Epri Capitulo12

EPRI Underground Transmission Systems Reference Book
12-1
Design engineering for an underground transmission cable system has been called a busi-
ness of details. No single activity is especially complex, but long-term, reliable cable oper-
ation requires attention to a number of details at every stage of system design.
This chapter addresses many of the common design considerations for designing an
underground cable circuit: cable type selection, route considerations, surveying require-
ments, soil thermal properties, mechanical considerations, installation modes, and design
calculations.
Related information may be found in other chapters. Chapter 2 describes cable system
types and their common applications. Chapter 11 covers ampacity principles and calcula-
tion procedures. Chapter 13 provides detailed descriptions of installation procedures.
Chapter 7 describes special application cables, such as submarine, gas-insulated transmis-
sion line, and superconductors. Chapter 10 discusses grounding of underground cable
systems and cathodic protection.
Dennis Johnson is a Senior Project Engineer with POWER Engineers,
Inc. (POWER). He received a BSEE degree from Brigham Young Uni-
versity (1985). He joined Arizona Public Service in 1986, where he ini-
tially was responsible for distribution system design. He later worked
for Black & Veatch from 1988 to 2001 as a design and project engineer
on numerous substation, overhead, and underground transmission
projects at voltages ranging from 69 to 345 kV. Mr. Johnson has worked
as a Senior Project Engineer in POWERs underground transmission
design group since 2001. Mr. Johnson is a Member of the IEEE Power Engineering Soci-
ety and a Voting Member of the IEEE Insulated Conductors Committee (ICC). He is a
member of various ICC subcommittees that are developing guides and standards for
high-voltage underground cable systems. He is a registered professional engineer in the
states of California, Connecticut, Iowa, Kansas, Kentucky, Nevada, New York, Pennsyl-
vania, Texas, Vermont, and Virginia.
Deepak Parmar is President of Geotherm Inc. He holds a B.Sc. in civil
engineering from Woolwich Polytechnic (1966, London, U.K.) and a
Diploma in management studies (1972, Slough Polytechnic, U.K.).
During undergraduate studies, Mr. Parmar worked at Soil Mechanics
Ltd. in London. From 1973 to 1978, he worked with two civil engineering
consulting firms in Canada, specializing in geotechnical and material test-
ing, and instrumentation of tunnels and shafts. In 1978, Mr. Parmar
founded Geotherm Inc., specializing in underground cable-related soil
and backfill testing. Under contract with Ontario Hydro Research for two years, Mr. Par-
mar worked on EPRI-funded projects for the design and development of the Thermal
CHAPTER 12 Installation Design
Authors: Dennis E. Johnson, POWER Engineers, Inc.
Deepak Parmar, Geotherm, Inc.
Reviewer: John S. Rector, Black & Veatch
Chapter 12: Installation Design EPRI Underground Transmission Systems Reference Book
12-2
Property Analyzer and thermal aspects of soil and
backfills. Since 1980, Mr. Parmar and Geotherm have
undertaken numerous research contracts for Canadian
government agencies as well as for electric utilities and
cable manufacturers in Canada, North America, and
overseas. Mr. Parmar and Geotherm have performed
more than 400 route thermal survey projects for under-
ground and submarine cable crossings. Mr. Parmar was
a member of the team of experts retained to investigate
the cable failures in Auckland, New Zealand, and he has
conducted numerous seminars and courses on under-
ground T&D cable ampacity. Mr. Parmar is a member
of the IEEE Power Engineering Society and Insulated
Conductors Committee, Canadian Society for Civil
Engineers, Canadian Geotechnical Society, Tunnelling
Association of Canada, Canadian Electrical Associa-
tion, and CIGR. He has authored and coauthored
numerous papers on the application of thermal parame-
ters of soils and backfills, and was the principal author
for the IEEE Guide for Soil Thermal Resistivity Testing.
12-3
EPRI Underground Transmission Systems Reference Book Chapter 12: Installation Design
12.1 INTRODUCTION
Design of an underground electric transmission cable
system installation involves anticipating a host of issues
related to the systems planning, construction, and oper-
ation. While every installation will inevitably confront
site-specific issues, there are, nevertheless, a number of
common factors that can be expected with most system
installation designs.
This chapter describes the common challenges of instal-
lation design. It covers three primary areas. Section 12.2
reviews typical design considerationsincluding selec-
tion of the cable type, choosing the best route, surveying
practices, transitions to overhead systems, soil thermal
properties, mechanical considerations during cable
installation, and monitoring systems.
Section 12.3 describes the main modes of installation,
which are trenching (directly buried or in conduit),
trenchless installations, bridges, and underwater instal-
lations.
Section 12.4 briefly explains the series of calculations
that must be performed, principally related to cable
pulling tension and sidewall pressure.
12.2 COMMON DESIGN CONSIDERATIONS
12.2.1 Cable-Type Selection
Often the primary procedure in installation design is to
decide on which type of transmission cable system to
use. Pipe-type cables have been the most commonly
used cable system at higher voltages in North America
until recently. Now, with the advances in purity of
extruded-dielectric cables and the general reluctance to
install fluid-filled systems in some parts of the United
States, extruded-dielectric cable systems are becoming
much more common at increasingly higher voltage lev-
els and longer circuit lengths. XLPE-insulated cables are
being designed and installed at voltages up to 230 kV
and 345 kV in long lengths. Self-contained fluid-filled
(SCFF) cables are generally only used in specialized
extra-high-voltage applications. Conditions on individ-
ual utility systems vary, which results in different cable
types being a better choice on one system than another.
Additional discussion on cable system selection is pro-
vided in Chapter 2.
Urban
Extruded-dielectric cables installed in duct banks and
pipe-type cables are frequently used in urban environ-
ments in the United States because of their ruggedness
and the ability to install short lengths of pipe or duct at
one time in city streets with trench openings of only
300600 ft (90180 m). The cable is installed in a sepa-
rate operation, with minimal traffic disruption.
Suburban
In suburban locations, extruded-dielectric, pipe-type,
and SCFF cable can be used, depending upon the spe-
cific application. Careful route-specific technical analy-
sis and costing are required to determine the best cable
type for the application.
Rural
Rural installations are, by their nature, well-suited for
most cable systems. Low traffic volume and long trench
openings allow flexibility for the designer to consider all
the different types of cable systems.
Special Cases
Long underwater crossings often dictate use of SCFF or
extruded-dielectric cables, primarily because it is diffi-
cult to make pipe-type splices underwater. SCFF or
extruded-dielectric cables are generally preferred for
bridge crossings, because the weight and expansion
characteristics of pipe-type cable systems require resolu-
tion of the bending forces, which tends to complicate the
design. Short sections of cable requiring high power
transfer, such as substation ties, are often best served
with compressed-gas-insulated cable systems. For short
underground dips, extruded-dielectric cables are better
suited. Generally, pipe-type cable systems are not eco-
nomical in short lengths.
12.2.2 Route Considerations
A utility should investigate several alternative routes for
an underground cable. The optimal route can save sub-
stantial time and project cost, and in many U.S. states,
the siting commission requires that at least two routes
be investigated. Overhead routes are typically not the
best route for an underground cable.
In general, the shortest route is typically the least expen-
sive. In urban and suburban environments, the allowable
routes are usually limited to existing thoroughfares, and
existing rights-of-way. Frequently, the cable route may be
the same as the route determined suitable for an overhead
line. Detailed route considerations are discussed below.
Environmental
A primary routing consideration is to avoid activity,
where possible, in or near the following types of environ-
mentally sensitive areas. If an environmentally sensitive
area cannot be avoided, consideration should be made
to limit or mitigate the impact.
12-4
Wetlands
Wetlands provide habitat for many plant and wildlife
species in addition to providing a method for replenish-
ing the earths reserve of fresh water. Three characteris-
tics define a wetland: (1) standing water, (2) aquatic
vegetation, and (3) hydric (or hydraulic) soils. If any of
these three characteristics are present, a high probability
exists that this location will be designated a wetland
area. In North America, these types of areas are feder-
ally protected and require a special permit for cable sys-
tem crossings.
Wetlands should be avoided if at all possible by choosing
an alternate route. Depending on the type of wetland,
open trenching may be used, but more likely other meth-
ods of installation such as horizontal directional drilling
may need to be considered when crossing these areas.
Aesthetics
One of the main reasons for selecting an underground
cable system is the aesthetic benefit. The line is placed
underground, so the public cannot see the transmission
line until it exits the ground at the substation or at an
overhead-to-underground transition structure. Aesthet-
ics should be considered when locating and designing
transition structures.
Archeological/Historical
Construction activity in and around known archeologi-
cal sites is typically regulated in the United States by the
state through a State Historical Preservation Office
(SHPO). If the SHPO knows or suspects that an archeo-
logical site exists at or near the site of the planned facil-
ity, further investigation may be required. An
archeologist may need to be engaged to deal with the
process of working in or around these sites.
If an area is archeologically sensitive, the best option is
to reroute the line. There are no construction methods
that will limit the impact to an archeologically sensi-
tive area.
Electromagnetic Fields
Electromagnetic Fields (EMF) may be a significant fac-
tor in the regulatory and environmental process for site
selection. Even though there is no federal standard
defining allowable EMF levels, local jurisdictions may
have restrictions on EMF levels related to the siting of
power lines near schools, daycare centers, childrens
playgrounds, and residential areas.
As with overhead lines, underground lines generate
magnetic fields. However, because the cable shielding
contains the electric field within the cable, no external
electric field is generated.
Magnetic field readings directly over a cable are gener-
ally higher than directly under an overhead transmis-
sion line, depending on the type of cable system and the
overhead conductor arrangement, generally because the
cables are closer to the measuring device. However, the
magnetic field strength falls off faster as you move away
from the centerline of the cables, as compared to an
overhead line. Similar to overhead transmission lines,
magnetic fields produced by cables systems can be
reduced by various mitigation methods. Chapter 16 dis-
cusses various mitigation methods.
Land Use
When routing an underground line, it is important to
know the type of area that the line will be traversing. The
type of land use often determines the amount of ease-
ment or right-of-way available for the underground line.
Underground lines are frequently located in existing
roadway rights-of-way. Typically, no easement is
required to install the underground line in public rights-
of-way; however, the owner of the road usually reserves
the right to have the utility relocate the underground
line if a future conflict occurs. For this reason, some
utilities prefer to install underground circuits within a
dedicated easement adjacent and parallel to the public
right-of-way, and to accept the added cost.
Urban
Urban areas are becoming more and more congested
with traffic and underground utilities. This makes the
installation of a new underground transmission line dif-
ficult. When choosing routes in urban areas for new cir-
cuits, extreme care is required to locate the existing
underground facilities. The typical location for a new
underground circuit in an urban area is within the road
right-of-way. There is usually very little undeveloped
land available that could be used for installing an under-
ground line. Major thoroughfares should be avoided
because of the large amount of traffic that would have
to be controlled. The designer should be aware that a
significant cost of installing circuits in urban locations is
traffic control.
Suburban
Suburban areas, like urban areas, are becoming con-
gested with traffic and construction activities. Schools,
churches, and homes will likely be located along the
route selected through suburban areas, requiring addi-
tional safety considerations during construction. These
areas should be avoided, if possible. During construc-
tion, the entire road may need to be closed to provide
sufficient working space for the installation of the
underground cable system.
12-5
Rural
Rural areas are generally easier locations in which to
construct underground lines because they usually have
fewer existing underground utilities; however, they also
lend themselves to overhead transmission lines more
easily than suburban and urban areas.
Construction and Maintenance Considerations
Regardless of whether the cable is installed in urban,
suburban, or rural locations, the cable route must be
patrolled routinely, and access maintained for quick
repairs in the event of a failure. An open space of
1520 ft (4.56 m) is required for vehicle access along
the route (Blau 1975). During construction, an addi-
tional temporary easement of approximately 3050 ft
(915 m) is needed to provide sufficient space for the
construction equipment along the route. Figures 12-1
and 12-2 show typical access requirements.
Recently Paved Streets
Metropolitan areas frequently have moratoriums on
new construction within newly paved or resurfaced
roadways. Sometimes it is possible, through long-range
planning and coordination with the local officials, to
install cable pipes or ducts in advance of when the cable
is actually needed, prior to repaving activities. Where
this is not possible, selection of another route is often
the best option.
Obstacles
Obstacles such as rivers, major highways, and railroads
should be avoided where possible. Crossing these obsta-
cles adds significantly to permitting requirements, con-
struction cost, and installation time. Additionally,
crossing these obstacles usually requires installing the
underground cable deeper underground, which results
in reduced ampacity.
Other Utilities
In many areas, especially well-established cities, there
are usually extensive underground utilities already in
place. Avoiding these utilities can be costly and time-
consuming. In addition, if the cable must frequently dip
beneath other utilities, pulling lengths between man-
holes may need to be decreased, and ampacity may
decrease because of the greater burial depth. Although
it is generally not possible to avoid these utilities com-
pletely, some routes may have fewer subsurface obstruc-
tions than others. Where this cannot be avoided, it may
be desirable to install the cable system shallower, above
the existing utilities.
Heat Sources
Routes having other heat sources, such as multiple dis-
tribution circuits or steam mains, should be avoided.
These are usually found in downtown urban locations.
The designer should try to maintain at least 12 ft (3.6 m)
clearance from these other heat sources to avoid
decreasing the cables ampacity or installing a larger
conductor. Thermal barriers and heat pipes have been
used in special cases, although at a significant added
cost (Iwata et al. 1991). Each situation should be evalu-
ated individually.
Traffic Control
Traffic control is becoming a significant consideration
for street construction in all parts of the country. Most
jurisdictions require that a complete traffic control plan,
on a block-by-block basis, be provided during the per-
mitting process. Busy thoroughfares may require elabo-
rate traffic control measures, including severe
restrictions on the hours during which construction may
take place. Smaller streets usually have fewer restrictions
and may even allow closure for short periods of time.
The selected route must provide access for construction
material and equipment both during the installation and
after, should there be a cable failure.
Permitting
Longer underground cable routes frequently pass
through several different jurisdictions, adding to the
complexity of obtaining permits. There are some routes,
even in metropolitan areas, where construction is dis-
couraged or prohibited because of traffic congestion,
tourist attractions, etc. Zoning classifications affect the
allowable construction hours and noise levels, and dif-
ferent routes may have varying degrees of public opposi-
tion. Areas with potential archeological and historical
significance should be avoided wherever possible.
Figure 12-1 Typical and minimum access requirements for
urban and suburban areas.
Figure 12-2 Typical and minimum access requirements for
unpaved areas.
12-6
Soil Types
Trenching costs vary significantly as a function of soil
type, amount of pavement, amount of rock, amount of
fill material, passage through old landfill areas, and
amount of contaminated soil (e.g., near old gasoline
stations with leaky underground tanks). Some soils
require more shoring than others, and extensive dewa-
tering may cause subsidence in some soil types.
Although there may not be many alternatives, the
designer should consider avoiding difficult trenching
areas. Detailed geotechnical evaluation is required in
the early stages of project design.
Soil Thermal Resistivity
Soil thermal resistivity can also vary widely along differ-
ent routes. Initial soil thermal studies should be made at
the same time as the geotechnical evaluation. The cost
of corrective backfill must be considered when evaluat-
ing different routes. A more detailed discussion of this
subject is included in Section 12.2.6.
12.2.3 Surveying
An accurate route survey, showing both above-grade
and below-grade structures and features, is a very
important design element. The survey can be either
land- or aerial-based; however, land-based surveys gen-
erally provide better detail of the surface features and
the below-grade obstacles that the designer will need to
know. Aerial-based photography and surveys provide a
good overall picture, but lack the detail and accuracy
considered necessary for an underground design.
Level of Survey
For an underground line survey, it is important to get the
most detailed planimetric and topographical survey pos-
sible. Even though a route centerline may have already
been determined, it is important to make certain that
sufficient area on either side of the centerline is surveyed
to identify open areas in case the cable system must be
rerouted during the detail design and construction stage.
Typically, a distance of about 50 ft (15 m) on either side
of the route centerline is sufficient for this purpose.
Pre-Design Survey
The planimetric and topographical survey should
include the following information:
All property lines and street/utility rights-of-way.
Curb lines, street lanes, manholes, streetlights, fire
hydrants, traffic lights, driveways, and all pertinent
physical information.
Location of storm sewer inlets including widths,
depths, and elevations.
All sewer, water, communication, electric, and gas
manholes and valve covers.
Elevations to water/gas valve nuts or top of pipe, size
of pipes.
Culverts and invert elevations on open pipes.
In addition, as part of the planimetric and topographi-
cal survey, the following steps should be taken:
Ground elevations should be shot across the width of
the surveyed area to allow a 3-D Triangular Irregular
Network (TIN) model (file) to be generated. This will
allow the route to be relocated and a new profile to be
easily developed.
The local underground locating service should be
notified to have all existing underground facilities
located prior to the survey commencing, so this infor-
mation can also be documented by the surveyor.
Efforts should be made to locate other unmapped,
abandoned, and private underground facilities and
structures (fuel tanks, private lines [gas, electric, com-
munication, and water], old foundations, etc.).
Contact all the local underground utilities and obtain
their facility drawings. This is a crucial step in verify-
ing the accuracy of the survey, and identifying data
missing from the survey.
Underground Locating
In addition to the field markings, in highly congested
areas, it may be necessary to field-locate existing under-
ground facilities to determine actual location and
depth. The American Society of Civil Engineers
(ASCE) has developed a guide for collecting existing
subsurface utility data (ASCE 2003). The ASCE guide
has established four levels of data to determine the
accuracy of the underground facility locations. These
levels are as follows:
Utility Quality Level APrecise horizontal and
vertical location of utilities by the actual exposure
and subsequent measurement of the existing subsur-
face facility.
Utility Quality Level BInformation obtained
through the application of appropriate surface geo-
physical methods to determine the existence of sub-
surface utilities. Methods such as electromagnetic,
magnetic, elastic wave, and other high-cost special-
ized methods.
Utility Quality Level CInformation obtained by
surveying and plotting visible above-ground utility
features.
Utility Quality Level DInformation derived from
existing records or oral recollections.
12-7
At a minimum, the designer should develop drawings
using information up to Level C for all underground
projects. If the project is in an urban area, or there is
evidence of a significant amount of existing under-
ground facilities, it is recommended that the documen-
tation be to Level B or A. If exact location is needed,
Level A is required.
The ASCE guide indicates that the total savings on a
typical underground project using Quality Level B and
A data may range from 10 to 15% compared with costs
from a project using Quality Level C and D data
(ASCE 2003).
Post-Design Survey
Once the design is completed, it is important to have
the route resurveyed to determine if any conflicts exist.
This step is usually performed as part of the construc-
tion survey.
12.2.4 Rerouting Existing Lines
Occasionally, an existing underground circuit must be
relocated or rerouted to accommodate new infrastruc-
ture or the addition of a new substation.
General
Regardless of the type of cable system installed, rerout-
ing or relocating an existing underground circuit is
costly, time-consuming, and usually requires an
extended outage. If only small adjustments of a few
inches are needed, it may be possible to expose the
underground line by excavation and raise or lower the
pipe or ducts, as required. If more than a few inches are
needed, a new length of cable must be installed. Because
most utilities do not have large quantities of transmis-
sion cable in stock, they are unlikely to be able to per-
form the relocation quickly. The time required for the
repair is normally based on how quickly a utility can
obtain the required skilled labor and materials. Depend-
ing on the cable type and its location, repairs can take
from 1 month, if the materials are on hand, to as much
as 6 months if the cable has to be manufactured.
In general, the relocation or rerouting process consists
of the following activities.
Installation of new manholes around the existing
cable.
Installation of new pipe or duct between the new
manholes.
Installation of the new cable.
Splicing of the new cable to the existing cable.
Removal of old pipe or duct.
Extruded-dielectric Cables
Direct-buried extruded-dielectric cables require that a
splice pit be opened at each end of the relocated area. A
trench is excavated between the two pits, and the new
cable is placed in the trench. The existing cable is cut,
and the new cable is spliced to the existing cable. Some-
times manholes are used in place of a splice pit, but are
not required.
If the cable is installed in a duct bank, the relocation
would follow the process described above. Special care
should be taken to protect the existing cable while the
duct bank concrete is being removed.
Pipe-type Cables
A gas-filled pipe-type cable system would follow a simi-
lar procedure to the one described above. The main dif-
ference is the need to reduce the gas pressure prior to
cutting into the cable pipe and replacing the gas after
the cable has been spliced.
For a fluid-filled pipe-type cable system, the main con-
cern is controlling the flow of the dielectric fluid. If the
line length is relatively short, it may be cost-effective to
remove the dielectric fluid from the pipe. If removing the
fluid is not practical, the most common method of con-
trolling the fluid is by freezing the fluid, thus solidifying
the fluid. This is commonly known as a pipe freeze.
Chapter 13 discusses the method of installing a pipe
freeze. The pipe freeze is placed on either side of the
new splice points. Once the freeze is in place, the fluid
between the freezes can be removed, and the pipe and
cable can be cut. After the new installation is complete,
and prior to releasing the pipe freeze, the new pipe is
filled with fluid.
SCFF Cables
As with pipe-type cable, the dielectric fluid in an SCFF
cable must be controlled. This is commonly done by the
freezing method similar to the pipe-type cable system.
Otherwise, the relocation process is similar to the pro-
cess associated with extruded dielectric cable.
12.2.5 Overhead-to-Underground Transition
The connection of an underground circuit directly to an
overhead line at a transition structure is becoming more
common. The designer should give considerable atten-
tion to the design of the transition structure.
General
The two most common ways of making an overhead-to-
underground transition is to use a single-shaft structure,
or a small transition site. When planning for an under-
ground-to-overhead transmission line transition, the
designer must consider several issues relating to the sit-
12-8
ing of the structure or station. Some of the issues to be
considered are:
Environmental/community impacts, such as wetland,
archeological, land contamination, public lands,
noise, aesthetics, and EMF.
Physical site considerations, such as topography, geo-
logical, access, and overall land use.
System requirements, such as switching, outage dura-
tion, and maintenance.
Equipment requirements, such as load-break discon-
nect switches, breakers, and/or reactors.
In general, a complete system study should be per-
formed to determine the effect that the new under-
ground line will have on the overall transmission system.
Chapter 16 discusses the effect that an underground
cable system can have on a utilitys electrical system. In
addition, an analysis should be made on the equipment
to determine if the switches or breakers are capable of
switching out the cables given the anticipated charging
currents, and withstanding the anticipated switching
surges and overvoltages.
Single-structure Design
Design of the transition structure must provide the
proper electrical clearances defined in the latest version
of the National Electric Safety Code (NESC). These
clearances address the electrical clearance from the con-
ductors, jumper loops, equipment, and other energized
parts to the surfaces of the supporting structures and to
the spaces that will be designated for climbing and
working on the structures. Consideration should be
made to provide adequate climbing and working spaces
for safely performing maintenance on energized over-
head conductors using hot-line tools. The need to pro-
vide hot-line climbing and working areas should be
reviewed on a project-by-project basis, and the require-
ment waived where line configuration and installed
equipment on the structure would not permit hot-line
work or where the structure will not be climbed or
maintained with the line energized.
Structures must be designed for the wind, ice, equip-
ment, seismic, and code loads. Structure loads should
include all typical attachments for the structure with
conductors, static wires, and insulators.
Support arms must be designed for personnel loading
and all construction/installation loads. The attachment
plate for the terminations should have an open side to
prevent circulating currents from flowing in the plate.
Vertical steel members may be required below support
arms to provide support to the cable.
As with overhead line structures, various types of struc-
tures have been used for the transition pole, including
the following: self-supporting steel, guyed direct-
embedded steel, guyed wood, and guyed laminated. The
most common type of transition pole is self-supporting
steel. A self-supporting steel structure eliminates the
potential clearance issue that would exist if the structure
needed to be guyed. Since steel typically lasts a long
time, it eliminates the potential problem of needing to
replace a wood pole after a few years due to normal
deterioration. Additionally, wood structures dry out and
shrink, resulting in loose hardware fittings and guy wires.
Preferably, the cable should be installed on the outside
of the structure. There are special instances where the
cable may be required to be installed on the inside of the
structure by local ordinances. Special design and detail-
ing are required for a structure in which the cables pass
through the base plate and are installed internal to the
structure.
Transition Site Design
The layout of a transition site is determined by the
amount of equipment that is needed, such as disconnect
switches, reactors, breakers, control house, etc. The
design of this site is similar to a small switching station.
The utility should use their normal substation design
specifications and standards.
Transition structures for voltages up to 230 kV are usu-
ally a single-pole structure, an H-frame structure, or an
A-frame structure. The single-pole structure is common
on rights-of-way, whereas the other structures are com-
monly found within substations. At voltages higher than
230 kV, extruded-dielectric cables generally transition to
overhead within a transition station.
Extruded-dielectric cables can be installed on the inside
or outside of the transition structure, and terminations
are placed on support arms, as shown in Figure 12-3.
Most manufacturers recommend clamping the cable on
the vertical face of the structure at 5- to 6-ft (1.5-1.8 m)
intervals. A support channel from the pole to the base
of the terminator is sometimes installed to provide addi-
tional support to the cable. Most terminations require
two cable clamps immediately beneath the terminator
base plate. Basket-weave grips could also be used to
support the section of cable from the pole to the sup-
port arm. Basket-weave grips may also be used at inter-
mediate points to support long vertical runs of single-
conductor cable.
12-9
Most cable manufacturers and cable installers prefer
having the cable installed on the outside of the structure
because of the ease of the installation.
Pipe-type Cables
Pipe-type cable systems must have a pressurization plant
at one end of the cable system. If the new underground
line is in the middle of the overhead line, a transition site
is required at one end of the underground line to accom-
modate the pressurization plant. Pipe-type cable termi-
nations can be placed on single-pole structures, but they
would normally incorporate a small fenced area around
the facility. The reason for this is to provide protection
to the public in the unlikely event of a riser pipe failure.
Another design consideration when installing a pipe-
type cable on a single-pole structure is to limit the
mounting height of the terminations. The terminations
should be kept as low as possible for hydraulic reasons.
SCFF Cables
Most SCFF cable systems require a transition site
because of their pressurization system requirements. An
SCFF cable could be placed on a single-pole structure,
but would require the pressurization reservoirs to be
mounted on the structure.
12.2.6 Soil Thermal Properties and Special
Backfill
Importance
The earth portion of the thermal circuit is responsible
for the greatest percentage of thermal resistance for bur-
ied cables, often accounting for more than half of the
total resistance. The geometric component of resistance,
which is depth of burial, can be accurately determined
during route surveys. The intrinsic material component,
soil thermal resistivity (commonly called Thermal Rho
or TR), is the most variable of the components, chang-
ing with both time and distance along the route. Accu-
rately determining soil thermal resistivity permits much
more accurate ampacity calculations, and special back-
fill allows higher ampacity of the circuit. Thermal diffu-
sivity is a measure of the ability of soil to undergo
temperature change. Thermal diffusivity is important
for transient calculations and for calculating the effects
of daily load cycles as described in other sections below.
Factors Affecting Thermal Resistivity and Diffusivity
Several factors affect soil thermal resistivity and ther-
mal diffusivity. Figure 12-4 shows the effects of two of
the major determinants, soil composition (soil type) and
moisture content, for the most common soil types and
for some corrective thermal backfills. Similarly,
Figure 12-3 Extruded-dielectric cable installed on a steel
transition structure (courtesy POWER Engineers, Inc.).
Figure 12-4 Soil thermal resistivity as a function of soil type
and moisture content (courtesy Geotherm Inc.).
0
50
100
150
200
250
300
350
400
450
0 5 10 15 20 25 30 35
Moisture Content (% by dry weight)
T
h
e
r
m
a
l

R
e
s
i
s
t
i
v
i
t
y

(
o
C
-
c
m
/
W
)
Soft Organic Clay
Clay
Silt
Silty Sand with Gravel
Uniform Sand
Stone Screenings
FTB
12-10
Figure 12-5 shows the effect of moisture content on
thermal diffusivity for some common soil types.
Moisture Content
Even soils that have poor composition can have accept-
able thermal resistivity if the moisture content is kept
high. For most soils, thermal resistivity increases only
marginally when moisture decreases from a wet to a
moist condition. If the moisture content drops below
the critical moisture content level, the thermal resistivity
increases substantially. Moisture moves away from
higher temperatures and tends to migrate from the
cable/soil interface to ambient earth. If the moisture
content for material other than clay is above the critical
value (the knee of the curves in Figure 12-4), moisture
moves back toward the cable by capillary action quickly
enough to prevent drying. Below the critical moisture
content, moisture does not return as rapidly, and the
soil will continue to dry. If this condition persists, a ther-
mal runaway condition can occur. Thermal runaway
happens when lower moisture content causes an
increase in thermal resistivity, which causes a higher
cable temperature rise, which in turn drives more mois-
ture away, causing further increases in thermal resistiv-
ity, etc. until there is a cable failure. Ambient moisture
content can be very low for sands in dry areas, giving
high thermal resistivities. Moisture content under paved
surfaces generally is higher than under unpaved sur-
faces. Clays fall in a special category, since field mois-
ture content is very high, typically 2030%, and clays
dry and rewet at a much slower rate than other soils.
Moisture content is expressed as a percentage of the
weight of water to the dry weight of soil solids, as deter-
mined by oven drying at 105C.
Soil Composition and Texture
Soil is a composite material consisting of solid particles,
water, and air. Heat flows through soil primarily by con-
duction from particle to particle. In some cases, it is by
conduction and/or convection through the moisture or
air that occupies the pore space between solid particles.
Even highly compacted soil can have up to 30% free vol-
ume that is occupied by moisture and/or air. In a totally
dry condition, this volume is filled with air, which has a
very high thermal resistivity (about 4000 C.cm/watt
(40 K.m/W)), and thus the resulting soil resistivity is
high. A good thermal backfill is made up of well-graded
solid particles that provide more points of inter-particle
contact for conduction of heat. Two benefits result:
resistivity is low even at low moisture contents, and
moisture is less likely to migrate away from the cable/soil
interface. These benefits are attributed to the tight pack-
ing of the particles (high density, low porosity, and low
hydraulic conductivity). Figure 12-6 shows the mecha-
nism of heat transfer through wet and dry soil.
Specifications for controlled thermal backfill often
include a sieve analysis to ensure that the material falls
within specified limits of gradation characteristics. Fig-
ure 12-7 shows an envelope with upper and lower limits
for a good thermal backfill. This type of granular mate-
rial is made up of sound (nonporous) fine aggregate;
when installed at its optimum moisture content and at
95% standard Proctor density, it will give low and stable
thermal resistivity. The widespread use of the Thermal
Figure 12-5 Soil thermal diffusivity as a function of soil
type and moisture content (courtesy Geotherm Inc.).
0
0.002
0.004
0.006
0.008
0.01
0.012
0.014
0 2 4 6 8 10 12 14 16
Moisture Content (% by dry weight)
T
h
e
r
m
a
l

D
i
f
f
u
s
i
v
i
t
y

(
c
m
2
/
s
)
stone screenings
sandy silt
sand
clay
Figure 12-6 Heat flow through soilthe effect of moisture
content on thermal resistivity (courtesy Geotherm Inc.).
12-11
Property Analyzer (TPA) (EPRI 1981) has reduced the
reliance only on gradation characteristic (sieve analysis),
because users can readily make direct, accurate measure-
ments of thermal resistivity rather than inferring values
based only on sieve analysis. Soil texture refers to soil
grain size, shape, and particle size gradation. In engi-
neering applications, a soil is often qualitatively catego-
rized by a visual description using accepted adjectives to
indicate the fractional amount of each component (i.e.,
gravel, sand, silt, clay, etc.).
The 5% moisture content is a typical dry-season level
for soils in many parts of North America. The first four
materials listed are very stable and will not reach zero
moisture content unless heat flux is very high. Ambient
moisture contents for clays are often greater than 15%,
but if subjected to high heat flux, they can become ther-
mally unstable. The numbers in Table 12-1 are only rep-
resentative. Any backfill material considered for use on
a cable system should be tested thoroughly. Soil particle
soundness (porosity), mineral type (limestone, granite,
basalt, mica, etc.), and organic content also affect the
thermal resistivity to some degree. Porosity and organic
content higher than 4% can increase the dry thermal
resistivity by as much as 15%. The user should evaluate
the cost of changing routes to avoid poor soils such as cin-
ders or organic material, versus installing special backfill
in the trench.
Soil Compaction (Density)
Once installed, the density of a backfill does not change;
it is only the moisture content that may change due to
various factors. Backfill material must be carefully com-
pacted. Compaction decreases thermal resistivity, espe-
cially at low moisture contents, since the interparticle
heat path is improved. An increase in density of one
pound per cubic foot (16 kg/m
3
), which is approximately
1% additional compaction, can give a 23% decrease in
resistivity (EPRI 1977). This improvement is especially
important at low moisture levels. Compaction is best
carried out at moisture contents of 812% for granular,
noncohesive soils, and much higher for fine-grained
(silty-clayey) soils. Great care should be taken to obtain
good compaction along the entire cable route to prevent
hot spots.
Figure 12-7 Gradation limits for granular-type thermal backfill (courtesy of Geotherm, Inc.).
GRAIN SIZE DISTRIBUTION
0
10
20
30
40
50
60
70
80
90
100
0.001 0.01 0.1 1 10 100
Grain Diameter (mm)
P
e
r
c
e
n
t

P
a
s
s
i
n
g
0
10
20
30
40
50
60
70
80
90
100
GRAVEL SAND FINES
GRADING LIMITS FOR GRANULAR TYPE
THERMAL BACKFILL
COARSE FINE COARSE MEDIUM FINE SILT CLAY
unified soil classification system
3" 1" 3/4" 1/2" #4 #8 #16 #30 #50 #100 #200 U.S. standard sieve
Table 12-1 Representative Thermal Resistivity Values in
C.cm/watt (K.m/W)
5% Moisture 0% Moisture
Fluidized thermal backfill 40 (0.4) 80100 (0.81.0)
Concrete (no air) 30 (0.3) 70 (0.7)
Stone screenings 40 (0.4) 100 (1.0)
Thermal sand 50 (0.5) 100 (1.0)
Uniform sand 70 (0.7) 200 (2.0)
Clay 100 (1.0) 250 (2.5)
Lake bottom (organic) 100
a
(1.0)
a. 50% moisture.
>300 (>3.0)
Highly organic soil >300 (>3.0) >600 (>6.0)
12-12
Compaction is commonly specified in percentage of
Proctor Density (ASTM 1978). A Proctor Density of
100% gives the highest density possible for a given mate-
rial at the optimum moisture content. Proctor densities
of 90 to 95% are commonly specified for backfill in
cable trenches. In addition to improving thermal proper-
ties, high compaction reduces the chances of the pave-
ment settling. Many municipalities specify high Proctor
Densities to prevent the settling of road-base material.
Compaction (density) and moisture content of the
material affect both thermal resistivity and diffusivity.
Care should be taken to maintain the field (in situ) con-
ditions for the laboratory tests. A utility evaluating con-
trolled backfills may have laboratory tests performed at
several compaction levels to permit specification of field
compaction (density). Figure 12-8 demonstrates the
relationship between soil density and thermal resistivity.
Factors Affecting Thermal Stability
Unlike thermal resistivity and diffusivity, soil thermal
stability is a system-driven parameter. In other words,
soils can have widely differing thermal stabilities
depending on a number of operating and installed con-
ditions. Two criteria to reduce chances of dryout and to
evaluate thermal stability are discussedsoil interface
temperature and heat flux.
Soil Interface Temperature
For many soils, if the temperature at the interface
between the heat source and the soil is kept below
5060C, a balance will be maintained between mois-
ture migration away from the heat source and capillary
action, which returns the moisture. The soil will, there-
fore, remain stable. Published ampacity tables for distri-
bution cables list three values of allowable ampacity:
that governed by conductor temperature, and those
(generally lower) governed by 50 and 60C soil inter-
face temperatures (IPCEA 1976). Selection of the higher
interface temperature may be done only after testing the
soil; otherwise, the lower value should be used.
Heat Flux
Moisture migration actually occurs as a function of heat
flux at the surface of the heat source. Heat flux is deter-
mined by heat input in watts/foot (W/m) and cable
diameter in inches (mm). One approach to evaluating
soil stability is to energize a thermal probe using the
same heat input per unit length as the cable system (e.g.,
30 W/ft [98 W/m]) and measure thermal resistivity as a
function of time (Martin et al. 1981; Hartley and Black
1979). The time at which the slope changes (increase in
thermal resistivity) gives an indication of time to dryout.
The corresponding dryout time for the cable is obtained
from Equation 12-1:
(min) 12-1
Where:
t
c
= time for soil near cable to dry, min.
t
p
= time for soil near probe to dry, min.
D
c
= diameter of cable/earth interface, in.
D
p
= diameter of probe, in.
Figure 12-9 gives typical test results for sandy silt. A
dryout time of 20 minutes for a 157.5-mil (4-mm) probe
corresponds to a dryout time of about 112,500 minutes
or 78 days for an 11.81-in. (300-mm) cable pipe. Calcu-
lated dryout times for the cable system should be several
weeks to ensure stability over the summer season. This
approach can be conservative since cyclic loading and
rewetting due to rainfall serve to extend dryout time.
Soil thermal stability may also be evaluated during field
thermal surveys. A thermal probe may be energized at a
nominal heat input (typically 10 W/ft [33 W/m], which is
3050% of typical cable heat input). After the slope of
temperature vs. log time has stabilized, the heat input is
increased tenfold. If a new, constant slope is achieved,
the soil is stable. If the slope continues to increase, the
Figure 12-8 Effect of soil grain size and distribution on
density and thermal resistivity (courtesy Geotherm Inc.).
2
=
p
c
p c
D
D
t t
12-13
soil is not considered stable. Figure 12-10 shows a typi-
cal curve for this test.
Concrete duct bank encasement and controlled backfill
both have excellent thermal stability. Their use transfers
the interface temperature and heat flux concern to the
interface with the native soil. Temperatures and heat
flux are low at that interface, and stability of the native
soil is not generally a problem.
Laboratory Thermal-Resistivity Measurement
Undisturbed soil samples retrieved in thin-wall Shelby
tubes or bulk soil samples should be tested for thermal
and geotechnical parameters by a qualified and experi-
enced laboratory. The tests should include natural mois-
ture content, density, sieve analysis, compaction, thermal
resistivity, thermal stability, and thermal resistivity vs.
moisture content (dryout tests). Nominal 3-in. (76-mm)
diameter Shelby tubes should be used to retrieve undis-
turbed samples. Care should be taken to maintain the
natural moisture and density of the samples during
retrieving and transportation. Figure 12-11 shows vari-
ous types of samples being tested in the laboratory for
thermal resistivity using a Thermal Property Analyzer.
A laboratory-type thermal needle, 0.125-in. (3.2-mm)
diameter, is inserted into the test sample and connected
to a Thermal Property Analyzer (TPA). A series of ther-
mal-resistivity measurements is made for various mois-
ture contents by stage-drying the sample. Enough time
should be allowed to ensure moisture redistribution
throughout the sample at the start of each new stage.
When testing disturbs bulk samples or granular soils, a
test mold similar to a Shelby tube can be used to repack
the soil at a specified density and moisture content. The
curves shown in Figure 12-4 were developed from these
laboratory tests.
A similar setup can be used to determine the thermal
diffusivity of the soil. The test sample, either taken as a
section from a Shelby tube or reconstituted in a mold, is
allowed to equilibrate at a constant elevated tempera-
ture overnight in an oven. This temperature should only
be about 810C above the ambient room temperature.
The warm sample is immediately transferred to a circu-
lating water bath at a constant room temperature, and
Figure 12-9 Soil thermal stability for sandy silt.
Figure 12-10 Typical thermal stability test. Figure 12-11 Various types of samples being tested in
the laboratory (courtesy Geotherm Inc.).
12-14
the temperature of the thermal needle is monitored with
time. The resultant plot of temperature vs. log time is
analyzed by the theory of Shannon and Wells (Shannon
and Wells 1947) to calculate the thermal diffusivity. Fig-
ure 12-12 shows this setup.
Field Thermal-Resistivity Measurement
In situ tests provide rapid and accurate values of soil
resistivity along a cable route. The thermal probe has
been used for many years for this purpose. It consists of
a stainless-steel tube that contains a precisely wound
heater and a temperature sensor of high resolution and
accuracy. The annular space is filled with a thermal
epoxy, and all components are electrically insulated.
Commonly used field thermal probes are 0.25 in.
(6 mm) in diameter and 12 in. (300 mm) long. Probes of
larger diameter and lengths up to 6.5 ft (2 m) with multi-
ple temperature sensors have also been used successfully
in special conditions.
The probe head is designed for easy adaptation to con-
ventional soil-drilling equipment. The probe is simply
inserted into the undisturbed native soil by pushing or
driving it to the desired depth at a test location. Power is
applied to the probe heater, and temperature rise is
monitored with time. The slope of temperature rise vs.
log time determines the value of soil thermal resistivity.
Figure 12-13 shows a drill rig setup for in situ thermal
resistivity testing and soil sampling. A TPA powered by
a portable generator or a power inverter is shown in use.
It is important to perform the tests at the expected cable
depth. Soil thermal resistivity is usually lower at greater
depths because of higher soil moisture content.
The TPA (EPRI 1981) was developed in the late 1970s
to automate field thermal-resistivity tests for greater
accuracy, simplicity, and reproducibility of test results.
When a thermal probe is used with a TPA, a laptop com-
puter controls the test input parameters. Test results
including probe power, time, temperature, and thermal
resistivityare all displayed and stored during a test run.
The TPA also facilitates the thermal-stability measure-
ments by extending the standard test over longer time
periods or by ramping the probe power to a higher value.
IEEE Standard 422-1981, IEEE Guide for Soil Ther-
mal Resistivity Measurements, (IEEE 1981) should be
followed for soil thermal-resistivity measurements.
Prior to conducting a field thermal survey, any available
soil data along the proposed cable route should be
reviewed. Data may be available from government agen-
cies, other utilities, or borehole logs for existing over-
head power line tower foundations. Soil thermal-
resistivity measurements should be made at least twice
in each soil formation, and at least every 0.5 mile
(800 m) depending on the length of the cable route. At
least two samples of each soil type should be tested in
Figure 12-12 Shannon and Wells thermal diffusivity test
applied to a Shelby tube sample.
Figure 12-13 In situ thermal resistivity testing
and soil sampling (courtesy Geotherm Inc.).
12-15
the laboratory for thermal stability and resistivity vs.
moisture content. Since thermal-resistivity values are
extremely moisture-dependent, it is best to make field
measurements during the driest times of the year. Esti-
mates of moisture content at other times of the year can
be used to extrapolate thermal resistivities based on
curves similar to those shown in Figure 12-4.
Often a soil-testing firm breaks pavement and augers to
the desired depths where thermal-resistivity measure-
ments are to be taken, and the probe is inserted into
undisturbed soil below the augured hole. Soil samples
for laboratory testing are normally retrieved from the
same borehole, at depths in-between the test depths. At
locations where a soil drill-rig cannot be used, in situ
thermal testing and soil sampling can be conducted in
test pits dug by a backhoe.
Controlling Thermal Resistivity
It is not generally feasible or practical to modify the
thermal resistivity of native soil surrounding the cable
trench. Selected materials having good thermal resistiv-
ity and thermal stability can be placed in the trench,
however, to reduce overall thermal resistivity. The entire
trench width is filled with special backfill to a height
typically about 1 ft (300 mm) above the cable. In some
instances, the trench must be widened or entirely filled
with special backfill to offset the effects of poor
native soil.
Backfill used for this purpose, commonly termed con-
trolled backfill, is made of well-graded crushed stone
screenings or well-graded sands with some fines that can
achieve high densities at relatively low moisture con-
tents. There are several sources of controlled backfill:
Suitable natural sands having the required gradation
characteristics can be found in many parts of the
country. The material must be well compacted in the
trench, in layers of 612 in. (150300 mm) or so. A
1-ft (300 mm) layer compacts to about 8 in.
(200 mm). Standard civil engineering (construction)
practice must be followed where applicable.
In other areas, well-graded crushed limestone screen-
ings are used to give good thermal resistivities.
The screenings must also be well compacted as
described above.
Granular backfills of this type should be tested in the
laboratory to evaluate maximum dry density and opti-
mum moisture content. In situ density of the backfill
when placed in the trench should be about 95% of this
maximum dry density and at the optimum moisture
content. A quality control/assurance program should be
implemented for all field applications.
The addition of small quantities of cement to granular
materials, such as well-graded sands and screenings, can
provide excellent thermal resistivity and stability (Sandi-
ford 1981). The cement creates interparticle bonding
and lowers thermal resistivity by reducing contact resis-
tance and filling fine air voids. This type of cement-
based mixture must be moisture-conditioned to its opti-
mum value and should be placed by a compaction
method similar to that for granular-type backfill.
Fluidized Thermal Backfill
(FTB
) is a concrete-like
material that has been used extensively over the past
20+ years. It is a mixture of natural mineral aggregate,
sand, cement, water, and a fluidizer that is formulated
to meet specific thermal and strength requirements. Fly-
ash is a commonly used fluidizer. However, if flyash is
not readily available, a water-soluble resin or slag can be
used. FTB can be mixed in a regular concrete truck, and
most ready-mix concrete suppliers can handle this
material without any problem. FTB is delivered to the
site in a fluid state and can be poured or pumped into
trenches very easily using conventional concrete place-
ment equipment. FTB is often an ideal choice for cable
trench backfill application because of the advantages
that it offers over granular-type backfill. Figure 12-14
shows the application of FTB on a typical duct bank
project. The process is the same for pipe-type or direct
buried applications.
Preliminary work has been done to develop special
additives that may be injected into native soils to lower
thermal resistivity and increase thermal stability (EPRI
1977). Soil-interface temperature and heat-flux values
are seldom problems if special backfill is used. A typical
backfill thermal resistivity requirement is the range of
5070 C-cm/W at 34% moisture content.
Figure 12-14 Fluidized Thermal Backfill
(FTB
) being
applied to a duct bank (courtesy Geotherm Inc.).
12-16
It is important that the backfill in the immediate vicinity
of the cables is installed at the best possible density in
order to achieve a low thermal resistivity. Figure 12-15
shows that the temperature drop is inversely propor-
tional to the radial distance from the cable surface.
Calculating Effective Thermal Resistivity
Placing special backfill in the cable trench lowers
the effective thermal resistivity compared to native soil.
The heterogeneous earth portion is conveniently han-
dled by using Equations 11-58 to 11-75, as detailed in
Chapter 11. This treatment is considered sufficiently
accurate for ratios of controlled backfill with long
dimensions to short dimensions up to 3/1.
More sophisticated approaches, such as finite element
analysis (Soulsby and Donovan 1981), provide greater
accuracy at the expense of more complicated calculation
procedures.
Thermal Resistivities for Submarine Cables
Sub-bottom marine sediments are usually fully satu-
rated, so in situ thermal resistivities are generally low if
the material is nonorganic. However, depending upon
the composition of the material (very soft clay or mud),
the thermal stability may be poor. Heat generated dur-
ing normal power cable operation can cause the mois-
ture to migrate from the cable vicinity, creating areas of
very high thermal resistivity. The dryout problem is
especially severe when the bottom sediments are highly
organic. Fully saturated organic sediments with high
moisture contents are known to have thermal resistivi-
ties greater than 150 C.cm/watt (1.5 K.m/W). It is very
difficultand in most cases, impractical or expensive
to install controlled thermal backfill in submarine appli-
cations. Thus, the cable design must accommodate the
bottom conditions with respect to thermal resistivity.
Care must be taken in areas where silting takes place.
One meter of cover today might be four meters of cover
in a few years, with the thermal capability of the cable
suffering accordingly. Problems associated with cables
sinking in soft sediments should also be addressed.
Special attention must be paid to soil thermal resistivity
and stability measurements for most underwater cross-
ings, and specialized instrumentation approaches are
required. Reference (Radhakrishna and Stienmanis
1981) provides a good description of test equipment and
procedures. Figure 12-16 shows a drill rig on a barge
used for submarine thermal surveys.
Figures 12-17 and 12-18 show a vibra core sampler and
its adaptation to thermal probe for in situ thermal resis-
tivity measurements.
Thermal Resistivities for Cables in HDD (Horizontal
Directionally Drilled) Installation
Installation of cables or casing containing power cables
in directionally drilled bores is quite common when
Figure 12-15 Thermal gradients in a radiant flow field are
inversely proportional to the distance from the heat source
(courtesy Geotherm Inc.).
Figure 12-16 Drill rig on a barge for submarine thermal
surveys (courtesy Geotherm Inc.).
Figure 12-17 Vibra core sampler (courtesy Geotherm Inc.).
12-17
crossing railroads, highways, and rivers. For such instal-
lations, the geotechnical as well as thermal, properties
of the soil at the cable/casing elevation must be investi-
gated. This is conducted in the same manner as for shal-
low borings using a conventional soil drill rig. The
major difference is that, for river crossings, the cable ele-
vation may be below the water table, and bedrock may
be encountered at other locations.
Below the water level, soil is saturated with water, and
the thermal resistivity is not of concern unless organic
material or peat is encountered. This is quite common in
the upper soil layers below the bottom and near the river
banks. It is impossible to install corrective backfill at
deep elevations, and therefore, measurement of ambient
temperature and thermal resistivity is very important.
Most bedrock has fairly low thermal resistivity ranging
between 40 and 65 C.cm/W (0.400.65 K.m/W). The
annular space within the casing and around the cables
or cable conduits must be filled with a pumpable ther-
mal slurry of low thermal resistivity.
Ambient Earth Temperature
Figure 12-19 depicts the variation in the earth ambient
temperature at depths of up to 10 ft (3 m), as a function
of the time of year. Air temperature near the surface
changes significantly and quickly as a function of time,
whereas the temperature at depths is slow to react to
this change. Although this data was collected for
Toronto, Canada at latitude 40, similar trends hold
true for other locations. In southern states, the varia-
tions in the upper and lower limits are smaller, and the
difference between the peaks at different depths (lag
time) is shorter.
12.2.7 Mechanical Considerations
The principal mechanical considerations for under-
ground transmission cables consist of pulling tensions
and sidewall pressures during cable installation,
as described in Section 12.3. Other considerations
include Thermomechanical Bending (TMB), proper
manhole racking, supports for vertical cable sections,
and vibration.
Thermomechanical Bending
Cables expand and contract with load cycling. This
motion is usually accommodated by cable snaking
within the pipe of a pipe-type cable, or within the
duct for single-core duct systems. However, for large
conductor and cable sizes, it is critical to make an in-
depth investigation into this issue. For directly buried
single-core cables, the earth constrains most motion,
except for the small amount that may occur in man-
holes, if present.
There have been a number of splice failures caused by
excessive unrestrained TMB forces within the manholes.
There has been significant discussion on the best way of
handling the TMB for large single-core cables. The com-
mon practice for determining the required duct size for
a particular installation is to size the inside diameter of
the conduit 1.5 times larger than the outside diameter of
the cable. For example, if the cable has an outside diam-
eter of 3.5 in. (89 mm), then the minimum-sized conduit
ID would need to be 5.25 in. (133 mm), resulting in a
6-in. (166-mm) duct requirement. Historically, no major
circuit failures have been caused by thermomechanical
forces when this practice has been followed. However, if
this practice is not followed, a detailed TMB analysis
should be made.
Figure 12-18 Vibra core sampler with thermal probe for
testing (courtesy Geotherm Inc.).
Figure 12-19 Variations in earth temperature at
various depths as a function of time of year (courtesy
Geotherm Inc.).
-10
-5
0
5
10
15
20
25
0 30 60 90 120 150 180 210 240 270 300 330 360
Day
T
e
m
p
e
r
a
t
u
r
e

(
o
C
)
Air temperature
1.5 m depth temp.
3.0 m depth temp.
Average ground temp.
Freezing temperature
Min. & max. temp.
12-18
The following variables affect the overall TMB design,
and each should be considered when determining the
best and most cost-effective design.
Size of conduit
Size of cable
Size of the manhole
Strength and number of cable clamps
Route geometry
Strength and design of the cable splice
Cable restraint and support brackets
There are two common methods for handling the TMB
forces resulting from the cable load cycling: the first is
the straight-through approach using sufficient clamping
to force the cable expansion back into the duct; the sec-
ond is to install an S-bend within the manhole and allow
the S-bend to flex and absorb the expansion. With the
straight-through approach, the size of the conduit plays
an important part in the overall thermomechanical
design. When the cable is fully restrained and no move-
ment of the cable is permitted in the manhole, the
thermal expansion or contraction must be accommo-
dated in the duct. This could result in high internal com-
pressive and tensile forces. This design allows the
manholes to be shorter.
With the S-bend approach, compressive and tensile
forces are not developed within the cable. However,
because the cable is allowed to flex, special consider-
ation must be made to prevent the cable from exceeding
its bending limits. This results in the manhole needing to
be wider and longer.
Another factor that affects the overall TMB forces is
the number of vertical and horizontal bends that the
cable will go through between the manholes. If the route
has many bends, then the resulting forces will be smaller.
If the route is flat, the resulting forces will be signifi-
cantly higher.
Pipe-type Cables
Early 345-kV pipe-type cables have experienced TMB in
the splice area, where the pipe diameter increases in the
joint casing (Aabo and Moran, Jr. 1988; Moran et al.
1984). Early designs had long unsupported cable lengths
in this area, and the joints could flex too easily. These
early joints were retrofitted with additional joint sup-
ports, and new designs include the proper supports.
SCFF Cables
SCFF joints have experienced mechanical fatigue in
manholes. Sheaths have cracked at the transition to the
joint sleeve, and soldered connectors have pulled out
after many years of load cycling. Proper reinforcement
of the sheath/sleeve transition and the use of pressed
connectors have solved these two problems.
Manhole Racking and Sizing
Manhole racking is only required for SCFF and
extruded-dielectric cable. For pipe-type cables, no rack-
ing is typically needed, because the steel cable pipe/joint
sleeve usually supports itself across the manhole. If the
manhole is too long, the designer may want to provide
some type of support.
As discussed previously, there are two common methods
of installing cable in a manhole: the straight-through
method and the S-bend method. The main design con-
cern with manhole racking is to protect the splice from
experiencing excessive thermomechanical forces.
Regardless of the installation method, this is accom-
plished by placing cable clamps on either side of the
splice to prevent the cable from moving near the splice.
When designing the cable supports for the straight-
through method, it is important to calculate the
expected TMB forces and design the cable clamp and
support hardware accordingly. This method allows the
manholes to be smaller. Sizing of the manhole is based
on the splice design. There must be enough space within
the manhole to allow the splice sleeve to slide out of the
way during splicing procedures. The typical inside
dimensions for a 138-kV manhole are 6 ft wide by 20 ft
long and 7 ft high (1.8 x 6.1 x 2.1 m). Figure 12-20
shows a typical manhole clamping arrangement.
For the S-bend design, the manholes must have enough
room to allow the cable to move smoothly to accommo-
date thermal expansion and contraction. This can gen-
Figure 12-20 Typical cable clamping arrangement for
cables passing straight through a manhole (courtesy
POWER Engineers, Inc.).
12-19
erally be accomplished by having the cable ducts enter
the center of the manhole and then train the cables to
their proper position. The cables form two reverse
bends, as shown in Figure 12-21, as the joints are racked
on the manhole sidewall. The typical inside dimensions
for a 138-kV manhole with this arrangement are 8 ft
wide by 25 ft long and 7 ft high (2.4 x 7.6 x 2.1 m). Some
utilities make a full 360 loop in the manhole to ensure
that movement is not concentrated in one location,
which would require an even larger manhole. A straight
distance of at least two cable diameters should be left at
each end of the reverse bend. Enough straight distance
must be provided near the joint to allow the joint sleeve
to slide out of the way during splicing procedures.
Pipe-type Cables
Since there is typically no support requirements, the siz-
ing of the manhole for a pipe-type cable system is based
on the joint-casing design. A sufficient straight distance
from the face of the reducer to the manhole end wall
must be provided to allow the joint sleeve to slide out of
the way during splicing procedures. The joint casings
can be designed in multiple telescopic sections. This
design allows the manhole length to decrease. However,
additional welding is required with multiple sleeves
while the cable is in the pipe. The recommendation is to
limit the number of welds, if possible.
Supports for Vertical Cables
Special consideration of conductor stresses and
methods to support cable weight is required for cable
systems designed to be installed on steep slopes or verti-
cal applications.
Extruded-dielectric cables can be installed on the inside
or outside of single-pole deadend structures. The cables
are typically clamped to the pole, as shown in
Figure 12-22. Cable clamps or basket-weave grips can
be used at intermediate points to support long vertical
runs of single-conductor cable. Thermal expansion and
contraction can be accommodated by cable snaking
between the clamps.
If an extruded cable is installed along a long steep slope,
an intermediate clamping manhole may be needed to
prevent the cable from ratcheting down the slope.
Pipe-type Cables
For vertical and steep slope designs, a riser cable design
is used for pipe-type cables. Long-lay stainless-steel
tapes are placed under the skid wires. These ribbons
support most of the cable weight. Anchor joints are
placed at the top of the slope, and skid joints are placed
at the bottom. Chapter 8 describes these types of joints.
SCFF Cables
The method for installing SCFF cables would be similar
to the methods described above for extruded-dielectric
cables. However, special consideration should be
given to the hydraulic issues associated with large eleva-
tion differences.
12.2.8 Monitoring
Monitoring systems can help to ensure long, trouble-
free cable life, and rapidly determine and locate troubles
on the cable system. Monitoring systems that are part of
cable system design are described below.
Temperature Monitoring
For many years utilities have installed thermocouples on
cable pipes several hundred feet (50100 m) on either
side of manholes. The thermocouples are terminated in
boxes in the manholes, and can be read as desired. In
the 1970s, elaborate temperature monitoring was
applied as part of early dynamic rating systems (Patton
et al. 1979). Temperature monitoring is part of more
sophisticated dynamic rating systems designed in the
early 1990s (Engelhardt and Purnhagen 1991). Distrib-
uted fiber-optic temperature monitoring offers the pos-
sibility of monitoring temperatures along the entire
Figure 12-21 Training cables in a manhole (courtesy
Figure 12-22 Cables clamped to a transition pole (courtesy
12-20
length of a cable. These systems are suitable for all cable
system types.
Distributed Temperature Monitoring
A distributed temperature monitoring system (DTS)
consists of installing a fiber optic cable, which acts as a
temperature sensor, along the entire length of a cable
and some type of monitoring. The monitoring system
reads the temperature of the fiber. The monitoring sys-
tem can be permanent or portable. Whether the moni-
toring system is permanent or temporary should be
based on the needs of the utility. If real-time monitoring
is desired, a permanent installation is necessary. If only
historical or benchmark information is desired, a porta-
ble unit would be sufficient. With a portable system,
multiple systems could be periodically monitored.
For the most accurate temperature reading, the best
location to install the temperature-sensing fiber is inside
the conductor itself. This method eliminates outside
interference and any misinterpretation of results. How-
ever, there is currently no manufacturer that can accom-
plish this type of installation.
The second best choice is to place the temperature-sens-
ing fiber, during cable manufacturing, under the outside
jacket. The calculation of the conductor temperature
based on the measured temperature is straightforward in
this case, since the thermal characteristics of the cable
materials are well known. When a temperature-sensing
fiber is integrated into the power cable, it may be part of
the power cable screen, laid close to the screen or situ-
ated in the interstices of a number of cables (e.g., a trefoil
arrangement). The optical fibers may be installed inside
a plastic or metal tube. A disadvantage of this method is
that it requires splicing the fiber at each manhole.
The most common method in the majority of existing
and new cable installations is to install the temperature-
sensing fibers nearby the cables. This can be accom-
plished by either attaching the optical fibers to the cable
using cable ties or installing the optical fibers in an adja-
cent conduit. If the cable surface is accessible along the
entire length, then the sensor can be fixed to the outside
surface of the power cable. This would be the preferred
method for direct-buried cables. There will, however, be
uncertainty in knowing what is being measured: the sur-
face of the cable or the layer of air or protection around
the fiber in its vicinity. When a temperature-sensing
fiber is attached to a power cable, it may be installed in a
plastic or metal tube, or it may consist of an optical
fiber cable with additional strength members which may
include metallic parts. If metallic parts are included,
then consideration needs to be given to the possibility of
induced circulating currents generating additional heat-
ing effects and causing temperature measurement errors.
In addition, the bonding and grounding of metallic
parts needs to be considered. If the cable with the tem-
perature-sensing fiber attached is to be pulled into a
duct, special consideration needs to be given to the
forces that the fiber will encounter during the pulling of
the power cable.
Table 12-2 Advantages and Disadvantages of Fiber Optic Cable Location for Temperature Sensing
Advantages Disadvantages
Integrated into the power cable
Provides a better indication of the conductor core temperature Manufacture of power cable is more complex
More responsive to current loading Optical splicing at each cable joint is required and is more complicated
Fiber is protected by the power cable More fiber splices are needed (determined by power cable drum length),
resulting in higher dB losses.
Well-suited where power cable pulled into ducts May require more monitoring loops.
Fiber needs to be replaced and respliced in the event of a cable failure.
Externally attached to the outside of the power cable
Fiber cable can be run in long lengths without splicing, resulting
in lower dB losses
Less responsive to load changes than integrated design
If installed in a tube, fibers can be blown out and new ones
blown in, should a fiber failure occur
Less representative of conductor temperature
Can be easily attached to a conventional cable in an open
trench (before backfilling)
Increases installation work
Can be easily attached to a power cable in a tunnel, or to the
roof or floor of a tunnel
Not suitable for attaching to power cables that will be pulled through
long ducts
In a separate duct or close to the power cable
Relatively easy to install or may already exist Least responsive method to load changes
Fiber cable can be retrofitted Least representative of conductor temperature remote from conductor
Can be run in long lengths without splicing, resulting in lower
dB losses
Can use spare fibers in an already installed telecom cable
12-21
In a duct bank, any empty duct can be used to house the
temperature-sensing fiber. However, the farther the tem-
perature-sensing fiber is from the hottest conductor, the
more uncertainty is introduced in the correlation
between the measured temperature and the actual tem-
perature of the conductor, and thus the rating. When a
sensor fiber is separately installed and positioned a
short distance from the power cable, the cable construc-
tion can be of various types. The cable may be part of a
telecommunication system and one or two fibers made
available for temperature sensing. However, typically a
separate fiber cable is installed.
The advantages and disadvantages of using optical fiber
sensors in different configurations are given in Table 12-2.
Leak Detection
Rapid detection of a fluid leak reduces the amount of
fluid loss. Leak detection is especially important for
high-pressure fluid-filled (HPFF) cables, because of the
large quantities of fluid involved. Fluid loss is also
important in SCFF cable due to the significantly lower
reserves of dielectric fluid and the possibility of mois-
ture ingress and ultimate failure of the cable. As
described in Chapter 9, pressurizing plants contain sim-
ple leak detection in the form of fluid-level alarms on
the storage tanks or frequent-operation alarms on
pumps. Cumulative flow meters add further sophistica-
tion and reduce the amount of fluid loss before detec-
tion. More elaborate systems, combining software that
evaluates expected fluid expansion/contraction as a
function of cable loading, provide even further sensitiv-
ity (Engelhardt and Purnhagen 1991). Incorporating
these devices in the cable design and installations stages
is more effective and less costly than adding them after
the line is in service.
Leak Location
Once a leak is detected, rapid location minimizes fluid
loss. An EPRI project in the 1980s investigated many
leak location methods (Williams et al. 1983). Chapter 15
addresses the operating procedures for leak location and
repair. The engineer can take several steps at the design
stage to speed future leak location efforts.
Probes have been developed that can detect the very
slow fluid flow in the direction of a leak (Ghafurian et
al. 1989). These probes must be inserted into the cable
pipe. Valves can be added at any time. However, if they
are installed at the top of joint casings or perhaps on the
line pipe in cable manholes before cable installation, the
cost of installing them is lower, and there is no chance of
damage to the cable. These valves are also useful for
fluid sampling for dissolved gas analysis.
Leak location cables were developed in the late 1980s
(Ghafurian et al. 1989). These cables are sensitive to
dielectric fluids, and may be placed in the trench adja-
cent to a cable pipe or an SCFF cable. They indicate
that a leak has occurred and locate a leak to within
about 6 ft (2 m).
12.2.9 Grounding/Cathodic Protection
Another important design consideration is the type of
grounding needed to ensure the proper operation of an
underground cable system. A detailed discussion of the
grounding and cathodic protection for underground
cable systems is included in Chapter 10.
12.3 INSTALLATION MODES
12.3.1 Splices: Directly Buried vs. in Manholes
Cable splices are required to join sections of cable. The
distance between splices is governed by pulling tensions
or the length of cable that can fit on a reel, and varies
from less than 1000 ft (300 m) to more than 7000 ft
(2100 m). Splices are described in detail in Chapter 8.
Splices must be made in a clean, dry environment.
Humidity control must be provided for higher-voltage
cables, especially paper-insulated cables. Splicing is
generally done in a permanent manhole, although splic-
ing requirements can be met using a temporary man-
hole, which is then backfilled, thus creating a directly
buried splice.
Most pipe-type cable splices are installed in permanent
manholes. Directly buried splices are often used with
directly buried SCFF or extruded-dielectric cables.
Some users provide permanent manholes for easy access
to splices even if the cables themselves are directly bur-
ied. Permanent manholes are almost always used for
cables in duct. Because of the increasing importance of
dissolved gas analysis, more utilities are installing sam-
pling valves on splice casings for HPFF and SCFF
cables (or are using the evacuating/filling valves that are
already in place), so that permanent manholes are nec-
essary for access to the valves.
A temporary manhole can consist of a precast or
poured concrete pit, with plywood walls and tarpaulins,
or a temporary chamber placed over the pit to protect
the cables during splicing. Typical dimensions are
610 ft wide x 1520 ft long (1.83.0 x 4.66.1 m). After
the splice is complete and all tests are done, the pit is
filled with clean sand. Pavement is restored if necessary,
and a marker is placed so the splice can be located
quickly if repairs are needed.
12-22
Permanent manholes are typically 8 ft wide x 18 ft long
x 7 ft high (2.4 x 5.5 x 2.1 m), although sizes can vary
depending upon the number of circuits, cable type, and
voltage. Manholes can be precast, which are often less
costly if facilities exist nearby for casting and if many
manholes are ordered. Alternatively, field-poured man-
holes may be installed. Field-poured manholes are espe-
cially useful in areas where other utility lines are nearby,
preventing the installing of a precast manhole.
Almost all transmission cable manholes have two
accesses (chimneys) to facilitate cable installation. Man-
holes include significant steel reinforcement and the cor-
rect grade of concrete to ensure that there is no problem
with heavy truck traffic, and reinforcements are pro-
vided to permit installing cable-pulling accessories such
as sheaves. Figure 12-23 shows a precast manhole being
lowered into place.
12.3.2 Trenching
General
The most common method for constructing an under-
ground cable system is by open-cut trenching. This con-
sists of using a backhoe to remove the concrete, asphalt
road surface, topsoil, and subgrade material to the
desired depth, as shown in Figure 12-24. The material
can be stockpiled near the trench, if acceptable by the
local authorities, or removed to an appropriate off-site
location for disposal or used for fill as appropriate. The
pipe/duct is installed in the trench and the trench back-
filled. Fill materials can be thermal sand and/or a con-
crete mix. Depending on the depth, shoring of the
trench may be required. The U.S. Occupational Safety
and Health Administration (OSHA) requires any trench
deeper than 5 ft (1.5m) to be shored unless a competent
person determines that the soil conditions do not
require shoring. Along roadways, shoring may be
needed to maintain the trench sides and prevent damage
to the rest of the roadway.
During the construction of the underground transmis-
sion line, obstacles, such as major roadways, large utili-
ties, and railroads may be encountered that require
alternatives to open trenching. In cases where open
trenching is not an option, jacking and boring or hori-
zontal directional drilling (HDD) may be used. Discus-
sion of these installation methods are provided later in
this chapter.
Beginning in the mid-1980s, utilities began installing
fiber-optic cables or spare ducts for future fiber-optic
cables. The fiber-optic cables are used for cable system
communication or other utility requirements. In some
cases, the utility may make the fiber-optic cables avail-
able for outside communication company use. Recently,
additional ducts were added for temperature-monitor-
ing purposes.
Extruded-dielectric cables can be installed either directly
buried, as illustrated in Figure 12-25, or in duct, as illus-
trated in Figure 12-26. General practice in North Amer-
ica is to install the cables in ducts to facilitate removal
and replacement in event of cable failure. However,
directly buried cables have 1015% higher ampacities
than duct cables because there is no dead air space to
impede heat transfer to the earth. Reference (Tarpey et al.
1991) describes an extruded-dielectric cable installation.
The advantages of direct-buried cable systems are that
they are significantly less expensive to install than duct
Figure 12-23 Precast manhole being lowered into place
(courtesy POWER Engineers, Inc.).
Figure 12-24 Backhoe excavating a trench (courtesy
12-23
bank cable systems. Direct-buried cable systems can
require that significant lengths of trench be left open
until the cable is installed. Repairs to direct-buried cable
systems require the trench to be opened and the cables
exposed, which in itself may further damage the cable or
even the adjacent good cables. When the trench is
located within commercial or residential property, land-
owners can expect significant disruption to their busi-
ness or property due to the excavation required to
expose the cable system. Any type of cable can be direct
buried, provided that it is designed appropriately for
direct-buried applications.
Directly buried cables should be protected from above
to reduce the chances of dig-in. Precast or free-poured
concrete slabs are typically placed 1 ft (30 cm) or so
above the cables. Trench widths typically vary from 36
to 100 in. (91 to 254 cm), depending upon the number of
circuits and spacing among cables, and depths vary
from 46 to 52 in. (117 to 132 cm). Trenches for directly
buried cables often have controlled backfill installed to
improve ampacity. Spacing between adjacent phases is
usually 912 in. (2330 cm), and typical spacing
between circuits is 24 in. (60 cm) or greater. Fiber-optic
cables can also be installed. In addition, ground conti-
nuity conductors can be installed in the trench to carry
fault currents if required by the bonding scheme.
It is possible to install single-conductor cables in duct
using the flat configuration shown in Figure 12-25.
Ampacity will be reduced because of the dead air space
in the ducts.
Duct banks designed for the utility industry are fre-
quently located within or near public rights-of-way such
as along roads and highways, beneath sidewalks, or in
green spaces within the rights-of-way that are assessable
to the general public. It is common practice in these
cases to encase the duct system in high-strength con-
crete of at least 3000 psi for protection from mechanical
damage. The concrete encasement also provides a mea-
sure of safety for the personnel who may be installing
another facility in close proximity to the duct system in
the future. The latest edition of the National Electric
Safety Code should be consulted to ensure that all
applicable code requirements for minimum burial depth
are met.
In some cases, the municipal authorities may allow the
installation of a duct bank system backfilled with a spe-
cial soil or thermal sand instead of high-strength con-
crete. In this instance, the utility should install and
maintain a High-Voltage Buried Cable warning sys-
tem of signs indicating the hazard potential, and the
utility should regularly patrol the cable route, especially
in areas where excavation is most likely. A good practice
to protect the ducts from mechanical damage is to place
4-in.-thick (10-cm) preformed concrete slabs on top of
the thermal sand backfill directly over the ducts, or to
simply pour a 3- to 4-in. (7- to 10-cm) layer of concrete
directly from the concrete truck about 1 ft (30 cm) above
the cable. Ducts installed in this fashion without con-
crete encasement should be at least Schedule 40 in thick-
ness. Thinner walled ducts may become oval in shape
when encased in concrete if proper pouring practices are
not followed.
Figure 12-25 Typical directly buried
trench detail (courtesy POWER Engineers, Inc.).
Figure 12-26 Typical duct bank trench detail
12-24
The designer should consider including one additional
duct per circuit as a spare conduit to be used in the
event that a new replacement cable needs to be pulled in.
Should a cable experience a fault, the heat and duration
of the fault may bond the cable to the duct material,
prohibiting the removal of the cable. If there is a possi-
bility of a future circuit being installed along the same
alignment, consideration needs to be given to the
cost/benefit of installing additional ducts during the ini-
tial installation. The incremental cost of including addi-
tional ducts during the design stage of the project is
small compared to the initial cost of excavating the
trench and repaving or relandscaping. If future circuits
are planned but not going to be installed at this time, the
ampacity requirements for the ultimate circuit build-out
needs to be incorporated into the duct bank design.
A t ypi cal duct bank confi gurat i on i s shown i n
Figure 12-26. The ducts are installed in a vertical
stacked arrangement. Spacers hold the ducts in position
until a concrete envelope is poured around the ducts.
The concrete envelope provides mechanical protection
and aids heat dissipation. Since concrete has a low ther-
mal resistivity, controlled backfill is not generally
needed outside the envelope; clean fill is usually ade-
quate. The spare ducts may be used for fiber-optic
cables, ground continuity conductor, or replacement
power cable.
The duct materials primarily used today are polyvinyl
chloride (PVC), polyethylene (PE), and reinforced ther-
mosetting resin (fiberglass). Because transmission volt-
age cabl es are l arger and general ly heavi er than
distribution cables, and because it is economically justi-
fiable to design longer pulls rather than shorter pulls,
the coefficient of friction of the duct material becomes
an important consideration in the selection of duct
material. The designer must consider the type of cable
jacket that will be used on the cable to avoid selection of
two materials (cable jacket and duct material) that pro-
duce a higher coefficient of friction, thus limiting the
length of the cable pull and potentially requiring an
additional manhole. For instance, a PE jacket and a PE
duct material would produce a higher coefficient of fric-
tion than a PE jacket in a PVC duct.
Most concrete-encased duct bank systems and direct
buried conduit systems use PVC duct. This material
comes in straight sections and, when joined together,
makes for an extremely straight duct system without
undulations. On the other hand, PE duct can generally
be delivered to the site on reels. This material has suffi-
cient elastic memory to cause the individual duct runs to
undulate, which increases pulling tensions. Although
this is acceptable for lightweight distribution cables, it is
not satisfactory for transmission voltage cables.
Although more expensive that PVC, fiberglass duct has
the advantage of being less brittle and generally stiffer
than PVC, making it suitable for nonburied duct runs
that are supported at regular intervals between hanger-
type supports without significant bowing between sup-
ports. Fiberglass is also a good alternative for installa-
tions that will be above grade and exposed to the
elements. Fiberglass is highly resistant to weathering
with good UV radiation stability (sunlight resistance).
Fiberglass also has a lower coefficient of friction, allow-
ing longer cable pulls with less reliance on lubricants.
Another advantage of using fiberglass conduits is that
fiberglass has a lower coefficient of thermal expansion,
which makes it an excellent option for bridge attach-
ments and other exposed installations.
PVC conduits are available in several ratings: DB or
Schedule 20, Schedule 40, and Schedule 80. Schedule 40
is the most commonly used PVC rating, although many
utilities will allow DB-rated PVC if it is encased in con-
crete. DB is thinner walled and is more easily broken
during construction activities, especially by workers
stepping on the conduit and by concrete dropping onto
it during encasement. Schedule 80 is considered heavy
wall and is often used where above-grade protection of
the cable is important, such as cable risers on termina-
tion structures. Be aware that, without special formula-
tion for UV resistance, PVC will grow brittle with
extended exposure to UV light.
Further discussion on conduit installation is provided in
Chapter 13, Section 13.4.2.
Trench widths can vary from 12 to 54 in. (46 to 137 cm),
depending upon the configuration of the cable or con-
duit. Occasionally, wider trenches are used to keep the
depth of the trench as shallow as possible. Most utilities
require the top of the concrete to be a minimum of 36 in.
below grade. Typically, the minimum spacing between
the edges of the ducts is 3 in. (76 mm). There are spacer
designs with this spacing. If a greater distance is needed,
the spacer would have to be special ordered. The mini-
mum distance from the duct to the edge of trench
should be 3 in. (76 mm). A distance greater than 3 in.
(76 mm) from the duct to the edge of trench may be nec-
essary, due to the trench depth, to facilitate the installa-
tion of the duct into the trench, or to provide sufficient
space for reinforcing steel. Reinforcing steel may be
installed if additional duct bank strength is needed, but
should never metallically encircle the duct bank.
12-25
Pipe-type Cables
The cables for high-pressure fluid-filled (HPFF) and
high-pressure gas-filled (HPGF) pipe-type cable systems
are installed in steel pipes, which withstand pressure,
provide protection, and facilitate replacing failed cables.
The pipe can be installed directly buried or encased in a
low-strength thermal concrete, as shown in Figure 12-27.
Many utilities install two cable pipes per trench, with
one line able to carry the full load for the duration of a
repair on a failed companion circuit. Reference (Hatcher
et al. 1966) describes pipe-type cable installations.
Trench widths can vary from 18 to 54 in. (46 to 137 cm),
depending upon pipe size, whether there are one or two
pipes per trench, and whether welding is done alongside
or in the trench. Occasionally, wider trenches are used
for three-pipe installations, or for circuits containing
fluid-return pipes for forced cooling. Typically, mini-
mum 24-in. (61-cm) spacing is maintained between
cable pipes to permit working on the pipes during instal-
lation and during possible future repairs. Greater spac-
i ng al so reduces mutual heati ng effects, thereby
improving ampacity.
Trench depth is selected to give a cover over the pipe of
3642 in. (91407 cm) if possible, since a shallower
depth provides greater ampacity. At these depths, shor-
ing may not be required, since the trench will be less
than 5 ft (1.5 m) deep. In many cases, however, the depth
must be greater to avoid existing utilities, or to maintain
sufficiently large bending radii, etc. Intersections require
greater depths because of the need to pass under exist-
ing utilities in the cross streets.
The trench is commonly filled with a special low-ther-
mal resistivity controlled backfill to aid heat dissipation
and improve ampacity. The amount of backfill varies
with native soil conditions. Many utilities standardize
on 12 in. (30 cm) above the cables. See Section 12.2.6 for
a discussion of controlled backfill. The material above
the controlled backfill may be native soil if it has suit-
able thermal and mechanical properties. If the line is
under pavement, roadbed sub-base is then installed, fol-
lowed by the base pavement and final pavement.
SCFF Cables
SCFF cables are installed in a similar manner as
extruded cables. Reference (Kozak et al. 1965) describes
an extensive SCFF cable installation.
12.3.3 Trenchless Installations
During the process of planning a route layout for an
underground cable system, extensive research must be
performed to locate any encumbrances that would pro-
hibit open trench excavation and force the utility into
using various trenchless technologies: horizontal direc-
tional drilling, jack and bore, and tunneling/microtunnel-
ing. These encumbrances could be due to the following:
Numerous subsurface services that require a protec-
tive barrier between them and the cable system to
be installed.
Major street intersections that cannot be excavated
by the open trench method during certain times.
Some city governments do not allow any open trench
excavation or work during normal business or com-
muter hours in particular areas of their city. All work
must then be scheduled for the night hours.
Interstate or state highway systems including major
city parkways and boulevards where open trenching
is not permitted.
Railroad systems where there is no available existing
underpass that could be utilized.
Water or wetlands crossings.
Trenchless installation costs are generally greater than
costs for open trenching. However, it is quite possible
to reduce some installation costs by comprehensive
planning. When roads, highways, parks, airports, rail-
roads, etc., are built or rebuilt, utility planners should
take the opportunity to install bores, tunnels, man-
holes, or casings while open trenching is permitted,
Figure 12-27 Typical pipe-type cable trench cross section
12-26
thus reducing future installation costs. Being aware of
current and planned civil construction projects and
comparing them to anticipated cable routings can result
in significant savings, depending on the present worth
of future construction.
Descriptions of the various trenchless technologies are
given in the following subsections.
Horizontal Directional Drilling (HDD)
Horizontal boring operations have become quite popu-
lar with utilities. This began with small directional bor-
ing operations for utility distribution services (electric,
gas, and telephone). Large directional boring machines
were then developed for long lengths of gas and oil
transmission pipe systems. They were previously only
used for river or water crossings where the cost of exca-
vation would exceed that of a directional bore, or where
excavation of the river bottom would present environ-
mental problems. This type of boring is now used by
electric utilities for underground transmission projects
where conventional trenching is not an alternative. The
North American Society for Trenchless Technology
(NASTT) has prepared a guide entitled Horizontal
Directional Drilling Consortium HDD Good Practices
Guidelines (NASTT 2004).
HDD can be a preferable construction method because
of its minimal impact to the surrounding environment.
One of the drawbacks to using a HDD is the impact
on the circuit ampacity. A typical HDD installation
results in the cable being installed at least 20 ft (6.1 m)
below grade.
HDD installations are used for both pipe and duct
installations. Typically, the longer installations (greater
than 1.5 miles, or 2.4 km) use pipe-type cable because of
two main factors: (1) the pulling forces are distributed
over three cables, instead of just one, and (2) pipe-type
cables frequently have a smaller diameter, thus, allowing
more cable to be carried on one reel; therefore longer
distances can be installed.
General
The HDD method consists of three processes. First, a
small-diameter pilot hole is drilled from entry to exit.
Next, the pilot hole is enlarged by reaming. Finally, the
product pipe is pulled into the enlarged hole. Each indi-
vidual process is described in detail.
There are several factors to be considered in the design
process of an HDD crossing. Among these are depth of
cover requirements, underlying soil conditions, existing
utilities, entry and exit angles, radius of curvature, and
compound curves and available drilling and pull back
space. Each individual design factor is discussed in
detail below. Equations used in the design process are
also provided.
The first to be discussed is the depth of cover. A plan
and profile view of a typical HDD crossing is presented
in Figure 12-28.
The underlying soil strata of an HDD should be investi-
gated to determine the suitability of the HDD method.
Soil samples should be taken at each end of the crossing
and at the center point, and/or the deepest point, along
the crossing. This information will allow the designer to
determine if there are any unusual or unexpected mate-
rials that the drill string will encounter. This informa-
tion will enable the designer to plan for the appropriate
equipment and drill head, select the casing material, and
understand the risks of the operation before mobilizing
to the site.
Soil conditions such as silts, sands, and clays tend to be
well suited to directional drilling. Soils containing gravel
and cobble make all phases of the HDD process much
more difficult. Even though facilities have been installed
in numerous types of rock formations, HDD in rock is
oftentimes much more costly than a comparative instal-
lation in soil. This is due to the need for specialized
rock-drilling tools. These tools are discussed in detail in
a later section. During the design phase, a geotechnical
exploration of the subsurface should be investigated to
determine the suitability of the HDD method. The
depth of the installation is determined primarily by the
Figure 12-28 Plan and profile view of a horizontal directional drill (courtesy POWER Engineers, Inc.).
12-27
type of soils encountered. One of the primary design
concerns is the potential of frac out, the loss of drill-
ing fluid, during the drilling process. Loss of drilling
fluid can cause significant environmental problems. The
deeper the installation, the less likely a frac out will
occur, but the greater the impact on circuit ampacity.
Existing utilities to be crossed and other underground
facilities and structures in the proximity of a planned
HDD crossing must be identified, just as with any
underground digging activity.
The entry angle of the directional drill needs to be
understood when planning the route and the set-up
area. The entry angle is defined as the angle that the bit
enters the ground relative to horizontal. Most direc-
tional drilling rigs are capable of an entry angle of
between 8 and 16, although many drilling companies
prefer an entry angle of 11 to 12. A steeper entry angle
allows the drill path to reach its maximum depth rela-
tive to the entry point in a shorter horizontal distance.
The exit angle must also be considered. Exit angles typi-
cally range from 6 to 20 or more relative to horizontal.
Shallower exit angles are generally reserved for larger-
diameter steel pipeline installations. Shallow exit angles
(6 to 10) reduce the amount of pipe-handling require-
ments (cranes, side booms, etc.) on the exit side. The
downside is that a greater horizontal distance is
requi red when smal l exi t angl es are used, whi ch
increases the length of the crossing. Steeper exit angles,
defined as in excess of 15 are used where distance con-
straints are a prime consideration in the design process.
Steep exit angles also increase the need for additional
pipe-handling equipment.
One of the most important factors in the design of an
HDD crossing is the radius of curvature. This is of
greater importance when steel pipe is used. A general
rule-of-thumb in the design process is 100 ft (12 m) of
radius per inch (cm) of outside diameter (OD) of prod-
uct pipe. For example, the radius for a nominal 10-in.
OD (25.4 cm) steel pipe would be 1000 ft (304.8 m).
Larger radii are desirable if sufficient room is available
in the length of the crossing. For high-density polyethyl-
ene (HDPE) pipe, much tighter radii can be used. As a
rule-of-thumb, the minimum radius of curvature for
HDPE pipe is given as 50 times the outside diameter.
Radii this small are generally not used, because the steel
drill pipe used to drill the pilot hole is incapable of toler-
ating relatively small radii. Equations used in the design
of HDD drill paths are presented below.
Horizontal Distance of Curve = 12-2
Where:
R = Radius of Curvature
= entry or exit angle (degrees)
Vertical Distance of Curve = 12-3
Where:
Length of Curve = 12-4
Where:
Compound curves should be avoided in the design
of HDD crossings whenever possible. A compound
curve occurs when there is a change in direction in both
the horizontal and vertical planes simultaneously. In
certain cases, it is an unavoidable situation. A method
for estimating the compound radius is given in the fol-
lowing equation:
R
c
= 12-5
Where:
R
c
= Compound Radius of Curvature
R
v
= Vertical Radius of Curvature
R
h
= Horizontal Radius of Curvature
Site Layout
In addition to the necessary space required to set up the
drilling machine, sufficient space is needed for preparing
the duct/pipe for pull back. This pullback area can be
narrow, but can be extremely long. Optimally, the length
of the pull back area should equal the length of the drill;
however, sometimes that may not be possible. Another
thing to consider is the weight of the pipe/conduit that is
being installed.
The size of the drilling area is dependent on the type of
drill rig required. HDD rigs are available in many sizes,
ranging from 7000 lbs (31,150 N) of pullback force to
1,000,000+ lbs (4450 kN). Also, the rigs have rotary
torque rangi ng from approxi mately 4000 ft-l bs
(5420 N-m) to in excess of 80,000 ft-lbs (108,465 N-m).
The ancillary equipment that is used with the rig is a
mud pump, mud-cleaning system, drill pipe, a tempo-
rary site office, a control unit for the rig, power unit for
the rig, water supply, and a tool/spares container. A typ-
) (sin R
) cos 1 ( R
180
R
2 2
2 2
) ( ) (
) ( * ) (
h v
h v
R R
R R
+
12-28
ical HDD site layout is shown in Figure 12-29. A typical
footprint would be about 150 x 250 ft (45.7 x 76.2 m).
A photo of an HDD rig is presented in Figure 12-30.
Extruded cables are always installed in inner ducts
within a bore, both with and without casings. Though
technically feasible, a cable system has never been
installed directly into a bore: most owners consider this
risky since the cable could be damaged during installa-
tion, and there would be no opportunity to pull out
damaged cable in the event of a failure. Historically, the
method of installing single-core cable with HDD tech-
nology was to install an outer casing with the individual
conduits installed inside.
Installing a cable system without a casing has some
advantages, including avoiding the cost of the casing
and possibly smaller trenchless bending radii, perhaps
meaning the length of the bore is shorter since the con-
duit system can get down to depth more quickly without
as much setback. When no casing is used, the conduits
or cable pipes are installed directly during pullback, and
there is no opportunity to place an engineered grout
backfill around the conduits or cable pipes. If the
ampacity goals can be achieved, this is typically the low-
est-cost alternative. In addition, there is one less pull-
back step since the inner ducts or cable pipes do not
need to be pulled back into the casing pipe after the cas-
ing is installed.
Casing Design Considerations
A number of design considerations are associated with
installing a casing using the HDD method. The first
consideration is to determine the casing material. A
common casing material is HDPE. Steel casings are also
used, but use of steel casings may result in a significant
reduction in ampacity due to additional losses from hys-
teresis and eddy current losses in the steel pipe. Also an
HDPE casing is less likely to affect the tracking of a
parallel bore as compared to a steel casing. The next
step is to determine the size and thickness of the casing.
The diameter of the casing is determined by the number
and size of the ducts that will be installed in the casing.
The casing thickness is determined by the calculated
stresses associated with the HDD design, as discussed
previously. A plastic casing offers some advantages over
steel in that the material is softer and more flexible, so
bending radii may be smaller than a comparable steel
casing. However, a steel casing has a higher tensile load
limit for pullback.
When pulling the duct bundle into a casing, special
spacers with rollers are usually used to maintain the
conduit spacing and lower the forces required to pull in
the duct package. The most common material used for
ducts within casings is HDPE, although PVC (polyvinyl
chloride) and fiberglass can also be used. Once the cas-
ing and ducts have been installed, it is recommended
that the casing be filled with a thermally approved grout
to eliminate the air pocket inside the casing and improve
the ampacity of the cable circuit. Depending on the
length and depth of the HDD installation, this can be
very problematic. The main design concern is the grout
pressure, associated with filling the casing with grout.
The designer needs to be sure that the duct material can
withstand the grout pressure. If grouting is not used, the
designer must consider the derating effect of the air
pocket on the circuit ampacity.
Figure 12-29 Typical HDD site layout (courtesy POWER Engineers, Inc.).
12-29
Bundle Design Considerations
The main design consideration for installing the ducts
as a bundle is to adequately size the ducts to withstand
the installation stresses. This normally requires a larger
wall thickness for the duct than what is needed for a
normal trench installation. Because the larger wall
thickness affects the inside diameter, a larger duct may
be needed. This will cause some problems with the inter-
face of the HDD duct to the normal trench duct. Either
the HDD duct is taken directly into a manhole, or a spe-
cial coupling is needed to make the transition. The con-
cern with the transition coupling is to make sure the
interface is smooth to prevent damage to the cable dur-
ing installation. The primary duct material used for this
type of HDD installation is HDPE due to the strength
of the joint. The HDPE is usually butt-fused together,
resulting in a very strong joint. Depending on the length
of the drill and the type of geography, the designer may
consider pressure-grouting the hole to improve the ther-
mal environment. If grouting is not used, the designer
must consider the thermal properties of the drilling mud
to determine the impact to the circuit ampacity.
Pipe-type Cables
Generally, an outer casing is not needed for pipe-type
cables, unless specifically required by local regulations.
For all cable systems, but particularly for pipe-type
cables, the casing offers protection and a degree of sec-
ondary containment in the event of a dielectric fluid
leak. If no casing is required, the cable pipe is pulled
directly into the hole similar to a casing. The thickness
and coating of the pipe are the main design concerns.
Depending on the length of the drill, the pipe thickness
may need to be increased to withstand the installation
stresses. This may cause some problems when the
thicker HDD cable pipe is connected to the normal
cable pipe. It is recommended that the HDD cable pipe
be extended to the manholes. Otherwise, a special cou-
pling is needed.
The recommended coating for an HDD cable pipe is a
fusi on-bonded epoxy wi th an external POWER-
CRETE. This provides a stronger coating to with-
stand the abrasi ons that coul d occur duri ng the
pullback. Another option is to increase the normal
coating thickness.
SCFF Cables
SCFF cables are installed similar to the method
described above for the extruded-dielectric cable.
Because of the associated depth, it is important to con-
sider any pressure-related issues.
Jack and Bore
Another trenchless alternative is boring, which involves
augering a holes and jacking a casing into the hole
simultaneously. The boring method is generally used to
provide an opening for cable systems under roadway
crossings, street intersections, and railroad crossings,
where no bends are necessary for the cable route. Bores
are normally less costly to install than other trenchless
methods, because less manual labor is required and they
can be installed faster.
General
One of the main design considerations with a jack-and-
bore operation is to ensure that there is adequate space
to perform the bore. To initiate a casing installation, a
pit having a minimum size of 40-ft (12.2-m) long by 10-
to 15-ft (3- to 4.6-m) wide is required by the boring
equipment and for placing and welding 20-ft (6.1-m)
sections of casing pipe. Also, prior to starting the boring
process, an exit pit approximately 10 ft (3 m) in length
should be excavated. Since a bore is basically at a
greater depth than an open trench, the entrance and exit
pits require shoring (and possibly tight sheeting) in
accordance with OSHA regulations. Secondly, if the
boring is to be done in poorly consolidated soil (soil that
begins to slough or flow if unsupported after a few min-
utes), solid sheeting is required to enclose the entire
entrance pit, allowing only an opening for inserting and
installing the casing and auger.
Casing sizes can vary for cable systems from 14 to 84 in.
(36 to 213 cm), depending upon the type of cable sys-
tem, number of circuits being installed, and length of
the bore. It is suggested that the design engineer select a
casing size with sufficient space or clearance to permit
an easy and thorough backfilling operation after the
cable system has been installed. Generally, casings
Figure 12-30 Horizontal directional drilling rig (courtesy
12-30
should be sized 8 in. (20 cm) larger than the OD of the
cable or cable pipe. Borings should be limited to lengths
less than about 400 ft (122 m). The longer the bore, the
more difficult it is to control the direction of the bore
and to fill the casing after the pipe/duct installation.
Selection of the casing material is also very important.
Historically, most bores have been installed with steel
casings, but there has been a trend to nonmetallic cas-
ings such as concrete, HDPE, fiberglass-reinforced con-
crete, and composites. The designer must consider the
derating effect on cable ampacity when using a steel cas-
ing or a concrete casing containing steel reinforcing.
Using a steel casing may reduce the circuit ampacity by
510%, for non-pipe-type cable systems. Each of these
materials has its own advantages and disadvantages and
may not be applicable to a particular application. Addi-
tionally, the owners of the facilities granting the crossing
permit will frequently have specific requirements for the
casing materials, thus, requiring the designer to use a
specific material. Contractors specializing in jack-and-
bore installations are the best source of information
regarding the application of the various casing materials
and their limitations in particular soil types.
Casing diameters for typical cable systems will usually
be between 14 and 84 in. (36 and 210 cm), depending on
the type of cable system, number and size of inner con-
duits, number of circuits being installed, and length of
the bore. Additionally, the wall thickness of the casings
will vary considerably, depending on the casing material
selected. For instance, a polyethylene casing will typi-
cally have a wall thickness of 2 to 3 in. (5 to 7 cm),
whereas a typical steel casing may have a wall thickness
of 0.5 in. (1.25 cm). Therefore, the resulting inner diam-
eter of the casing may be smaller for a nonsteel casing
than for a steel casing. It is suggested that the design
engineer select a casing size with sufficient space or
clearance to permit an easy and thorough backfilling
operation after the cable system has been installed. Gen-
erally, casings should be sized 8 in. (20 cm) larger than
the outside diameter (OD) of the cable or cable pipe.
Once the casing is installed, the pipe/duct is installed
using specially designed spacers. The pipe should then
be filled with a thermally approved grout. Since the
grout needs to be pumped into the casing, consideration
must be made to ensure that the grout is pumpable and
still retains good thermal properties.
Since the number and size of ducts vary between
projects, the duct spacers are typically designed on a
project-to-project basis. The spacers are needed to
ensure the proper separation between the ducts. Spacers
with rollers are recommended to allow for easier instal-
lation of the conduit bundle. Detailed ampacity calcula-
tions should be made to determine the optimal duct
spacing for each installation. Typically, a minimum sep-
aration of 3 in. is recommended.
Pipe-type Cables
Normally, only one cable pipe is installed in a single cas-
ing. Polymer spacers are needed to protect the pipe coat-
ing by keeping it off the casing. The spacer runners can
be from 0.75 to 12 in. (19 to 305 mm) in length. With
longer runners, other ducts can be attached to the pipe
and pulled in with the cable pipe. These ducts are typi-
cally located on the outside of the spacer between the
runners and tied to the cable pipe using nylon ties.
SCFF Cables
described above for extruded-dielectric cables.
Tunneling/Microtunneling
Other trenchless alternatives are tunneling or microtun-
neling, which involve digging and building the tunnel
simultaneously. These methods may be selected if the
size of the opening is too large to be accommodated by
either HDD or jack and bore.
Conventional tunneling is similar to pipe jacking. The
tunnel is excavated manually or mechanically, and the
spoil is removed using an auger system, carts, or con-
veyor. The difference is in how the pipe is installed. In
the case of tunneling, the pipe or tunnel liner is added at
the face of the tunnel. A temporary casing precedes the
liner plates, which are assembled at the face of the tun-
nel and added on. The temporary casing is then jacked
forward against the liner, as opposed to the entire length
being jacked forward. Additional pieces are added, and
the procedure continues. The diameters of the casings
are usually greater than 3.5 ft (1.2 m), with essentially
no upper limit. The driving distance also is unlimited.
This method may also be selected if cooling of the cir-
cuits is required to achieve the desired loading require-
ments. This method is very costly and should only be
used when the other methods cannot be used effectively.
Microtunneling is a horizontal boring technique that is
used in instances where the diameter of the final casing
is greater than what is customarily used in traditional
jack-and-bore operations or where rock is encountered.
Microtunneling uses highly sophisticated, laser-guided,
remote-controlled equipment. This enables the user to
monitor its precise location and to ensure accurate line
and grade. This system typically consists of a steerable
tunneling machine that is jacked from a shaft ahead of
the permanent pipes or a temporary casing. The spoil is
removed by using either the auger method or the slurry
12-31
method. The auger method uses a continuous flight of
augers to remove the spoil. The auger is surrounded by
a casing placed inside the permanent pipes or the tem-
porary casing. The slurry method uses a slurry mixture
to pump the spoil back to the surface.
Any of the cable types may be used in tunnels. Com-
pressed-gas-i nsul ated cabl e can be an attractive
approach for tunnel installations, especially those with
large elevation changes, since there is no fire danger and
no problem with hydrostatic pressures. Ventilation is
important should a gas leak occur. Thermal expansion
is an important consideration since the aluminum enclo-
sure has a high coefficient of thermal expansion. It may
be necessary to anchor every three to five 4060 ft
(1218 m) sections, and to provide expansion relief in
the joints and enclosure between anchors. Further dis-
cussion on tunneling and microtunneling is provided in
Chapter 13, Section 13.3.1.
General
It is important to note that, because of the size of the
tunnels, it is not practical to fill the tunnels, so special
consideration needs to be taken to calculate the ampaci-
ties of the circuits based on the cables being installed in
air. Forced ventilation may be needed to achieve the
desired ampacities.
Leak detection and fire protection are very important
for tunnel installations of HPFF and SCFF cables. Leak
detection wires may be installed alongside the cables,
and fluid sensors can be installed in joint bays. Smoke
and fire monitoring can be provided using conventional
detectors or newly developed, fiber-optic temperature-
monitoring cables. For extensive tunnel installations in
Japan, the cables are typically placed in sand-filled
troughs, which provide cable restraint, contain possible
leaks, and provide fire protection (Nakagawa et al.
1973). Concrete protection has been provided in other
tunnels (Chaaban et al. 1991). Figure 12-31 shows an
SCFF installation in a tunnel.
A design concern for installing a single-core cable in a
tunnel is to determine how the cable is to be supported.
Extruded-dielectric cables are normally installed in tun-
nels on support brackets spaced at 5 to 10 ft (1.5 to 3 m)
intervals, attached to the sidewall or to floor-mounted
stanchions. By allowing the cables to hang in a catenary
between supports, the TMB forces are relieved.
Pipe-type Cables
As with the single core, a major design concern is the
method used for supporting the cable pipe within the
tunnel. Because of the cable coating, it is important to
design a system that allows the cable pipe to expand and
contract, but not damage the coating. Analysis should
be made on the cable pipe to determine the design limi-
tations. The pipe of pipe-type cables can be restrained,
and expansion can be accommodated, but restraining
the cables themselves is difficult in applications where
there are large elevation changes. Riser-type cable,
which has long-lay stainless-steel outer tapes to hold
most of the cable weight, may be required. As described
in Chapter 8, special anchor and skid joints are required
with the riser cable.
SCFF Cables
described above for extruded-dielectric cables.
12.3.4 Bridges
Existing or new bridges can be an attractive alternative
to submarine cables for some water or railroad cross-
ings. When considering installing cable on a bridge, it is
very important to contact the owner of the bridge and
obtain all available design drawings. Some existing
bridges are not designed to accommodate additional
attachments. So, it is very important to have an existing
bridge analyzed to determine whether it can handle the
additional load. Another concern is the possibility that
the bridge may need to be rebuilt in the near future. This
may cause additional cost to the utility, because the util-
Figure 12-31 Cable installation in a tunnel (Ray et al. 1974)
(courtesy IEEE).
12-32
ity may be required to relocate the cable. An additional
alternative to attaching to an existing or new bridge is to
install a utility bridge. Some design considerations for a
separate utility bridge are the following:
Protection from sun. Solar radiation will derate
the cable.
Physical protection. Consideration should be made
to protect the cable system from potential
external damage.
Cable/Duct radii. In many cases, sharp radii may be
required to attach to a separate utility bridge. This, of
course, will affect the pulling tensions and sidewall
pressures on the cable.
General
Cable/pipe clamps, cable weight, vibration, and methods
to account for bridge expansion joints are the major
design issues. Fairly elaborate schemes have been
employed to account for the large motion that occurs at
bridge expansion joints (Minemura and Maekawa 1989).
Compressed-gas-insulated cables are not generally
considered for bridge use because of thei r l arge
space requirements.
Single-conductor cables are more easily installed on
bridges than pipe-type cable, since bending can be
accomplished by cable snaking, and elaborate expan-
sion joints are not required. If the bridge is short, ducts
may be used. The advantage of ducts is that they pro-
vide added mechanical protection and make installing
the cable easier. The main concern for installing ducts is
to have a duct that will not degrade over time due to UV
rays. Fiberglass duct is the preferred duct material for
bridge applications. Figure 12-32 shows a fiberglass duct
bridge crossing.
Pipe-type Cables
Many short, pipe-type cables have been installed on
bridges in North America (Sheirer and Winistsky 1958;
Gillette 1959). Pipe expansion is accommodated by 90
or S-bends, or less commonly by expansion joints. Cable
weight, pipe expansion, bridge expansion joints, and
corrosion protection from road salts are design consider-
ations. Additionally, where the pipe is allowed to move
through its expansion range, Teflon-coated skid plates or
similar material can be used to protect the pipe coating.
SCFF Cables
There are significant numbers of SCFF installations on
bridges outside North America (Marquez et al. 1981),
including a 14-mile (22-km), 500-kV line installed in the
late 1980s (Minemura and Maekawa 1989). During
winter months, SCFF cables can experience excessive
pressures when load is restored after the cables are
de-energized, because of high fluid viscosity at low
ambient air temperatures on the bridge. Lower viscosity
fluids and larger fluid areas are used to reduce the pres-
sure problems.
described above for extruded-dielectric cable.
12.3.5 Underwater Installations
Underwater power transmission cables have been success-
fully installed for over 65 years. This experience has
included all of the transmission cable types (with the
exception of compressed-gas-insulated and superconduct-
ing cables) in a wide range of installation environments.
While the longest submarine cables are dc, primarily
because of charging-current limitations associated with
long ac cables, the majority of submarine cable installa-
tions are ac. Chapter 7 provides a review of worldwide
submarine cable experience. Submarine cable installa-
tions involve a variety of potential hazards during
installation, operation, and even during retrieval for
repair. They are far more expensive to install than land-
based cables and extremely expensive to repair. Thus,
careful planning and thorough engineering of underwa-
ter cable installations are essential.
Installing cables underwater involves many technical
and economic design considerations beyond those in a
conventional underground cable installation. Selecting a
cable type and designing it specifically for underwater
installation involve more than defining the power trans-
mission capability and the desired transmission voltage.
Requirements for installation, operation, effects of
marine life and environment, fault location, cable
retrieval, and cable repair, as well as the effects of the
installation on the environment, are all factors to be
Figure 12-32 Bridge crossing with fiberglass duct (courtesy
12-33
studied and evaluated in the planning stage of an under-
water installation. This is necessary to select the correct
cable, cable configuration, route, handling equipment,
and installation method.
Typical Design Aspects of Submarine Cables
Cables for underwater applications are similar in their
basic electrical design to conventional underground
cables, but they differ in their exterior mechanical
design, sheath requirements, and electrical bonding
methods. The most significant difference involves the
addition of armor, both to protect the cable from exter-
nal mechanical damage and to provide mechanical
strength to allow for cable installation and retrieval. The
cable jacket and sheath design also have a critical role in
ensuring the integrity of the cable. A more detailed dis-
cussion on submarine cable system design is provided in
Chapter 7. The following are key design aspects.
Spacing
Mechanical strength and protection
Sheath/jacket requirements
Corrosion protection
Sheath bonding
Cable losses
Types of Submarine Cable
The cable/utility industry has had successful submarine
experience with a variety of cable types, including
high-pressure fluid-filled pipe-type cables (HPFF), self-
contained fluid-filled cables (SCFF), conventional
solid-type mass-impregnated paper-insulated cables,
paper-impregnated gas-pressure-assisted cables, and
extruded-dielectric insulated cable. Selection of cable
type depends on the desired transmission operating
voltage and power transfer level, and the depth, length,
and profile of the circuit. The alternative cable types are
briefly described below.
Fluid-Filled, Fluid-Pressurized Cables
The highest-voltage submarine cables are those that are
impregnated and pressurized with dielectric liquid (pipe-
type and self-contained fluid-filled). Pipe-type and
SCFF cables are considered suitable for voltages up to
765 kV. These cables possess three advantages: their
high power transfer; their ability to match overhead
transmission voltages, which means transformers and
other substation equipment could be eliminated at the
land /submarine transitions; and the fact that they are
self-protecting and provide warning in case of damage
to the sheath or pipe. In case of limited sheath damage,
the pressurized systems will pump dielectric fluid out at
the site of the damage, minimizing water and other con-
taminants from entering and traveling up into the cable.
Fluid pressure drop or reduced fluid levels in the pres-
surizing plant tank or reservoir can provide a warning of
potential sheath damage. However, the loss of dielectric
fluid is an environmental concern. These cable types are
only feasible for moderate water depths and short-to-
moderate lengths. Flow rate and pressure drop con-
straints related to transferring fluid in and out of the
cable due to load changes preclude using these cables for
really long lengths, generally more than 30 miles
(48 km). As the installation depth increases, the cable
must be maintained at a higher and higher pressure to
prevent the ingress of water should a leak occur, since
the specific gravity of the dielectric fluid is somewhat
lower than water. The above-surface portion of the cable
and its accessories must be designed to withstand the
fluid pressures necessary to provide slightly positive
internal pressure at the deepest point in the submarine
installation. For example, an 80 psi (550 kPa) pressure is
needed for the 1312-ft (400-m) depth of the Vancouver
Island cables.
High-Pressure Fluid-Filled Pipe-Type Cables
Pipe-type cable for submarine application has the same
construction as for land-based cables. Three impreg-
nated-paper-insulated conductors are installed in a pipe,
which is filled with dielectric fluid and pressurized.
Design differences for submarine application may
involve size and strength of skid wires, pipe wall thick-
ness, and concrete pipe coating. Heavier-walled pipes
are often used (0.5-in. or 0.375-in. vs. 0.25-in, wall
[12.7-mm or 9.5-mm vs. 6.4-mm wall]), and provide a
number of benefits, including increased mechanical
strength, greater weight to reduce buoyancy, better abil-
ity to withstand a corrosive or electrolysis condition in
the event of a coating failure, and reduced probability of
burn-through during a fault. A concrete coating is often
added to the exterior of the pipe for mechanical and
corrosion protection and for negative buoyancy.
The advantage of the pipe-type design is that the pipe
provides mechanical protection for the cables and allows
the three phases to be installed in close proximity. This
minimizes the installation trench size, spoils removal,
and construction time, and reduces environmental
impacts during construction. Pipe-type cable is not nor-
mally considered for long underwater crossings, because
it is limited by the mechanical forces required to pull the
cables into the pipe. The installed length cannot contain
field splices; splices do not fit in the pipe, nor do they
have the strength to be pulled into the pipe. In the event
of an electrical failure, the cables can be pulled out and
replaced from the land ends, with no need to deploy sur-
face repair ships or perform at-sea splices. The longest
section of submarine pipe-type cable, installed across
Long Island Sound near New York City, is 6900-ft
(2100-m) long (Bazzi 1982). One of the more ambitious
12-34
projects was a 230-kV, 2-mile (3.2-km) crossing of the
Baltimore Harbor with five cable circuits and two
splices per circuit, which were made on platforms in the
harbor (Ruekert and Bien, Jr. 1979).
SCFF Cables
Single-conductor cables of this type utilize a conductor
with a hollow core for feeding dielectric fluid to the
cable, which is insulated with paper tapes impregnated
with a low-viscosity cable fluid. Three-conductor
designs, which are not normally utilized above 138 kV,
provide ducts in the spaces between conductors for flow
of dielectric fluids. The fluid can be pressurized to over
200 psi (1380 kPa) during operation. The insulated con-
ductor is covered with a lead-alloy sheath, insulating
jacket, bedding, and armor. This design can be installed
to depths of 3280 ft (1000 m). The length of the cable
depends on dielectric fluid-feeding limitations. With
very low-viscosity fluids, large duct size, and high fluid-
feed pressures, this cable could be installed in lengths of
over 20 miles (32 km).
Impregnated-Paper, Nonpressurized, Solid Cables
This cable type consists of a conductor insulated with
paper tapes impregnated with a very viscous dielectric
fluid. The insulated conductor is covered with a lead-
alloy sheath to prevent water ingress. An insulating
jacket, bedding, and armor wire are applied outside the
lead sheath. This cable is considered suitable for opera-
tion up to 45 kVac and 250 kVdc. Without changes to
the conductor and armoring design, this cable is limited
to installation in water depths of approximately 1500 ft
(500 m) (Bazzi 1982). Its greatest advantage is for use on
long submarine cables. It can be installed in very long
lengths, because it is nonpressurized, and thus limita-
tions regarding fluid-feeding pressures and fluid chan-
nel size do not pertain. The disadvantage of the
nonpressurized insulation is that the maximum operat-
ing voltage is limited, due to ionization limitations, par-
ticularly with alternating current. Also, its maximum
operating temperature must not exceed approximately
60C or the impregnant in the cable will drain, causing
dielectric failure.
Gas-Filled, Pressurized Cables
This cable design utilizes a hollow-core, copper conduc-
tor wrapped with paper insulation that is impregnated
with a high-viscosity (jellylike) liquid. The cable is pres-
surized with gas. The advantages of this design, com-
pared to liquid-pressurized cables, are that the duct size
can be very small, thus reducing overall cable size and
weight, and extremely long lengths of cable are feasible,
since no dielectric liquid-feeding limitations exist.
Installation depth is limited to approximately 1300 ft
(400 m), based on gas pressure limitations. Below that
depth, the cable could be crushed due to water pressure.
This is only true if gas pressure is lost; the cable must be
designed for this contingency. The maximum voltage for
this cable is approximately 275 kVac and 250 kVdc
(Bazzi 1982).
This cable type has been used commercially for land
installations at 230, 345, and 400 kV, and short sections
of cable are presently operating under test at 500 kV. It
consists of a conductor insulated with extruded solid
dielectric, such as cross-linked polyethylene or ethylene-
propylene rubber (up to 138 kV). The advantage of this
construction is that the dielectric is truly self-contained.
Without the need for internal pressurization, a heavy
wall sheath is not required, further reducing the cable
size and weight. Length is also not limited by fluid-
feeding limitations. However, the dielectric performance
of cross-linked polyethylene is significantly weakened in
the presence of even microscopic amounts of water.
Thus, although there is no internal pressurization, a
watertight metallic sheath is still critical for cross-linked
polyethylene cables. The major disadvantage is the basic
performance of the insulation. Unlike laminar impreg-
nated-paper dielectrics, extruded dielectrics are not for-
giving of defects in the insulation. One contaminant or
void can cause an eventual breakdown of the cable.
Thus, if extruded insulation is to be used for a subma-
rine transmission cable, great care should be taken to
ensure that manufacturing quality and factory testing
are performed using state-of-the-art materials, equip-
ment, and techniques.
Splicing Submarine Cables
The designs of splices for submarine cables differ from
those for land-based cables due to the mechanical
requirements associated with installation and retrieval,
as well as operation under high water pressures. The
required depth of installation and the method to be used
during installation are critical. For example, installa-
tions where the splice must pass over a capstan ten-
sioner, through a linear cable tensioner, or over a
moderately sized overboarding sheave, often place
severe design constraints on the maximum diameter and
length of a splice. Therefore, SCFF and extruded-dielec-
tric submarine cables are usually supplied in a single
length, with smaller factory splices being used to join
factory lengths of conductor. In three-conductor sub-
marine cable, these factory splices are usually staggered
over a length of 2550 ft (7.615.2 m) to keep overall
dimensions to a minimum. Shallow-to-moderate water
depths, where the splice can be lowered off the side of
the cable installation vessel, would not necessarily
restrict the splice size in the same way, nor would the
mechanical stresses be the same.
12-35
The electrical design of a splice for submarine cable is
no different from that of a land-based cable in terms of
maximum allowable radial and tangential electrical
stresses. Requirements of the cable-handling equipment
tend to be primary factors in the configuration of the
splice. For example, limitations of the cable-tensioning
equipment make it desirable for the submarine splice to
be the same or close to the diameter of the cable.
Whereas splices for land application typically have a
maxi mum i ns ul at i on out er di amet er ( OD) of
1.52.0 times the cable outer diameter, splices for deep-
water application are often designed as a reconstruc-
tion of the cable insulation, matching the cable insula-
tion outer diameter. Thus, the length of splice tapers,
the dielectric tapes used, and the method of application
are defined in large part by the desired splice OD.
The mechanical design of the splice is also determined
primarily by how it is to be lowered to the sea bottom.
Splices that must pass through or over a cable-tension-
ing device are designed to be on size with the cable.
These cable splices often use armoring similar to that of
the cable. The armor wires from one side of the splice
are hand-applied to the splice area and individually
welded to the cable armor to effectively reconstruct
the cable.
Cable splices to be installed by lowering over the side are
often installed in a submarine splice casing. Making the
splice is similar to that for a land-based cable, including
installing a lead sleeve over the complete splice. The cas-
ing has clamps on both sides, both to hold the cable
armor wires and to take the mechanical load from the
armor wires. After the splice is made and installed inside
the casing (lead sleeve and all), the casing is filled with
semi-hard insulating compound to provide further
assurance that the splice is mechanically protected and
water ingress is prevented.
Generally, it is desirable to minimize or eliminate alto-
gether the use of splices in submarine cables, because a
splice adds a site of potential weakness in the electrical
insulation and mechanical structure. Further, if splices
are to be made at sea, exposure time to possible storms
and shipboard failures is obviously increased. Installa-
tion and handling are made more complex depending
on the splice design. Submarine cable circuits are more
reliable if they do not contain splices, at least not in the
lead sheath, the reinforcement, or the armor. Most
major underwater cables have been produced splice-free
in long lengths stretching from shore to shore. Where
necessary, splices have been laid successfully in water of
moderate depth such as the English Channel (Gazzana-
Priaroggia and Mascio 1973).
There are three different types of splices to be consid-
ered: flexible factory splices, field splices (at sea), and
repair splices. Flexible factory splices are provided to
allow the maximum length of cable to be installed with-
out the need for splicing at sea. Factory splices are supe-
rior to field splices in that their armor can still be applied
continuously, splice free, and they are made in highly
controlled conditions with trained factory technicians.
At-sea splices present greater risks, in that the splicing
must be accomplished on the vessel in varying weather
and sea conditions, and at-sea splices involve the recon-
struction of conductor, insulation, shields, jackets, and
armor. Extra cable is needed to allow the cable to be
brought up to the vessel, resulting in a U-bend in the
cable. While splicing is performed in a temperature- and
humidity-controlled enclosure, the cable-tensioning
equipment and vessel-positioning systems must be capa-
ble of holding the cables steady during the splicing oper-
ati ons. Heavy seas and severe deck moti ons can
obviously affect the quality of the splicing or, at least,
extend splicing time. In very heavy seas, it is sometimes
necessary to cut the cable free to protect the cable vessel
and its crew. At-sea splices should be completed
as quickly as possible to minimize exposure to bad
weather and sea conditions; if possible, they should be
avoided entirely.
Even if a submarine cable is designed and installed with-
out splices, a repair splice design must be available in
case of damage to a cable in mid-run, unless replace-
ment of the total length of submarine cable, shore to
shore, is economically and logistically feasible. There
have been cases where the submarine cable has been
damaged during the initial installation, in one case by
the anchor of the installation vessel. Thus, the repair
splice design should be completed before the installation
begins, and a suitable number of repair kits should be
available at the time of installation.
Underwater Installation Environment
Designing an underwater crossing requires consider-
ation of a number of factors related to the installation
environment. These factors affect the selection of the
type of cable, detailed cable construction, overall cable
transmission system configuration (i.e., how many cir-
cuits, how many spare phases, etc.), and the route of the
cable crossing. Obviously, length of the cable and maxi-
mum water depth are key technical issues, but addi-
tional factors must also be evaluated.
Underwater Profile and Bottom Topography
The underwater profile and bottom topography can be
significant factors in selecting the type of cable, the
route, and the method of installation. A mild, uniform
12-36
elevation change for a cable crossing is generally desir-
able. Securing cables on steep embankments usually pre-
sents some difficulties, and once installed, they could be
subject to problems such as impregnant migration with
solid paper cables, or creepage of the insulation. Cables
on steep embankments are also more prone to damage
from submarine landslides. Many methods have been
used to secure cables on steep embankmentsincluding
concreting or grouting in place, anchoring, snaking,
etc., all of which appear to have worked, but require
periodic inspection. The submarine cable design engi-
neer should select the most appropriate method based
on site conditions.
A rough, erratic bottom makes cable installation diffi-
cult and time consuming, and therefore expensive. If the
cable is to be laid without being embedded, it is impor-
tant for the cable to lay flat on the bottom. If the bot-
tom is such that lengths of the cable are unsupported
and underwater currents are present, these sections of
the cable could be subject to damage due to cable move-
ment. Damage due to vibration from underwater cur-
rents (called strumming) can include abrasion, chafing,
and sheath fatigue. If the cables are to be embedded, a
rough bottom makes the job of trenching and placing
the cables far more difficult. Some waterways have sig-
nificant numbers of man-made obstacles such as sunken
barges, anchor lines, etc.
Thus, bottom soundings along potential routes should
be taken to obtain accurate profile information. If the
bottom conditions are not uniform along the route,
more detailed bottom information should be obtained.
Serious problems with bottom conditions can warrant
the consideration of alternate routes and, at the least,
dramatically increase the cost for installation.
Surface and Subsurface Currents
Surface currents must be evaluated in determining the
method and equipment to be used to install the subma-
rine cable or retrieve it. Fast surface currents might
require special anchoring or active propulsion systems
for positioning a cable ship or barge, or might preclude
the use of surface-ship cable installation methods alto-
gether. Subsurface currents can also cause a variety of
problems. If the cable is laid on the bottom without
embedment, and it becomes suspended (for example,
due to the sediment below the cable being transported
downstream), the currents can cause the cable to
vibrate. As noted above, this in turn causes chafing dam-
age to the cable armor and jacket and fatigue to the lead
sheath. In rivers and streams, it is common practice to
lay cable in an arc with the bow upstream. If the cable
does move due to the currents eroding the bottom, slack
is provided, so that the cable does not become sus-
pended. Fast subsurface currents can actually move the
whole cable downstream. In severe conditions, anchor
systems have been installed to hold the cable in position
(Edison Electric Institute 1957). These currents can also
cause movement of thermal backfill and silt, burying a
cable beneath piles of soil.
Water Temperatures and Bottom Soil Thermal
Characteristics
Water temperature and soil thermal characteristics must
be considered very early in the design stages of subma-
rine cables. Just as thermal characteristics of the soil
have a significant effect on the ampacity of land-based
cables, the same is true with submarine cables. Poor
thermal conductivity of river and lake beds is not
uncommon. Further, silt can dry out if it is in contact
with a hot cable, and as it dries, it can form a thermal
insulation that is capable of initiating thermal runaway
and breakdown of the cable insulation. The 1969 failure
of a 115-kV fluid-filled cable crossing Lake Champlain
was attributed to thermal breakdown due to high ther-
mal resistivity of the bottom soil conditions (Electrical
World 1971). Additional discussion of soil thermal char-
acteristics is provided in Section 12.2.6.
Water temperatures in deep waters might be quite cool,
but improved cable ampacity over the total circuit due
to the cold-water environment cannot automatically be
assumed. Water temperatures at the shore ends are usu-
ally higher. Further, the positive effect of cold waters
can be eliminated by the bottom sediments that cover
the cable.
If poor thermal bottom characteristics or warmer
waters present an ampacity constraint for only a portion
of an underwater crossing, such as at either end
approaching landfall, an alternative is to use a larger
conductor size in only those areas to eliminate the ther-
mal restriction.
Man-made Facilities
Man-made facilities that may represent hazards to the
cable system should also be evaluated when determining
the suitability of a route. For example, submarine cable
installations near sanitary sewers or chemical discharges
might increase corrosion problems with the armor. Pil-
ings, bridge abutments, and other structures built in
waterways often cause turbulence nearby, particularly
during high-water conditions. Turbulence can cause
cable vibration or chafing. Ice jams at bridges, piers, and
other man-made facilities can also cause scouring of a
river bottom with related damage to cables. Thus, cable
routes should avoid these structures if possible. Finally,
areas involving construction of other facilities should be
avoided simply to reduce the possibility of damage to
the cable by dig-ins, etc.
12-37
Ship Traffic and Commercial Fishing Activities
The greatest threat to the reliability of submarine cable
systems are people. Externally caused damage from
trawling, dredging, or ship anchors is the most common
problem for cables. Thus, commercially active water-
ways should be avoided if possible. If the cable route
must cross an active waterway, the risks to the cables
must be minimized. Embedding the cable to protect it is
one option to minimize risk of cable damage. Widely
separating the cable phases (with single-phase cables) to
minimize the possibility of damaging multiple cables
during a single incident should be considered to improve
circuit availability. Installing additional spare cables or
cable circuits is another alternative.
Underwater or Shore Environmental Sensitivity
Although submarine power cables have not been found
to present significant environmental problems in the
past, environmental sensitivity, both underwater and on
shore, must be considered carefully in selecting the
route, the installation method, and the cable system
itself. For example, installation can be scheduled during
seasons that are environmentally safer. Installation tech-
niques that minimize bottom soil disruption should be
evaluated. If cable fluid loss is a concern, for example,
flow limiters, special jacketing, or solid cable designs
could be evaluated. Environmentally sensitive areas
should simply be avoided if at all possible.
Tides, Weather, and Seasonal Constraints
Tides, weather, and seasonal constraints can limit the
time periods during which the cable can be installed and
when access to the cable for retrieval and repair is possi-
ble. Limited access to the cable for retrieval and repair is
the more serious of the two issues. Constraints with
regard to seasonal periods when a cable system can be
installed can be accommodated with good construction
planning in most cases. If the cable system is inaccessible
for repair for a few months each year (due to freezing,
severe sea conditions, ice, bitterly cold weather, etc.), the
system must be designed accordingly. For example, spare
cables or additional cable circuits could be installed, so
that power transmission can be maintained until the
damaged cable can be repaired when the weather will
allow repair procedures to begin. Underwater cable cir-
cuits should be rated to allow for the loss of a phase or
the total circuit for an extended period of time. Equip-
ment might also be installed to minimize the damage if
the cable must remain unrepaired on the bottom for an
extended period of time, such as flow-limiting devices on
a fluid-filled cable system or double-lead sheaths to pro-
vide additional protection for the cable core.
Tides represent a significant factor affecting the installa-
tion of submarine cables. Installation and repair are
usually scheduled during neap tides, when the difference
between high and low tide is very small and tidal cur-
rents are at a minimum.
Regulatory Restrictions
Regulatory restrictions from both federal and local
authorities should be clearly understood. Blocking of
navigable waters is often an issue that directly affects
construction and repair. For example, a cable system
floated across a river, blocking shipping traffic for a
period of time, might be acceptable if the proper author-
ities are notified in advance, and if local weather and
recreation situations allow. Also, there are often con-
cerns and restrictions associated with trenching, bottom
disruption, and spoil removal, as well as maximum pos-
sible fluid-loss levels. In most cases where navigable
waterways are concerned, the U.S. Army Corps of Engi-
neers has the ultimate jurisdiction, and can be contacted
to assist in obtaining the proper permits and under-
standing the limiting factors.
Environmental Issues
A wide range of potential environment interactions are
evaluated for submarine cable installations. Typical top-
ics considered in an Environmental Impact Statement
(EIS) submittal are listed below.
Terrestrial ecology
Aquatic ecology
Turbidity and sedimentation
Spoil disposal
Chemical quality
Biological quality
Land traffic
Water traffic
Noise
Socioeconomics
Aesthetics
Recreation
Terrestrial ecology, land traffic, noise, socioeconomics,
and recreation are primarily associated with impacts at
the land connections and terminations. Aquatic ecology
has major relevance to submarine cable operation.
Many of the impact areas (land traffic, water traffic,
noise, socioeconomics, and recreation) are associated
with construction operations and are short term in
nature. This is not to say that the short-term construc-
tion impacts are not important. In fact, these typically
have more significance in obtaining construction per-
mits, Corps of Engineers approvals, and public accep-
tance than long-term operational aspects of the cable.
12-38
Underwater Cable System Impacts on Aquatic Ecology
Turbidity, sedimentation, and spoil disposal receive sig-
nificant attention in environmental evaluations. They
are generally related to construction rather than opera-
tion, short term in duration, and moderate to insignifi-
cant in magnitude unless previously deposited toxic
substances are disturbed. Impacts related to these fac-
tors are likely to be higher in smaller, restricted bodies
of water (i.e., small river and lake crossings), where
marine life would be subjected to prolonged exposure to
high concentrations of suspended sediment.
The cable system operational impacts that are associ-
ated with the chemical and/or biological quality of the
water and would typically be evaluated for a site-specific
EIS include:
Electric and magnetic field effects
Cable heat dissipation
Cable dielectric-fluid spills.
Although there is little data on the effects of electric and
magnetic fields on the aquatic environment and com-
passes, due to the rather low fields and the localized
impact area, these effects have been considered insignifi-
cant in magnitude (Dames and Moore 1981).
Heat dissipation from the power cables is low, as are
cable/seabed interface temperatures. The possibility of
cable heat dissipation increasing the overall temperature
of the water body is of interest because this could
decrease the dissolved oxygen level in the water. How-
ever, when cabl e desi gns have been eval uated i n
restricted bodies of water, calculations have shown that
water temperature would not be elevated due to the
waters high heat capacity, motion, and latent heat of
evaporation (Dames and Moore 1981). Any effects due
to cable heat dissipation are localized directly around
the cable.
Fluid-filled cables, either pipe-type or self-contained,
also require consideration of environmental impacts due
to possible release of the dielectric fluid. Although no
fluid would be released during normal operations, envi-
ronmental concerns relate to the possible release of
dielectric fluid during a fault or accident, a highly
unlikely event. Typically used cable impregnants include
highly refined napthenic mineral oils, synthetic poly-
butenes, and alkybenzenes. These fluids have been
found to be relatively nontoxic (Chern 1981; Dames and
Moore 1981; U.S. Army 1975; Leiboritz 1972), and, as
evaporation occurs rapidly and biodegradation occurs
within 30 to 60 days, long-term effects are not an issue.
Previous EIS documents prepared for restricted and
moderate-sized bodies of water have found the possibil-
ity of a cable fluid spill not to be of significant environ-
mental concern (Chern 1981; Dames and Moore 1981).
Environmental impacts in deeper open waters are even
lower, since the open waters should see higher fluid dilu-
tion rates due to winds and tidal mixing, and the dis-
solved oxygen levels of these waters are probably better
than the restricted waters.
12.4 CABLE SYSTEM DESIGN CALCULATIONS
Underground transmission cable design requires a series
of calculations that must be done by the utility, cable
manufacturer, or a consultant. Initial system calcula-
tions such as impedances, load sharing with other trans-
mission lines, insulation coordination, etc. are described
in Chapter 16. Detailed accessory calculations, such as
electrical stresses in splices, are performed by the manu-
facturers. Utilities may specify allowable levels, but they
seldom perform the calculations. Cable installation
design calculations typically done by utilities are sum-
marized below.
Extruded-dielectric cables
Steady-state and transient/emergency ampacity
Cable-pulling tension and sidewall pressure
Sheath voltages
Pipe-type Cables
Hydraulic requirements
Forced-cooling requirements
Self-contained fluid-filled cables
Hydraulic requirements
Sheath voltages
In some cases, the utility performs trench design
calculations for controlled backfill. The Underground
Transmission Workstation, developed by EPRI in the
early 1990s, to be upgraded in the near future, provides
applications programs for each of these calculations.
The calculation procedures are summarized on the fol-
lowing pages.
12.4.1 Steady-State, Transient, and Emergency
Ampacity
Steady-state ampacity refers to the maximum current
that a cable can carry for long times, for a defined load
shape. Transient ampacity refers to allowable currents
for very short periods where all heat is stored in the
cable system, and emergency ampacity refers to allow-
able currents for longer times when heat storage of the
12-39
earth enters into the calculations. These very detailed
calculations are at the heart of a cable system design.
Chapter 11 is devoted to ampacity calculations.
12.4.2 Cable-pulling Tension and Sidewall
Pressure
Although the maximum length between splices is some-
times limited by the length of cable that can fit on a reel,
for most routes it is governed by allowable tension on
the cable conductor as the cable is pulled into the duct
or pipe, or allowable sidewall pressure as the cable goes
around a bend. Allowable clearances between cable and
duct or pipe and jam ratio are also checked during pull-
ing-tension calculations. Route information necessary
to calculate pulling tensions and sidewall pressures is
summarized below:
Section length
Radius of bends and dips
Arc length or angle of bends and dips
Position of bends and dips
Required cable system information includes:
Cable conductor material and size
Cable outside diameter
Cable weight
Cable skid wire material, if present
Pipe or duct material, size, and internal coating,
if any
Pulling rope data, including diameter and weight
COF (Coefficient of Friction) between cable and pipe
or duct
COF between pulling rope and pipe or duct
Most of this discussion addresses three conductors in a
common pipe and follows the procedures published by
Rifenburg in 1953 (Rifenburg 1953). Single-conductor
pipe-type cables for riser sections and single-conductor
SCFF or extruded-di el ectri c cabl es i n duct are
addressed as well. Three SCFF or extruded-dielectric
cables in a common duct are seldom used at transmis-
sion voltage. If calculations need to be done for this
arrangement, the procedures for pipe-type cables may
be followed, and the user can refer to Reference (EPRI
1984) for additional information.
Clearances
A minimum clearance of about 0.5 in. (12.7 mm) should
be maintained between the top of a single-conductor
cable and the pipe or duct, or between the three cables
and the pipe for a pipe-type cable. Reference (Rifenburg
1953) recommends a minimum 0.25-in. (6.4-mm) clear-
ance, but most utilities today use the larger number. The
minimum clearance for pipe-type cables occurs when
the cables are in a triangular configuration. This clear-
ance, C, is calculated from Equation 12-6.
(in. [cm])
12-6
Where:
D = nominal pipe inside diameter, in. (cm)
d = nominal cable insulation shield outside
diameter + 1.5 times skid wire thickness,
in. (cm)
The clearance for three 2500-kcmil (1267-mm
2
), 345-kV
cables with a 3.654-in. (9.3-cm) shield diameter and a
0.10 in. (2.54 mm) skidwire in a 10.25 in. (26 cm) inside
diameter pipe is 2.53 in. (6.4 cm).
Jam Ratio
If the ratio of pipe inside diameter to cable diameter is
in the range of 2.9 to 3.1 (some designers use a smaller
range of 2.95 to 3.05, although the 1957 Underground
Systems Reference Book [Edison Electric Institute 1957]
uses 2.8), there is a chance that the cables can jam in the
pipe during pulling. This is a special problem as cables
are pulled through a bend in the pipe. Since the pipe
tends to flatten slightly when the bend is made, the
cables can cross, and then they can wedge as they enter
the circular cross section beyond the bend. If the jam
ratio is below about 2.9, the cables will not cross. If the
ratio is above about 3.1, they are unlikely to wedge
because of the loose fit. The jam ratio calculation is
given in Equation 12-7.
(dimensionless) 12-7
The jam ratio is 2.82 for three 2500-kcmil (1267-mm
2
)
345-kV cables in a 10.25-in. (26-cm) ID pipe.
Pulling Tensions
Although the term tension is commonly used by cable
engineers, and is the term used in Reference (Rifenburg
1953), we actually calculate a force in pounds force, lb
f
(kg
f
). This is the force required to pull the cables into the
pipe or duct. In keeping with common usage, this book
refers to pulling tensions.
The force to pull the cable is applied to the conductor,
and allowable pulling levels are based on the conductor
material and site conditions. Acceptable levels are
10 lb
f
/kcmil (8.9 kg
f
/mm
2
) for copper conductors and
2
) (
1 ) (
2
1
366 . 1
2

+ =
d D
d
d D d
D
C
=
d
D
Ratio Jam 05 . 1
12-40
6 lb
f
/kcmil (5.4 kg
f
/mm
2
) for aluminum conductor
cables (EPRI 1983; Aabo et al. 1979).
Therefore, a 2500-kcmil (1267-mm
2
) copper conductor
can accept a 25,000-lb
f
(11,400-kg
f
) tension. Experimen-
tal evidence from EPRI experiments has shown that a
conservative assumption is that two of the three cables
in a pipe-type system carry the full tension (EPRI 1983).
The total allowable force on three 2500-kcmil conduc-
tors is therefore 50,000 lb
f
(22,000 kg
f
).
The Rifenburg paper addresses copper conductors only,
and proposes a maximum stress of 10,000 lb
f
/sq.in.
(7850 lb
f
/kcmil; 7.0 kg
f
/mm
2
) (Rifenburg 1953). All
three conductors are assumed to share the tension, but
the 10,000 lb
f
/sq.in. is reduced by 15% to account
for unbalance in tension. Working through these num-
bers gives the same 50,000-lb
f
(22,000-kg
f
) tension for
2500-kcmil cables as the EPRI results.
The tension, T, for a single cable pulled into a straight
pipe or duct is calculated from
T = K
0
wL lb
f
(kg
f
) 12-8
Where:
K
0
= basic coefficient of friction, dimensionless.
w = total cable weight, lb/ft (kg/m).
L = length, ft (m).
Coefficient of Friction
The coefficient of friction (COF) between cable and pipe
or duct is a very important parameter for pulling-ten-
sion calculations. The addition of a lubricant can reduce
the COF significantly for SCFF or extruded-dielectric
cables in duct. For pipe-type cables, the lubricant is typi-
cally a small amount of the filling fluid that will be used
in the pipe. Many commercial lubricants are available
for distribution-voltage extruded-dielectric cables, and
most can be used for transmission cables. The user
should check with the cable manufacturer before using
any lubricant. For pipe-type cables, Table 12-3 gives
COF values for different skid wire materials in steel line
pipe having an internal epoxy coating. A range of COF
values at varying installation temperatures, resulting
from EPRI-sponsored research (EPRI 1984) on distribu-
tion cables in the 1980s, is given in Tables 12-4 and 12-5.
These figures assume a soap-and-water-based lubricant.
Table 12-3 Basic Dynamic Coefficient of Friction for
Different Skid Wire Materials in Coated Steel Pipe (Aabo et
al. 1979)
Skid Wire Material COF
Brass 0.17
Stainless Steel, Bright 0.13
Stainless Steel, Matte 0.11
Zinc 0.15
Table 12-4 Basic Dynamic Coefficients of Friction for
Straight Pulls and Bearing Pressures Less than 150 lb
f
/ft
(224 kg
f
/m) (Soap-and-Water-Based Lubricants)
Duct Material Cable Sheath/Jacket
COF
75F
a
a. Installation temperatures.
25F
a
PVC
XLPE
PE
PVC
Neoprene
CN
Lead
0.40
0.40
0.50
0.90
0.40
0.25
0.40
0.35
0.25
0.55
0.40
0.25
PE
XLPE
PE
PVC
Neoprene
CN
Lead
0.45
0.25
0.30
0.65
0.20
0.20
0.35
0.20
0.20
0.45
0.20
0.25
Fiber
XLPE
PE
PVC
Neoprene
CN
Lead
0.30
0.25
0.40
0.40
0.40
-
0.20
0.35
0.20
0.30
0.35
-
Concrete
XLPE
PE
PVC
Neoprene
CN
Lead
0.30
0.35
0.55
0.50
-
0.55
-
-
-
-
-
-
Transite
XLPE
PE
PVC
Neoprene
CN
Lead
0.70
0.70
0.70
1.00
-
-
-
0.35
0.35
0.35
-
-
Steel
XLPE
PE
PVC
Neoprene
CN
Lead
0.60
0.50
0.65
1.05
0.50
-
0.45
0.50
0.40
0.70
0.50
-
Table 12-5 Basic Dynamic Coefficient of Friction at Bends
Where Sidewall Pressure is 150 lb
f
/ft (224 kg
f
/m) or Greater
(All Lubricants)
Cable Outer Covering Duct Material COF
XLPE, PE, N PVC, PE, Concrete 0.15
PVC PVC, PE, Concrete 0.30
XLPE, PE, N Steel 0.25
PVC Steel 0.30
Lead Steel 0.20
12-41
Better lubricants have recently become available.
Table 12-6 shows the anticipated COF for Polywater
J lubricants. This lubricant has been the mostly com-

monly used. Other lubricants with better or equal COF
values are also available. The lubricant manufacturers
should be consulted for updated COF values when the
utility is designing a specific installation. No COF val-
ues are available for fiberglass duct. When performing
pulling calculation for fiberglass duct, the designer
should consult with the lubricant manufacturers.
Weight Correction Factor
The weight correction factor, W
C
, accounts for the
wedging and resulting increase in effective COF that
takes place when three cables are pulled into the pipe.
(Three cables are rarely installed in the same duct for
SCFF or extruded-dielectric transmission cables.) The
weight correction factor is a function of cable and pipe
diameters. A triangular configuration is assumed for
ratios of pipe inside diameter to cable outside diameter
of 2.5 or lower, and many utilities use the triangular
configuration for all calculations. Some users average
the numbers for cradled and triangular configuration
when the ratio is greater than 2.5. W
C
for cables in trian-
gular configuration is given in Equation 12-9.
The weight correction factor for cables in cradled con-
figuration is given in Equation 12-10.
For a 2500-kcmil (1067-mm
2
), 345-kV cable with a
3.804-in. (9.67-cm) diameter in a 10.25-in (26-cm) ID
pipe, W
C
is 1.24 for triangular configuration and 1.46
for cradled configuration.
The effective COF, K, equals W
C
K
O
, and the allowable
pulling tension is shown in Equation 12-11.
12-11
Where:
K
O
= basic coefficient of friction (dimension-
less).
w = total cable weight (lb/ft [kg/m]).
L = length (ft[m]).
Dips and Bends
Equation 12-11 gives the pulling tension for straight
pipe sections. Longer equivalent lengths are calculated
to account for dips and bends. Figure 12-33 shows a
typical pull. The increase in length, L
1
, for major hori-
Table 12-6 Basic Lubricated and Unlubricated Coefficient of
Friction for Polywater
J
Conduit Jacket
Lubricated
COF
Unlubricated
COF
Aluminum
CPE
HDPE
Hypalon
LLDPE
Nylon
PVC
Polypropylene
XLPE
0.12
0.07
0.06
0.11
0.13
0.08
0.08
0.08
0.58
0.14
0.75
0.35
0.38
0.48
0.42
0.38.
EMT
CPE
HDPE
Hypalon
LLDPE
Nylon
PVC
Polypropylene
XLPE
0.42
0.19
0.38
0.24
0.28
0.43
0.24
0.25
0.65
0.42
1.10
0.43
0.33
0.80
0.45
0.42
IMC
CPE
Hypalon
LLDPE
Nylon
PVC
Polypropylene
XLPE
0.09
0.07
0.06
0.12
0.07
0.07
0.09
0.80
0.90
0.30
0.28
0.60
0.38
0.58
PVC
CPE
HDPE
Hypalon
LLDPE
Nylon
PVC
Polypropylene
XLPE
0.35
0.09
0.15
0.17
0.24
0.10
0.09
0.14
0.85
0.15
0.58
0.30
0.38
0.60
0.33
0.23
RIGID
STEEL
CPE
HDPE
Hypalon
LLDPE
Nylon
PVC
Polypropylene
XLPE
0.30
0.21
0.18
0.33
0.40
0.15
0.13
0.29
0.85
0.35
0.90
0.68
0.50
0.55
0.50
0.50
C
2
1
W =
d
1-
D-d

2
3
4
1

+ =
d D
d
W
c
C O f f
T = W K wL or T = KwL lb (kg )
Figure 12-33 Typical section for calculating equivalent length
of horizontal bends (Rifenburg 1953) (courtesy AIEE).
12-42
zontal bends such as Bend 1 in the figure, results in a
new length, L
2
, as calculated in Equation 12-12.
12-12
Where:
L
1
= equivalent length at entrance to bend,
ft (m).
L
2
= equivalent length at exit from bend, ft (m).
K = effective COF, dimensionless.
R = radius of bend, ft (m).
= angle of bend or dip, in degrees.
No correction is needed if the radius of a dip is less than
the coefficient of friction times the length of the dip. The
increase in length, L
2
, for more severe vertical dips, as
shown in Figure 12-34, is given in Equation 12-13.
12-13
The total equivalent length is the straight-line length,
plus increases due to major dips and bends. Also, an
equivalent length of about 50 ft (13 m) is added to
account for the force needed to pull the cables off the
reel; see Equation 12-14.
12-14
Appendixes IIII of Reference (Rifenburg 1953) provide
a series of sketches and equations for calculating cable-
pulling tensions in many typical dips and bends.
Sidewall Pressure
Sidewall pressure (SWP) is the crushing action that
occurs when cables are pulled around bends. SWP is
defined as the tension, T, in lb
f
(kg), divided by the bend
radius, R, in feet (m) as shown in Equation 12-15.
12-15
SWP is affected by triangular or cradled cable configura-
tion and the ratio of cable to pipe diameter. SWP for tri-
angular configuration is divided equally between the two
bottom conductors, and is shown in Equation 12-16,
where W
C
is calculated from Equation 12-9.
12-16
The SWP calculation for cradled configuration is given
in Equation 12-17, where W
c
is calculated from Equa-
tion 12-10.
12-17
The cable manufacturer should be consulted before
sidewall pressure calculations are finalized. A range of
typical allowable values is given in Table 12-7.
As noted in the EPRI research (EPRI 1983), these side-
wall pressures assume a smooth pipe. Ripples on the
inside of a field bend can cause cable damage at a lower
SWP than the values listed in Table 12-7.
When the cable tensions and sidewall pressures are cal-
culated, the effect of the pulling rope should also be
evaluated. The tension measured during a pulling oper-
ation is the pulling-rope tension, so the rope tension
also needs to be estimated to use calculated values as a
criterion for evaluating the actual cable-pulling tension.
For pipe-type cable installation, a steel rope COF of
0.250.5 is often used. The effect of the rope on the pipe
or duct is discussed in Chapter 13.
Hydraulic Calculations
Fluid volume and pressure calculations are important
for HPFF and SCFF cable design and operation.
Proper fluid pressures must be maintained over the full
length of the circuit for a wide range of loading condi-
tions and rates of load changes. The fluid expelled from
the cable system during increasing-load periods must be
stored in a reservoir tank, and the tank capacity must be
sufficient to provide fluid during cable contraction as

2
2
1 2 1
R
L = L coshK + L + sinhK ft (m)
K
Figure 12-34 Typical section for calculating equivalent
length of vertical dips (Rifenburg 1953) (courtesy AIEE).
4K 4K 3K K
4 1
R
L = L e + e - 2e + 2e -1 ft (m)
K

tot 1 2 3 n
L =L +L +L +...L + 50 ft (13m) ft
f f
lb kg T
SWP =
R ft m

C f f
W lb kg T
SWP = 3
2 3R ft m

( )
f f
C
lb kg T
SWP = 3 W 2
3R ft m

Table 12-7 Range of Allowable Sidewall Pressures for
Different Cable Types
lb
f
/ft Kg
f
/m
Impregnated Paper 1000 1490
Laminated Paper-Polypropylene 1000 1490
Crosslinked Polythylene
a,b
a. With overall jacket.
2000 2980
Ethylene-Proplylene Rubber
a,b
b. Reduce to 1500 lb
f
/ft (2235 kg
f
/m) if jacket with a wire or
tape shield is not applied tightly.
2000 2980
12-43
well as to supply a possible leak. Refer to Chapter 9 for
typical reservoir tank sizing as a function of fluid vol-
ume changes.
Fluid demand requirements, fluid flows, and pressure
changes are complex calculations, which require the use
of empirical data, usually in the form of curves. Detailed
calculation procedures for pipe-type cables are given in
a 1955 AIEE committee report (AIEE 1955), and those
for SCFF cable are given in a 1956 paper (Buller et al.
1956). This book reviews the general calculation proce-
dures, but the two referenced papers, and their back-
ground references, should be used for comprehensive
hydraulic calculations.
High-Pressure Fluid-Filled Cables
The fluid requirement for an HPFF cable pipe is calcu-
lated from Equation 12-18. Volume in liters may be
obtained by multiplying gallons by 3.785.
12-18
Where:
L = line length, ft.
D = pipe inside diameter, in.
d
S
= cable shield diameter, in.
This volume is typically increased by 510% to account
for extra fluid in joint casings and piping, fluid absorp-
tion by the cable as pressure is increased, and tempera-
ture changes as the fluid leaves tank trucks and enters
the cable pipe. The amount to fill the reservoir tank to
the desired level, plus a buffer in the tank for possible
fluid leaks, must be added to this volume.
Equation 12-19 gives the volume change, V, with tem-
perature changes for a pipe-type cable:
V = Cd
c
2
T
c
+ 0.012(d
s
2
d
c
2
) T
i

+ 0.0070(D
2
3d
s
2
) T
o

0.0003D
2
T
p
in
3
/ft 12-19
Where:
C = constant, 0.0060 for concentric stranded
conductors, or 0.0036 for compact seg-
mental conductors.
d
c
= conductor diameter, in.
d
s
= cable shield diameter, in.
T
c
= conductor temperature change, C.
T
i
= insulation temperature change, C.
T
o
= fluid temperature change, C.
T
p
= pipe temperature change, C.
The last term in Equation 12-19 is negative, since pipe
expansion with increasing temperatures offsets some of
the fluid expansion. The temperature differences are
usually taken from the coldest ambient earth tempera-
ture to the maximum operating temperature for each
component. Calculation procedures in Chapter 11 may
be used to calculate conductor, insulation, fluid, and
pipe temperatures for a given load condition. The safest
approach is to use the highest emergency operating tem-
perature, which is a 105C conductor temperature.
Dielectric losses do not enter into normal daily volume
changes since dielectric losses are present whenever the
line is energized. However, dielectric losses are taken
into account when sizing pumps, reservoirs, and pipe
sizes for pressure drop considerations, since the conser-
vative assumptions are either de-energizing a fully
loaded cable, or energizing a cold cable to full load.
Volume changes calculated from Equation 12-19 are
given in cubic inches per foot of cable length. For U.S.
units, total volume change is obtained from multiplying
by the length in feet. Number of gallons is obtained by
dividing cubic inches by 231, and number of liters is
obtained by multiplying gallons by 3.785.
Fluid demand, a, determines the required pump capac-
ity to ensure that fluid can be supplied during rapid
cool-down. Fluid demand also determines the required
free fluid area in the pipe to ensure that pressure
increases are not excessive during rapid heating. Fluid
demand is determined by differentiating Equation 12-19
with respect to time, as shown in Equation 12-20. Vol-
ume change in liters/meter-sec may be obtained by mul-
tiplying Equation 12-20 results by 0.054.
12-20
Solving Equation 12-20 requires knowledge of tempera-
ture changes with time for each cable system compo-
nent. Reference (AIEE 1955) provides curves for several
cable and pipe sizes. Rigorous temperature change cal-
culations may be performed using procedures described
in Reference (Buller et al. 1956). Some commercial
ampacity programs are able to calculate fluid demand,
and the EPRI Underground Transmission Workstation
includes hydraulic calculations including fluid demand.
Maximum fluid demand values given in Reference
(AIEE 1955) range from 30 to 45 x 10
-6
in.
3
/watt-sec per
foot (1.62.4 l/watt-sec per meter) for HPFF cables up
to 230 kV. Fl ui d demand per mi l e i s, therefore,
0.33 gal/min for a typical ohmic loss of 8 watts per con-
ductor foot (demand per km is 0.75 l/min for 26 W/m).
Pump capacity should be greater than the calculated
f l ui d demand. Some ut i l i t i es s t andar di z e on
1012 gal/min (3845 l/min) pumps, since there is little
2 2
s
12
V=L (D 3d ) gal
4 231

3
d
a = Vin. /ft-s
dt

12-44
disadvantage to oversizing pumps, and larger pumps
provide the opportunity for slow circulation to reduce
hot spots along the route.
The design engineer should also calculate fluid pres-
sures along the route for all loading conditions to ensure
that pressures are not too low during cooling transients,
or too high during heating transients. Most utilities
maintain a nominal 200 psi (1380 kPa) pressure at the
high point of the circuit. Pressure at any other point
must also include the hydrostatic head, P
h
, due to eleva-
tion changes as calculated from Equation 12-21.
P
h
= 0.434 h psi 12-21
Where:
= specific gravity, dimensionless.
h = elevation difference, ft.
Pressures in kPa may be obtained by multiplying psi
by 6.895.
Additional pressures arise from dielectric fluid motion
during fluid expansion and contraction. This friction
pressure, P
f
is a function of fluid viscosity, distance,
cable/pipe size, and cable position in the pipe, as shown
in Equation 12-22.
12-22
Where:
x = distance from closed end, ft.
= fluid viscosity, centipoise.
a = fluid demand, from Equation 12-20,
in.
3
/ft-s.
K = constant, depending upon cable/pipe
diameter ratio.
Values for the constant, K, can be obtained from
Figure 12-35. Fluid viscosity should be taken from Fig-
ure 3-35, for the coldest temperature anticipated during
cable system operation. Note that frictional pressure
drop is much less of a concern with the low-viscosity
pipe-filling fluids used today, compared to the much
higher viscosity fluids used in the 1950s.
Pressure drop due to rapid dielectric fluid circulation
will be in addition to the values calculated above. It is
possible for a differential pressure of 200 psi (1380 kPa)
to be required to force a 300 gal/mm (1135 l/min) flow in
a cable pipe. Refer to Chapter 9 for these calculations.
Self-Contained Fluid-Filled Cables
Calculation principles and procedures for SCFF cables
are quite similar to those for HPFF cables, and the pro-
cedures of Reference (Buller et al. 1956) closely parallel
those of (AIEE 1955). For single-conductor SCFF
cables, fluid expansion and contraction occur princi-
pally in the conductor core. The fluid requirement for
one phase of a SCFF cable is calculated from
Equation 12-23. Volume in liters may be obtained by
multiplying gallons by 3.785.
12-23
Where:
L = line length, ft.
d
O
= cable core diameter, in.
This calculated volume is typically increased by about
25% to account for fluid in joints, plus fluid in the insu-
lation and under the sheath. The amount required to fill
the reservoirs to the desired level and to provide a buffer
in the reservoir for possible fluid leaks must be added to
the calculated volume. The sum of these volumes is the
total fluid volume for one conductor in an operating
cable system. The total volume for three phases is three
times greater than that calculated for a single cable.
However, each conductor typically has its own oil reser-
voir. In addition to this amount, several times as much
fluid may be used during cable installation to provide
fluid flow during splicing and terminating. The cable
manufacturer should be consulted when determining
the fluid required during installation.
2 6
f 4
K
P 35.5x a x 10 psi
D

=

Figure 12-35 Factor K as a function of cable/pipe diameter
(From Figure 3, Reference [AIEE 1955]) (courtesy AIEE).
2
o
12
V L d gal
4 231

=

12-45
Fluid demands and pressures are calculated using pro-
cedures similar to those described previously for HPFF
cables. Refer to Reference (AIEE 1955) for detailed
equations and empirical curves.
Trench Design for Corrective Backfill
Because all soils do not have the same thermal resistiv-
ity, an underground transmission line designer must
evaluate the effect that the native thermal resistivity will
have on the circuit ampacity. For example, if the native
soil has a thermal resistivity of 150 C.cm/watt
(1.5 K.m/W), the resulting ampacity for a given circuit
will be lower compared to an identical circuit placed in a
native soil of 90 C.cm/watt (0.9 K.m/W). The main way
to improve the overall effective thermal resistivity seen
by a cable system is to install a thermal backfill material
around the duct/pipe. Equation 12-24 can be used to
calculate the effective thermal resistivity for various
widths and depths of backfill (Edison Electric Institute
1957, pp. 1015). This equation is good for a trench
width-to-depth ratio of about 3 to 1.
Effective =
f
+ (
o
f
)
-
12-24
Where:
w = width of trench.
h = depth to the bottom of the structure.
= natural log.
f
= of fill.
o
= of original.
Figure 12-36 shows the effect of changing the width of
the controlled backfill.
Additional calculation procedures for the design of a
trench with corrective backfill are given in References
(Brookes and McGrath 1960; Williams et al. 1993). An
effective overall soil thermal resistivity is assumed, and
trench size is determined along the different sections of
the route having varying thermal resistivities of the
native soil.
h
w
Figure 12-36 Effect of changing the width of a trench (courtesy POWER Engineers, Inc.).
12-46
REFERENCES
Aabo, T., J. A. Moran, Jr., and J. R Shimshock. 1979.
Increased Pipe Cable Section Lengths. Proceedings of
the 1979 IEEE T&D Conference. April 16.
Aabo, T., and J. A. Moran, Jr. 1988. Thermomechanical
Bending Effects on Extra-High Voltage Pipe-Type
Cables. July. EPRI Report no. EL-5880.
AIEE. 1955. Oil Flow and Pressure Drop Calculations
for Pipe-Type Cable Systems. AIEE Committee
Report. AIEE Transactions. April. pp. 251259.
ASCE 2003 Standard Guideline for the Collection and
Depiction of Existing Subsurface Utility Data
CI/ASCE Standard 38-02.
ASTM. 1978Moisture-Density Relations of Soils,
Using 5.5-lb (2.5 kg) Rammer and 12-in. (304.8 mm)
Drop. ASTM Standard D 698.
Barnes, C. C. 1977. Submarine Telecommunication and
Power Cables. Peter Peregrinus Ltd.
Bazzi, G. 1982. Advanced Technologies in HVAC and
HVDC Submarine Power Cable Transmission. Pre-
sented at Fourth Conference on Electric Power Supply
Industry (CEPSI). Bangkok. November 2226.
Blau, D. 1975. Study of Environmental Impact of
Underground Transmission Systems. May. EPRI Report
no. 7826.
Brookes, A. S., and M. H. McGrath. 1960. Soil Ther-
mal Characteristics in Relation to Underground Power
CablesPart V. Practical ApplicationsTrench Design
and Construction. AIEE Transactions. No. 51. Decem-
ber. pp. 792856.
Buller, F. H., J. H. Neher, and F. O. Wollaston. 1956.
Oil Flow and Pressure Calculations for Self-Contained
Oil-Filled Cable Systems. AIEE Transactions on Power
Apparatus and Systems. No. 23. April. pp. 180194.
Chaaban, M. et al. 1991. Evaluation of HVDC Cables
for the St. Lawrence River Crossing of Hydro Quebecs
500 kV DC LinePart III: Thermal Behavior. Pro-
ceedings of the 1991 IEEE T&D Conference. Dallas, TX.
September 2227. pp. 117121.
Chamberlin, D. M. 1980. Operating Experience with
The Norwalk, CTNorthport, NY 138 kV Submarine
Cable. Presented at Transmission & Distribution Com-
mittee Meeting No. 57. Engineering Operations Divi-
sion. Electric Council of New England. May 22.
Chern, C. 1981. Ocean Thermal Energy Conversion
(OTEC) Project, Bottom Cable Protection Study, Envi-
ronmental Characteristics and Hazards Analysis.
Washington, D.C. Ocean Engineering and Construction
Project Office. Chesapeake Division. Naval Facilities
Engineering Command. February 19.
CIGRE. 2002. Working Group 19 of CIGRE Study
Committee 21. General Guidelines for the Integration
of a New Underground Cable System in the Network.
March.
Dames and Moore. 1981. Draft-Study Documentation
Report Ashe-State Environmental Impact Statement
Crow Butte Crossing. Prepared for Bonneville Power
Administration. March.
Edison Electric Institute. 1957. Underground Systems
Reference Book.
Eich, E. D. 1976. Installed Cost Comparison for Self-
Contained and Pipe-Type Cable. November. EPRI
Report no. EL-935.
Electrical World. 1971. Underwater Cable Fault Due to
Thermal Runaway. May 1. pp. 5253.
Engelhardt, J. S., and D.W. Purnhagen. 1991. Dynamic
Rating and Underground Monitoring System. July. EPRI
Report no. EL-7341.
EPRI. 1977. Backfill Materials for Underground Power
Cables. Phase I, EPRI Report no. EL-0506.
EPRI. 1981. Thermal Property Analyzer. EPRI Report
no. EL-2128.
EPRI. 1983. Increasing Pipe Cable Section Lengths.
March. EPRI Report no. EL-2847.
EPRI. 1984. Maximum Safe Pulling Lengths for Solid
Dielectric Insulated Cables. February. EPRI Report no.
EL-3333.
EPRI. 1997. Soil Manual for Underground Transmission.
EPRI Report no. TR-108919.
12-47
Foxall, R. G., K. Bjorlow-Larsen, and G. Bazzi. 1984.
Design, Manufacture and Installation of a 525-kV
Alternating Current Submarine Cable Link from
Mainland, Canada to Vancouver Island. CIGRE
21-04. 10 p.
Gazzana-Priaroggia, P., J. H. Piscioneri, and S. W. Mar-
golin. 1971. Long Island Sound Submarine Cable
Interconnection. IEEE Spectrum. Vol. 8. No. 10. Octo-
ber. pp. 6371.
Gazzana-Priaroggia, P., and C. Mascio. 1973. Contin-
uous Long Length AC and DC Submarine HV Power
Cables, The Present State of The Art. IEEE Paper T-
73-127-8. Presented at IEEE Power Engineering Society
Winter Meeting. January 28February 2.
Ghafurian, A. R., H. Chu, et al. 1989. Detection and
Location of Dielectric Fluid Leaks on Pipe-Type Cables
Systems. IEEE Transactions on Power Delivery. July.
pp. 14991503.
Gillette, R. W. 1959. Installation of 69-kV High-pres-
sure Gas-filled Pipe-type Feeders on Queensboro
Bridge, New York City. IEEE Insulated Conductors
Committee Minutes. Spring. Appendix F.
Hatcher, C. T., R.W. Gillette, and R.W. Burrell. 1966.
345-kV Underground Transmission on the Consoli-
dated Edison Company of New York System. IEEE
Transactions on Power Apparatus and Systems. Vol. 85.
No.4. April. pp. 353360.
IEEE. 1981. IEEE Standard 422-1981. IEEE Guide
for Soil Thermal Resistivity Measurements. The Insti-
tute of Electrical and Electronics Engineers, New York.
IEEE. 1987. IEEE Standard 575-1988. IEEE Guide
for the Application of Sheath-Bonding Methods for
Single-Conductor Cables and the Calculation of
Induced Voltages and Currents in Cable Sheaths.
The Institute of Electrical and Electronics Engineers.
New York.
IPCEA. 1976. Insulated Power Cables Engineers
Association. Ampacities Including Effect of Shield
Losses for Single-Conductor Solid-Dielectric Power
Cable, 15-kV through 69-kV. IPCEA Publication
P-53-426. May.
Iwata, Z. et al. 1991. Heat Pipe Local Cooling System
Applied for 145 kV Transmission Line in Copenhagen.
Proceedings of the 1991 T&D Conference. Dallas, TX.
September 2227. pp. 5260.
Kozak, S., J. T. Corbett, and F. J. Bender. 1965. Fea-
tures of the New 138-kV Self-Contained Oil-Filled
Cable Systems. IEEE Transactions on Power Apparatus
and Systems. Vol. 94. No. 3. May.
Luoni, C., and P. Anelli. 1976. Armor Corrosion in
Single Core Submarine AC Cables. IEEE Paper No. A
76 190-9. Presented at IEEE Power Engineering Society
Winter Meeting & Tesla Symposium. New York. Janu-
ary 2530.
Marquez, R. D., et al. 1981. 230 kV Self-Contained
Oil-Filled Cable Line Installed Underneath a Bridge
Located in Maracaibo, Venezuela. IEEE Transactions
on Power Apparatus and Systems. Vol. 100. No. 7. July.
pp. 31533165.
Martin, Jr., M. A., R. A. Bush, et al. 1981. Practical
Aspects of Applying Soil Thermal Stability Measure-
ments to the Rating of Underground, Power Cables.
IEEE Transactions on Power Apparatus and Systems.
Vol. 100. No. 9. September. pp. 42364249.
Minemura, S., and Y. Maekawa. 1989. 500-kV Oil
filled Cable Installed on Bridges. Proceedings of the
1989 IEEE T&D Conference. New Orleans. Paper 89TD
354-2-PWRD.
Moran, Jr., J. A., T. Aabo, and J. F. Shimshock. 1984.
Thermo-Mechanical Bending Effects in EHV Pipe-
Type Cables. Proceedings of the 1984 IEEE T&D Con-
ference. Kansas City. April 29May 4.
Morgareidge, K. 1965. Feeding Studies with Poly-
vis-SH in Rats and Dogs. Food and Drug Research
Laboratories, Inc. March 12.
Nakagawa, H. et al. 1973. Installation of 275-kV
XLPE Cables in the Long and Steep Slope Tunnel.
Vol. 102. No. 12. December.
NASTT. 2004. Horizontal Directional Drilling Con-
sortium HDD Good Practices Guidelines. American
Society for Trenchless Technology.
Patton, R. N., S. K. Kim, and R. Podmore. 1979.
Monitoring and Rating of Underground Power
Cables. IEEE Transactions on Power Apparatus and
Systems. Vol. 98. No. 6. November/December.
pp. 22852293.
12-48
Radhakrishna, H. S., and J. E. Steinmanis. 1981. Ther-
mal Resistivity Survey of Lake Erie Sediments for the
Ontario Hydro-GPU Interconnection. Symposium on
Underground Cable Thermal Backfill
Ray, J. J., C. Arkell, and H. W. Flack. 1974. 525-kV
Self-Contained Oil-Filled Cable Systems for Grand
Coulee Third Powerplant Design and Development.
March/April. pp. 630639.
Rifenburg, R. C. 1953. Pipe-line Design for Pipe-type
Feeders. AIEE Transactions. Vol. 9. December.
Rueckert, J. J., and J. I. Bien, Jr. 1979. 230-kV Cable
Crossing of Baltimore Harbor. IEEE Power Engineer-
ing Society Transmission and Distribution Conference
and Expo. Atlanta, GA. April 16.
Saleeby, K. E., W. Z. Black, and J.G. Hartley. 1979.
Effective Thermal Resistivity for Power Cables Buried
in Thermal Backfill. IEEE Transactions on Power
Apparatus and Systems. Vol. 98. No. 6. Novem-
ber/December.
Sandiford, P. 1981. Cable Backfill MaterialsState-of-
the-Art. Symposium on Underground Cable Thermal
Backfill. September 17 and 18. pp. 39.
Shannon, W. L., and W. A. Wells. 1947. Tests for Ther-
mal Diffusivity of Granular Materials. ASTM Proce-
dures. Vol. 47. Pp. 10441055.
Sheirer, E. W., and L. Winitsky. Cable Installations
on Bridges. AIEE Transactions. Vol. 77. Part III.
pp. 3942.
Soulsby, D. R., and A. J. Donovan. 1981. The Effect of
Backfill of the Temperature Distribution in a Buried
Circuit. Symposium on Underground Cable Thermal
Backfill. September 17 and 18. pp. 167180.
Steinmanis, J. E. 1981. Thermal Property Measure-
ments Using a Thermal Probe. Symposium on Under-
ground Cable Thermal Backfill. September 1718.
Tarpey, J. W. et al. 1991. Installation of a Solid Dielec-
tric 138-kV Underground Transmission SystemFrom
Concept to Completion. Proceedings of the 1991 IEEE
T&D Conference. Dallas, Texas. September 2227.
U. S. Army. 1975. Final Environmental Statement,
Applicants Environmental Report, and Responses
of Commenting Agencies. Documents submitted to
the U.S. Army Engineer District, New York. For an
environmental statement related to the Con Edison
Long Island Lighting Companys Power Transmission
Cable Crossing of Long Island Sound, Hempstead
Harbor, New York.
Weedy, B. M. 1988. Thermal Design of Underground
Systems. Chapter 8. Chichester, England and New
York, NY: John Wiley & Sons, Ltd.
Williams, J. A., S. Kozak, and T. J. Rodenbaugh. 1983.
Leak Location Methods for HV Underground
Cables. IEEE Transactions on Power Apparatus and
Systems. Vol. 102. July. pp. 20292037.
Williams, J. A., D. Parmar, and M.W. Conroy. 1993.
Controlled Backfill Optimization to Achieve High
Ampacities on Transmission Cables. IEEE 1993 Win-
ter Power Meeting. Columbus, Ohio.

Epri Capitulo12

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Epri Capitulo12

Uploaded by

Copyright:

Available Formats

EPRI Underground Transmission Systems Reference Book

J lubricants. This lubricant has been the mostly com-

You might also like