The Need for NESC Loading & Strength Revision

By Richard F. Aichinger, PE, American Iron and Steel Institute, Steel Utility Pole Task Group

Abstract

The National Electrical Safety Code (NESC) through its identification of loading conditions and strength and loading factors has effectively established the design relationship of electrical support structures since the early 1900's. Yet, questions are being raised by utilities that ask why a disproportionate number of wood poles fail in comparison to steel poles after severe weather events. At least part of the answer will lie in the strength and loading factors used in the code. This paper examines the treatment of past and current factors used in the specification of strength and loading, compares it to "sound engineering principles", and concludes that the NESC has been flawed for some time. This examination will compare the "minimum yield strength" approach of steel design with the "mean Modulus of rupture" approach of wood design; the allowance of 33% degradation of wood strength; and Grade B and Grade C Construction. Revisions to the Code, such as those proposed in this paper, are required to provide consistency and reliability for support structures design, whether of wood or steel manufacture.

Introduction

The National Electric Safety Code (NESC) has provided a means for establishing minimum loading and strength for transmission and distribution lines and structures since the early 1900’s to the present time. For much of that time wood pole structures were the dominant material used which simplified the establishment of conditions "that work". The advent of steel and concrete poles for transmission necessitated the establishment of criteria for these materials while maintaining some relationship with existing criteria for wood poles. Design conflict in the Code may have appeared minimal, due to steel and concrete materials usage being effectively limited to transmission lines and Grade B Construction. However, the advent of steel material being used for distribution lines and potentially in Grade C Construction, highlighted the discrepancies in the criteria based on a relational comparison of wood and steel. The intention of this paper is to provide: some history and background on the NESC loading and strength evolution; some of the inherent discrepancies in the material relationship; and some recommendations for resolution of these discrepancies. Generally, comparisons will be based on wind loading.

Some NESC Background

Very early editions of the NESC utilized stress limitations and loading conditions as a means of qualifying the sizing of structures. Early editions used 8 psf wind with ice through the 4th Edition, released in 1927. Through this period, the material sizing selection was determined by specified allowable stresses for wood and steel. The 4th Edition specified southern yellow pine to have an ultimate strength of 6500 psi and the pole to be sized to an allowable stress of 50% under Grade B and 75% under Grade C. The same edition specified steel to have an ultimate strength of 55 to 65 ksi and a yield strength not less than 50% of ultimate and an allowable stress of 26 ksi for Grade B and 30 ksi for Grade C. It would appear that the general intent at the time was to have material limited to 50% of ultimate for Grade B. However, steel had an additional limitation of 80% of the presumed Yield strength.

The 5th Edition (1941) provided a major change in approach for loading and strength with the change from 8 psf to 4 psf wind under ice conditions. Additionally, overload factors were introduced for steel while wood remained with a limitation based on a percentage of the ultimate strength. Despite the apparent loading change, the intent of structure strength appeared to remain the same by factoring the limitations. This edition introduced the transverse factor for steel as 2.54 (Grade B). This factor’s derivation can be assumed by the following: since the apparent wind loading was halved, the factor needed to double, but the allowable stress was limited to 80% of yield so the ratio became . The wood allowable stress was reduced from 50% to 25% of ultimate to reflect the need to equate the strength to past practice. Similar ratios were performed for Grade C construction for steel and wood to arrive at a 2.2 factor for steel and 37.5% of ultimate for wood. An examination of these wood allowable percentages in terms of a factor designation method shows that the wood wind factor becomes 4.0 for Grade B and 2.667 for Grade C. The 5th Edition increased the ultimate strength of southern yellow pine to 7400 psi.

By the 1973 Edition of the NESC the wood ultimate strength was specified by reference to the latest ANSI 05.1 edition and this practice has been maintained to the present. The 1973 edition maintained the 2.54 and 2.2 overload factors for steel, but the allowable percentage of ultimate for wood poles was changed to 50% for Grade C and the pole replacement definition was added to reflect the minimum strength deterioration at which the pole must be replaced or rehabilitated. The replacement strength became 37.5% for Grade B and 75% for Grade C (compared to 25% and 50% respectively at new construction).

The 1977 Edition of the NESC changed the Grade B steel factor to 2.5 for wind. The wood was now changed to an overload factor designation of 4.0 for Grade B wind and 2.0 for Grade C wind, which reflect the percentage values of 1973. However, the replacement factors changed to the current requirement of replacement or rehabilitation when deterioration reduces the pole strength to 67% of that required when installed (both Grade B and Grade C construction, using Rule 250b).

The 1997 Edition of the NESC introduced the concept of separate Loading and Strength factors for wood and steel materials; however, wood maintained the previous single factor and an allowance was made to be able to use either factor system. This system change served to magnify the relationship in the representation of loading and strength for these materials. The following chart compares the factors of the 1997 Edition with the overload factors of the 1990 Edition:

1997 Edition, Grade B Construction1990 Edition
Strength Factor Equiv. O.L.F. Overload Factors  
Load Type Overload Steel Wood Steel Wood Steel Wood
Vertical 1.5 1.0 0.65 1.5 2.31 1.5 2.2
Transverse  
Wind 2.5 1.0 0.65 2.5 3.85 2.5 4.0
Tension 1.65 1.0 0.65 1.65 2.54 1.65 2.0

 

1997 Edition, Grade C Construction1990 Edition
Strength Factor Equiv. O.L.F. Overload Factors  
Load Type Overload Steel Wood Steel Wood Steel Wood
Vertical 1.5 1.0 0.85 1.5 1.76 1.5 2.2
Transverse  
Wind 2.2 1.0 0.85 2.2 2.06 2.2 2.0
Tension 1.1 1.0 0.85 1.1 1.53 1.1 1.3

The Relationship of Materials

The NESC would appear to have maintained a prescribed level of strength specification based on experience and assumed successful practice of pole structure selection and design. Although the design loading conditions were revised for separate reasons from an 8 psf to 4 psf wind, the strength of the pole design was deliberately maintained by the restriction of the allowable stresses in direct proportion. The specification of wood has remained relatively consistent over the years of NESC revisions. However, the insertion of steel in the Grade B and C Constructions has produced discrepancies in specification approach. Some of these discrepancies follow:

Allowable Strength of Steel Material

The allowable strength designation of the 3rd and 4th Editions (1920 and 1927) of the NESC may have properly reflected the state of the art of steel production at the time, but the steel material purchased today for utility pole production comes with steel mill certification and quality which assures the minimum strengths and properties required by the design and manufacturing needs. Manufacturers are virtually assured of receiving material with the properties and strength specifically ordered for these products. As a result, the 20% reduction in allowable strength versus the defined yield strength is not required to maintain control over strength adequacy. This 20% factor has been carried on into the current 1997 Edition and is reflected in the 2.5 overload factor for Grade B as shown earlier in this paper. Although the wood ultimate strength values are presumably updated through periodic evaluations of the wood product, the improved quality and certification processes for steel have not been accounted for.

Using the logic previously established for the relationship of overload factors prescribed through the NESC history, the steel’s wind overload factor should more reasonably be 2.0. This factor would reflect current practice and acceptance of the steel material design to the actual yield strength rather than a value of 80% of the expected yield strength.

Allowable Strength of Wood Material

The allowable strength used throughout the NESC editions has been based on the mean modulus of rupture of the fibers derived from testing of poles at various times by various agencies and organizations. The NESC has now established a practice of referencing the ANSI 05.1 for the definition of strength for the wood poles. As the wood strength changes over time due to "fast growth" techniques, the circumference definition of the poles should change to reflect these changes in relation to the load definition for Class 1, etc. New data from various sources indicate increasingly lower values for the strength of wood.

However, the use of mean values for wood strength specification and the allowance of future deterioration of the wood pole strength due to rot, insect, natural fiber strength loss from drying and age, etc. points to meaningful problems comparing it to materials such as steel:

  1. The mean value is based on the middle of the bell curve of the sampling taken. It has been assumed in the industry that the increased overload factor attributed to the wood poles over steel poles (Grade B Construction for wind loading specifies the relationship as 4.0 to 2.5) accounts for this difference in strength value definition for wood and steel. Steel is specified to a minimum yield strength which is verified by inspection and certification. The wood is specified to a mean ultimate strength which only represents the mean value of a sampling of actual poles (which may be presumed to be the best quality possible). In the latest edition of the NESC a Wood Strength Factor is introduced as 0.65 for Grade B Construction. First, since this value was presumably intended to represent the difference between wood and steel for Grade B the value should have been 0.625 (i.e. 2.5/4.0) to accurately represent the most common design loading condition (wind).

    Second, this value should represent an effective relationship between the "mean" value definition used for wood and the "minimum" value definition used for steel. A testing program applied to a large sampling of steel material ordered to a particular specification would produce results in a bell curve similar to the wood pole test. The steel pole design requirement limits the strength value to the minimum yield strength of the material, not the mean ultimate strength (see Figure 1). This specification for steel allows the factor to be applied to the loading which represents the possible conditions for a region while allowing confidence in the material used to resist those loadings. To provide equivalent design practice for wood and steel, the material factor for wood should translate the test results available (mean ultimate rupture strength) to the minimum strength of the sampling where fiber rupture begins. In steel yield strength determines stress levels that will induce material deformation. In wood this stress level is much more difficult to determine, but as an alternative, a stress level that represents the minimum value of the testing sample would be appropriate for specification of design strength limits. Therefore, the material factor used for wood should represent the factor required to convert the "mean" value to the "minimum" ultimate strength value.

  2. If the material factor represents the conversion of the "mean" to "minimum", then the issue of deterioration of the wood strength by 33% must be accounted for in the design philosophy. The design of steel poles provides for a strength that is known not only at the time of fabrication and installation, but also throughout its use as long as corrosion of the material is accounted for. Even with the corrosion aspect identified as a limitation of life, the steel pole design is not accepted with any degradation of cross-section or strength, much less 33% of it. Therefore, not only must the material factor for wood represent the conversion of the "mean" strength value to the "minimum" strength value, but it must additionally account for the "allowed deterioration" of the wood strength by 33%. In other words, if the 0.65 factor listed in the 1997 NESC represents the conversion factor to minimum only, then the real material factor for wood should be 0.43 (0.65 x 67%).

Grade C Construction

The specification of factors for Grade B and Grade C Constructions for wood and steel material has been flawed for many editions of the Code with the 1997 Edition only providing graphic representation of this through the newly introduced Overload and Strength Factor representations for wood and steel. The Grade B considerations are adequately represented in the discussion above. The Grade C considerations in factors are as follows:

  1. The material strength factor should not vary. The material has the same strength whether or not it is used in Grade B or Grade C Construction. If, however, it is intended to merely represent a willingness to account for less reliability in the design of the system, then any acceptable increase in wood stress over allowable should be matched for the steel material. This would account for an increase in the steel factor to 1.31 from 1.0 (based on the wood strength factor increase to 0.85 from 0.65), but the strength factor cannot reasonably be greater than 1.0.
  2. However, the accounting for reduced reliability is more appropriately provided for in the loading factors. The loading (and, by association, the loading factors) should not change because of the material used to resist it. If an overload factor of 1.75 is the appropriate value for wind in Grade C Construction (as opposed to an overload factor of 2.5 for Grade B), then that factor would be applied for the design of the support structure independent of the material used to resist those loads.

Load factors should be applied to loads to account for uncertainties in magnitude or for reliability requirements while strength factors should be applied to account for material issues only.

Recommendation

Although the National Electric Safety Code (NESC) has provided a means of controlling the strength and loading of transmission and distribution line design since the early 1900’s, the material used in the support of these lines has evolved beyond the system originally devised for wood pole material considerations. In transmission line construction the use of steel poles has become dominant or nearly so. In distribution line construction the use of steel poles is just beginning, but is being driven by the user, the electric utilities, in their search for a dependable, reusable, non-degrading, lighter, and cost effective material. However, the overload factor system and the overload and strength factor system provide inconsistencies in specification that limit the effectiveness of the steel pole alternative in a design conversion.

Considering the above discussion, the strength and loading factor charts shown on page 2 of this document can be revised to more appropriately reflect a truly common approach to design specification. The following stipulations are made:

  • it was earlier noted that the strength factor in Grade B for steel included a 20% reduction factor. To revise the strength factor for steel to eliminate this 20% would increase the factor greater than 1.0. Again, since this is not reasonable, the factor remains at 1.0 for steel;
  • the strength factor for wood is 0.625 (not 0.65) as shown earlier;
  • the stength factor in Grade B for wood (0.625) represents only the conversion of strength from "mean" to "minimum" and, accounting for the 33% deterioration, reduces the real strength factor for wood to 0.417 (0.625 x 0.667);
  • the reduced reliability intended for Grade C will be represented by the loading factors and the loading factors for wood (1997 Edition) are used;
Recommendation, Grade B Construction1990 Edition
Strength Factor Equiv. O.L.F. Overload Factors  
Load Type Overload Steel Wood Steel Wood Steel Wood
Vertical 1.5 1.0 0.42 1.5 3.60 1.5 2.2
Transverse  
Wind 2.5 1.0 0.42 2.5 6.0 2.5 4.0
Tension 1.65 1.0 0.42 1.65 3.96 1.65 2.0

 

Recommendation, Grade C Construction1990 Edition
Strength Factor Equiv. O.L.F. Overload Factors  
Load Type Overload Steel Wood Steel Wood Steel Wood
Vertical 1.5 1.0 0.42 1.5 3.60 1.5 2.2
Transverse  
Wind 1.75 1.0 0.42 1.75 42 2.2 2.0
Tension 1.3 1.0 0.42 1.3 3.12 1.1 1.3

Alternatively, if it is assumed that the Grade B strength factor for wood (0.625) represents both the conversions of strength from "mean" to "minimum" and the future deterioration of the wood strength by 33%, these same charts would be revised as follows:

Alternative Recommendation, Grade B Construction1990 Edition
Strength Factor Equiv. O.L.F. Overload Factors  
Load Type Overload Steel Wood Steel Wood Steel Wood
Vertical 1.5 1.0 0.625 1.5 2.4 1.5 2.2
Transverse  
Wind 2.5 1.0 0.625 2.5 4.0 2.5 4.0
Tension 1.65 1.0 0.625 1.65 2.64 1.65 2.0

 

Alternative Recommendation, Grade C Construction1990 Edition
Strength Factor Equiv. O.L.F. Overload Factors  
Load Type Overload Steel Wood Steel Wood Steel Wood
Vertical 1.5 1.0 0.625 1.5 2.4 1.5 2.2
Transverse  
Wind 1.75 1.0 0.625 1.75 2.8 2.2 2.0
Tension 1.3 1.0 0.625 1.75 2.08 1.1 1.3

Conclusion

The intent of the Code should be to provide guidelines for the design of safe and reliable electrical systems for the utility industry. Those design guidelines should reflect sound engineering principles. The strength factor associated with a material must be consistent and independent of loading. In like fashion, the overload factors associated with loadings must reflect the loading and the construction type, but be independent of the material type used for the design of the support structure.

Utilities are questioning the discrepancy in the failure rates of different structures. The aftermath of severe weather has caused a concern about the consistency of design and reliability. If support structures of various materials are governed by the same loading condition, grade of construction, and Code, the question asked is, "Why do some materials fail more often than others?" The answer would appear to be an inconsistency in the strength and loading requirements of the Code. although the above recommendations can be debated at the detail level of the specific factor values this approach does provide the relationship methodology and understanding required to provide consistency of design and strength reliability. As this discussion illustrates, the current approach of the NESC is flawed. To provide a minimum criterion for establishing consistency and reliability for support structures, a change must occur to bring the method in line with the intent.


REFERENCES

ANSI 05.1 (Various). "American National Standard for Wood Poles - Specifications and Dimensions". American National Standards Institute, New York, NY.

NESC, C2 (Various). "National Electrical Safety Code". Institute of Electrical and Electronics Engineers, New York, NY.

AUTHOR

Richard F. Aichinger is Manager of Design Engineering, Transmission and Distribution Structures in the Industrial Products Group, Valmont Industries, Inc. He received the BS Civil Engineering from the University of Minnesota in 1970. He is a Registered Professional Engineer in Minnesota, Nebraska, and Oklahoma. He is a member of ASCE and IEEE.