2064 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 25, NO.

4, OCTOBER 2010

Ampacity Calculations for Cables in Shallow Troughs

George J. Anders, Fellow, IEEE, Mark Coates, and Mohamed Chaaban, Senior Member, IEEE

Abstract—This paper presents the findings of an investigation sulting in two analytical expressions for cable rating calcula-
concerning a calculation method for current-carrying capability tions. One approach is based on the analysis of the work already
(ampacity) of cables located in a shallow trough. The investiga- done in this area and extraction of the most suitable model. The
tions proceeded in two parallel directions. On the one hand, an ex-
tensive search of published and unpublished records dealing with other approach involved numerical studies with finite element
the subject was performed. The second path of investigation in- analysis and development of a new rating formula.
volved studies applying finite element analysis. Sensitivity analysis The paper is organized as follows. Section II presents some
was also performed. In both cases, the recommendation is to use background on the existing empirical equation given in the
the same approach for determining the current rating of cables in standard IEC 60287 and outlines several other published
a trough. They both use the cable rating in “free air” as the basis
and add an extra thermal resistance to take into account the effect
approaches. One of these approaches is selected as a recom-
of the trough. Based on the investigation of the previous work and mended method of analysis of such installations. Section III
the finite element analysis, two new equations were developed for briefly describes the numerical studies applying finite element
the external thermal resistance. The proper choice would depend method and the resulting analytical developments. Section IV
on the field or laboratory tests that would confirm our findings or compares the results obtained with different methods. Section V
lead to the modifications of the rating formulae. In the absence of
contains a numerical example. Section VI provides comments
any additional information, it is recommended that the calculations
be performed using one of three methods listed below. The methods on possible methods for improving current ratings for cables in
are listed from the most to the least conservative. troughs.
1) method given by a new equation based on the finite element
analysis; II. PUBLISHED CALCULATION METHODS
2) method given in the IEC standard 60287; Since the IEEE standard 835-1994 [5] does not address the
3) method given by a new analytical equation presented in this
paper. issue of cable rating for trough installation, the investigations fo-
The comparison of the calculated results with the available test cused on the work done outside North America. The commonly
data shows that the method 3) above, the least conservative one, used approach is that published in the IEC standard 60287 [1],
gives a good agreement with the measured values. and this Chapter reviews the method presented there followed
Index Terms—Ampacities, cable rating, power cables, shallow by the discussion of other published approaches.
troughs, utility sidewalk.
A. IEC 60287
IEC 60287 is a standard, published by the International Elec-
I. INTRODUCTION trotechnical Committee that sets out methods for calculating
the current rating of cables under a range of different installa-
ISTRIBUTION cables within cities in the United States
D may be installed within covered troughs where the top
cover is flush with the surrounding surface. These sidewalk
tion conditions. Many of the equations given in the standard are
based on fundamental heat transfer theory, and others are em-
pirical or semi-empirical equations derived from test work. The
troughs are constructed of precast reinforced concrete placed standard includes an empirical approach to the determination of
on a 6-inch gravel bed. The trough has a 6-inch-thick concrete current ratings for cables in unfilled troughs where the top cover
lid designed for H20 traffic loading. The precast sections are is flush with the ground surface. The approach used is to calcu-
72 inches long and 48 inches wide with 6-inch-thick concrete late the current rating for the cable in air but with an addition to
walls 12 inches tall. A number of cables may be laid on the the ambient temperature that takes account of the effect of the
base of such a trough. trough. The increase in ambient temperature is given by
The purpose of this study was to develop a theoretical model
for determining the current rating of cables installed in a side- (1)
walk trough. Two complementary approaches were adopted, re-
where:
Manuscript received December 30, 2009; revised April 06, 2010. Date of pub-
lication August 23, 2010; date of current version September 22, 2010. Paper no. total power dissipation of all the cables in the
TPWRD-00970-2009. trough, per meter (in watts per meter);
G. J. Anders is with the Department of Electrical Engineering, Technical Uni-
versity of Lodz, Lodz, Poland (e-mail: [email protected]). perimeter of the trough that is effective for heat
M. Coates is with Cobham plc., London KT22 7SA, U.K. (e-mail: mark. dissipation (in meters).
[email protected]).
M. Chaaban is with IREQ, Montreal, QC J3X 1S1, Canada (e-mail (mo- If the trough is shaded, then the perimeter, , is given by
[email protected]). . If the top cover is exposed to solar radiation,
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org. then —that is, the top cover is not
Digital Object Identifier 10.1109/TPWRD.2010.2051239 considered effective for heat transfer.
0885-8977/$26.00 © 2010 IEEE
ANDERS et al.: AMPACITY CALCULATIONS FOR CABLES IN SHALLOW TROUGHS 2065

In calculating the current rating for each cable in the trough, C. ECRC Report
the ambient temperature , is replaced by . As the ERA Technology carried out measurements of temperature
maximum permitted power dissipation of each cable is a func- rise of two sizes of cable in two sizes of trough during 1968.
tion of its current rating, an iterative approach has to be used to This work was carried out for the Electricity Council Research
determine the current rating in a trough. Centre (ECRC), and reported in [3]. The Electricity Council
This equation was included in the first edition of IEC 60287 was a body that represented U.K. electricity distribution com-
in 1969 and was also given in the 1964 CIGRE report that was panies. A review of the contents of the report is presented in
the basis of the IEC standard. Appendix B.
The equation is credited to A. Morello of Pirelli Milan. In The report provides empirical equations for the temperature
his paper [2], he states that the empirical equation gives good difference between the cable surface and the inside of the trough
agreement with rating factors given in “IEE Wiring Regulations, and from the trough wall to ambient. The report includes a com-
13th Edition, 1955”. No details of the derivation of the empirical parison between the test results and those obtained by the cal-
equation are given in Morello’s paper. culation method given in IEC 60287. This comparison shows
that the IEC equation gave a current rating that was about 14%
lower than that obtained from the test results. It was concluded
B. IEE Wiring Regulations that the IEC method is very conservative.
The report also states that, if the trough is exposed to solar ra-
The IEE Wiring Regulations, BS 7671 1992, is a United diation, the ambient temperature used in the calculations should
Kingdom standard that sets out the requirements for low-voltage be increased by 9 C.
electrical installations in domestic, commercial, and industrial
properties in the U.K. The aims of BS 7671 are similar to D. Slaninka Paper
those of the National Electric Code in the United States in that A paper by P. Slaninka, [4] sets out a theoretical approach to
it sets out requirements to minimize the risk of danger from the calculation of the thermal resistance of a cable channel. As
electric shock or fire. Although the aims of the U.K. and U.S. with the IEC 60287 method, Slaninka’s approach leads to an ad-
documents are similar, the detailed requirements are different. ditional temperature rise, which is added to the ambient air tem-
BS 7671 contains tables of ratings and rating factors for com- perature to derive a rating for the cable in a trough. Slaninka as-
monly used types of low-voltage cables. Included in these is a sumed isothermal conditions for both the ground surface and the
table of rating factors for cables in closed troughs. Three sizes inner surface of the enclosure as well as made certain assump-
of trough are included: 18 in. (0.457 m) wide by 12 in. (0.305 m) tions that are only valid for troughs of roughly square cross-sec-
deep, 18 inches (0.457 m) wide by 24 in. (0.610 m) deep, and 24 tion. The resulting equation equivalent to (1) uses an increase in
in. (0.620 m) wide by 30 in. (0.762 m) deep. All three troughs the ambient temperature, , given by
have a 4-in.-thick concrete cover. Rating factors are given for
up to 3 circuits in the smallest trough and 12 circuits in the (2)
largest trough. The rating factors given are applied to the tab-
ulated rating for cables in free air. For the smallest trough, the where is derived in Appendix C.
cables are assumed to lie on the bottom of the trough, and for Slaninka compared his theoretical approach with the test re-
the other two troughs, they are supported on brackets fixed to sults and demonstrated that he obtained significantly better cor-
the side of the trough. The cables are assumed to be separated relation than was obtained with the IEC 60287 equation.
by at least 2 in.
E. Other Work
These derating factors were first introduced into the standard
in the 13th edition, which was published in 1955 and have re- Further work has been carried out to extend Slaninka’s so-
mained unchanged, other than metrication of the dimensions, lution to a situation with non-isothermal conditions and troughs
since then. The text associated with the derivation of the der- that are not roughly square. This work also attempted to develop
ating factors is given in Appendix A. a method for calculating the heat transfer between the cable sur-
The origin of the derating factors given in BS 7671 is not face and the inner surface of the trough. The method divided the
known. However, a draft document has been found that derives trough along a horizontal axis that ran through the centreline of
a simple equation for the calculation of the thermal resistance the cables. Heat transfer to the upper part of the enclosure was
of a trough. The equation is for troughs with a 4- inch-thick lid, taken to be by convection and radiation, while that to the lower
and the document has a graph of derating factors that aligns with part was taken to be by radiation only. Heat transfer by conduc-
the factors given in BS 7671. The document is not dated, and no tion for cables laid on the base of the trough was not considered.
author is given. However, from the content of the document, it The attempt to develop a general calculation method was
is believed to have been prepared by one or more U.K. cable abandoned, because it was considered too complex for normal
manufacturers. The document also contains a list of references. use as it required the use of many precalculated graphs and ta-
the latest of which is dated 1944. This fact suggests that the bles of factors as well as an iterative calculation.
document was prepared in the late 1940s or early 1950s. In turn,
this dating suggests that the calculation method given was the F. Proposed Analytical Solution Based on Previous Work
basis of the factors given in BS 7671. The equations given in An extension of the Slaninka’s method is set out below. The
this document are reproduced in Appendix A, in metric units. unpublished work compares the calculated results with those
2066 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 25, NO. 4, OCTOBER 2010

obtained from test work reported in ECRC Report ECRC/R219

and gives good agreement.
The thermal resistance of the trough [K m/W] is given by

(3)

where

(4)

in still air, indoors, or (5)

in moving air outdoors, where

is the wind speed, m/s (6)
(7) Fig. 1. Sketch of the installation under study. Cables are numbered from 1 to
6, from left to right.
(8)

where duct as compared to a shaded installation. Hence, the maximum

value of 15 C with the effect of solar radiation in a trough will
thermal resistivity of the cover material (in K.m/W); be on a safe side in places with high solar intensity.
thermal resistivity of the surrounding material (in
III. FINITE ELEMENT STUDY
K.m/W);
internal height of the trough (in meters); A. Introduction
internal width of the trough (in meters); This section describes the numerical analysis of heat dissipa-
tion from six cables installed on the floor of a shallow trough.
thickness of the cover (in meters). The cover of the trough, made of concrete, is exposed to the am-
The value computed in (3) is then used in (2) in the place of bient air and to the sun. The floor and the walls, also made of
. concrete, are in contact with compacted rock and native soil. A
The method given above takes no account of solar radiation. sketch of the installation is shown in Fig. 1.
However, the ECRC report states that, if the trough is exposed to
solar radiation, the ambient temperature used in the calculations B. Approach
should be increased by 9 C. A review of the report shows that The analysis of this installation is done by coupling the finite
this value was determined from the results of tests on a shallow element technique to solve for the heat dissipation in the solid
trough during a period when the intensity of the solar radiation region (soil, concrete, rock) with an analytical approach inside
ranged between 930 and 1020 W/m . The results from a trough the air cavity where the cables are installed. The solid domain
300 mm deep gave a temperature rise, due to solar radiation of is discretized into small elements to enable the solution of the
8 C over the same period. 2-D partial differential equation of heat conduction. Inside the
It is proposed that an increase in ambient temperature of 8 C cavity, several heat balance equations are written between the
should be applied to take solar radiation intro account. cables, the walls, and the air inside. This calculation leads to
This method gives a derating factor, for solar radiation of an algebraic system of equations that is solved, at each iteration
(time step), and the results are fed back as input into the finite
element analysis.

C. Results of the Analysis

The proposed temperature increase due to solar radiation
is taken from measurements made in southeast England. As The study was conducted in transient mode due to the fluctu-
such, it is considered that it can be applied in Northern Europe, ation of sun radiation and outside air temperature. Fig. 2 shows
Canada, and the northern United States. For areas having a the results, which give the required temperature distribution at
greater intensity and longer durations of solar radiation, it the quasi-steady-state conditions.
is suggested that temperature rise measurements are made
(but it is considered unlikely that a temperature rise of more D. Discussions
than 15 C will be encountered due to solar gain. The studies The results of the analysis are summarized in Table I.
performed by the authors showed that the temperature rise on To compare these results with the available guides and stan-
a cable exposed to 1000 W/mm of solar radiation can vary dard IEC-60287, both circuits were analyzed in free air at 40 C
between 10 C for cable in air and 17 C for cable installed in a and inside a trough with an “assumed” ambient temperature of
ANDERS et al.: AMPACITY CALCULATIONS FOR CABLES IN SHALLOW TROUGHS 2067

Fig. 3. Influence of the thermal resistivity on conductor temperature based on

finite element study.

Fig. 2. Temperature fluctuation at quasi-steady state. E. Proposed Analytical Solution Based on the Finite Element
Study
TABLE I It should be noted that the IEC formulas for calculating cable
SUMMARY OF CABLES’ DAILY TEMPERATURE VARIATION IN THE TROUGH temperature in free air are developed on the basis of numerous
tests and observations. The devised equations include many im-
portant factors that can influence heat transfer from the cable to
the surrounding air. It may, therefore, be expected that their re-
sults are, most of the time, sufficiently accurate.
The modification of the ambient temperature given by (1) for
the covered trough case is performed using only one equation,
for which the only parameters are the total thermal dissipation
from the cable and the length of the trough perimeter effective
for heat dissipation. It does not take into account the thermal
resistivity of the trough materials and the surrounding soil or
the placement of the cables inside the trough. For the worst-
case conditions normally used in rating calculations, we have
assumed throughout that the cables in each of the two circuits
are touching with the spacing between the circuits sufficient for
unimpeded heat dissipation by each circuit. To take into account
the effect of the soil thermal resistivity, a series of sensitivity
TABLE II studies was performed.
The sensitivity analysis performed using the coupled finite
MCM = I04 0 =
CABLE TEMPERATURE IN FREE AIR ACCORDING TO IEC-60287 (I-750
418 A, = 200 A) element—analytical solution permits investigation of the influ-
ence of the thermal resistivity of the trough materials and the
surrounding soil on the cable temperature. This influence is pre-
sented in Fig. 3.
In the analysis, the thermal resistances of the materials of the
trough (concrete and compacted rock) and the surrounding soil
were equal to each other and changed simultaneously. It can
be clearly seen that changing the thermal resistance results in a
significant change of temperatures of the cable. We should point
out that the surface emissivity of the cover also has an influence
on the temperature rise, especially when shaded. On the basis of
the above mentioned analysis, it is proposed that the influence of
the soil thermal resistivity be taken into account by modifying
64 C. This temperature is calculated by increasing the ambient (1) as follows:
using (1) by a value equal to the total losses (106.6 W/m) inside (9)
the trough divided by three times the perimeter without the top
cover (1.47 m). Table II shows a summary of these calculations where
for the hottest cable in each circuit. thermal resistivity of the material of the trough and
One can see that the formulation given by the IEC gives the the soil. Constant is equal to 0.077, assuming that
conductor temperatures that are very similar, but somewhat represents an effective thermal resistivity of the
smaller than those obtained from the FE analysis. trough and the soil. For the analyzed case, .
2068 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 25, NO. 4, OCTOBER 2010

TABLE III
CALCULATED THERMAL RESISTANCES OF THE TROUGH

term has a very small effect on the computed value of the

thermal resistance.
The comparison shows that the IEC 60287 equation is very
limited in that it does not take into account the thermal resistivity
of the surrounding soil and treats a tall narrow trough in the same
way as a shallow wide trough.1
As expected, the Slaninka 1 and 2 results are the same where
the lid and the soil have the same thermal resistivity. If the
thermal resistivity of the surrounding soil is increased, then the
result from Slaninka 1 is higher than that from Slaninka 2; this
is expected. The Slaninka 2 results are higher than that obtained
Fig. 4. Cable conductor temperature computed from the finite element analysis
and (9).
with the lower soil thermal resistivity; this is expected. Also
the wide shallow trough has a lower thermal resistance than the
narrow deep one, this is expected as well.
The results from the “recommended” method given by (3)
The added component in the square bracket will be equal to
form a similar pattern to those from Slaninka 2, except that the
0 for standard soil thermal resistivity of 0.9 K.m/W. For other
recommended results are consistently higher.
resistivities, it will appropriately shift the value of . As can
The method based on the finite element analysis (9) gives the
be observed in Fig. 4, the results from the finite element model
most pessimistic results but very close to the IEC method and
and (9) are very close.
to the Slaninka method 1 when the soil and lid resistivities are
the same.
IV. COMPARISON As both the recommended method (3) and that of Slaninka 1
The results given by the IEC 60287 method, Slaninka’s claim to give good agreement with test results, it is difficult to
method, and the proposed approach are given in Table III for decide which method should be adopted.
a range of conditions. The IEC equation has been modified to Slaninka’s method was compared to test work on a 2.1-m-
to give the additional thermal resistance, assuming wide 1.6-m-deep trough with a 0.35-m-thick lid. The calcu-
that there is only one cable in the trough. This approach allows lated thermal resistance of the trough by Slaninka 1, is 0.061
direct comparison with the other calculations. The column (K m/W), and the value derived from the test work ranged from
headed Slaninka 1 used the equation where the top sides and 0.068 to 0.075 (K m/W). The calculated value by using the rec-
bottom are considered together, Slaninka 2 is where the three ommended method (3) is 0.077 (K m/W). This result indicates
components are considered separately, that is, the thermal resis- that Slaninka’s method is slightly optimistic and the proposed
tivity of the cover can be different from that of the surrounding method slightly pessimistic, but both could be said to give rea-
soil. sonable agreement with the test results.
To fill in the last column in Table III, the expression in the 1In this comparison, solar radiation is disregarded, hence only the total
square bracket in (9) was computed. Omission of the constant perimeter length plays a role.
ANDERS et al.: AMPACITY CALCULATIONS FOR CABLES IN SHALLOW TROUGHS 2069

The method given by (3) was compared with test results for TABLE IV
a 0.61-m-wide 0.31-m-deep trough with a 0.075–m-thick lid. CALCULATED PARAMETERS
The calculated thermal resistance using the proposed method
is 0.101 (K.m/W), and the value derived from the test work is
0.103 (K.m/W). The value derived from Slaninka’s method 1 is
0.054 (K.m/W). This result suggests that Slaninka’s method 1
may not be appropriate for troughs that are not approximately
square or those that have thin covers.
The values corresponding to (9) are similar to the IEC
method, but they take into account the soil thermal resistivity.
These values are slightly larger than the IEC results because of
the design of (9), where an additional term is added.
From the above, it is suggested that the proposed method is
used for the shallow trough installations, but a further allowance
should be made to allow for heat transfer by direct conduction
to the base. Eq. (5)
It should be noted that, from the results shown in Table II, the
conductor temperature computed with the IEC method is below Eq. (4)
the range of the design value of 90 C. Since this method is
more pessimistic than the proposed approach, we can conclude Eq. (3)
that the cables in the utility sidewalk can be safely loaded to
their full design ampacity of 200 and 418 A for the 4/0 and 750 The thermal resistance of the trough is an additional external
MCM constructions, respectively. thermal resistance, and as such, is added to the value of given
However, the results obtained before should be verified by above for a cable in air. In this case, because there are three
either laboratory or field measurements. These tests would single-core cables in the trough, the value of has to be mul-
permit suitable adjustment of the proposed model parameters. tiplied by 3 to give a per-cable value. This is not necessary for
The measures that can be applied to increase current ratings are a three-core cable in a trough because the factor , number of
discussed in Section VI. cores, is included in the current rating equation. The cable rating
The following section presents a numerical example em- is then recalculated using this new value of
ploying the proposed approach.

V. EXAMPLE CALCULATION
The current rating then becomes
The following example is based on a single circuit, where
12.5 kV 750 MCM cable is selected. The cables are installed in
trefoil in a 0.3 m deep 0.6 m wide trough having a 0.1 m thick
concrete lid. The installation is not exposed to solar radiation,
and the wind speed is taken as zero. The thermal resistivity of
the concrete and the surrounding soil is taken to be 1.2 K m/W.
The ambient air temperature is taken as 30 C, and the ground If the trough was filled with cement bound sand having a
temperature as 20 C. thermal resistivity of 1.2 K.m/W, then the circuit in the trough
The basic equation for calculating the current rating of a cable could be treated as a direct buried circuit, and its current rating
is given in IEC 60287 as would be 556 A.

VI. INCREASING CURRENT RATINGS

A. Background
where the ac conductor resistance R, the various thermal resis-
tances T, and the loss factors are calculated using the equations The current rating of a cable is dictated by the maximum
set out in the different sections of IEC 60287. These values are permitted conductor temperature and the rate at which the con-
given in Table IV. ductor can dissipate heat to the environment. This rate of heat
The thermal resistance of the trough is calculated using the dissipation is governed by the temperature difference between
equations given in Section II-F as follows: the conductor and the environment and the thermal resistance
in the heat flow path. If the current rating of a cable is to be
Eq. (8) increased, for a given temperature difference, then the thermal
resistance must be decreased.
Eq. (8) For a cable in a trough, the thermal resistance can be divided
into four components:
1) internal thermal resistance of the cable (a function of the
Ea. (7) cable materials and construction);
2070 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 25, NO. 4, OCTOBER 2010

2) thermal resistance from the cable surface to the inner sur- gravel or other sharp stones should be included in the CBS mix.
face of the trough; Once it has cured, the CBS becomes a solid mass, which cannot
3) thermal resistance of the trough walls, base, and cover; simply be shovelled out of the trough. It has to be broken up like
4) thermal resistance from the outer surface of the trough to a weak concrete to remove it. The main disadvantage of using
the environment. CBS is this difficulty in removing it, and hence it would greatly
Since the internal thermal resistance of the cable is a function impede the installation of additional cables and the repair of any
of its design, this measure cannot be changed significantly if the faults in the existing cables.
cable is to remain capable of reliably distributing electricity. At first sight, the option of adding ventilation holes in the top
There are a number of possible methods of reducing the cover to allow hot air to escape would appear to be an option
thermal resistance from the cable to the inner surface of the to improve the heat transfer from the trough. This option is not
trough. These methods will be discussed later in this section. viable because hot air would only flow out if there was a means
The thermal resistance through the walls, base, and cover of for cold air to flow in to replace it. Test work was carried out
the trough is a function of the material used and its thickness. In some years ago to improve the current rating of cables installed
a concrete trough, the walls and base have a similar thermal re- in air but surrounded by a thermally insulating enclosure (in-
sistivity to the surrounding soil, and there is little to be gained by tended to protect the cables in the event of a fire). It was found
attempting to improve the resistivity of the walls and the base. that removing 50% of the top cover of the enclosure only pro-
However, as a significant proportion of the heat loss from the duced a small increase in the cable rating. It was necessary to
trough will be through the cover, there will be advantages in re- also remove 50% of the base of the enclosure to provide suffi-
placing the concrete lid with one having low solar absorptivity. cient air flow through the enclosure to significantly improve the
If the sidewalk is shielded from solar radiation, a cover with cable rating. Because this latter step would invalidate the fire
low thermal resistivity would improve heat dissipation. The ob- protection, the 50% derating factor caused by the enclosure had
vious choice in this case would be a fabricated steel lid. A steel to be accepted.
lid could be readily constructed to have the required strength. Forced ventilation of the trough is not considered a viable
However, a steel lid has a number of disadvantages when used option, because this would require the installation of fans or
in a public thoroughfare. It may create a noise nuisance by rat- blowers at numerous points along the route of the trough. There
tling when traffic passes over it. It will become slippery when would be significant costs associated with installing and main-
wet and cause a slip or skid hazard. taining such a system. Such costs may exceed the cost of in-
Alternatively adding cast in “fins” to the underside of the con- stalling additional cables in parallel with the existing ones to
crete trough cover should improve the heat transfer to the cover. share the load. Also, the higher current rating could not be main-
The extent of such an improvement has not been quantified. tained in the event of a failure in the forced ventilation system.
To reduce the external thermal resistance, it is necessary to Installing a circulating water system in the troughs to extract
reduce the thermal resistivity of the material surrounding the the heat has the same drawbacks as a forced cooling system,
trough. As the existing, nominal, thermal resistivity of the soil with the additional risk of a major water leak into the trough.
is taken as 0.9 K m/W, it is unlikely that this resistivity could be Natural ventilation can be used to ventilate cable tunnels.
improved. It is noted that the trough is laid on a gravel bed. The Ventilation is achieved by having two vents in the tunnel to
thermal resistivity of the gravel, if dry, would be expected to be chimneys of different heights. Natural convection will cause air
closer to 2 K m/W. Replacing the gravel with a weak sand/ce- to be drawn in through the lower chimney pass along the tunnel
ment/gravel mix would improve the heat transfer from the base and exit from the taller chimney. The spacing between the chim-
of the trough. However, this would also prevent water draining neys and their relative height are a function of the resistance
out through the opening in the base of the trough, and another to air flow offered by the chimney-tunnel-chimney system. Be-
approach would have to be adopted for drainage. cause a relatively small cable trough offers a higher resistance
to air flow than a large cable tunnel, this form of ventilation may
B. Cable—Trough Thermal Resistance not be a practical proposition. The required spacing and height
A significant proportion of the total thermal resistance be- of any chimneys have not been studied in detail. It is considered
tween the cable conductors and the environment is that between likely that this solution will not be practical for a “utility side-
the cable surface and the inner surface of the trough. Reducing walk” system.
this resistance will increase the current rating of the cable. Of the aforementioned options, only that of filling the trough
The simplest means of reducing this thermal resistance would with CBS is considered to offer a realistic method of improving
be to fill the trough with a material having a relatively low the cable rating.
thermal resistivity. One option would be to fill the trough with
a weak mix cement-bound sand, CBS. A 14:1 CBS mix is often VII. CONCLUSION
used as a bedding material for HV cable systems to improve The analytical methods reviewed in this paper generally use
the thermal properties of the material immediately around the the same approach for determining the current rating of cables
cable and reduce the risk of thermal runaway should the back- in a trough. They all use the cable rating in “free air” as the
fill around the cable dry out. The thermal resistivity of dry CBS basis and add an extra thermal resistance to take into account
can be better than 1.2 (K m/W) if the sand is carefully selected. the effect of the trough. There are two factors that this does not
CBS is usually compacted around the cables to give a dry den- take into account: the direct conduction of heat from the cable to
sity of at least 1600 kg/m . Due to the compaction required, no the base of the trough, and the restriction in natural convection
ANDERS et al.: AMPACITY CALCULATIONS FOR CABLES IN SHALLOW TROUGHS 2071

due to the trough. Since the first factor would increase the cable thermal resistance of the trough , converted into metric units,
rating and the second factor would reduce it, the two factors may is
balance out to some extent.
The finite element analysis considers heat transfer in all direc-
tions from the cables, and the radiation considers the presence
of walls and other cables. These types of studies are very in- where
volved and time consuming; therefore, only limited sensitivity
runs were performed. However, it appears that a simple analyt- internal height of the trough (in meters);
ical solution derived from these investigations may overcome
internal width of the trough (in meters).
some of the major limitations of the IEC method.
The amount of test data available relating to the temperature The cable derating factor is then given by
rise of cables in troughs is very limited. None of the test data
found relate to troughs that are relatively wide and shallow, as
is the utility sidewalk.
All of the methods that have been considered for improving the
current rating of cables in troughs have significant disadvantages. where
If a higher rating is required, increasing the conductor size
or using cables with copper conductors may be the best total thermal resistance of the cable in air (in K-m/W);
option. total number of cables in the trough.
The test work required to validate any theoretical model
It is stated in the document that the equation uses published
should be carried out on a full-size trough. Tests in a laboratory
data on heat losses through various types of floor, walls, ceilings,
are preferred, because this will eliminate the problems caused
and roofs. This suggests that data available on heat losses from
by changes in the environmental conditions during the tests. A
buildings were used. It is also stated that a ground temperature of
design of the tests can be discussed separately.
25 C was used, and it was assumed that the heat loss, per unit
area, through the trench walls is the same as that through the
base. It is clear that the equation assumes that the inside surface
APPENDIX A of the trough is an isotherm.
EXTRACT FROM BS7671:1992 Since the equation was based on heat flow data for buildings,
it is probably only applicable to troughs within buildings where
IEE Wiring Regulations: The IEE Wiring Regulations, BS conditions are generally drier, and thermal resistivities will be
7671, is a U.K. standard that sets out the requirements for low- higher, than for outdoor troughs.
voltage electrical installations in domestic commercial and in-
dustrial properties in the U.K. The aims of BS 7671 are similar
to those of the National Electric Code in the U.S. in that it sets
out requirements to minimize the risk of danger from electric APPENDIX B
shock or fire. Although the aims of the U.K. and U.S. docu- EXTRACT OF THE REPORT FOR ELECTRICITY COUNCIL
ments are similar, the detailed requirements are different. RESEARCH CENTRE, ECRC IN THE U.K.
BS 7671 contains tables of ratings and rating factors for com- The work performed at ERA was limited to tests on a 12-in.-
monly used types of low-voltage cables. Included in these tables wide 6-in. deep concrete trough containing three cables 2 in.
is a table of rating factors for cables in closed troughs. Three in diameter and tests on a 24-in.-wide 12-in.-deep trough, first
sizes of trough are included: 18 inches wide by 12 inches deep, containing three cables 2 in. in diameter and then containing
18 inches wide by 24 inches deep, and 24 inches wide by 30 three cables 4 in. in diameter. The test work was carried out out-
inches deep. All three troughs have a 4-in.-thick concrete cover. side between February and July 1968. The report indicates the
Rating factors are given for up to 3 circuits in the smallest trough difficulties of obtaining stable temperature results when there
and 12 circuits in the largest trough. The rating factors given are continuous changes in the ambient conditions, such as wind
are applied to the tabulated rating for cables in free air. For the speed, rainfall, temperature, solar radiation, and soil conditions.
smallest trough, the cables are assumed to lie on the bottom of The report provides empirical equations for the temperature
the trough, and for the other two troughs, they are supported on difference between the cable surface and the inside of the trough
brackets fixed to the side of the trough. The cables are assumed and from the trough wall to ambient. These equations are:
to be separated by at least 2 in. For 2-in.-diameter cable in the 12 6-in. trough, the cable
These derating factors were first introduced into the standard surface to trough wall
in the 13th edition, which was published in 1955 and have re-
mained unchanged, other than metrication of the dimensions,
since then.
The origin of the derating factors given in BS 7671 is not For 2-in.-diameter cable in the 24 12-in. trough, cable sur-
known. However, it is considered likely that they were derived face to trough wall
from equations given in an anonymous document probably pre-
pared in about 1950. The equation given in the document for the
2072 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 25, NO. 4, OCTOBER 2010

For 4-in. diameter cable in the 24 12-in. trough, cable sur- where
face to trough wall
thermal resistivity of the material surrounding the
sides of the trough (in K m/W)
For a 12 6-in. trough, trough to ambient

For a 24 12-in. trough, trough to ambient

internal height of the trough (in meters).

where 3) For the thermal resistance from the base to ambient soil
power dissipation from one cable (in watts per
meter);
temperature difference between the cable surface and where
the inner wall of the trough (in Kelvin); thermal resistivity of the material under the base of the
temperature difference between the inner wall of the trough (in K.m/W).
trough and ambient (in Kelvin). The overall thermal resistance of the trough is then given
It should be noted that the equations shown before were ob- by
tained by “curve fitting” of the best fit lines drawn through the
test results. Because of this, the equations may only be valid
for the test conditions at the time of the selected results. The se-
lected results were the averages of those recorded over a 24-hour
The above equations constitute the method that we refer to as
period on three days, one in April, one in May, and one in June
Slaninka 2.
1968. These days were chosen, because ambient conditions had
If the thermal resistivity of all materials can be taken to be
been stable for sufficient time, and the average wind speed was
equal, then is given by
less than 2 mph. The reported thermal resistivity of the backfill
around the troughs at these times was about 0.60 K m/W, and
that at different positions away from the troughs was between
0.45 and 1.03 K m/W.

This equation is referred to in the text as the Slaninka 1

APPENDIX C method.
SLANINKA’S METHOD
In this method, the thermal resistance of the trough can be REFERENCES
divided into three parts: 1) from the base of the trough to am- [1] Calculation of the Continuous Current Rating of Cables (100% Load
bient soil, 2) from the sides of the trough to ambient soil, and Factor), IEC Std. 60287 (1969, 1982, 1994), 1994–1995, 1st ed. 1969,
3) from the top of the trough to ambient air. This approach al- 2nd ed. 1982, 3rd ed.
[2] A. Morello, “The calculation of the current flow in new power trans-
lows different thermal resistivities to be used for the material mission cables,” L’Elettrotecnica, vol. 46, no. 1, p. 2, Jan. 1959.
surrounding each portion of the trough. The equations are [3] D. G. McCormick, “Cable ratings in concrete troughs,” ECRC/R219
1) For the thermal resistance from the inner surface of the Nov. 1969.
[4] P. Slaninka, “Thermal resistance of a cable channel,” Bull. Vuki, vol.
cover to ambient air 18, no. 5, pp. 212–221, 1965.
[5] IEEE Standard—Power Cable Ampacity Tables, IEEE Std. 835, 1994,
IEEE Press.

George J. Anders (F’99), photograph and biography not available at the time
where of publication.
thermal resistivity of the cover material (in
K m/W);
internal width of the trough (in meters);
Mark Coates , photograph and biography not available at the time of publica-
thickness of the cover (in meters). tion.

2) For the thermal resistance from the sides to ambient soil

Mohamed Chaaban (SM’00), photograph and biography not available at the
time of publication.