One of the more menacing challenges transmission operators face today is identifying and preventing wind-induced conductor motion damage.

Transmission lines are unique in that no other large structure has as much of its mass in a highly flexible form and is so continuously exposed to the forces of the wind. This makes transmission lines susceptible to vibration, galloping or other types of movement. Because conductors are supported and supplemented by thousands of pieces of hardware, numerous opportunities for damage arise during these motions.

The damage is insidious, however, because it is typically very difficult to perceive at any given moment. Often, damage can only be truly identified when the conductor is taken out of service and evidence such as broken strands is discovered.

Given today's budget and manpower limits, there is a growing tendency for vibration-caused problems to go undiagnosed, even when they result in outages. Crews are dispatched to the outage to repair or replace the failed line component on a “like-for-like” basis. Thus, the cause of the problem may not be investigated.

This article describes vibration problems at two U.S. utilities — Bonneville Power Administration (Portland, Oregon) and Arizona Public Service (Phoenix, Arizona) — that are, in different ways, representative of these sorts of problems.

BONNEVILLE POWER ADMINISTRATION

by Jerry Reding, Bonneville Power Administration

BONNEVILLE POWER ADMINISTRATION (BPA) is addressing a significant conductor vibration issue characteristic of many utilities in North America: the replacement of failing vibration control devices installed 25 to 40 years ago. The challenge is that, while failure of the devices can damage conductors and lead to outages, utilities have only limited resources to apply to the replacement of a large number of devices.

BPA has used the EPRI Orange Book since its publication, both in the design of new lines and for forensic analysis. In addition, with the book as a background source, BPA has developed its own custom tools for modeling and analysis.

To control conductor vibration, BPA uses a variety of common devices. For single overhead shield wire and all-dielectric self-supporting fiber-optic cables, we install spiral vibration dampers (SVDs); for single-phase conductors, we use stockbridge dampers; and for bundle conductors, we use either stockbridge dampers with spacers or spacer dampers. Two types of the spacers are failing: the spring-type spacer dampers and steel-coil twin spacers. The spacer dampers are on 500-kV triple-bundled conductors, predominantly triple Bunting and triple Chukar. The devices, which were installed in the 1970s and 1980s, have a steel spring mounted inside an aluminum-plate housing. Over time, on many units, the steel springs have moved around, worn through the aluminum plate and released the primary pins that hold the assembly together. This failure can allow the assembly to come apart, which is like losing the bolts out of your car.

Failure of the spacer dampers can lead to several types of damage. Because the conductors are no longer properly spaced, collateral damage to adjacent subconductors is possible. The failed spacers can severely damage the adjacent conductors. BPA linemen have observed conductors damaged to the point that all four layers of aluminum are abraded and the steel core is visible. Also, because the damping quality has been lost, eventually there is the threat of conductor fatigue damage at the suspension clamp.

In 1999, when BPA first noticed the failed spacer dampers, it had 130,000 units in the field. Initially, the failed units were spotted by helicopter or walking patrols. BPA addressed the problem with an “identify-and-replace” approach, but soon found it difficult to go to a single spacer damper in the span and replace it efficiently. As a result, BPA decided to replace all spacer dampers in a span. When BPA crews removed units that did not appear to be damaged, they discovered that the units were significantly worn and would have soon failed. To date, BPA has replaced about 30,000 of the spacer dampers.

Another device that is failing, the steel-coil twin spacer, was installed on 500-kV twin Chukar conductors in the mid-1960s and 1970s. The spacers have aluminum clamps on both ends of a coiled-steel spring. They are used in conjunction with stockbridge dampers to control vibration and to space the subconductors. Failures of these spacers were also noted in 1999, at which time BPA had 190,000 units installed. Five years ago, BPA initiated a formal replacement program for the spacer dampers and spacers, and is now replacing about 26,000 per year.

For BPA, replacing the spacer dampers and the spacers is difficult for a variety of reasons. Because live-line work is not permitted, lines must be de-energized for the devices to be replaced. This means scheduling outages at a time of high demand. Replacement of the devices is also a matter of time and money. BPA has 115 linemen for more than 15,000 miles (24,140 km) of lines. There are about 80 units per mile (50 units per km) on a single-circuit line. It takes a seven-person crew two weeks to replace the spacer dampers or spacers on 12 miles (19 km) of line. The rate at which BPA is currently replacing these units extends beyond what the units can reasonably survive. Eventually, BPA will have to develop a more aggressive solution.

---

Jerry L. Reding received his bachelor's degrees in electrical engineering and computer science in 1973 from Oregon State University. He joined Bonneville Power Administration in 1974 in the line design section. Reding is the principle design engineer for BPA's overhead power and fiber-optic lines with expertise in areas relating to the design, operation and maintenance of those lines with a concentration in cable design and selection, conductor ratings, connecting hardware, and insulator mechanics, and field problems related to stringing and sagging cables. He is a registered professional engineer in the state of Oregon and a senior member of IEEE PES. jlreding@bpa.gov

ARIZONA PUBLIC SERVICE

by Mark Orth, Arizona Public Service

ARIZONA PUBLIC SERVICE (APS) faced a mystery that did not initially seem to be related to vibration at all. Although APS persevered to determine the cause of failures, our case illustrates the challenges involved and the reason that many utilities repair failures without ever identifying vibration as the root cause.

In the mid- to late-1990s, APS began using polymer line post insulators on new overhead lines. We began to see the failure of insulator stud bolts on 12-kV and 69-kV polymer post insulators. Initially, the number of failures was relatively low, 10 to 15 over three years. However, despite these low numbers, the utility committed itself to a root-cause analysis because of the serious nature of the failures. The broken bolts typically led to an outage and a potential safety hazard to the public and utility personnel. In addition, the utility was installing a significant number of insulators yearly. (Eventually, APS would install more than 80,000 units.)

All the failed bolts exhibited signs of fatigue cracking. Vendors supplying the insulators and those providing dampers and damper installation requirements were contacted. Among the possible causes first investigated were defective studs, improper installation and overtensioning of conductors. But metallurgical tests showed that the bolts were not defective. Installation errors were ruled out due to the consistent failure points (cyclic loading of loosened insulator studs resulted in fatigue at other stud locations), and conductor tensions were shown to be below those tensions recommended as “final tension” in the EPRI Orange Book.

Aeolian vibration was the next suspected cause. Many of the failures occurred on 31 miles (50 km) of line in flat, open areas with minimal or no structures or vegetation to act as windbreaks.

Two state-of-the-art line vibration recorders were used under vendor supervision. The strain and frequency data accumulated over several months (and using the methodology used in the Orange Book for rigid supports) was analyzed. The conclusion was: “This force is dependent on frequency, but for loop lengths associated with about 40 Hz, a force of less than 10 lb [4 kg] would be required to cause [this] strain in the strands at the support points. The force would be lower for lower frequencies.”

APS sent several of the studs to an outside laboratory to determine the strength of the studs under cyclic loads. The lab technicians mounted the insulator with studs and mounting plate on a rigid support, and attached the line end of the insulator to an electro-hydraulic actuator, which physically moved the insulator up and down at a given load. A crude extrapolation of the results (the data acquisition required to develop a statistically defendable endurance limit would have taken more months of testing at prohibitive costs) showed that it required a 50-lb (23-kg) cyclic load for approximately 3 million cycles to break the studs.

This was a dead end that just deepened the mystery. If 50 lb was required to break the studs, why were they failing with conductor loads of 8 lb to 10 lb?

Since this was far higher than the loads predicted by the industry data (taken in general from lines supported by rigid porcelain line posts) in the Orange Book, the APS team decided to measure the actual conductor vibration in the field. In addition, strain gauges and a triaxial accelerometer were installed on the insulators themselves. These instruments measured strain in the insulator (FRP portion) near the base of the insulator (stud location) and the velocity at the end of the insulator, not on the conductor. The line recorders confirmed that conductor bending strain was as before (that resulted in estimated loads of 8 lb to 10 lb based on the clamp, as well as the insulator being stationary).

However, the strain gauges and accelerometers sensed loads on the insulators during late evening and early morning hours corresponding to static loads measured in the laboratory of 80 lb to 120 lb (36 kg to 54 kg). These loads are more than enough to break the studs if experienced throughout the evening for several months of the year over a period of years. These lines had been installed two to three years prior to the stud failure; by the time the testing was performed, the conductor tensions had relaxed to tensions at or near final tension. This meant that loads on new lines could be 150 lb (68 kg), which would account for studs on new lines, close to initial tensions, breaking even sooner.

At the same time, the number of failed stud bolts began to increase up to 35 per year. The most cost-effective solution was to install dampers on all of the spans (many less than 275 ft [84 m]) on the 31 miles (50 km) of lines where the stud failures had occurred. Installation of the dampers was shown to reduce loads to 7 lb to 8 lb (3 kg to 4 kg). APS conductor tensions were below those recommended in the Orange Book. Further reduction in conductor tension was not a cost-effective means of resolving the problem; therefore, APS adopted a conservative approach and increased damping beyond vendor recommendations at that time and has eliminated nearly all the bolt failures since the change was made.

Using the Orange Book allowed APS to identify what typical loads were anticipated on the insulators based on conductor tension. Once we measured the strain on the actual conductor and the load required to break the studs, we could see that those two didn't match. If we didn't know they didn't match, the root-cause evaluation would have been much more difficult.

Preliminary investigations indicated that the mystery — the discrepancy between the predicted loads from the Orange Book and the measured values — may be explained by the fact that the clamp, as well as the insulator, is not stationary as assumed in the Orange Book. They will, in general, move in response to the conductor vibration.

---

Mark Orth received his bachelor's degree in nuclear engineering from the University of Arizona in 1980. He worked for Bechtel Power Corp., Sargent and Lundy Engineering, and as a consulting engineer in the nuclear power plant industry until joining APS in 2000. Orth is currently a senior engineer for APS in the Generation Engineering department. Mark.Orth@aps.com

THE ORANGE BOOK

by John Chan, EPRI Transmission Capacity Program

In 1979, EPRI published the Transmission Line Reference Book: Wind-Induced Conductor Motion as was one of a series of EPRI overhead reference books. Because of the book's bright orange cover, it quickly became known in the industry as the “Orange Book.” It was a state-of-the-art reference guide to conductor motion. The book enabled several generations of overhead line designers to anticipate the circumstances in which cyclic conductor motion might be expected, to become familiar with protection methods and to refine their in-house design practices.

Twenty-five years since its publication, the Orange Book is still the industry standard. It is still commonly used to diagnose and solve conductor motion issues. However, over the years, considerable additional progress has been made in this area. As a result, EPRI has assembled a team of international experts who is currently working on an update. Publication is scheduled for 2007.

Examples of this new information include: In the area of Aeolian vibration, progress has been made in the analysis of wind excitation data, the behavior of new conductor designs, improved laboratory measurements using laser technology, the interpretation of vibration records and the modeling of vibration behavior. Regarding conductor fatigue, there have been considerable developments on inspection tools and fatigue endurance of conductors and clamps. With galloping, field studies have led to improved knowledge of galloping amplitudes, with and without control devices, for single and bundle conductor lines and the refinement of application techniques.

Originally, the book covered three primary types of motion: Aeolian vibration, conductor galloping and wake-induced oscillation. There was detailed information on causes, mechanisms, incidence, influencing factors, resulting damage and protection methods for each motion. New information will also be included on the vibration behavior of fiber-optic cables and available control devices. Research results will be provided on transient motions not previously covered in the original edition, including short-circuit forces, bundle rolling, ice drop and gust response. Accompanying the book will be applets — small calculation programs — to assist users in understanding and applying concepts.