1

Minimum Voltage Rating of Sheath Voltage

Limiters in Underground Cable Systems: The
Influence of Corrugated Cable Sheaths.
Phillip Nichols, University of Newcastle, Australia.

Abstract—Sheath voltage limiters (SVL) are a critical compo-

nent of an underground cable’s protection system. SVLs must
be sized such that, under steady state fault conditions, no
conduction through the arrester occurs. This could otherwise
lead to a catastrophic failure. A SVL failure [1] under steady
state fault conditions prompted an investigation into the adequacy
of traditional approaches used to model cable sheaths and
underground cable networks. Injection tests have demonstrated
that the type of sheath used has a large influence over the sheath
voltage profile under faulted conditions. This is particularly
evident with corrugated sheaths. In this paper, two common types
of corrugated sheaths are investigated - the helically corrugated
sheath (HCS) and the annularly corrugated sheath (ACS). The
paper establishes improved models for each type of sheath and Figure 1. A Cable Sheath Exhibiting Helical Corrugations
compares these models to sheath voltage profiles obtained from
injection tests on three different types of underground cable
systems.
Index Terms—Corrugated Sheath, EMTP-RV, Underground The main geometric characteristic of the HCS is that it
Cables, Sheath Voltage Limiter maintains a uniform cross section in any plane perpendicular
to its axis [2]. A typical cross section of a HCS is displayed
in Figure 2. The figure shows that the sheath cross section is
cylindrical and contacts both the primary insulation and the
serving of the cable at a single point. Moving longitudinally
I. I NTRODUCTION along the cable, the sheath rotates around the core, always
maintaining a point of contact on the serving and primary
PPROPRIATELY sizing sheath voltage limiters (SVL) is insulation. Hence the sheath maintains a constant inner and
A becoming increasingly more important as transmission
networks become more complex. When SVL ratings are
outer radius and has a constant offset from the axis of the core.
The inner radius of the sheath is therefore equal to the average
exceeded under steady state fault conditions, the results can of the outer radius of the primary insulation and the inner
be catastrophic, potentially endangering the public. Injection radius of the cable serving, minus half the sheath thickness.
testing on different underground cable networks has shown The thickness of the sheath is added to the inner radius to
that both the sheath bonding methodology and the type of obtain the outer radius.
sheath used strongly influence the minimum voltage rating of
the SVL. The purpose of this paper is to analyse two types of Insulation
corrugated sheaths which are in widespread use in the power Serving
system. These are the helically corrugated sheath (HCS) and
the annularly corrugated sheath (ACS). The major difficulty
associated with modelling corrugated sheaths is the conversion
of the non-cylindrical shape into an equivalent cable model Core
for programs such as EMTP-RV. Traditionally, an averaging
technique, such as the geometric mean radius (GMR), has
been applied to convert the corrugations into an equivalent
cylinder. The primary objective of this paper is to establish the Sheath Position
accuracy associated with the traditional model and, in cases Moves
where this approach is deficient, to propose new steady state as the Cable
models. To quantify the accuracy limits of each model, sheath Aluminium Sheath Progresses
voltage profiles for both the proposed and traditional models
are compared to injection test results on three different cable Figure 2. Axial Cross Section of a Helically Corrugated Sheath
networks, containing cables with HCS and ACS.

The constant sheath cross section and corrugations cause the

sheath to have a longer effective length than that of the core
II. C ORRUGATED S HEATH T OPOLOGIES or the serving of the cable. The type of bonding used (that is
whether current flows in the sheath or not) influences how the
A. Helically Corrugated Sheaths length extension affects the impedance and admittance of the
cable model. This effect is described in Section III.
A sheath with helical corrugations is shown in Figure 1.

Phillip Nichols is completing a part time PhD with the University of

Newcastle, Australia. He also works full time with the System Plan- B. Annular Corrugated Sheaths
ning section at Ausgrid, Australia. e-mail: [email protected] or
[email protected]. A sheath with annular corrugations is shown in Figure 3.
 2

VLn = [Zno + Zn,n+1 + Zn+1,i ] lILn − Zt,n lILn (1)

VLn = Pn,n+1 ILn (2)

The earth return impedance Ze , used in the models proposed
by this paper, is given by the finite length expressions given in
[7], [8], as this model has been found to be in better agreement
with field measurements [7], [8]. The EMTP-RV models used
for comparison utilise the infinite expressions derived by
Pollaczek [9]. As the radii of the insulators between conductive
Figure 3. Cable Sheath Exhibiting Annular Corrugations shells are changed in this paper, the relative permittivity of
the equivalent insulator must be adjusted to obtain the correct
values of capacitance and conductance. Relative permittivities
The longitudinal cross section of an annular sheath differs in this paper are determined using measurements of the cable
from the helical in that it is coaxial with the core. This means capacitance and loss factors. Procedures such as those in [9],
that the inner and outer radii of the sheath are continually vary- [10] should be used when measurements are not available.
ing along the length of the cable. The continuous variability
of the radii has implications with regard to the application
of the length extension caused by the corrugations. These B. Proposed Model For Sheaths With Helical Corrugations
implications will be further explored in Section III.
As the inner and outer radii of the HCS remain constant
relative to the core, the impedances in the inner loop are also
III. M ODELLING C ORRUGATED S HEATHS constant over the corrugation period. The sheath is, however,
not coaxial but slightly off centre. This means that the theory
The difficultly in modelling corrugated cable sheaths in for an off centre cable must be employed for the inner loop.
programs such as EMTP is that cables are modelled as a series The inner loop impedance for an eccentric conductor of length
of coaxial shells. Non-cylindrical sheaths must therefore be L, surrounded by a finite thickness sheath, is given by the
converted into an equivalent coaxial model for use in these following [6], [12], [13] (where the radii in the following
simulation tools. To the author’s knowledge, there is very formulae are given in Figure 5. The constant b is the radial
little information concerning the accuracy of models involving distance between the centre of the core and the centre of
corrugated sheaths. Most works such as in [3] quote that the the sheath).√Kn (x) and In (x) are modified Bessel functions
GMR of the sheath should be used. This section provides a and p = jωµ0 µr σ. The relative permeabilty µr and the
methodology to convert both helical and annular corrugated conductivity σ are properties of the sheath.
sheaths into equivalent coaxial cylindrical sheaths, taking into
account the geometric properties given in Sections II-A and
II-B. ZL = [Z1o + Z12 + Z2i ] LIL1 − Zt LIL2 (3)
!
2
A. EMTP Cable Models jωµ0 ri,s − b2
Z12 = log +
The current EMTP-RV model must be examined in order to 2π ri,s rc,o
create an equivalent coaxial model for each corrugated sheath. ∞  n
EMTP-RV utilises the model in [4] to calculate the component 1X b
internal impedances and potentials that make up a cable. The [Fn In (pri,s ) + Cn Kn (pri,s )] (4)
component impedances form loop impedances and potentials, σ n=1 ri,s
which in turn can be transformed into phase quantities. Figure
4 shows the impedance and potential loops for a single core
cable with a core and sheath. The admittance can be obtained Fn =
by inverting the potential matrix. This technique has been  n  
p b (µr + 1)n
generalised to n loops in [5]. Kn (pro,s ) − Kn+1 (pro,s ) (5)
παDn ri,s pro,s
Ze
+ + Cn =
IL2 Z23 IL2 n 
P23
 
VL2 Z2o VL2 p b (µr + 1)n
- -
− In (pro,s ) + In+1 (pro,s )
Z2i Zt παDn ri,s pro,s
+ + (6)
VL1 VL1 IL1 P12
IL1 Z12
Z1o  
- - n
Dn = (µr,c − 1)In (pri,s ) − In+1 (pri,s )
Cable Loop Impedances Cable Loop Potentials pr
 i,s 
n
Figure 4. Cable Loop Impedances and Potentials (µr,c + 1)Kn (pro,s ) − Kn+1 (pro,s )
pro,s
 
The loop equations for the cable in Figure 4 are given by (1) n
and (2). In these equations Zn0 is the outer surface impedance − (µr,c − 1)Kn (pri,s ) − Kn+1 (pri,s )
of the nth shell, Zn,n+1 is the mutual impedance between pri,s
 
the core and the sheath. Zn+1,i and Zt,n and are the inner n
surface impedance and the transfer impedance between inner (µr,c + 1)In (pro,s ) + In+1 (pro,s ) (7)
pri,s
and outer surfaces of the nth shell respectively. There formulae
for each component impedance are given in [5], [6]. In these To convert the loop impedance in (3) into the model
equations n = 1 for the inner loop and n = 2 for the outer described by (1), the HCS must be changed into a cylindrical
loop. The impedance Zn+1,i for the outer loop is the earth shell. The resulting cylinder is coaxial and centred around the
return impedance and is denoted Ze . Similarly, the transfer GMR, if the corrugations are flattened longitudinally. How-
impedance Zt,n for the outer loop is zero. ever, the length of the sheath is now extended as is illustrated
 3

Figure 5. The effective extended length (EEL) is found through conductor. This requires two different models for the HCS
approximating the corrugations with the formula to be created, depending on whether current can flow in the
sheath or not.
f (x) = A sin(bx) (8)
where A is the amplitude of the corrugation, b = 2π The relative magnetic permeability of the insulator between
lp and the core and sheath is altered in order to equate the inner
lp is the length of the corrugation period. The EEL of the loop impedances in (10) to those in the equivalent model in
flattened corrugated sheath can be estimated by: (1). The EEL model that applies when current flows in the
Z bp sheath (crossbonded and solidly bonded systems) is given by:
L= 1 + (f 0 (x))2 dx (9)
a [Z1o,c l + Z12 l + Z2i L] IL1 − Zt LIL2 =
[Z1o L + Z12,e L + Z2i L] IL1 − Zt LIL2 (14)
Sheath Corrugations
Sheath Thickness Solving for therelative magnetic permeability in Z12,e =
Extended Effective jωµr,e µ0 r
Min Distance Avg Distance GMR ro,c ro,s ri,s Length. Flattening the 2π log ro,s
i,s
gives:
From Core From Core Corrugations
rc,o  
Core Core

Original Cable Length

Z1o,c l − Z1o L  l
Max Distance From Core ur,e =    +   Z12 (15)
ro,s r
Ljµ0 f log ri,s Ljµ0 f log ro,s i,s
Serving
Serving
The outer loop impedance for the EEL model is obtained
from:
Figure 5. Helical Sheath Model

As the lengths of the core and sheath are different, issues [Z2o L + Z23 l + Ze (L)] IL1 − Zt LIL2 =
arise with the mutual impedance between the core and sheath. [Z2o L + Z23,e L + Ze (L)] IL1 − Zt LIL2 (16)
Furthermore, corrections to the core impedance need to be
made in the loop impedance. For example, cables with sec- The equivalent
 relative magnetic permeability in Z23,e =
tored cores have a lower resistance than is calculated with jωµr,e µ0 ro,c
log ro,s gives:
formulae in [5], [6]. The correction required can be determined 2π
either from measurements, manufacturer’s specifications, or l
estimates such as [14]. The resulting corrected core impedance ur,e = ro,c Z23 (17)
is denoted by Z1o,c . The inner loop impedance of the equiv- jLµ0 f log( ro,s )
alent cable is given by
The inner loop impedance for the non-extended length
ZL = [Z1o,c l + Z12 l + Z2i L] IL1 − Zt LIL2 (10) (NEL) model, which applies when current does not flow in
the sheath (single point bonded systems), is described by:
Where the length, L is the EEL and l is the normal cable
length.
[Z1o,c l + Z12 l + Z2i L] IL1 − Zt LIL2 =
The outer impedance loop of the cable can be found in a [Z1o l + Z12,e l + Z2i l] IL1 − Zt lIL2 (18)
similar manner to that described above, except there are some Assuming that Zt l ≈ Zt L, the solution for the relative
important differences. The first is that the mutual impedance magnetic permeability in Z12,e gives:
between the sheath and earth is given by [6]:
!
∞  Z1o,c − Z1o
!
jωµ0 2
ri,s − b2 1X b
n 1
Z23 = log + Ln Kn (pri,s ) ur,e = ro,c + ro,c Z12
2π ri,s rc,o σ n=1 ri,s jµ0 f log( ro,s ) jµ0 f log( ro,s )
!
(11) L−l
+ Z2i ro,c (19)
  n  ljµ0 f log( ro,s )
1 2nµr Cn Kn (pα) Ip b
Ln = − (12) The outer loop impedance for the NEL model is:
Kn−1 (pα) pα πα α
For the mutual impedance calculation in (11), the earth sur-
rounding the cable is considered to be infinite in all directions. [Z2o L + Z23 l + Ze (l)] IL1 − Zt LIL2 =
The high resistivity of the soil and the weak eccentricity of the [Z2o l + Z23,e l + Ze (l)] IL1 − Zt LIL2 (20)
cable allows the proximity effect term to be neglected. This
simplifies the mutual impedance to: The outer insulators relative magnetic permeability in Z23,e
2
! is given by:
jωµ0 ri,s − b2 !
Z23 = log (13) 1 L−l
2π ri,s rc,o ur,e = ro,c Z23 + Z20 ro,c (21)
jµ0 f log( ro,s ) ljµ0 f log( ro,s )
A second difficulty in the outer impedance loop is the
application of the self earth return impedance. If the sheath is These models will be referred to as the EEL HCS approach
solidly bonded, current flows through the EEL of the sheath. and the NEL HCS approach in the following sections. As the
The interaction between the sheath and the earth return resulting relative permeability is a complex value, this cannot
therefore applies over the EEL. Hence the extended length is be directly entered into EMTP-RV. To cater for this input, a
used in the earth return impedance. If current is not flowing cable constants program which incorporates complex values
in the sheath (ie single point bonded systems), the induced for variables in the cable models has been written in Matlab.
voltage on the sheath is only attributed to currents flowing The resulting impedances and admittance matrices are then
in the core and in external cables. This means that the earth used in conjunction with steady state Pi models to simulate
return impedance only applies over the length of the core the cable network.
 4

C. Proposed Model For Sheaths With Annular Corrugations above each minor section. The top number represents the EEL
ACS have a continuous radial variation between the core and and the bottom number the normal cable length. The section
sheath. This causes the impedances associated with the sheath is not length extended if a single length is displayed.
to vary with length. In order to fit a cylinder to the corrugation
pattern, an average is taken over a corrugation period (i.e. a 334m 413m 382m
248m JB-1/2 307m JB-2/3 284m JB-3/4 490m JB-4/5 510m
GMR). Averaging the sheath corrugations causes the resultant
cylinder to be the same length as the core, with no length
extension required in any bonding arrangement. Consequently,
the proposed model for an ACS utilises the GMR approach, Cores + ECC - Solid Lines
ECC ECC
with half the thickness of the sheath placed either side of this Cable Sheaths - Dashed Lines
- 3kV SVL
radius. This model is shown in Figure 6. The radii in Figure 6 JB-3/4 Ze
ZS1 JB-4/5 Ze ZS2
are used in the impedance formulae in the following sections.
Figure 7. Hybrid Crossbonded & SPB Cable System
Sheath Corrugations
Sheath Thickness
The links of the hybrid crossbonded SPB system were re-
Min Distance Avg Distance GMR ri,s ro,s ro,c configured to create a single major crossbonded cable section.
From Core From Core This system is shown in Figure 8.
rc,o
Core Core
747m 1042m
Max Distance From Core Original Cable Length l 555m 774m 686m
JB-1/2 JB-2/3 JB-3/4 JB-4/5 510m

Serving
Serving
Cores + ECC - Solid Lines ECC ECC
Cable Sheaths - Dashed Lines
Figure 6. Model for a Sheath With Annular Corrugations JB-3/4 Ze JB-4/5 Ze ZS2
ZS1

However, the sheath resistivity must be corrected for the

EEL attributed to the corrugations in the DC condition. This Figure 8. Single Major Section Crossbonded Cable System
can be accomplished by using (8) and (9). Finally, if a
correction to the core impedance is required, this can also The cable model used in this study is given in Table I,
be incorporated into the relative magnetic permeability of the where Xtn denotes the relative permeability of the core sheath
insulation between the core and sheath. The impedance loop insulator for the EEL HCS model. The other insulator µr
equations for the equivalent EMTP-RV model in (1) and those describes the NEL HCS model. As the system contains both
conditions described above are given by: SPB and crossbonded cables, both the EEL and NEL HCS
models are used in this study.
[Z1o l + Z12,e l + Z2i l] IL1 − Zt lIL2 = Table I
[Z1o,c l + Z12 l + Z2i l] IL1 − Zt lIL2 (22) 430 MM2 CU1 OFAZV. C ABLE OUTER RADIUS IS 0.035 M
Solving (22), the expression for the relative magnetic per-
meability µre of the Property Core Sheath
 equivalent core sheath insulator, Z12,e =
jωµr,e µ0 ro,s ri (m) 0.006 0.0145
2π log ri,s is given by:
ro (m) 0.027 0.0291
(Z1o,c − Z1o ) Shell ρ 2.19 × 10−8 2.86 × 10−8
µre = µr,12 + r (23)
jµ0 f log( ro,s
i,s
) Shell µr 1 1
No correction is required for the outer loop for any bonding Insulator µr Xtn 0.67 + 0.23i 0.74
arrangement as the cable sheath is not extended. This model Insulator µr 1.01 − 0.59i 1.22 − 0.71i
will be referred to as the ACS approach in the following Insulator r 3.5 4
sections.

IV. S HEATH VOLTAGE P ROFILES ON H ELICAL An injection test simulating a single line to ground fault on
C ORRUGATED S HEATHS blue phase was performed on both of these systems, with an
injection current of 30A.
Injection tests were carried out on systems with single
point bonding (SPB) and crossbonding in order to assess
the accuracy of the traditional GMR model and the models B. Sheath Voltage Profile - Hybrid Crossbonded & Single
presented in Section III Point Bonded Cable System
The sheath voltage profiles and relative errors between the
A. Cable System Information predicted and measurements for each phase are shown in
Figures 9 and 10. The model incorporating both the EEL
The hybrid crossbonded SPB system consists of a single and NEL HCS models, was superior to the GMR model in
major crossbonded section and two cascaded single point predicting the measured sheath voltage profile for this cable
bonded sections. A 160mm2 CU1 earth continuity conductor system. The highest sheath voltage was measured on blue
(ECC) is installed in the centre of the cable trench in the phase at JB4/5B. The error between the HCS model and the
SPB section. The trench has two identical feeders placed in measured results at this JB4/5B is -3% compared with -6%
trefoil. The separation in the trefoil is 0.07m and is 0.432m for the GMR model. Both models are, however, still non-
between the feeders. The burial depth is on average 1.2m. conservative. Ideally, the SVLs in this system are sized off the
The system is shown in Figure 7. All cables in the system JB4/5B voltage. As such, the relatively small non-conservative
are 430mm2 CU1 OFAZV with HCS. The sheath amplitude error between the measured and predicted results from the
and pitch are 3.75mm and 18mm respectively. According to HCS models from Section III, would not have caused the
(9), this corresponds extending the cable length by a factor SVLs to be undersized, as appropriate safety margins would
of 1.35. In Figures 7 and 8, the section lengths are displayed be applied. This would also be true for the GMR approach.
 5

The major sources of error in both models are at S1 and JB1/2.

It is unclear what caused these errors. It is possible that the
large amount of underground infrastructure in the area could
be affecting the current distribution in the SPB section of the
feeder. Also, reinforced concrete cable covers run the length of
the cable route, potentially influencing the current distribution
in the cable system. This in turn would effect the voltage
distribution by increasing the voltage on the SPB part of the
system.

Figure 11. Single Major Crossbonded Section Sheath Voltage Profile

Figure 9. Hybrid Crossbonded & SPB Sheath Voltage Profile

Figure 12. Relative Errors Between Measured and Predicted Single Major
Crossbonded Section Sheath Voltage Profile

V. S HEATH VOLTAGE P ROFILES ON A NNULAR S HEATHS

An injection test was also undertaken on a system with

a cable with an ACS. The results of the injection test are
compared to both the model in Section III and the traditional
Figure 10. Relative Errors Between Measured and Predicted Hybrid GMR approach.
Crossbonded & SPB Sheath Voltage Profile

C. Sheath Voltage Profile - Sectionalised Crossbonded Cable A. Cable System Information

System
The sheath voltage profiles and relative errors between the The system consists of a four major section crossbonded
predicted and measurements for each phase are shown in cable system with two feeders in a common trench. The cables
Figures 11 and 12. The predicted results for the EEL HCS are installed in trefoil in 0.23m ducts. The spacing between
model show excellent accuracy, except for the red phase sheath circuits is 0.6m. The average burial depth is 1.2m. The cable
profile, which had a 60% conservative error. The GMR error bundles are installed in a vertical configuration in a short cable
for this case was 20% lower than the measured value. This tunnel for 100m of the 9.2km route. The depth of the tunnel
error can be attributed to the measured system impedance of is 7m. A block diagram of the system is shown in Figure 13.
the substation used in the study. The error in this impedance In this diagram, each block represents a single major section
affects the earth potential rise at the substation and hence the similar to that shown in Figure 8. The minor section lengths
sheath voltage profiles. As the red phase sheath was exposed are shown within the block diagram.
to the injection current for the shortest length, the error in the
EPR affects this sheath more than the others. The maximum
error at the worst case joint bay being approximately 1%. The Major Section 1 Major Section 2 Major Section 3 Major Section 4
maximum error associated with the GMR model was -27%. Minor 1 = 908m
Minor 2 = 900m
Minor 1 = 667m
Minor 2 = 665m
Minor 1 = 499m
Minor 2 = 488m
Minor 1 = 926m
Minor 2 = 900m
This is a substantial difference. As such, if the GMR model Minor 3 = 885m Minor 3 = 700m Minor 3 = 515m Minor 3 = 913m
is used, the SVLs would be substantially undersized, leading
to a potential failure under steady state fault conditions. From Zsub1 Ze2 Ze3 Ze4 Zsub2
the results, the EEL HCS model is the better predictor of the
measured sheath voltage profile. It is interesting to note that the
environmental factors affecting the hybrid crossbonded SPB Figure 13. Four Major Section Crossbonded Cable System
system are not present in this model. This is due to the much
higher coupling associated with crossbonded cable systems,
with less current returning through the earth. The minimum Two of the major sections have 2000mm2 CU1 GXQ-
voltage rating of the SVLs in the crossbonded system is also GALZYQ installed, and the other two have 1200mm2 CU1
much lower than that of the hybrid crossbonded SPB system, GXQGALZYQ. The EMTP-RV model for each of these cables
reducing voltage stresses on the cable serving. is given in Tables II and III:
 6

Table II
2000 MM2 CU1 GXQGALZYQ. C ABLE OUTER RADIUS IS 0.0695 M

Property Core Sheath

ri (m) 0 0.0274
ro (m) 0.0559 0.0586
Shell ρ 2.19 × 10−8 3.7 × 10−8
Shell µr 1 1
Insulator µr 1 + 0.082i 1
Insulator r 2.5 2.8

Figure 15. Relative Errors Between the Measured and Predicted Sheath Volt-
Table III age Profiles, Sectionalised Crossbonded Cable System: Four Major Sections
1200 MM2 CU1 GXQGALZYQ. C ABLE OUTER RADIUS IS 0.063 M

Property Core Sheath relative magnetic permeability of the insulation between the
ri (m) 0 0.0274 conducting shells in the cable. Results obtained from injection
ro (m) 0.0559 0.0586 tests were compared to the simulated results from both the
traditional GMR approach and the new models. The new
Shell ρ 2.14 × 10−8 3.52 × 10−8 models have proven to be more accurate and more reliable
Shell µr 1 1 than the GMR approach for both types of cable sheath and for
Insulator µr 1 + 0.0508i 1 all of the different bonding conditions tested. However, it was
also found that the GMR approach can be accurately applied
Insulator r 2.5 2.8 to cables with annularly corrugated sheaths and applied with
reasonable accuracy in SPB cable sections.
An injection test simulating a single line to ground fault on
yellow phase was performed on this system using an injection R EFERENCES
current of 20A. [1] P.Nichols, D. Woodhouse and J. Yarnold. “Effects of Earth
Potential Rise on Surge Arrester Specification in Specially
Bonded Cable Systems.” AUPEC 2008, December 2008.
B. Sheath Voltage Profile - Sectionalised Crossbonded Cable [2] IEEE Std 635. ”IEEE Guide for Selection and Design of
Aluminium Sheaths for Power Cables.” IEEE PES. May 2004.
System: Four Major Sections [3] IEEE Std 575. ”IEEE Guide for the Application of Sheath-
The sheath voltage profiles for each phase for the modelled Bonding Methods for Single Conductor Cables and the Calcu-
lation of the Induced Voltages and Currents in Cable Sheaths.”
and measured results, are shown in Figure 14. The relative ANSI/IEEE. 1988.
errors between the measured results and those predicted by the [4] S. A. Schelkunoff: “The Electromagnetic Theory of Coaxial
ACS and GMR models are displayed in Figure 15. All models Transmission Lines and Cylindrical Shields”, Bell Syst. Tech.
(ACS and GMR) show excellent accuracy, with a maximum J., Vol. 13, pp 532-579 (1934)
[5] T. Noda. “Numerical Techniques for Accurate Evaluation of
error of 4% at the worst case joint bay and an average error Overhead Line and Underground Cable Constants“ IEEJ Trans
of approximately 6% across all profiles. The largest errors 2008, 3: 549-559
in all models occurred at the earthing points (JB3/4, JB6/7 [6] A.Ametani: “A General Formulation of Impedance and Admit-
tance of Cables”, IEEE Trans. on Power Apparatus and Systems,
and JB9/10) along the sheath and can be attributed to the Vol. PAS-99, No. 3, May/June 1980
error in measurement of very low voltages present during the [7] P. Nichols “Implementation of Finite Mutual Impedances and
testing process. From the measured results, the ACS model its Influence on Earth Potential Rise Estimation Along Trans-
is the better predictor of the measured sheath voltage profile. mission Lines” UPEC 2010
[8] A. Ametani & A. Ishihara. “Investigation of Impedance and
However, the traditional GMR method also provided excellent Line Parameters of a Finite Length Multiconductor System.”
accuracy. This suggests that the GMR approach is applicable in IEEJ, Vol.114, No.4,1994
cable systems with cables with an annularly corrugated sheath. [9] F. Pollaczek. “Uber Das Feld Einer Unendlich Langen Wechsel
Stromdurchflossen Einfachleitung.” E.N.T, Band 3 (Heft 9).
pp339-360. 1926
[10] B. Gustavsen. “Panel Discussion on Data for Modelling System
Transients Insulated Cables. Proc. IEEE Power Engineering
Society Winter Meeting, 2001, 2, 718-723
[11] Gustavsen, B.; Martinez, J. A. & Durbak, D. Parameter determi-
nation for modeling system transients-Part II: Insulated cables,
Proc IEEE, 2005, 20, 2045-2050
[12] J.A Tegopoulos & E.E. Kriezis. “Eddy Current Distribution in
Cylindrical Shells of Infinite Length Due to Axial Currents.
Parts I & II.” IEEE Trans. Power Apparatus & Systems, PAS-
90, pp1287-1294, 1974.
[13] M. Kane, A. Ahmad & P. Auriol. “Multiwire Shielded Cable
Parameter Computation.” IEEE Trans. on Magnetics, Vol 31,
No. 3, May 1995.
[14] IEC 60287. “Electric Cables - Calculation of the Current
Rating.” 2006.

Figure 14. Sheath Voltage Profiles, Sectionalised Crossbonded Cable System:

Four Major Sections

VI. C ONCLUSIONS
In this paper, new models for annular and helical corru-
gated sheaths have been presented. The new models were
adapted to EMTP-RV style cable models by manipulating the