Вы находитесь на странице: 1из 10

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/238192671

Improved Design of Embedment Depths for Transmission Pole Foundations


Subject to Lateral Loading

Article  in  Practice Periodical on Structural Design and Construction · February 2010


DOI: 10.1061/(ASCE)SC.1943-5576.0000025

CITATIONS READS
4 999

2 authors, including:

Sivapalan Gajan
State University of New York (SUNY) Polytechnic Institute
38 PUBLICATIONS   590 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Sivapalan Gajan on 09 August 2017.

The user has requested enhancement of the downloaded file.


Improved Design of Embedment Depths for Transmission
Pole Foundations Subject to Lateral Loading
Sivapalan Gajan, Ph.D.1; and Cassie McNames, P.E.2

Abstract: Self-supported direct-embedded poles are widely used by the utility industry in the United States to support high-voltage
transmission lines. For a typical transmission pole, the lateral loads caused by wind and ice loadings govern the design of the pole and
foundation. Recent research findings reveal that the methods used in current practice do not yield consistently reliable pole foundation
embedment depths in all soil types and for all possible pole classes, lengths, species, and pole loading scenarios. In order to generate
improved design methods for transmission pole foundations, validated methods for analyzing laterally loaded piles have been incorporated
into the current study. New reliable methods to design safe and cost-effective transmission pole foundations, incorporating both soil and
pole properties, are proposed and recommendations are made in rigorous and simplified forms such that they can be easily adopted for use
in the utility industry. This study found that the current methods for determining embedment depths for relatively rigid transmission poles
significantly underestimate or overestimate the required embedment depths depending on the pole and soil properties by as much as
⫾60%. A field case study is presented, where direct-embedded round wood transmission poles fell over due to foundation failure, and the
results of this case study are compared to the methods presented in this paper. The proposed methods are also compared with PLS-Caisson
software predictions, and it is determined that the methods proposed in this paper are relatively easy to use and generate accurate and
reliable results.
DOI: 10.1061/共ASCE兲SC.1943-5576.0000025
CE Database subject headings: Piles; Electric transmission structures; Poles; Lateral loads; Embedments; Soils.
Author keywords: Piles; Transmission poles; Lateral loading; Embedment; Soil.

Introduction possible pole classes, lengths, species, and pole loading scenarios
共Keshavarzian 2002兲.
Self-supported direct-embedded poles are widely used by the util- When the embedment depth specified is not adequate for the
ity industry in the United States to support high-voltage transmis- combinations of applied loads and in situ soil conditions, the
sion lines. These poles are subjected to lateral loads caused by transmission poles will lean as the soil around the pole yields and
wind and gravity loads due to the weight of the wire and ice that deforms. In some cases, the poles may even fall over, causing
forms around the wire. For a typical tangent transmission pole, costly power outages and creating a potential risk to human life.
the lateral loads govern the design of the pole and foundation. In On the other hand, in competent soils, the current methods yield
the case of self-supported poles, these lateral loads are resisted by over conservative embedment depths that are unnecessary and
the soil along the depth of the pole below ground level. Although uneconomical. Despite the consequences of transmission pole
close attention is paid to the design of the pole itself, the design of foundation failures, only a few studies have been performed on
the foundation is often based only on empirical methods in cur- direct-embedded self-supported transmission pole foundations
rent practice. According to the Rural Utility Service 共RUS兲, the 共Seiler 1932; Kinney 1959; Meador 1997; Keshavarzian 2002兲
standard of practice, or rule of thumb 共ROT兲, is to embed the pole since a transmission structure is not held to the same standards as
10% of the total length of the pole plus 2 ft 共0.61 m兲 into the a typical building or bridge structure. However, several research-
ground 关Rural Utilities Service 共RUS兲 2004兴. The current design ers have studied pile foundations subjected to lateral loading 关e.g.,
methods for transmission pole foundations do not take into con- Matlock and Reese 共1960兲; Broms 共1964a,b, 1965兲; Reese 共1970兲;
sideration the pole or soil properties. It has been shown that the Poulos and Davis 共1980兲; Fleming et al. 共1992兲; Prasad and Chari
methods used in current practice do not yield consistently reliable 共1999兲兴. The concepts of laterally loaded piles and the lateral
pole foundation embedment depths in all soil types and for all pressure distribution in soil can be adopted to determine appro-
priate embedment depths for transmission poles that are subjected
1
Assistant Professor, Dept. of Civil Engineering, North Dakota State to lateral wind and ice loadings.
Univ., 1410, 14th Ave. N, Fargo, ND 58105 共corresponding author兲. The main focus of this paper is to provide utility engineers
E-mail: s.gajan@ndsu.edu with reliable methods to design safe and cost-effective founda-
2
Structural Engineer, Ulteig Engineers, 3350 38th Ave. S., Fargo, ND tions for transmission poles. To achieve this goal, reliable meth-
58104. ods for analyzing laterally loaded piles have been incorporated
Note. This manuscript was submitted on August 27, 2008; approved
into the current study to develop improved methods to design
on March 30, 2009; published online on April 1, 2009. Discussion period
open until July 1, 2010; separate discussions must be submitted for indi- transmission pole foundations. The proposed relationships and
vidual papers. This paper is part of the Practice Periodical on Structural recommendations are made in rigorous and simplified forms such
Design and Construction, Vol. 15, No. 1, February 1, 2010. ©ASCE, that they reasonably incorporate both soil and pole properties and
ISSN 1084-0680/2010/1-73–81/$25.00. that they can be easily adopted for use in the utility industry. A

PRACTICE PERIODICAL ON STRUCTURAL DESIGN AND CONSTRUCTION © ASCE / FEBRUARY 2010 / 73

Downloaded 16 Mar 2010 to 134.129.122.119. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
Table 1. Details of the Western Red Cedar Transmission Poles
Pole class Range of L 共ft兲 Range of D 共ft兲 P 共kips兲
2 ft 2 40–125 1.13–1.75 3.7
P 1 40–125 1.19–1.87 4.5
H1 40–125 1.27–2.00 5.4
H2 40–125 1.35–2.10 6.4
H3 40–125 1.42–2.20 7.5
Note: L, D, and P are illustrated in Fig. 1.

L cohesionless soils, an average unit weight of 110 pcf 共17.3


kN/m3兲 and friction angles 共␸兲 varying from 20 to 40° are se-
lected. Saturated and unsaturated conditions were analyzed for the
case of cohesionless soils.
For a particular type of pole 共given L and D兲 and design lateral
load 共given P兲, the required embedment depth of the pole, z, can
D be calculated by using lateral soil pressure distribution along the
z depth of embedment and moment equilibrium of the soil-pole
6 ft system. The ANSI standard dimensions and design loads for west-
ern red cedar transmission poles of different classes are given in
Table 1. Table 2 presents the length and diameter of Class 2 and
Fig. 1. Schematic of a round wood transmission pole Class H3 western red cedar transmission poles, as most of the
results presented in this paper are obtained using these two pole
properties.
case study is presented, where direct-embedded round wood
transmission poles fell over due to soil failure, and the results of
this case study are compared to the methods presented in this Assumptions and Simplifications
paper and the PLS-Caisson software predictions. It is illustrated
how using the ROT method for foundation design can signifi- It is assumed that the embedded portion of the transmission pole
cantly underestimate or overestimate the required embedment behaves as a short-rigid pile since the ratio of depth of embed-
depth depending on the pole properties and soil types. Thus, the ment to diameter of the pole is typically smaller than 10 共z / D
methods proposed in this paper will not only lead to safe founda- ⱕ 10兲. The stiffness of the wood pile itself is typically greater
tion design; they will also provide suggestions for more efficient than the surrounding soil such that the soil yields before the pile
and economical designs. does. The stiffness of the pile does not have a significant effect on
the lateral resistance of the soil-pile system 共Fan and Long 2005兲.
For the purpose of this study, the term transmission pole is
Problem Definition used to describe a tangent single pole structure that is self-
supported. In practice several transmission poles in sequence are
Fig. 1 shows a schematic of a self-supported transmission pole
along with the important dimensions. The total length of the pole
is L, diameter of the pole at 6 ft 共1.83 m兲 from the base of the pole Table 2. Dimensions of Pole Classes 2 and H3 of Western Red Cedar
is D, lateral load applied at 2 ft 共0.61 m兲 from the top of the pole Transmission Poles
is P, and the depth of embedment of the pole is z. Transmission L 共ft兲 Class 2 D 共ft兲 Class H3 D 共ft兲
poles are classified based on the minimum diameter 共measured 6
40 1.13 1.42
ft from the base of the pole兲 required to support a horizontal load
45 1.18 1.49
applied 2 ft from the top of the pole as determined by American
National Standards Institute 共ANSI兲 O5.1 共ANSI 1987兲. 50 1.23 1.55
There are 15 classes of round wood poles ranging in size from 55 1.29 1.62
Class 10 to Class H6, with Class 10 being the smallest and Class 60 1.33 1.67
H6 being the largest 共ANSI 1987兲. The pole types studied in this 65 1.37 1.72
paper include western red cedar wood poles of classes 2, 1, H1, 70 1.41 1.78
H2, and H3, which are commonly used in the design of transmis- 75 1.45 1.82
sion lines. Pole heights ranging from 40 to 125 ft 共12.2–38.1 m兲 80 1.49 1.87
are included in this study as these pole heights cover the wide 85 1.51 1.91
range of poles used for transmission lines. Homogeneous cohe- 90 1.55 1.95
sive and cohesionless soils are chosen for this study. The diameter 95 1.58 1.99
of the poles considered in this study varies from 1.13 to 2.2 ft 100 1.62 2.03
共0.34–0.67 m兲 and the design load of the pole 共P兲 varies from 3.7 105 1.64 2.07
to 7.5 kips 共16.5–33.4 kN兲 as can be seen from Table 1. Both total 110 1.67 2.11
stress and effective stress analyses are performed to incorporate
115 1.70 2.14
the undrained and drained behavior of soils. Undrained shear
120 1.72 2.18
strengths, cu, ranging from 200 to 4,000 psf 共about 10–200 kPa兲
125 1.75 2.20
were used to calculate embedment depths in cohesive soils. For

74 / PRACTICE PERIODICAL ON STRUCTURAL DESIGN AND CONSTRUCTION © ASCE / FEBRUARY 2010

Downloaded 16 Mar 2010 to 134.129.122.119. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
1.5D

2cUD

3D
z z

9cUD 9cUD
9cUD 9cUD

POINT OF ROTATION POINT OF ROTATION

BROMS' METHOD FOR FLEMING ET AL. METHOD FOR


COHESIVE SOILS COHESIVE SOILS

z z

2
3γDzKp Kp γDz

POINT OF ROTATION POINT OF ROTATION

BROMS' METHOD FOR FLEMING ET AL. METHOD FOR


COHESIONLESS SOILS COHESIONLESS SOILS

Fig. 2. Schematics of soil pressure distributions in laterally loaded short-rigid piles

strung together with conductors. The size and quantity of conduc- soils were used to determine the necessary embedment depths of
tors, the location of the conductors, and the magnitude of the transmission poles. When subjected to lateral loading, short piles
loads applied to the structure vary for each project and location. behave as rigid bodies such that the entire pile rotates about a
For the purpose of this study, the loading associated with each pivot point below ground. The applied lateral loads are resisted
transmission pole is determined by the maximum design load 共P兲 mainly by the passive soil pressures developed in front of the pile
associated with the pole classification as listed in the ANSI O5.1 above the point of rotation. The rigid body displacement of pile at
standard 关American National Standards Institute 共ANSI兲 1987兴 the ground surface has to be such that the passive pressures are
共Table 1兲. However, the proposed new relationships can be used developed along the length of the pile. Typically for sandy soils,
to design foundations for various magnitudes of lateral loading these displacements are up to 1% of the embedment depth and for
共either working or ultimate兲 on transmission poles. the case of clays, the displacement at the ground surface is ex-
The diameter of the pole varies along the length of the pole, pected to be up to 5% of the embedment depth 共Das 2007兲. Two
though the variation in diameter is negligible along the depth of reliable and validated methods to analyze laterally loaded piles
embedment. For western red cedar transmission poles, the varia- are chosen from the literature 共Broms 1964a,b; Fleming et al.
tion in diameter is equal to 0.12 in./ft 共10 mm/m兲 length of the 1992兲. Fig. 2 shows the lateral soil pressure distributions along
pole 关American National Standards Institute 共ANSI兲 1987兴. The the depth of embedment of pole used in different methods of
diameter of the pole along the depth of embedment is assumed to analyses for cohesive and cohesionless soils.
be uniform; the diameter at 6 ft from the base of the pole is used
in calculations 共as shown in Table 1兲. The soil profile along the
depth of embedment is assumed to be uniform and homogeneous.
However, as shown later, a weighted average-strength parameter Cohesive Soils „Total Stress Analysis…
can be used to design embedment depths in layered soils. Un- In 1964, Broms proposed a solution for ultimate load capacity of
drained shear strength 共cu兲 for cohesive soils 共total stress analysis兲 laterally loaded rigid short piles embedded in cohesive soils
and friction angle 共␸兲 for cohesionless soils 共effective stress 共Broms 1964a兲. Broms 共1964a兲 assumed that the soil surrounding
analysis兲 are used in the analyses. However, if drained loading is laterally loaded piles has a rectangular load distribution with a
critical for cohesive soils 共for clays with relatively small friction maximum lateral pressure of 9cuD at a depth of 1.5D below
angles兲, the depth of embedment should be obtained as the critical ground surface 共see Fig. 2兲, where cu is the undrained shear
depth of embedment using the results obtained from both effec- strength of clay and D is the pile diameter. He also suggested that
tive stress and total stress analyses. the actual soil distribution increases linearly from a value of 2cuD
to 9cuD near the ground surface; however, he chose to neglect the
soil resistance at the top of the pile since gaps can form behind
Theory: Determination of Embedment Depths the pile 共Broms 1964a兲. Using this pressure distribution and mo-
ment equilibrium of the pole-soil system about pivot point, the
Well-established methods for determining the lateral load capac- required embedment depth 共z兲 can be obtained by solving the
ity of short, rigid piles embedded in cohesive and cohesionless following equation:

PRACTICE PERIODICAL ON STRUCTURAL DESIGN AND CONSTRUCTION © ASCE / FEBRUARY 2010 / 75

Downloaded 16 Mar 2010 to 134.129.122.119. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
P pressure distribution and moment equilibrium of the pole-soil sys-
1.5 ⫻ D + tem, the required embedment depth can be obtained from the
9 ⫻ cu ⫻ D

冑 冋 册
following equation:
P


P ⫻ L − 2 − z + 1.5 ⫻ D +
9 ⫻ cu ⫻ D 3
6 ⫻ P ⫻ 共L − 2兲
+ −z=0 z= 共5兲
2.25 ⫻ cu ⫻ D K2P ⫻ ␥ ⫻ D
共1兲 For effective stress analyses 关Eqs. 共4兲 and 共5兲兴, no added clear-
Fleming et al. 共1992兲 proposed a method similar to the method ance for depth of embedment is provided, as the correlation used
proposed by Broms 共1964a兲 to analyze laterally loaded piles in for K P 关Eq. 共3兲兴 is a lower bound of the actual passive pressure
cohesive soils. Fleming et al. 共1992兲 noted that the actual lateral coefficient and slightly conservative. It should be noted that both
pressure distribution increases linearly from 2 · cu · D at the ground methods by Broms 共1964b兲 and Fleming et al. 共1992兲 converge to
surface to a value of 9 · cu · D at a depth of 3 · D 共see Fig. 2兲. Below the same solution when the friction angle of the soil approaches
a depth of 3 · D, they used a constant pressure distribution equal to 30°, which represents a majority of the average-strength cohe-
9 · cu · D 共Fleming et al. 1992兲. Using this pressure distribution and sionless soils 共i.e., for ␸ = 30°, K P = 3兲. For soils with friction
moment equilibrium of the pole-soil system, the required embed- angles less than 30°, the method by Broms 共1964b兲 will yield
ment depth 共z兲 can be obtained by solving the following equation: smaller embedment depths and for soils with friction angles
greater than 30°, the method by Fleming et al. 共1992兲 will yield
4.5 ⫻ cu ⫻ D ⫻ z2 − 10.5 ⫻ cu ⫻ D2 ⫻ z + 10.5 ⫻ cu ⫻ D3 − P smaller embedment depths. The method by Broms 共1964b兲 was
tested by Poulos and Davis 共1980兲 and comparisons with field test
⫻ 共L − 2兲 = 0 共2兲
results show a tendency for the measured capacities to be under-
The method by Fleming et al. 共1992兲 result in marginally estimated by about 30% 关as cited in Fleming et al. 共1992兲兴. Based
higher failure loads compared to Broms 共1964a兲 due to the differ- on the experiments performed by Barton 共1982兲, it was deter-
ence in assumed pressure distribution near the ground line. Thus, mined that the method by Fleming et al. 共1992兲 tended to under-
Fleming et al. 共1992兲 method will result in smaller required em- estimate the actual measured failure loads by about 6% 共Fleming
bedment depths than those calculated by using Broms 共1964a兲 et al. 1992兲. Fleming et al. 共1992兲 concluded that their method
method. Since Eq. 共2兲 was determined by taking the moment will be more accurate than the method proposed by Broms
about the base of the pole to simplify the calculations, the embed- 共1964b兲 for naturally occurring sand, where K P is greater than 3.
ment depths determined from Eq. 共2兲 were increased by 20% to Detailed derivations of Eqs. 共1兲–共5兲 are presented in McNames
account for the length of the pole below the point of rotation. This 共2008兲.
is a valid assumption for sheet piles 共Powrie 1997兲 and has been
adopted in this study for determining the embedment depth for
Rural Utility Service ROT Method
the method by Fleming et al. 共1992兲 for laterally loaded piles.
The 2004 RUS Bulletin 1724E-200 recommends a ROT method
for determining the required embedment depth for “most wood
Cohesionless Soils „Effective Stress Analysis…
pole structures in good soils and not subjected to heavy loadings”
Broms 共1964b兲 suggested that the lateral resistance per unit length as 10% of pole length plus 2 ft 关Rural Utilities Service 共RUS兲
of the pile embedded in cohesionless soil is 3 · D · ␥ · z · K P; where 2004兴. Though this method is crude and does not take into ac-
␥ is the effective unit weight of the soil and K P is the lateral earth count the pole and soil properties, it is widely used in current
pressure coefficient in passive condition. As shown in Fig. 2, as practice in the United States. The RUS also includes another em-
depth increases, the lateral pressure increases from zero at the pirical equation found in the Wood Preserving News for deter-
ground surface to a maximum value of 3 · D · ␥ · z · K P 共Broms mining appropriate embedment depths for round wood poles 关as
1964b兲. The passive lateral pressure coefficient, K P, can be cor- cited in Rural Utilities Service 共RUS兲 2004兴. This empirical equa-
related to the friction angle of soil 共␸兲 using Rankine theory: tion uses a soil constant 共Se兲 to represent a wide range of cohesive

冋 册
and cohesionless soils 共140 for good soils, 70 for average soils,
␲ ␾ and 35 for poor soils兲 关Rural Utilities Service 共RUS兲 2004兴. The
K P = tan2 + 共3兲
4 2 results presented in this paper are compared only with the ROT
Using this pressure distribution and moment equilibrium of the method used in current practice.
pole-soil system, the required embedment depth can be obtained
from the following equation:
Results and Discussions
z= 冑
3
2 ⫻ P ⫻ 共L − 2兲
KP ⫻ ␥ ⫻ D
共4兲
Effect of Soil Properties
Fleming et al. 共1992兲 proposed another method of calculating Fig. 3 presents the results for required depths of embedment cal-
the load capacity of laterally loaded piles in cohesionless soils. culated using Eqs. 共1兲–共5兲 as a function of total length of the pole.
Fleming et al. 共1992兲 suggested that the pressure distribution is Embedment depths for Class 2 poles in cohesive soils 共cu = 500
linear as proposed in Broms 共1964b兲, although they state that the and 2,000 psf兲 and cohesionless soils 共␸ = 20 and 40° and effec-
pressure is directly proportional to K2P rather than 3K P based on tive unit weight 110 pcf兲 are presented in Fig. 3. Also plotted in
the centrifuge experiments performed by Barton 共1982兲 关as cited Fig. 3 are the embedment depths calculated using the ROT
in Fleming et al. 共1992兲兴. As shown in Fig. 2, as depth increases, method 共10% of pole length plus 2 ft兲. The selected soil properties
the lateral pressure increases from zero at the ground surface to a represent the approximate lower and upper limits of the shear
maximum value of D · ␥ · z · K2P 共Fleming et al. 1992兲. Using this strength parameters of natural soils. Thus, the results presented in

76 / PRACTICE PERIODICAL ON STRUCTURAL DESIGN AND CONSTRUCTION © ASCE / FEBRUARY 2010

Downloaded 16 Mar 2010 to 134.129.122.119. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
ROT (10% + 2') Broms (Class H3, cu = 500 psf)
o
Broms (cu = 500 psf) ROT (10% + 2')
Fleming et al. (Class H3, cu = 500 psf) Broms (Class H3, φ = 20 )
Broms (φ = 20o) Fleming et al. (Class H3, φ = 20 )
o
Fleming et al. (cu = 500 psf) Broms (Class 2, cu = 2000 psf)
Fleming et al. (φ = 20o) Broms (Class 2, φ = 40 )
o
Broms (cu = 2000 psf) Fleming et al. (Class 2, cu = 2000 psf)
Broms (φ = 40o) o
Fleming et al. (Class 2, φ = 40 )
Fleming et al. (cu = 2000 psf) Fleming et al. (φ = 40o)
80

Error in Rule of Thumb (ROT) Method (%)


Cohesive Soils Cohesionless Soils
0.30
Pole Class 2
60
Pole Class 2
Normalized Depth of Embedment [z / L]

Cohesive Soils Cohesionless Soils


40
0.25
20

0.20 0

-20
0.15
-40

0.10 -60

-80
0.05
40 60 80 100 120 40 60 80 100 120
40 60 80 100 120 40 60 80 100 120
Length of Pole, L (ft) Length of Pole, L (ft)
Length of Pole, L (ft) Length of Pole, L (ft)
(a) (b)
(a) (b)

Fig. 5. Percentage error in depth of embedment calculated using


Fig. 3. Effect of soil types and soil properties on depth of embed-
ROT method: 共a兲 cohesive soils; 共b兲 cohesionless soils
ment: 共a兲 cohesive soils; 共b兲 cohesionless soils

Fig. 3 represent the practical lower and upper bounds of the re- cohesive soils 共cu = 1 , 000 psf兲 and medium dense cohesionless
quired embedment depths 共about 7–30% of the length of the pole兲 soils 共␸ = 30°兲. Also plotted in Fig. 3 are the embedment depths
of Class 2 poles in cohesive and cohesionless soils. As can be calculated using the ROT method. Embedment depths required
seen from Fig. 3, the ROT method, which does not take into for other commonly used pole classes 共classes 1, H1, and H2兲 fall
account the soil properties, significantly underestimates the em- in between the embedment depths bounded by pole classes 2 and
bedment depths in weak soils and overestimates the embedment H3. Thus, the results presented in Fig. 4 envelop the required
depths in strong soils. For cohesive soils, the method by Broms embedment depths of commonly used pole classes in medium
共1964a兲 is slightly conservative than the method by Fleming et al. strength cohesive and cohesionless soils 共about 9–28% of the
共1992兲 共by about 15%兲. For dense cohesionless soils, the method length of the pole兲. Class H3 poles require bigger embedment
by Broms 共1964b兲 is slightly conservative than the method by depths than those of Class 2 poles because the design lateral load
Fleming et al. 共1992兲; however, in weak cohesionless soils, the of a Class H3 pole 共7.5 kips兲 is bigger than that of a Class 2 pole
method by Fleming et al. 共1992兲 is more conservative 共by about 共3.7 kips兲. As can be seen from Fig. 4, for the most part in me-
15%兲. dium strength cohesive and cohesionless soils, the ROT method
underestimates the required embedment depths. Similar to the ob-
servations made in Fig. 3, results presented in Fig. 4共a兲 also re-
Effect of Pole Classes
veal that the method by Broms 共1964a兲 is slightly conservative
Fig. 4 presents the effect of pole class on required embedment than the method by Fleming et al. 共1992兲 共by about 15%兲 in
depths as a function of pole length. Both methods by Broms cohesive soils. As expected, both methods by Broms 共1964b兲 and
共1964a,b兲 and Fleming et al. 共1992兲 were used to calculate the Fleming et al. 共1992兲 yield the same embedment depths in cohe-
embedment depths for classes 2 and H3 poles in medium stiff sionless soils with 30° friction angle.

Discussion on Current Rule of Thumb Method


ROT (10% + 2')
Broms (Class H3)
Fleming et al. (Class H3) ROT (10% + 2') Fig. 5 presents the results for percent error in depth of embedment
Broms and Fleming et al. (Class H3)
Broms (Class 2)
Fleming et al. (Class 2) Broms and Fleming et al. (Class 2)
calculated using ROT method for a wide range of soil properties
0.30
and pole classes. Percent error in embedment depth is defined in
Cohesive Soils Cohesionless Soils the following equation:
Normalized Depth of Embedment [z / L]

cu = 1000 psf o
φ = 30
0.25
共Proposed Method − Rule of Thumb兲
%Error = ⫻ 100 共6兲
0.20 Proposed Method
where proposed method refers to either by Broms 共1964a,b兲 or by
0.15
Fleming et al. 共1992兲. It should be noted that the results plotted in
0.10
Fig. 5 were obtained using the combinations of stronger poles
supported by weak soils 共Class H3 poles in cohesive soils with
0.05
cu = 500 psf or in cohesionless soils with ␸ = 20°兲 and poles with
40 60 80 100 120 40 60 80 100 120 smaller design lateral loads supported by strong soils 共Class 2
Length of Pole, L (ft) Length of Pole, L (ft) poles in cohesive soils with cu = 2 , 000 psf or in cohesionless
(a) (b) soils with ␸ = 40°兲. Thus the results presented in Fig. 5 represent
the upper and lower limits of the percent errors in embedment
Fig. 4. Effect of pole classes on depth of embedment: 共a兲 cohesive depths calculated by ROT method for commonly used round
soils; 共b兲 cohesionless soils wood poles supported by natural soils.

PRACTICE PERIODICAL ON STRUCTURAL DESIGN AND CONSTRUCTION © ASCE / FEBRUARY 2010 / 77

Downloaded 16 Mar 2010 to 134.129.122.119. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
0.5
Broms (Class H3, φ = 20o)
cu - Undrained Shear Strength of Soil
Fleming et al. (Class H3, φ = 20o) D - Diameter of Pole
Broms (Class 2, φ = 40o) 0.4 L - Total Length of Pole
M - Ground Line Moment on Pole
Fleming et al. (Class 2, φ = 40o) z - Required Depth of Embedment
Error in Rule of Thumb (ROT) Method (%) 80
0.3
Saturated Sandy Soils

[z / L]
"Rule-of-Thumb"
60 Method
0.2
40

0.1
20
Using Broms (1964a)
Using Fleming et al. (1992)
0 0.0
1 10 100 1000
-20 . . 2
[cu D L / M]

-40 Fig. 7. Variation of required embedment depths in cohesive soils


40 60 80 100 120 with pole and soil properties
Length of Pole, L (ft)

Fig. 6. Percentage error in depth of embedment calculated using


ROT method in saturated sandy soils SLP _ S = f共L,D,␥,␾,M兲 = 再 D ⫻ L3 ⫻ ␥ ⫻ 关tan共␾兲兴1.5
M
冎 共8兲

where M = ground line moment on pole. These parameters are


defined in such a way that they increase with increasing soil
strength and pole size and they decrease with the increase in
As can be seen from Fig. 5, the percent error varies from applied moment. Thus, one would expect that the required nor-
+60% to about ⫺60% depending on the type and length of the malized depth of embedment 共z / L兲 would decrease with increas-
pole and soil properties 共⫾60%兲. A positive percent error indi- ing SLP_C and SLP_S.
cates that ROT method underestimates the embedment depth and, Figs. 7 and 8 plot the results for normalized embedment depths
hence, the design is unsafe while a negative percentage error in- as a function of strength-loading parameters for cohesive soils
dicates that the ROT method overestimates the depth of embed- and cohesionless soils, respectively. Results presented in Figs. 7
ment and, hence, the design is uneconomical. Fig. 6 plots the and 8 were obtained from different pole heights 共from 40 to 125 ft
percent errors in embedment depths in saturated cohesionless in 5-ft increments兲 of five different pole classes 共classes 2, 1, H1,
soils calculated using ROT method. As expected, the percent er- H2, and H3兲 embedded in cohesive soils with undrained shear
rors increase in the case of saturated sandy soils as compared to strengths 共cu兲 varying from 200 to 4,000 psf in 200-psf incre-
dry soils due the reduction in effective unit weight of soil. More ments and in cohesionless soils with friction angle 共␸兲 varying
results and comparisons of embedment depths for a variety of from 20 to 40° in 1° increments. Both dry and saturated cohesion-
pole classes and soil properties can be found in McNames 共2008兲. less soils are included in Fig. 8. Results obtained from both meth-
It should be noted that very weak soils, such as slurry, expansive ods by Broms 共1964a,b兲 and Fleming et al. 共1992兲 are plotted in
soils, and collapsible soils have not been considered in this study, Figs. 7 and 8 along with the range of embedment depths calcu-
as those soil types need case by case analysis to determine the
embedment depths.
0.5
φ - Friction Angle of Soil
γ - Effective Unit Weight of Soil
Proposed Relationships for Embedment Depth 0.4 D - Diameter of Pole
L - Total Length of Pole
M - Ground Line Moment on Pole
z - Required Depth of Embedment
Dimensionless Parameters 0.3
[z / L]

In order to generate general relationships to determine the embed- "Rule-of-Thumb"


Method
ment depths of self-supported transmission poles, regardless of 0.2
the pole and soil properties and loading magnitude, dimensionless
parameters have been developed. For both cohesive and cohesion-
less soils, dimensionless parameters that include the effects of 0.1
Using Broms (1964b)
pole properties, soil properties, and applied loading magnitude are Using Fleming et al. (1992)
developed. These parameters are termed as combined “strength- 0.0
loading parameters” 共SLP_C and SLP_S兲 and are defined in the 1 10 100 1000
following equations: . 3. .
[D L γ (tan (φ)) 1.5
/ M]

冋 册
Fig. 8. Variation of required embedment depths in cohesionless soils
cu ⫻ D ⫻ L2 with pole and soil properties
SLP _ C = f共cu,D,L,M兲 = 共7兲
M
78 / PRACTICE PERIODICAL ON STRUCTURAL DESIGN AND CONSTRUCTION © ASCE / FEBRUARY 2010

Downloaded 16 Mar 2010 to 134.129.122.119. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
Cohesive Soils Cohesionless Soils
[z / L] = a.[cu.D.L2 / M]b [z / L] = a.[D.L3.γ.(tan (φ))1.5 / M]b
0.5
Broms (1964a) Broms (1964b)
a = 0.8, b = -0.46 a = 0.73, b = -0.34
0.4 Fleming et al. (1992) Fleming et al. (1992)
a = 0.73, b = -0.48 a = 0.86, b = -0.38

0.3
[z / L]

0.2

0.1

0.0
1 10 100 1 10 100 1000
Fig. 10. Collapse of a 34.5-kV double circuit transmission line due
[cu.D.L2 / M] [D.L3.γ.(tan (φ))1.5 / M]
to foundation failure in Hancock County, Iowa 共picture courtesy of
(a) (b)
Hasib Bhuiyan of ITC Midwest兲
Fig. 9. Proposed relationships for determining embedment depths of
self-supported transmission pole foundations in: 共a兲 cohesive soils;
共b兲 cohesionless soils conditions in the field. Also, a factor of safety for the depth of
embedment can easily be incorporated in the proposed relation-
ships by using a strength reduction factor 关for cu and tan 共␸兲兴 or a
lated using the ROT method. Once again, it is clear that the ROT load factor 共for M兲. The proposed relationships are applicable to
method underestimates the embedment depths in weak soil con- not only wood poles but also to other types of poles as long as the
ditions and overestimates the embedment depths as the normal- pole is rigid compared to the stiffness of soil. These relationships
ized strength of soil-pole system increases. Results plotted in can also be used to back-calculate the moment capacity of general
Figs. 7 and 8 reveal that the correlations between normalized laterally loaded piles, given the depth of embedment.
depth of embedment and the strength-loading parameters are
unique regardless of the pole dimensions, soil strength, and ap-
plied loading and these correlations can be used to develop gen- Comparison of Proposed Relationships with Field
eral relationships to determine the embedment depths in cohesive Case History Data and PLS-Caisson Software
and cohesionless soils.
In early 2004, several transmission poles used to support a double
circuit 34.5-kV power line fell over in Hancock County, Iowa
Proposed Normalized Relationships 共Wen 2004兲. Fig. 10 shows a picture of the site, where the trans-
Based on the results presented in Figs. 7 and 8, normalized rela- mission pole foundation failures occurred. The poles used on the
tionships are developed to determine the embedment depths for line were western red cedar 55-ft Class H1 poles and were found
self-supported transmission pole foundations. The depth of em- on saturated cohesive soils. The embedment depth used in the
bedment can be determined using the relationships provided in design was based on the ROT method such that the poles were
Eqs. 共9兲 and 共10兲. For cohesive soils embedded in the ground 7.5 ft. Upon investigation, it was deter-

冋 册
mined that it was not the poles that failed but the foundations
z cu ⫻ D ⫻ L2 b
failed due to the inadequate embedment depths 共Wen 2004兲. This
=a⫻ 共9兲
L M foundation failure is a good practical example that the ROT
method does not work in weak soils conditions.
and for cohesionless soils

再 冎
Required embedment depths, for Hancock County site trans-
z D ⫻ L3 ⫻ ␥ ⫻ 关tan共␾兲兴1.5 b mission pole foundations, are calculated using the proposed rela-
=a⫻ 共10兲 tionships 关Eq. 共9兲 and Fig. 9兴 based on the data obtained from soil
L M
borings performed in the site 共Wen 2004兲, dimensions of the pole,
The coefficients a and b in Eqs. 共9兲 and 共10兲 are obtained from and estimated lateral load on poles based on the guidelines avail-
best fit regression analysis of the data presented in Figs. 7 and 8. able in the National Electrical Safety Code 共IEEE 2006兲. The
For cohesive soils, the method by Broms 共1964a兲 yields a = 0.8 calculated embedment depths are compared with the actual em-
and b = −0.46 共with R2 = 0.98兲 and the method by Fleming et al. bedment depths provided in the field and the embedment depths
共1992兲 yields a = 0.73 and b = −0.48 共with R2 = 0.99兲. For cohe- calculated using PLS-Caisson program 关as presented in Wen
sionless soils, the method by Broms 共1964b兲 yields a = 0.73 and 共2004兲兴. PLS-Caisson program is a computer software that can be
b = −0.34 共with R2 = 0.98兲, and the method by Fleming et al. used to determine proper embedment depths of transmission poles
共1992兲 yields a = 0.86 and b = −0.38 共with R2 = 0.98兲. embedded in layered soils 共Caisson, version 4.91 2002兲, although
The proposed relationships are plotted in Fig. 9 for both cohe- it is not widely used in practice. The PLS-Caisson program uses a
sive and cohesionless soils. Given the dimensions of pole 共L and modified version of the method by Broms 共1964a兲 to determine
D兲, soil properties 共␸ and ␥ or cu兲, and the applied loading 共M兲, embedment depths 共Caisson, version 4.91 2002兲. PLS-Caisson
the designer can obtain the required depth of embedment 共z兲 program uses a uniform pressure distribution for cohesive soils to
using either Fig. 9 or Eq. 共9兲 or 共10兲. These relationships can be be equal to 8cuD and for cohesionless soils, the pressure distribu-
applicable to a wide range of pole and soil properties and loading tion proposed by Broms 共1964b兲 is used 共Caisson, version 4.91
conditions. It should be noted that the ground line moment 共M兲 2002兲. Based on the data obtained from soil borings performed in
on pole can be estimated using either the design lateral load 共P兲 the site and using PLS-Caisson program, the required embedment
of pole 共for ultimate load conditions兲 or the actual working load depths of poles were recalculated 共Wen 2004兲. More detailed cal-

PRACTICE PERIODICAL ON STRUCTURAL DESIGN AND CONSTRUCTION © ASCE / FEBRUARY 2010 / 79

Downloaded 16 Mar 2010 to 134.129.122.119. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
Table 3. Details of the Field Case History Data from Hancock County, ships require only two soil strength parameters; undrained shear
Iowa strength for cohesive soils and friction angle for cohesionless
Parameters Boring # 4 Boring # 5 soils. Although it is cumbersome and costly to obtain soil borings
for each site or each individual pole location, there are less eva-
Diameter of pole, D 共ft兲 1.45 1.45
sive methods available for determining these soil shear strength
Length of pole, L 共ft兲 55 55 parameters.
Undrained shear strength of clay, cu 共psf兲 255 607 One simplified method for soil exploration uses a soil probe to
Ground line moment on pole, M 共kips-ft兲 189.5 189.5 obtain an estimate of the equivalent field standard penetration
Field embedment, z 共ft兲 共using ROT method兲 7.5 7.5 number. CHANCE makes a soil probe test that is portable and
Required depth of embedment, z 共ft兲: easy to use 共Hubbell Power Systems 2004兲. The soil probe test
1. Using PLS-Caisson software 19.4 12.4 measures the torque required to penetrate the soil which is then
2. Using proposed relationship 关Eq. 共9兲兴 converted to an equivalent standard penetration number 共Hubbell
a. Using Broms 共1964a兲 method 19.2 13.2 Power Systems 2004兲. Once the standard penetration number is
b. Using Fleming et al. 共1992兲 method 17.1 11.6 obtained, it can be converted to the shear strength of soils, such as
undrained shear strength or friction angle. Several correlations
exist for determining the friction angle and undrained shear
culations, estimated loading, PLS-Caisson program report, and strength from the standard penetration number 共Das 2007兲. There-
soil boring data summary sheets can be found in McNames fore, the proposed methods can easily be adopted in practice,
共2008兲 and Wen 共2004兲. instead of the crude ROT method and more expensive and time
Table 3 and Fig. 11 present the field case history data and the consuming computer programs.
comparisons of embedment depths calculated using proposed
methods with the embedment depths calculated using PLS-
Caisson program and ROT method. It is clear that the ROT Conclusions
method underestimates the depth of embedment in weak soil con-
ditions 共by about 60% in this case兲. The embedment depths cal-
Reliable and validated methods for analyzing laterally loaded
culated using the proposed relationships compare well with the
piles have been incorporated into the current study to generate
PLS-Caisson program predictions. As expected, PLS-Caissons
improved design methods for self-supported transmission pole
predictions are closer to the Broms 共1964a兲 method; the Fleming
foundations. The methods analyzed include Broms 共1964a兲 and
et al. 共1992兲 method deviates by about 12%. Even though the
Fleming et al. 共1992兲 for laterally loaded piles in cohesive soils
undrained shear strength was obtained by using weighted average
and Broms 共1964b兲 and Fleming et al. 共1992兲 in cohesionless
method 共McNames 2008兲, the proposed relationships still predict
soils. The self-supported transmission poles are assumed to act as
a reliable and accurate design. This suggests that the proposed
short-rigid piles in homogeneous soil foundations. When the re-
methods can be used in place of or in combination with the PLS-
sults of the methods presented in this paper are compared to the
Caisson program because the proposed methods are simple, easy
current design ROT method, it is evident that the ROT method
to use, and not time-consuming.
can significantly underestimate or overestimate the required em-
bedment depth depending on the pole class, pole dimensions, soil
properties, and applied load by as much as ⫾60%. A positive
Practical Implications percentage error indicates that ROT method underestimates the
embedment depth and, hence, the design is unsafe while a nega-
Based on the results presented in this paper, it is evident that the tive percentage error indicates that the ROT method overestimates
current ROT method can underestimate or overestimate embed- the depth of embedment and, hence, the design is uneconomical.
ment depths depending on the pole class, pole height, soil type, The main focus of this paper is to provide utility engineers
and soil strength. The proposed relationships, taking into account with reliable and accurate methods to design safe and cost-
pole and soil properties and applied loading, predict a much more effective foundations for transmission poles. Improved relation-
reliable and accurate embedment depth. The proposed relation- ships to determine the embedment depths for self-supported
transmission pole foundations are proposed. The proposed rela-
tionships take into account pole and soil properties and applied
0.5 Proposed method
using Broms (1964a) loading and provide normalized equations and graphs to deter-
Proposed method
using Fleming et al. (1992)
mine the embedment depths for a wide range of pole classes, soil
0.4 Depth of embedment calculated
using PLS-Caisson software
properties, and loading scenarios. The methods presented in this
Depth of embedment in field paper were compared to a case study where several transmission
0.3 based on rule of thumb method
poles fell over due to poor soil conditions and inadequate foun-
[z / L]

dation design. It is determined that the normalized methods pre-


0.2 sented in this paper predict embedment depths similar to the PLS-
Caisson program. This suggests that the proposed methods can be
0.1 used in place of or in combination with the PLS-Caisson program
Boring # 4 Boring # 5 because the proposed methods are simple, easy to use, and not
0.0 time-consuming. The methods proposed in this paper will not
1 10 100
only lead to safe foundation design they will also provide sugges-
[cu.D.L2 / M] tions for more efficient and economical designs. In a broader
picture, the proposed methods can also be used to analyze and
Fig. 11. Comparison of proposed relationships with PLS-Caisson design embedment depths for general cases of short-rigid laterally
software predictions and field case history failure embedment depths loaded piles.

80 / PRACTICE PERIODICAL ON STRUCTURAL DESIGN AND CONSTRUCTION © ASCE / FEBRUARY 2010

Downloaded 16 Mar 2010 to 134.129.122.119. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
Acknowledgments Hubbell Power Systems. 共2004兲. CHANCE encyclopedia of anchoring,
bulletin 01–9401, rev. B, Hubbell, Mo.
IEEE. 共2006兲. “National electrical safety code.” C2-C2007, New York.
The writers thank the following companies who contributed to
Keshavarzian, M. 共2002兲. “Self-supported wood pole fixity at the ANSI
this research at one point or another: Alliant Energy, ITC Mid- groundline.” Pract. Period. Struct. Des. Constr., 7共4兲, 147–155.
west, LLC, Power Line Systems, Inc., and Ulteig Engineers, Inc. Kinney, E. E. 共1959兲. “Correct embedment for pole structures.” Wood
The writers thank Henry Wen, Hasib Bhuiyan, Michael Hook, and Preserving News, American Wood Preservers’ Association, October
Otto Lynch for sharing their research findings and experiences. 1959.
Matlock, H., and Reese, L. C. 共1960兲. “Generalized solutions for laterally
loaded piles.” J. Soil Mech. and Found. Div., 86共SM5兲, 63–91.
McNames, C. 共2008兲. “Improved design of embedment depths for round,
References wood transmission pole foundations subject to lateral loading.” MS
thesis, College of Engineering and Architecture, North Dakota State
American National Standards Institute 共ANSI兲. 共1987兲. “American na- Univ., July 2008.
tional standard for wood poles—Specifications and dimensions.” Meador, N. F. 共1997兲. “Mathematical models for lateral resistance of post
ANSI O5.1, American National Standards Institute, New York. foundations.” Trans. ASAE, 40共1兲, 191–201.
Barton, Y. O. 共1982兲. “Laterally loaded model piles in sand: Centrifuge Poulos, H. G., and Davis, E. H. 共1980兲. “Section 7: Ultimate lateral
tests and finite element analysis.” Ph.D. thesis, Cambridge University, resistance of piles.” Pile foundation analysis and design, Wiley, New
U.K.
York.
Broms, B. 共1964a兲. “Lateral resistance of piles in cohesive soils.” J. Soil
Powrie, W. 共1997兲. Soil mechanics: Concepts and applications, E & FN
Mech. and Found. Div., 90共SM2兲, 27–63. Spon, London.
Broms, B. 共1964b兲. “Lateral resistance of piles in cohesionless soils.” J. Prasad, Y., and Chari, T. R. 共1999兲. “Lateral capacity of model rigid piles
Soil Mech. and Found. Div., 90共SM3兲, 123–156. in cohesionless soils.” Soils Found., 39共2兲, 21–29.
Broms, B. 共1965兲. “Design of laterally loaded piles.” J. Soil Mech. and Reese, L. C. 共1970兲. “Lateral load test on piles—Arkansas River project.”
Found. Div., 91共SM3兲, 79–99. J. Soil Mech. and Found. Div., 96共SM5兲, 1583–1604.
Caisson, version 4.91. 共2002兲. Power Line Systems, Inc. 共PLS兲, Wis. Rural Utilities Service 共RUS兲. 共2004兲. “Design manual for high voltage
Das, B. M. 共2007兲. Principles of foundation engineering, 6th Ed., Brooks/ transmission lines.” RUS Bulletin 1724E-200, U.S. Dept. of Agricul-
Cole Publishing, Pacific Grove, Calif. ture.
Fan, C. C., and Long, J. H. 共2005兲. “Assessment of existing methods for Seiler, J. F. 共1932兲. “Effect of depth of embedment on pole stability.”
predicting soil response of laterally loaded piles in sand.” Comput. Wood Preserving News, American Wood Preservers’ Association, No-
Geotech., 32, 274–289. vember 1932.
Fleming, W. G. K., Weltman, A. J., Randolph, M. F., and Elson, W. K. Wen, H. 共2004兲. “Hancock 34.5 kV Windfarm line.” A memorandum to
共1992兲. Piling engineering, 2nd Ed., Blackie & Son, New York. Bob Vosberg, July 12, 2004.

PRACTICE PERIODICAL ON STRUCTURAL DESIGN AND CONSTRUCTION © ASCE / FEBRUARY 2010 / 81

View publication stats Downloaded 16 Mar 2010 to 134.129.122.119. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright

Вам также может понравиться