The Pacific Northwest boasts many tall trees. Oregon in particular is known for its Douglas fir forests. When combined with wind, snow and ice, these trees can be a significant source of distribution feeder outages. A utility in Oregon, Portland General Electric (PGE) has recently undertaken a project to review the application of covered wire to reduce the frequency and duration of tree-related outages on its system.

Reliability is the Prime Mover

Maximizing system reliability is every utility's mission. Today, with increasing customer expectations and added factors such as regulatory requirements, mandated service-quality standards and state laws, maintaining system reliability is more than a challenge, it is a requirement.

With its ongoing commitment to reliability performance, PGE was ranked first in power quality and reliability for Western Region utilities in a 2009 J.D. Power study of business customers. PGE also has numerous programs in place to prevent outages, including underground cable replacement, overhead switch exercise and inspection, padmount switch inspection, wood pole inspection, infrared inspection and routine vegetation management.

To further address reliability needs, the utility assembled a project team to evaluate tree-caused outages not preventable by routine vegetation management. Tree outages continue to be one of the leading contributing factors affecting system reliability, especially during storm conditions.

Although PGE's routine tree maintenance program has been effective at preventing “grow-ins” from trees in close proximity to power lines, it does not completely prevent outages from trees off of the right-of-way. Vegetation management is a temporary solution that, in most cases, accelerates tree regrowth. With approximately 8340 circuit miles (13,422 circuit km) of overhead primary lines in its distribution system, PGE had a desire to address this problem from a construction standards approach using targeted covered-wire conversions.

Tree-Failure History

PGE tracks outage data including weather conditions to distinguish storm-related outages from non-storm outages. Over the last five years, 94% of all fair-weather tree-related outages have been caused by either broken branches or uprooted trees located outside of the right-of-way. Short of clear-cutting near power lines, no amount of pruning reduces outages caused by trees off of the right-of-way.

Over the last 20 years, PGE foresters have reviewed fair-weather tree-caused outages to identify potential tree-failure risk and collect data on specific tree defects and environmental conditions that contributed to the tree failures. Tree-failure profiles were developed from the data to assist the foresters and tree crews in accurately identifying tree-failure risk indicators by specific tree species and tree proximity to conductors. This effort has helped PGE's vegetation management crews eliminate many potential failure risks during routine tree maintenance.

Genetically, Douglas fir limbs are weakly attached and can break readily in winds greater than 40 mph (64 kmph), even on perfectly healthy trees. These breaking limbs are a leading cause of tree-related outages on PGE's distribution lines. Limbs that break out from high up in a tree's crown can sail in the wind as far as 75 ft (22.86 m). The abundance of Douglas fir trees and frequent occurrence of distribution lines adjacent to or downwind from these trees present a significant reliability challenge.

To address this challenge, tree-caused outages on all of the underperforming feeders over the last five years have been plotted on feeder maps to accurately identify feeder sections with a history of tree-failure problems. Feeders are considered underperforming if they exceed established reliability index thresholds for system average interruption duration index (SAIDI), system average interruption frequency index (SAIFI) or momentary average interruption frequency index (MAIFI) based on their classification as urban, rural or remote. Portions of the feeders that have a specific history of broken tree limbs were identified for possible covered-wire conversion. Other distribution feeder sections also were identified with similar site characteristics consistent with Douglas fir stands that have a history of limb failures upwind from the line.

With the credible tree-caused outage/failure database established, the forestry department can team up with distribution engineering to incorporate tree-failure information into future distribution feeder and tap line reliability studies. When applied in conjunction with current distribution reliability programs, this approach will help identify those areas that would see the most benefit from selective covered-wire conversions.

Innovative Application

Covered primary wires have been used for decades to prevent intermittent arcing of conductors due to storms and windblown debris. Early types of covered primary used braided fibers coated with weather-resistant compounds. Modern covered conductors have replaced the coated fibers with a polyethylene jacket. While this polyethylene jacket is sufficient for the intended purpose of preventing arcing due to intermittent contact, covered wire is not considered to be insulated.

One of the challenges given to PGE's engineering organization was to find ways to maximize the benefits of installing covered wire within a fixed budget. A significant breakthrough came when a member of the Standards & Reliability Engineering group suggested that many of the benefits of covered wire could be achieved by covering only the center phase.

Experience has shown that approximately 75% of all tree-related outages involving windblown Douglas fir limbs only impact two phases. The remainder is typically a result of larger limb failures due to ice and snow. In these cases, the distribution feeder typically is located below the tree canopy, where large limbs are more likely to drop and span all three phases.

Field Tests

To further evaluate the effectiveness of covering only the center phase of a distribution feeder, the team scheduled two field tests. The first of these tests involved installing covered wire on the center phase of a rural fused tap line. An electronic recloser also was installed ahead of the test location to provide controlled clearing in the event of a fault and prevent other customers from experiencing any effects of the staged faults. The covered wire used in the field tests was weatherproof cross-linked polyethylene conductor manufactured by Alcan Cable, although several manufacturers offer weatherproof conductor.

Different framing configurations were evaluated during this field test. The first involved standard distribution framing on an 8-ft (2.4-m) wood crossarm with polymeric clamp-top insulators. With a Douglas fir branch spanning the center phase and one outer phase, the circuit was energized using the installed recloser. No fault occurred in this configuration.

Next, a Douglas fir branch was laid across all three phases and the circuit was energized. After approximately 5 minutes, smoke began to emerge from the branch. After 18 minutes, a carbon path developed across the entire branch, resulting in a phase-to-phase fault.

The phase-to-phase distance was then increased using a 10-ft (3-m) wood crossarm to see if the resulting voltage gradient was large enough to prevent a fault from branches spanning all three phases. The same tests were repeated first using a branch spanning the center and outer phases, followed by a branch spanning all three phases. The branch spanning the covered center phase and one of the outer phases did not result in a fault. The branch spanning the covered center phase and both bare outer conductors again resulted in a phase-to-phase fault between the base phases after approximately 18 minutes. This test was repeated with a Cottonwood branch spanning all three phases. The branch burned clear without resulting in a fault

A second field test was performed to evaluate additional framing configurations with covered wire installed on the center phase only. The first configuration used a standard 8-ft wood crossarm with a 30-inch (0.8-m) elevated center pin and polymeric clamp-top insulator. The second configuration used a 10-ft wood crossarm with a 30-inch elevated center pin. The third configuration used a 5 ft, 7-inch (1.7-m) crossarm with an elevated center phase and condensed spacing on the outer phases to create a more compact delta configuration.

In the first two configurations, a heavy branch was able to pull down the center phase, allowing the branch to contact all three phases. In the compact delta configuration, a Douglas fir branch was not able to contact all three phases simultaneously. With only two phases contacted by the branch, the circuit was able to remain in service indefinitely without a resulting fault.

Cost Comparison

A detailed evaluation of three consecutive years of underperforming feeder data was performed. Of the 18 underperforming feeders evaluated, 10 of the feeders identified trees as leading contributors to either SAIFI or SAIDI. A sample feeder with a high tree-related SAIDI contribution was selected for further evaluation using the forestry outage database.

A five-span section of primary wire was identified as a candidate for potential covered-wire conversion. Although making such conversions is rarely justified solely on the basis of cost, covered wire in this specific section would provide potential savings and benefits through reduced momentary and sustained outages, less storm damage, reduced vegetation management frequency and improved customer satisfaction.

Design Toolkit

The results of the field tests led to the development of a set of preferred design alternatives and design process improvements in the application of covered wire. This design toolkit has been adopted companywide as a tool to assist designers in the application of covered wire to improve distribution system reliability.

The first design alternative is to cover the center phase only. This approach can be effective where a few spans are at high risk of being contacted by blowing Douglas fir limbs. The team recommended using PGE's standard construction method of framing a deadend structure at the transition points of covered conductor to bare conductor and limiting the length of covered conductor to only the areas where tree exposure exists. This approach not only reduces exposure to tree-related outages, it also provides greater avian protection.

The second alternative is to cover the center phase and raise the center phase on a 30-inch fiberglass pin. This alternative can be effective in preventing faults where trees are near the end of the line and the line is located below a canopy of heavy overhanging limbs. Typically, limbs would need to be longer than 7 ft (2.1 m) in length to span the outer phases. For areas where there is not a major reconductor project planned, it is generally more cost effective to construct a deadend and limit the length of covered wire installed to only those spans where there is tree exposure.

For areas where a capacity-addition-caused reconductor project is planned, covering all three phases and increasing the length of covered wire installed to existing deadend transitions is often the preferred approach. In these cases, the incremental material cost to cover all three phases is typically lower than the cost of pulling in multiple conductor types and constructing additional deadends to limit the length of covered wire.

Finally, for areas that have a high probability of large trees being uprooted, installing underground conductors continues to be the preferred design alternative.

While it is understood that construction complexity varies from job to job, these general guidelines help to ensure consistent application of covered-wire solutions across the service territory.

As an aid to designers and engineers, the team developed an engineering and construction work practice as a tool for developing the most effective solution for a given set of field conditions. In addition to outlining the available construction alternatives, the work practice recommends that the designer engage the regional forester in assessing the probability of limb contact for the lines of interest. By incorporating the tree-failure profiles with resources such as annual feeder reliability reports, the team can identify those areas where covered-wire installation will have the greatest impact on reliability.

On any given job, there are many variables beyond the cost of the primary conductor and the mechanical loading conditions, for example, terrain, easements, existing construction, anchor easements and tree exposure. These additional variables and factors add to the complexity of construction option decisions and support the need to perform additional evaluation on a job-by-job basis.

Solution Now in Practice

Given the reality of abundant trees, frequent storms, regulatory demands, increasing customer expectations for reliability and limited budgets, an innovative approach to the application of covered wire was needed. Through the collaboration of foresters, distribution engineers, standards engineers and distribution designers, PGE has developed an innovative and cost-effective approach to applying covered wire to improve distribution reliability. Only after bringing these perspectives together could the team achieve the most effective and cost-efficient solution.

No single design approach is appropriate for all circumstances. The creation of a design toolkit and work practice allows for different solutions based on the conditions at a given location, improves coordination between foresters and design staff, and results in the more efficient use budget dollars to improve customer reliability. As these new design and construction practices were only recently developed, the long-term impact of this program is still something that bears watching.


The authors wish to acknowledge the contributions and support of this article provided by Steve Hawke, Bill Nicholson, Dave VanBossuyt, Jay Landstrom, Tim Sullivan, Craig Newman, Ed Burns, Dary Ebright, Rob Annen, Adolfo Tolento, Doug Kirk as well as the Southern and Western Region line crews.

Steve West ( is a standards engineer at Portland General Electric. He has 40 years of experience in distribution design and standards. His primary area of work is overhead distribution systems and materials. He is a graduate of Oregon State University, a registered professional engineer in Oregon and a member of IEEE.

Jeff Chittick ( is a distribution engineer at Portland General Electric. He has 39 years of experience in distribution. He holds bachelor's degrees in electric power technology and business from Oregon State University and is a registered professional engineer in Oregon.

David A. Johnson ( is the senior forester and program director for Portland General Electric's vegetation management program. He has been active in urban forestry efforts in many Oregon communities throughout PGE's service territory, bringing better awareness of proper planning, planting and caring of trees.

Richard Goddard ( is supervisor of the Standards & Reliability Engineering group at Portland General Electric. His responsibilities include managing the development of design and construction standards, root cause assessments, and T&D quality and reliability programs. Previous management responsibilities include transmission engineering, project management and economic development. He holds a BSEE degree from Washington State University, a MBA degree from Portland State University and is a registered professional engineer.

Companies mentioned:

Alcan Cable

J.D. Power

Portland General Electric