SEVERE WINDSTORMS OCCUR IN ALL PARTS OF THE UNITED STATES AND AROUND THE WORLD. Since 2004, the Southeastern United States has been hit with nine major hurricanes causing billions of dollars in damage to electrical and telecommunications infrastructure. And in December 2006, the Pacific Northwest experienced its worst windstorm in more than a decade, knocking out power to more than 1.5 million homes and businesses, and killing at least six people. State utility commissions are increasingly investigating the response of utilities after a storm and investigating infrastructure damage that occurs during the major storm. Utilities are taking notice and are beginning to consider the possibility of exceeding minimum safety standards so that structures will be less likely to fail during extreme winds.

In August 2006, Midwest Energy (Hays, Kansas, U.S.) experienced a windstorm that blew down a 2-mile (3.2-km) stretch of wood transmission and distribution poles in western Kansas. All poles broke at the ground line. Thousands of customers experienced outages.

After these failures, Midwest Energy had many questions, primarily related to the transmission-pole failures. Should these transmission poles have failed? Should the failures have occurred at the ground line? Did these transmission poles exceed their design criteria? Should inspections have identified these poles as being at risk for failure? Are there other issues related to the pole failures that are not typically considered during design and inspection?

STRENGTH VERSUS LOADING

Midwest Energy is a customer-owned electric and gas utility that serves about 80,000 customers in central and western Kansas. It is located in what the National Electrical Safety Code (NESC) designates as the Heavy Loading District. This requires all structures to comply with winter-storm loading criteria for heavy combined ice and wind loading conditions: 40 mph wind with 0.5 inches (64 km/h with 13 mm) of radial ice on the conductors. In addition, when this particular transmission line was designed, the NESC required structures more than 60 ft (18 m) tall to withstand extreme summer wind loading: 90 mph (145 km/h) wind on bare conductor. The loadings are then factored differently for the grade of construction. (Note: A new design under the 2007 NESC would not be subject to greater loadings.)

The transmission line with the failed poles was originally designed in 1983. It consists of 70-ft (21-m) Class 2 Douglas Fir poles set 9 ft (2.7 m) deep. Conductors are 477 kcmil ACSR attached at 57, 45 and 37 ft (17, 14 and 11 m) above ground with a 0.375-inch (9.5-mm) steel shield wire at the top. To meet Grade B strength requirements, span lengths for this design are limited to 425 ft (130 m). The average span length is approximately 388 ft (118 m).

For most short poles, the theoretical point of maximum stress will occur at or below the ground line. However, if most of the wires are attached near the top of a taller pole, the maximum stress point can be above ground line. As the location of the wire attachments spreads down from the top of taller poles, the maximum stress point moves toward the ground line.

An obvious question was whether there was deterioration at the ground line, reducing the strength at this location and creating a point of failure. These poles had been regularly inspected and treated, and visual examination did not reveal the presence of significant decay. A forensic loading analysis was performed to test where the transmission lines should have broken assuming no deterioration. This was done in two stages: a pole strength calculation and a pole loading calculation.

The pole strength is computed at each pole height. This is first done at the ground line using standard methods. The strength at all other heights can easily be computed knowing that the strength of the pole at a given height is proportional to the cube of circumference at this height. Pole strength versus height corresponds to the top of the green area in the graph on page 37.

The wind force on the pole at each height is computed using standard methods. This force is then used to compute the corresponding bending moment at each position below. The sum of all moments at each point corresponds to the total bending moment at that point. Assuming a 122-mph (196-km/h) wind, pole loading versus height corresponds to the top of the yellow area in the graph.

This process was performed iteratively and through it Midwest Energy found that a 122-mph wind is required to blow down the pole, assuming the pole is at 100% strength. The graph shows that the pole strength exceeds the pole loading above the ground line. This means that a ground-line break can be expected for this particular design.

DETERIORATION ANALYSIS

The NESC requires that wood structures be replaced or rehabilitated when deterioration reduces the structure strength to two-thirds of that required when installed. Grade B construction requires that a wood pole withstand at least four times (overload factor) the loading. Replacement or reinforcement is required when deterioration reduces the overload factor below 2.67. Field measurements of several failed poles indicate that the effective overload factor may have been reduced to 3.5. Further review of pole inspection reports revealed that no failed transmission poles were recommended for replacement or rehabilitation.

Assuming a one-third loss of strength, analyses indicate a ground-line break would happen at 99-mph (160-km/h) winds. The NESC summer wind criterion is 90 mph (145 km/h). Recorded maximum wind-gust speed, measured several miles from the line failure, was 88 mph (142 km/h). This still raises questions about whether or not these poles should have fallen down.

RADIAL BORING EFFECTS

The poles installed on the Midwest Energy line were bored near the ground line to allow the initial preservative treatment a more effective penetration. Approximately 24 borings with a diameter of either 0.25 inches (6.4 mm) or 0.313 inches (7.9 mm) extend 4.5 inches (114 mm) toward the center of the pole. These borings impact pole ground-line strength, and this strength reduction is not accounted for because of the benefit of improved decay prevention.

Strength-reduction estimates have been made assuming a 47-inch (1.2-m) ground-line circumference with 24 borings around the circumference with a repeating pattern that extends 2 ft (0.6 m) above and 4 ft (1.2 m) below ground line. Calculations show that the remaining strength from the 0.25- and 0.313-inch borings is approximately 83.5% and 79.8%, respectively. Because the maximum stress point occurred at the ground line of the failed poles, the radial borings likely affected the failure wind speed.

Through boring is another option for an improved initial preservative pole treatment. The borings run parallel to the direction of the wires and go all the way through the pole. A 2005 study by Oregon State University determined that the impact strength with through boring seems less than radial boring. Thus, through boring may be an option for utilities to consider in the future. Results will be reviewed at the 2007 annual meeting of the ANSI O5 committee to determine if the impact on new-pole strength needs to be addressed in the ANSI O5.1 new-pole specifications.

WOOD-POLE MAINTENANCE

Midwest Energy regularly inspects and treats its wood poles. When necessary, treatment is accomplished by boring holes near the ground line and inserting an aluminum tube with the fumigant methylisothiocyanate (MITC). After the 2006 storm, the utility raised the question as to whether the borings from the maintenance program were having a measurable impact on pole strength. For the transmission structures in question, four borings are spaced around the pole circumference 90 degrees apart so that the entire cross section can be treated more effectively. Each boring is 6 to 8 inches (150 to 200 mm) higher on the pole, creating a spiral pattern.

Only one out of the four borings will impact pole-bending strength for a specific wind direction. It is estimated that the reduction in strength resulting from one boring will reduce the pole's transverse bending strength by less than 5%. This is not significant, because the average pole strength has a variation of approximately ±20%. The inspection and maintenance program will continue so that supplemental preservatives can help extend a pole's useful life.

DISTRIBUTION POLES

The distribution poles that blew over were right across the street from the transmission lines. This presented the opportunity to examine the ability of these structures to withstand high winds and to compare these results to the transmission structures.

Both Class 4 and Class 5 distribution poles failed. All failed poles were 35 ft (10.7 m) tall Southern Pine wood, set 5.5 ft (1.7 m) deep, and had 8-ft (2.4 m) crossarms supporting three primary conductors. A loading analysis similar to the transmission analysis showed, assuming no strength deterioration, a maximum wind speed of 144 mph (232 km/h) for the Class 4 poles and 128 mph (206 km/h) for the Class 5 poles. Assuming a one-third strength reduction, maximum wind speeds correspond to 118 mph (190 km/h) and 105 mph (169 km/h).

Based on the distribution-pole analysis and conservative assumptions, wind-gust speeds were highly likely to be greater than 118 mph (190 km/h). This is close to exceeding the design strength of the transmission structures assuming no deterioration (122 mph), and exceeds the design strength of the transmission structures when accounting for a 20% strength reduction due to radial boring. In this case, it seems likely that the wind loadings simply exceeded structure design strength.

Recall that local weather stations only recorded maximum wind-gust speeds of 88 mph (142 km/h) in the area. An important lesson is that localized weather effects can result in winds much stronger than those recorded at nearby weather stations. In this case, it is possible that a microburst occurred, where precipitation-cooled air rushes rapidly toward the ground and then spreads out at very high speeds. Damage from microbursts can be similar to damage from tornadoes.

SYSTEM HARDENING

Hardening infrastructure for high wind is an emerging but important topic. Ideally, a utility can compute the expected damage that will occur in future storms, compute the cost of various hardening options and compute the expected damage reduction that will result from each of these options. This process allows for decisions to be made based on quantifiable costs and benefits, and goes far beyond the design of a structure to a specific wind speed. Possibilities for strengthening transmission and distribution lines include stronger poles, upgraded poles, shorter spans, smaller conductors, storm guying, push braces, less pole-mounted equipment, fewer third-party attachments, aggressive tree removal and a multitude of other options.

Achieving the proper level of infrastructure performance during high winds at the lowest possible cost is a daunting task, but will increasingly be demanded of utilities by their regulators and customers. Now is the time for utilities to start collecting detailed data on failures that occur during extreme weather. Now is the time for utilities to perform forensic investigations on storm damage. Now is the time for utilities to question whether the use of existing design standards as the basis of structure strength will result in adequate performance, or whether hardened systems are in their future.


Michael V. Engel is the vice president of engineering and asset management for Midwest Energy. He is former chair of both the Midwest Energy Association Electric Activities and the IEEE Planning and Implementation Committee. Engel earned his BSEE degree from Kansas University and his MBA degree from Fort Hays State University. mengel@mwenergy.com

Richard E. Brown is the vice president of operations for InfraSource Technology. He is an IEEE fellow, vice chair of its Planning and Implementation Committee, and author of the book Electric Power Distribution Reliability. He earned his BSEE, MSEE and PhD degrees from the University of Washington and his MBA degree from the University of North Carolina.
richard.brown@infraSourceinc.com

Edmund G. Phillips is an advisor for InfraSource Technology. He earned his BSEE and MSEE degrees from Florida International University, where he was awarded Electrical Engineering Graduate of the Year.
edmund.phillips@infraSourceinc.com

Nelson G. Bingel is the vice president of engineering for Osmose Utilities Services. He is chairman of the ANSI O5 Committee, which is responsible for new wood-pole specifications, and is a principal member of the Strength and Loading Subcommittee of the National Electrical Safety Code. He is a graduate of Purdue University.