Design Lines to Mitigate Wildfire Risk

The power sector is actively undergoing research and testing to provide insight and solutions into how to mitigate fire risk and other complications across the energy system. One such effort is centered around overhead distribution lines and the risk of local vegetation and their impact on the system. These lines are vulnerable to tree and branch strikes, which could lead to power  interruptions, outages or fire.

The two modes of failure related to tree-conductor conflicts are mechanical damage and electrical faults. Typically, only a small percentage of tree strikes mechanically damage the system; in contrast, the dominant mode of failure is an electrical fault. This occurs when branches contact the line and cause phase-to-phase or phase-to-ground faults.

A high percentage of tree-initiated faults on the three-phase distribution system cause no mechanical damage to infrastructure. In these cases, a branch spanned one or more conductors, providing a fault pathway between two differing electrical potentials (voltage).

Branches that strike and remain in contact with conductors can provide a short-circuit fault pathway. That fault may result in an interruption and subsequent generation of burning material, potentially acting as a source of ignition. If that material fell from the line into combustible material below, it could cause a wildfire. A best-case scenario occurs when utility crews must remove branches that do not self-clear, but this still consumes time and limited resources.

Researching Branch Strikes

One option to reduce interruptions caused by falling branches or leaning trees is to implement distribution designs that are less likely to catch and retain falling branches or have sufficient clearance to avoid electrical breakdown of the woody tissue of the branch. While research has been performed to assess voltage gradients capable of producing a low-impedance, high-current fault, little experimental work has been done to determine which designs are prone to capturing branches.

Several factors contribute toward branch retention including the branch’s form, mass distribution, force of impact and trajectory. Factors related to the line include phase configuration, impact location along the span and the conductor tension. The Electric Power Research Institute (EPRI) is actively exploring these factors at the EPRI Power Delivery Laboratory in Lenox, Massachusetts. The objective of this research project was to design and execute laboratory testing to quantify branch-capture performance of various distribution designs.

Previous testing demonstrated that phase spacing as related to the length of the fault pathway provided by a branch is an important consideration. Earlier work by EPRI evaluated four different three-phase overhead distribution line designs and demonstrated significant differences in the likelihood that a branch strike would be retained and remain in contact with conductors. The designs tested included horizontal, vertical and staggered offset phase configurations. The difference in the likelihood of branch retentions was found to be due to both the spacing and orientation of the individual phase conductors. Given the importance of phase geometry as a factor of branch retention, Xcel Energy and EPRI evaluated design alternatives that could reduce the likelihood of tree-conductor conflicts.

Experimental Method

A fundamental component of the research is to understand and characterize the risk posed by trees to overhead three-phase distribution lines. The risk assessment considered several factors as shown in Table 1.The project team used tree-caused interruption data provided by Xcel Energy to identify branch characteristics of relevance and to define prototypical branch specimens for use in testing. Conifers in general and deciduous species within a common genus (populus) were selected for testing. The testing was conducted in Western Massachusetts.Two species within these two groups were selected in order to locally source plant material. Eastern cottonwood (populus deltoides) is a common tree in the Northeast, and Colorado blue spruce (picea pungens) is used as a landscape tree throughout the region. The vegetation management staff at National Grid USA identified sources and provided the spruce branches used as test specimens. Twelve specimens were used in this investigation.

Structure Geometry

Xcel Energy is in the process of upgrading its electric distribution system in the wildfire risk areas. Over the next five years, the utility will rebuild more than 300 miles of single-phase and three-phase lines. Electric distribution lines were prioritized by wildfire risk, small conductor size (e.g., #6 solid copper, #4 solid copper, etc.), condition of line and tree coverage.

For three-phase lines that met the replacement criteria, hardware, conductor and conductor configuration that exceeded the performance of existing lines was needed. Stronger conductor (2/0 ACSR) was selected for the phase conductors and neutral conductor, reducing the chance of downed conductors. Neutral-high conductor configuration was used to absorb tree and branch strikes before contact with phase conductors occurred. Clamp-top insulators that allow the conductor to slip while keeping the conductor contained in the insulator and in the air were included.

After reviewing several options for the phase configuration, including flat conductor configurations (would catch branches between phase conductors) or vertical conductor configurations (hard-to-work-in mountain areas), an “A” configuration (neutral conductor at top of the A, two outside conductors on the crossarm at the base of the A) with a vertical offset for the center phase was chosen. The optimal vertical offset distance was unknown. A higher center phase would make for more difficult line work when making equipment connections. A lower center phase might make it easier to catch branches across phase conductors. The branch retention testing was performed to answer this question.

• Three variations of the new distribution line design standard were provided by Xcel Energy, each having a different center-phase height.
• The position of the middle phase above the crossarm depicted in the first figure on the facing page was used in the initial round of tests.
• The middle phase was elevated in the second round of tests (+8 in.) and raised further (+16 in.) above the crossarm in the final round of branch drop tests. Phase spacings were derived from the design drawings to determine voltage gradients between conductors.

Test Setup

EPRI built a test span at EPRI’s Lenox test facility. Linemen framed the two structures on either end of the test span and tensioned the conductors per the design standard provided by Xcel Energy. The utility also supplied the materials used to fabricate the structures and span. Once initially framed, shown in photo above, the location of the middle phase was altered in each of two subsequent rounds of tests.

Branch strike testing involved dropping branch specimens on the test span of overhead distribution line. The protocol started with 10 tests for each of the two branch types being used. If all 10 tests of a set resulted in the same outcome, the test was terminated; if not, an additional ten (N=20) drops were performed.

Branch drops were from 15 ft above the static shield conductor. The basket on the aerial lift was generally positioned  72 in. away from Center Line (C/L) of the design (the neutral conductor). The Center of Gravity (CG) was marked with flagging ribbon. Each drop was controlled for the position of conductor strike to the neutral (C/L) relative to the CG. Previous EPRI research demonstrated how branch rotation imparted due to conductor impact affects the likelihood of retention. This is a function of where along the branchthe strike occurs relative to its center of mass and whether the strike occurs on one or multiple conductors at nearly the same time.

• Butt first = Center of Gravity
      (CG) of branch strike with
      neutral on side being evaluated.
• Tip first = Center of Gravity
      (CG) of branch strike with
      neutral on opposite side of
      what is being evaluated.

Findings

Test results demonstrated a difference between the three designs in terms of the likelihood of a falling branch being captured and remaining in contact with conductors, as shown in the chart to the left, includes comparative data from previous testing to illustrate the relative performance of the Xcel Energy design.

The highest number of branches that fell clear happened on the side of the structure that did not include the center-phase conductor. However, the configuration with the lowest number of phase-to-phase captures included the center phase located at its highest position, 16 in. above the crossarm. This conductor configuration created a sloped geometry that the branch could slide off of.

General Observations

As a result of this project, a number of observations were made, including:

• The dynamic motion imparted on branches with CGs not aligned with the shield wire upon impact was significant. The elevated position of the neutral helped initiate rotation. That was consistent with previous EPRI research. This increased rotation made it more likely that the branch would fall clear. In many cases, branches that fell clear slid by and  between conductors with limited potential of only making intermittent contact with phases as they fell away. This was due to the increased phase-phase spacing.
• The importance of branch length was demonstrated repeatedly. Shorter branches were less likely to be captured by these designs.
• The importance of considering branch architecture was clearly demonstrated over the course of this work. The loss of lateral branches over the course of a series of branch drops resulted in a branch being less likely to be retained. It is important also to recognize that branch architecture would vary within species and between species. Ponderosa and lodge pole pine perform differently than blue spruce.

Identifying Phase Configurations

Branches that remain in contact with conductors can create a high-impedance fault. If the voltage gradient between areas of unequal potential is high enough, the pathway may evolve into a low-impedance path, resulting in a high-current fault and a subsequent interruption or outage. Utilities that experience outages caused by branches should consider identifying phase configurations designed to reduce the likelihood of branch retention. This approach could be especially helpful for utilities in high-risk wildfire locations.

Xcel Energy optimized one design through laboratory testing that demonstrated how shifting the location of one phase affects branch retention. By creating a sloped conductor configuration, Xcel reduced the observed number of phase-to-phase branch retention outcomes from 15% to 11%. This project demonstrated the importance of considering phase spacing and orientation when evaluating the vulnerability of overhead multi-phase distribution lines to tree strikes.
Editor’s Note: For more information on this project, and similar research conducted by EPRI, visit: https://distribution.epri.com/resources/applications/structure-testing/

John Goodfellow ([email protected]) has 40 years of experience in the utility industry and is an authority on utility vegetation management and reliability.  He currently manages an active portfolio of vegetation management related research projects and serves as the chair of the Right of Way Steward Council’s Technical Advisory Committee, which established accreditation requirements for IVM on electric transmission and pipeline systems in North America. He is the lead author of the new ISA BMP Utility Tree Risk Assessment (2019).

Dr. Joe Potvin ([email protected]) is a senior project manager at the Electric Power Research Institute (EPRI) and leads overhead distribution
asset research. His research activities focus on improving the reliability and resiliency of overhead distribution systems through enhanced line design
and equipment performance. Dr. Potvin is currently investigating structural failure mechanisms of overhead power distribution systems. This work is
identifying areas of mechanical weakness as well as causes of prolonged outages and improving upon the processes and devices used in restoration. Dr. Potvin received BS, MS, and PhD degrees in electrical engineering from Clarkson University in Potsdam, New York.

David Flaten ([email protected]) works in the Electric Distribution Standards Department for Xcel Energy. He earned his bachelor’s degree in mechanical engineering from the University of Minnesota. He spent the last 42 years working on overhead standards, design, structures and material.