IEEE Transactions on Power Delivery, Vol. 13, No.

4, October 1998 1385

APPLICATION OF THE RULING SPAN CONCEPT FOR OVERHEAD LINES

IN MOUNTAINOUS TERRAIN.

L.M. Keselman Y. Motlis, Senior Member

Consultant, Tel Aviv, Israel Ontario Hydro, Toronto, Canada

Abstract. The ruling span method is used by line engineers to longitudinal displacement. However, tension will change differently
calculate the approximate tension in a multi-span dead-ended line in spans of unequal length, and the insulator swing does not
section. Providing satisfactory results for level spans, it may, completely equalize the tension in adjacent spans.
however, result in an error of over 30% if it is used to calculate The method of "local" tensions (sometimes called "true" tensions)
tensions in significantly long and steep spans e.g., in mountainous predicts tensions in individual spans more accurately than the
terrain. Following the ruling span concept in general, this paper traditional ruling span method because it takes into account the
presents equations to calculate tensions for inclined single spans or longitudinal insulator swing effect due to changes in load on
multi-span line sections based on calculation of conductor length in conductor (temperature or ice)[3].
three-dimensional space. These equations include the equivalent In inclined spans, the freedom of displacement of the suspension
parameters such as equivalent inclined span, equivalent slope, insulator strings toward the lower structure is somewhat limited
equivalent load, equivalent effective tension - all needed to take into compared to the insulator swing in level spans. For lines crossing
account the effect on tension of the changing geometry and length of rough terrain, there may be spans of significantly different length
conductors in significantly inclined spans. The numerical examples in which wind load can be unequal (the spatial effect). In
show the effect of span inclination on conductor tension calculated mountainous areas, the ice load in the adjacent spans with large
using these equations. A simple and accurate equation (without difference in elevations can also be unequal. For all cases of
hyperbolic functions) is derived to calculate conductor length in an unequal load on conductors (ice and/or wind) in the adjacent spans,
inclined span. This paper also addresses calculation of sags in non- it is possible to calculate the equivalent loads for non-level spans
level spans. The presented method integrates the practicality of the using the three-dimensional method of vectors presented in [4,5].
ruling span concept with the accuracy of the threedimensional vector
method for analysis of inclined spans. This method is most efficient For the inclined line sections, the challenge is to propose a method
for overhead lines in mountainous terrain. to calculate conductor tension that could be used as a ruling span
tension, and would match closely the "local" tensions in those
Key words: overhead line, ruling span, sag, tension, equivalent spans. The method of equivalent parameters presented in this
parameters, three-dimensional method, mountainous terrain. paper is capable of meeting this challenge using the ruling span
concept as the base. The distinctive feature of this method is the
I. INTRODUCTION introduction of the equivalent inclined (ruling) span through the
equivalent parameters that are based on the accurate vector analysis
Transmission line engineers have revealed a long time ago some of the inclined spans in three-dimensional space [4,5]. The
significant errors when using the traditional ruling span method to calculation of tension is based on conductor length that changes due
calculate conductor tensions for spans having large differences in to the change of its geometry in the long and steep spans especially
elevations and span lengths [1,2]. when affected by strong side winds.
Conductor tensions in a multi-span line section can be calculated
using the ruling span method or the method of "local" tensions. The 11. NOTATION AND UNITS.
ruling span method assumes that the horizontal tension is equalized
in all spans of a dead-ended line section. In accordance with the A = total area of conductor, mmz
traditional ruling span method, the ruling span length is defined by 01 = temperature coefficient of linear expansion, 1/OC
(cf,3/c[xi)0'5 where fxiis the horizontal projection of the (i)
E = modulus of elasticity, N/mm2
eqv. = index denoting equivalent parameters
span. Although this equation allows for many assumptions and
f, f,, fy, f, = sag and its projections
simplifications (hence, an approximate method), it serves line
engineers well for design analysis of lines at conductor temperatures
6 = angle between axis T and deflected conductor plane AB
below 9OoC. It assumes that the insulators are effectively of an L = conductor length, m
infinite length, therefore the suspension points between the spans are = length of the chord in inclined span, m
free to move without restraint regardless of the magnitude of the m, n = indexing the existing and new state of conductor
respectively.
N, R = angular coefficients in the conductor plane
PE-199-PWRD-0-11-1997 A paper recommended and approved by
the IEEE Transmission and Distribution Committee of the IEEE Power p = resulting load per unit length of conductor, Nlm
Engineering Society for publication in the IEEE Transactions on Power To = tension at "zero" point, N
Delivery. Manuscript submitted July 16, 1997; made available for t = conductor temperature, OC
printing November 1 1, 1997. To = abscissa of "zero" point in conductor plane
T = abscissa of any point along the conductor
x,y,z = coordinates in the three-dimensional system of coordinates.

0885-8977/98/$10.00 0 1997 IEEE

1386
111. THE THEORY. A. Calculation of conductor length in inclined sDans.
Based on the state-of-equilibrium equation for a catenary in t h r e e
The purpose of this paper is to derive the equations and their dimensional space [4], the conductor length in a single inclined
components in order to calculate tension at the "zero" point (the low span can be determined as:
point for conductor in a vertical plane) on the inclined catenary. The
equation for conductor length is used to calculate the change in To Pro P(lR-qJ
conductor tension due to the change in loading conditions. This paper L=-[sinh-+si* 1 (1)
presents a simple (without hyperbolic functions) equation for P TO TO
conductor length in inclined spans the accuracy of which is
acceptable for most of the practical applications. where 'CO is derived in [4,5] as following:

Figure 1 shows an inclined span of conductor in three-dimensional PlN

space deflected by wind. ro=-IR +--arcsinh
To
plR (2)
2 P 2T0sinh---
2TO

ARer transforming Eq.1 and combining Eq.1 and Eq.2, obtain

Eq.3 for conductor (exact) length in a single inclined span:

-
R=CQS6, N = (1 COS^^)^.^ = S h 6 , where R and N are the
angular coefficients[5] contributing to the high accuracy of Eq.3
and Eq.4. ARer transforming Eq.3, the approximate conductor
length in a single inclined span can be calculated using Eq.4:

Eq.4 is called an approximate equation to distinct from the exact

method involving hyperbolic functions. This equation is simpler
Figure 1. Inclined Span of Conductor Deflected by Wind and more efficient for inclined spans compared to the traditional
ruling span equations because it is derived using a more accurate
It is worth clarifying the meaning of the "equivalent" and "effective" vector method for three-dimensional analysis of inclined spans.
tensions used in the equations below: "Equivalent" denotes conductor
tension equalized in all spans between two dead-ends. "Effective" is B. Calculation of conductor tension.
defined as that value of the tension along the conductor in a single
span which satisfies Hooke's law, i.e., it is the tension resulting in 1. Single span.
the same elastic deformation of the conductor as the elastic The change-of-state Eq. 5 is us& to calculate tension in a single,
deformation due to the variable tension at any point along the level or inclined, span:
conductor. The term "effective equivalent" tension is used in this
paper to define the effective tension averaged for the equivalent Tefm (5)
inclined (ruling) span, and it reflects the integral affects on the Ln[l---a(t,-to>l=L,[l---a(f~-fo>]
AE AE
conductor along the dead-ended line section [1,2].

For level spans, the effective tension iS Usually assumed as the Substituting L, and L, from Eq.4, Eq.5 is transformed into Eq.6;
tension at the conductor low point. For spans with small inclinations,
the effective tension is assumed sometimes at the mid-point of the Z&- 2 3 4 -I,Tefm p,Z,R,
p,Z,R, 2 3 4
span. The three-dimensional method [4,5] for design analysis of - - - - - - - - ~ " ~ ~ ( ~ ~ - ~ ~ - ~ ~ ~ - - ~ ~ , + ~ ~ ~ - ~ ~ (6)
inclined line sections assumes that the effective tension is at the uT; I*
"zero" point in each span under consideration.
For a single dead-end span, 4'" =I, =e, Then:
 1387
b) Equivalent load, Peqv:

(14)

For most practical applications, the effective tension can be

calculated using Eq. 8 (a short form): c) Equivalent parameter (pPR'),,:
T,, = T,*R' (8)
Equation 8 for inclined spans gives the same result as T,, = To for
level spans.
Combining Eq.7 and Eq.8, obtain Eq.9:

When the loads in the spans are equal (pi is constant in Eq.15),
the equivalent parameter (PR*)eqv.is calculated using Eq. 16:

where T, and T, are the tensions at the "zero" point on the catenary.
The integral form of the effective tension as defined in [1,5] can be
calculated using the following Eq.10:
d) Equivalent inclined ruling span, I,, :

(101

Substituting Eq.2 into Eq.10, obtain Eq.11:

e) Effective equivalent tension, Tel.egv

:

After a series of transformations of Eq.11, and using:

sinhk=k+k3/6, coshk=1+k2/2, Eq.12 (a long form) is derived as: For a dead-ended multi-span line section, Eq.6 should b e written
for each span of the line section under consideration, and all the
respective components in these equations summed up. The sum of
the last components on the right side of the equations is very small
and can be neglected. Assuming T, and T, as equal respectively in
all spans, and neglecting the difference in the span length for old
24T; 24Tt and new state of conductor, P, =e,, Eq.6 is rewritten to obtain
Eq.19:
2 3 4
2. Multi-span section.
'(1jTefJ
--- E UP& -
To derive the change-of-state equation and to calculate conductor AE1i 24c Eli
tension for inclined dead-ended multi-span line section, the following
equivalent parameters are introduced for each inclined span (i),load
(pi), slope (ej); and tension (TJ which is equal at the "zero" points
in all spans:
a) Equivalent slope, Itegv.,or Reqv.,,,:
P 1
Combining Eq. 15, Eq.18 and Eq.19, a general form of the change-
of-state Eq.20 is obtained for an inclined multi-span section:
1388
projections of sag can be determined as:

f,=fp,/p; fy=fpy/p; and f,=fp,/p, and the abscissa of the

"zero" point (x,,) is equal to:
Eq.20 is similar to Eq.7 but it includes the equivalent parameters to
take into account the integral effect of the inclined span components
(span length and inclination) on the effective tension.
Indirectly, the above equations take into account the effect of wind
L10
direction because the angular coefficients N and R depend on the
components of the load p (px, pr, p,) which can de determined for
any wind direction using the equations in [SI. When wind is absent,
the conductor is located in a vertical plane, and for ice load equal in
each span, the equations can be simplified by replacing:
N = sin0 = !,/e; R = cos0 = P,/P; p = p,.

C. Conductor sag in inclined soans.

In this paper, a conductor sag is defined as a difference of ordinates
along axis "n"between conductor and chord AB. Figure 2 shows
sags (f) in a deflected plane and their projections on y- and z- axes
of the Cartesian system of coordinates. The sags are calculated for
each span and do not require use of the equivalent parameters.
Therefore, the sags can be calculated using the catenary equations.
Figure 2 shows the conductor sag deflected by wind in an inclined
span, and the related parameters to calculate the sags using the
following equations.
a)sag (f) for any point on the conductor:

Figure 2. Conductor sag deflected by wind in inclined span

b)sag (f,,) for the mid-span point:
IV. NUMERICAL EXAMPLES.
f-=-[eosh---
To $ 7 0 -cos IN
1-7 Case 1.
P To TO General input data: conductor "DRAKE", 795 kcmil, ACSR, 26/7;
span length = 100 m; slope=OO(level span); radial thickness of ice
The abscissa T of any point in the conductor plane is tied to the on conductor = 25.4 mm; f = - l0C;p,=53.6 N/m;
abscissa (x) in the Cartesian system of coordinates as: E=74460 N/mmz; a=18.9*1Oe6 U0C, A=468.5 mm'.
The assumed initial condition is: horizontal conductor tension of
21991 N at tm=16'C, conductor weight p1 = 15.9 N/m. For a
level span, R=cosB=l, P=P,=100 m, and using Eq.9:

--T, 53.62*10@*74460-21991 15.~*10@*74460-

468.5 24*c 468.5 24*219912
If wind is normal to a span, the px=O, then: x = T /PR e, or:
1: = xPR I e,.In the absence of wind, R = cos6 = e,/$;
N = !,/e, =x, r0 = x ~ , then:
find T,= 45632 N

P Case 2.

When conductor is in a vertical plane (no wind), the sag is vertical. The input data is the same as for Case 1 but the slope is 300.
When the sag is not vertical e.g., inclined spans and side wind, the Then, P,=lOO m, P,=57.735 m, R=[l-@z*P,/p*l)3°..5 =
=P,/P=100/115.47=0.866; T, = 21991 N.Using Eq. 7:
 1389

--Tejn 53.62*115.472*0.8664*74460-
468.5 24*e

---
-26006 Te$ogv.n - 24.72*474.336L*'74460
15.~*115.4P*0.8664*74460_18~9~,0-6~74460(-17~, -- ---- - 25538
468.5 24 *219912 468.5 24*c 468.5

find T,=42067 N, and T,,=48612 N , where TCt, is calculated using

Eq.12 as following: 15'92*396'8392*74460-18.9 * 106*74460 *(16- 16),
24 *219912
42067*0.866 53.62*115.472*0.8662)+
[(I +
T&= 1+5 3 . 8 * 1 15.472*0.8664 12*42067'
24 *42067'
Substituting the calculated equivalent parameters into Eq.20,
find T, = 38450 N, and Tsf.eqv.n=40802 N.

OS2( 1 + 53.62*115.4?*0.868) V. DISCUSSION ON T H E EFFECT OF SPAN INCLINATION

8 *4206? ON CONDUCTOR TENSION.
+ =48612N
0.866'(1+ 53.62 * 115.4?*0.868)
Reviewing the summary of calculations given in Table 1 it is
24 *420672
important to note the following:
with wind load on conductors, tension is increased by 34%when
For comparison, using the short format (Eq.8), T , , = T,/R = span inclination is increased;
42067/0.866 = 48576 N that is close to the result (48612 N) using 0 with ice load on conductors, tension is decreased by 14%when
the long format (Eq.12). span inclination is increased;
at very low temperatures, tension is decreased by 10% in span
Case 3. of 100 m, when span inclination is increased. However, conductor
Line section consisting of two spans, 100 m and 500 m; slope in each tension is increased in the two-span section where the effect of the
spanis equal30'; wind loadpy=18.9 Nlm; t,=tm=16'C; Tm=21991 500 m span prevails.
N; p,=15.9 N/m, and p=24.7 N/m. the largest difference between the equivalent and local tensions
Determine parameters for each span: e,, = 100 m; e,, = 57.735 m; takes place for the following cases: a) very short insulator strings,
,'k = 115.47 m; less than 0.5 m; b) side wind, different than 90' transverse;
N, = (18.9*0+15.9*57.735)/(24.7*115.47)=0.32186; c) unequal wind load in spans with different inclinations; d) very
R, =(1-N2)"' = 0.94679. large difference in span length (more than 5 times);
Px2 =500 m; lz2= 288.675 m; P2 = 577.35 m; the largest difference in tension is for the slope of 40' and
Nz =(15.9*288.675)/(24.7*577.35)=0.32186; R2 = 0.94679. temperature of O'C in the presence of wind, and it is associated
Because the loads in both spans are equal, Eq. 13 and Eq.16 are used with the movement of the "zero" point toward the mid-span point.
to find the equivalent parameters r(ePy,,,,, and (1R2),qv.n,m
respectively. For the conductor in the vertical plane, the "zero" point is more
distanced from the mid-span point and the tension is smaller.
115.47+577.35 =o.94679 as the length of the inclined span and of the conductor is
-+-115.47 577.35 increasing, there is an obvious increased effect of the additional
0.94679 0.94679 load on conductor and decreased effect of the temperature; or
0 as the span inclination is decreasing, the effect of the additional
load on the conductor is decreasing, and the effect of the
temperature in increasing.
Sensitivity analysis were done for spans of 100, 500, and 1000
115.47+577.35 =o.866 m, and for inclinations of 0 ' and 30'. It was found that for spans
Reqv.v.n=
115.47
-+- 577.35 up to 500 m, the error is less than 1%if Eq.8 is used in place of
0.866 0.866 Eq.12. Therefore, for inclined spans up to 500 m, it is more
convenient to calculate (TJ using Eq.8 (a short form). For spans
longer than 500 m or for shorter spans with reduced tension
(increased sag), use of the more accurate Eq. 12 ( a long form) is
1~ 4 *0.946794
7 ~ +577.353*0.94679 =474.336 recommended.
(lRz),= 0 To verify the accuracy of the presented method of equivalent
115.47+577.35
parameters, the tensions calculated using this method were
1390

compared with the method of "local" tensions that takes into account 6. The presented method integrates the practicality of the ruling
the longitudinal swing effect. A two-span line section consisting of span method and the accuracy of the method of "local" tensions
100 m and 500 m spans was considered, insulator string length of 3 with the vector method [SI of design analysis in three-dimensional
m with a weight of 1000 N was assumed. space which is most efficient for overhead lines crossing
-
The errors, i.e. (TeqvTlocal)/Teqv, are as following: in the 500 m mountainous terrain.
span, from (0.01 to 0.33)% without wind, and up to 0.65% with
wind. In the 100 m span, from (-1.22 to +2.32)% without wind, and VIZ. REFERENCES.
up to 7% with wind for an inclination of 40'.
[l]. P.F. Winkelman, "Sag-Tension Computations and Field
Table 1. Summary of calculations, Cases 1,2, and 3. Measurements of Bonneville Power Administration", N E E
Transactions, Vo1.78, pp.78, pp.1532-1548, February 1960.
[2]. L.M. Keselman, Fundamentals of Mechanical Design of
Overhead Lines, in Russian, Moscow, Energoatomizdat, 1992.
[3]. "Limitations of the Ruling Span Method for Overhead Line
Conductors at High Operating Temperatures", Draft-Report of the
IEEE Task Force "Bare Conductor Sag at High Temperature",
discussed at 1997 IEEE/PES Summer Meeting, Berlin.
[4]. L.M. Keselman, Design Method for Overhead Line
Conductors in Mountains, in Russian, Moscow, Energoatomizdat,
1981.
[5]. Y. Motlis, L.M. Keselman, "Enhanced Analytical Design
Method for Overhead Line Conductors in Non-Level Spans", Proc.
IEEE/PES T&D Conference, pp.359-365, Los Angeles, 1996.

100 26765 27009 27773 29192 31558 L.M.Keselman was bom in Ukraine in
1 1.0491 1.0376 1.0907 1.1791
I 1 1922. He holds a M.Sc. in power system
electrical engineering from the Tashkent
Industrial University (1943), Ph.D from
Moscow Polytechnical University (1968),
and D.Sc. from the Moscow University
of Transport Engineers (1986). He has
45 years of experience in design of
52641 52378 51535 49989 47590
1 0.995 0.979 0.9496 0.9041
transmission lines including work from
I I I E 1969 until 1992 as a Head of the R&D
500 24315 24395 24638 25126 26002 Laboratory for overhead lines in
1 1.0154 1.0133 1.0333 1.0694
mountainous areas. He is an author of over
100+500 24802 24896 25191 25777 26817 90 publications including two books and holding 26 patents in
1 1.0038 1.0157 1.0393 1.0812 transmission lines area. Since 1994, he is a Consultant for the
Israel Electric Corporation, Tel Aviv, Israel.

Yakov Motlis (M85, SM90) was bom

VI. CONCLUSIONS. in Stalingrad, Russia, in 1942, received
his M.Sc. degree and engineering
1. This paper supports that the ruling span method is applicable also diploma in power systems electrical
for significantly non-level spans with some corrections that take into engineering from Lviv Polytechnical
account the integral effect on tension of the equivalent parameters University, in 1963, and has worked
(inclined spans, load, slopes, and effective tensions). for 34 years in design and construction
2. Equations are given to calculate the equivalent parameters and of overhead and underground,
tensions at the "zero" point using the three-dimensional vector distribution and transmission lines up
analysis method for inclined spans. The equation to calculate the to 500 kV in Ukraine, Israel, and
appmximate length of oonduotor in an inclined span is approaching Canada. He is a Professional Engineer
the accuracy of the catenary equation, of Ontario, Chairman, IEEE TF "Bare
4. There are sometimes significant (up to 34%) differences in Conductor Sag at High Temperature", Member of the Canadian
tensions compared to those calculated using the traditional ruling span National Delegation IEC/TCl 1, Canadian Representativeat CIGRE
method without insulator swing effect. WG12-1. For the last 16 years he is a Design Engineer - Specialist,
5. Comparison of the presented method with the "local" tensions Ontario Hydro, presently with TRANSCO. H e is a winner of
method, shows that in most cases the errors are within the acceptable Ontario Hydro two New Technology Awards, 1990 and 1994; and
practical limits. co-author of over 20 technical papers.