costs has resulted in a net 70 percent decline in probe acquisition costs over the past 10 years.

System Robustness

In addition to ruggedized probes, system robustness has been improved. No longer just modified laboratory instruments, state-of-the-art systems now use long-life, solid-state LED light sources, not halogen bulbs, and come available with industry-standard NEMA enclosures. Critical accessory items, such as the fiber optic feed through, have also seen recent improvements aimed at eliminating oil leakage from the tank.

System Costs Have Dropped

As with most new technologies, when first released, fiber optic hot spot monitors were quite expensive. A system like Luxtron’s WTS-22, including instrument, feed throughs, extensions and probes for four measurement points installed would typically add 3 percent to 5 percent to the cost of a 100-MVA transformer. Because of the high cost, use of these systems has mostly been restricted to the larger (>100 MVA) or other critical (i.e. mobile substation, industrial) transformers.

However, in just the last five years, system costs have declined 65 percent as manufacturers have released newer designs into the market. In the case of the lowest cost options, these require the transformer OEMs to integrate the modules into their control schemes, but a number of OEMs are starting to do just that.

In fact today, the cost of fiber optics has fallen to the point where commercially available control schemes with fiber optics are being offered to the market at the same cost as those without. Fiber optic transformer hot spot monitoring has truly come of age.

Buying Future Flexibility

With more than 20 years of experience, field rugged systems, and competitive costs, many utilities have started specifying fiber optic systems for their new transformers. Even if the utility is not prepared to invest in the instrument, an increasingly smart practice is to at least have the probes installed. This serves two purposes.

First, with probes installed, fiber optic hot spot monitors can be used in heat run testing to ensure the new transformer performs to its nameplate specification.

Second, once the probes are installed, instruments can always be field retrofitted at a later date should future flexibility in capacity be required or loading practices updated. Utilities looking at this possibility must take care to select probes from the manufacturer from whom they wish to buy instruments in the future, as probes from different manufacturers are not interchangeable.


Fiber optic transformer hot spot monitoring is a proven technology available to improve reliability of the T&D grid today. Advancements in probe and system designs have resulted in systems rugged enough for field use, as evidenced by hundreds of installations worldwide over the past 20 years. Costs have fallen to the point where commercially available control schemes with fiber optics are being offered to the market at the same cost as those without and can easily be justified on transformers as small as 25 MVA. ࢝®à¢®

Surinder Sandhu received his B. S. degree in mechanical engineering from Punjab University, India, in 1972 and received an M.S. degree in mechanical engineering from Washington University in Saint Louis in 1976. In 1976 he joined Westinghouse Electric Co.’s Power Generation Division and worked for 17 years in various positions. Since 1996, he has been working at Luxtron Corp. of Santa Clara, Calif. He first joined as a senior engineer in fiber-optic temperature applications group and currently manages Luxtron’s application engineering department. Sandhu has served several years on ASME Power Test Code Committee PTC-6 and is active in the IEEE Transformers Committee in general and its task force on Winding Temperature Indicators in particular.

Edwin Oh is President and CEO of Luxtron Corp. Prior to that, Oh was vice-president and general manager of Metcal Inc. (Menlo Park, CA), a supplier of soldering and PCB assembly equipment used in electronics manufacturing. He has nearly 10 years experience as a senior manager in Silicon Valley high-technology companies. Oh has an MBA and M.S. in chemical engineering from Stanford University, and B.S. degrees in chemical engineering and chemistry from the Massachusetts Institute of Technology.


Previous articleELP Volume 83 Issue 6
Next articleNew analysis finds circuit breaker market future bright

No posts to display